Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
The Indian Ocean Tsunami
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
BALKEMA – Proceedings and Monographs in Engineering, Water and Earth Sciences
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
The Indian Ocean Tsunami
Edited by Tad S. Murty Department of Civil Engineering, University of Ottawa Ottawa, Canada
U. Aswathanarayana Mahadevan International Centre for Water Resources Management Hyderabad, India
N. Nirupama Atkinson School of Administrative Studies, York University, Toronto, Canada
LONDON / LEIDEN / NEW YORK / PHILADELPHIA / SINGAPORE
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Taylor & Francis is an imprint of the Taylor & Francis Group, an informa business This edition published in the Taylor & Francis e-Library, 2007. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.”
© 2007 Taylor & Francis Group, London, UK All rights reserved. No part of this publication or the information contained herein may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, by photocopying, recording or otherwise, without written prior permission from the publishers. Although all care is taken to ensure integrity and the quality of this publication and the information herein, no responsibility is assumed by the publishers nor the author for any damage to the property or persons as a result of operation or use of this publication and/or the information contained herein. Published by: Taylor & Francis/Balkema P.O. Box 447, 2300 AK Leiden, The Netherlands e-mail:
[email protected] www.balkema.nl, www.taylorandfrancis.co.uk, www.crcpress.com Library of Congress Cataloging-in-Publication Data The Indian Ocean Tsunami / editors, Tad S. Murty, U. Aswathanarayana, N. Nirupama. p. cm. Includes index. ISBN-13 978-0-415-40380-1 (hardcover : alk. paper) 1. Indian Ocean Tsunami, 2004. 2. Tsunamis–Indian Ocean. 3. Tsunamis. I. Murty, T. S. (Tadepalli Satyanaraynan), 1938- II. Aswathanarayana, U. III. Nirupama, N. GC222.I45I53 2006 551.46’37156–dc22 2006020259 ISBN 0-203-96443-8 Master e-book ISBN
ISBN-13: 978 0 415 40380 1 (Print Edition)
Cover Illustration
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Schematic sketch map showing the pathway of the tsunami (thicker lines indicate higher intensity), and the projected location of the automated Deep Ocean Tsunami buoys (courtesy: Prem Kumar, NIOT, Chennai, India).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
In Remembrance
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
To remember the enormity of the human suffering involving the families of hundreds of thousands of people killed in the tsunami, the Editors could do no better than to reproduce the poignant note of Carol Amaratunga and the photograph (courtesy: Dr. Paul Gully) of the physical destruction wrought by the tsunami.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
A Note of Reflection and Dedication
A few days before the Asian Tsunami of December 26, 2004, I was in southern Sri Lanka attending a conference on Sri Lankan Studies at the University of Ruhuna. I took a few quiet moments at sunrise to observe artisanal fishing boats negotiating the waves near the Matara Rest House – an old colonial beach bungalow less than 30 m from the sea. The wrap-around Sea and azure sky that early morning were breath-taking. Nevertheless, for a fleeting moment, I was chilled by the realization of how completely exposed we were to a vast ocean stretching from the southern tip of Sri Lanka to Antarctica. The moment passed, and I rationalized that the bungalow had occupied this shoreline for the better part of a century. Five days later, the Matara Rest House was hit by one of the strongest tsunamis ever recorded, along with hundreds of homes, schools, businesses, and places of worship. The coastal devastation stretched for virtually hundreds of kilometers. In the weeks following the tsunami, Sri Lanka reported 31,229 confirmed deaths, 4093 people missing and more than 15,686 people injured. In this age of globalization and holiday eco tourism, nationals from 73 countries died. Millions of coastal residents were displaced in the 12 countries directly affected by the tsunami. The long-term psychosocial and intergenerational impact of the Asian Tsunami will be experienced for decades. This book is dedicated to the memory of citizens of Matara, Sri Lanka, and to the thousands of women, children and men whose lives were lost that day. This work is also dedicated to those who survived and prevailed. It is with great courage and resiliency that they rebuild their lives, families and communities. Carol Amaratunga University of Ottawa, Canada
Dedication
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Dedicated to Dr. R.A. Mashelkar, F.R.S., in appreciation of his indefatigable services in advancing applied science in India and the Developing countries.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Contents
List of figures List of tables Foreword Prologue Preface
xv xxvii xxix xxxi xxxiii
PART 1
1
1 A Historical Account of the Earthquakes and Tsunamis in the Indian Ocean b.k. rastogi
3
Geostructural Environment of Tsunami Genesis
2
Impact of Coastal Morphology, Structure and Seismicity on the Tsunami Surge k.s.r. murthy, v. subrahmanyam, g.p.s. murty, and k. mohana rao
19
3 Tsunamigenic Sources in the Indian Ocean: Factors and Impact on the Indian Landmass r.k. chadha
33
4
Paleo-Tsunami and Storm Surge Deposits k. arun kumar, h. achyuthan, and n. shankar
49
5
Overview and Integration of Part 1 u. aswathanarayana (editor)
57
PART 2
61
6 A Review of Classical Concepts on Phase and Amplitude Dispersion: Application to Tsunamis n. nirupama, t.s. murty, i. nistor, and a.d. rao
63
Modelling of Tsunami Generation and Propagation
7 A Partial Explanation for the Initial Withdrawal of the Ocean during a Tsunami n. nirupama, t.s. murty, a.d. rao, and i. nistor 8 The Energetics of the Tsunami of 26 December 2004 in the Indian Ocean: A Brief Review n. nirupama, t.s. murty, i. nistor, and a.d. rao
73
81
xii
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
9
Contents Possible Amplification of the Tsunami Through Coupling with Internal Waves n. nirupama, t.s. murty, i. nistor, and a.d. rao
91
10
Numerical Modeling of the Indian Ocean Tsunami z. kowalik, w. knight, t. logan, and p. whitmore
97
11
Modelling Techniques for Understanding the Indian Ocean Tsunami Propagation v.p. dimri and k. srivastava
123
12 Validation of Tsunami Beach Run-up Height Predictive Model Based on Work–Energy Theorem g. muraleedharan, a.d. rao, t.s. murty, and m. sinha
131
13
Normal Modes and Tsunami Coastal Effects n. nirupama, t.s. murty, a.d. rao, and i. nistor
143
14
Helmholtz Mode and K–S–P Waves: Application to Tsunamis n. nirupama, t.s. murty, i. nistor, and a.d. rao
151
15
Numerical Models for the Indian Ocean Tsunami of 26 December 2004: A Brief Review p. chittibabu and t.s. murty
159
16 The Cauchy–Poisson Problem: Application to Tsunami Generation and Propagation n. nirupama, t.s. murty, i. nistor, and a.d. rao
175
17 A Review and Listing of Tsunami Heights and Travel Times for the 26 December 2004 Event i. nistor, k. xie, n. nirupama, and t.s. murty
185
18
209
Overview and Integration of Part 2 n. nirupama (editor)
PART 3
Tsunami Detection and Monitoring Systems
213
19
Satellite Detection of Pre-Earthquake Thermal Anomaly and Sea Water Turbidity Associated with the Great Sumatra Earthquake a.k. saraf, s. choudhury, s. dasgupta, and j. das
215
20
Possible Detection in the Ionosphere of the Signals from Earthquake and Tsunamis t.s. murty, n. nirupama, a.d. rao, and i. nistor
227
21
Seismo-electromagnetic Precursors Registered by DEMETER Satellite a.k. gwal, s. sarkar, s. bhattacharya, and m. parrot
22 Web-Enabled and Real-Time Reporting: Cellular Based Instrumentation for Coastal Sea Level and Surge Monitoring a. joseph and r.g. prabhudesai
235
247
Contents xiii
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
23
Methodologies for Tsunami Detection t.s. murty, n. nirupama, a.d. rao, and i. nistor
259
24 Tsunami Travel Time Atlas for the Indian Ocean p.k. bhaskaran, s.k. dube, t.s. murty, a. gangopadhyay, a. chaudhuri, and a.d. rao
273
25
293
Overview and Integration of Part 3 n. nirupama (editor)
PART 4
Biophysical and Socio-Economic Dimensions of Tsunami Damage
295
26
Performance of Structures Affected by the 2004 Sumatra Tsunami in Thailand and Indonesia m. saatcioglu, a. ghobarah, and i. nistor
297
27
Field Observations on the Tsunami Impact Along the Kerala Coast, Southwest India n.p. kurian, t.n. prakash, and m. baba
323
28
Ecological Impact of Indian Ocean Tsunami c.s.p. iyer
339
29 Tsunami Damage to the South Eastern Coast of India n. chandrasekar and r. ramesh
351
30
365
Hydrophysical Manifestations of the Indian Ocean Tsunami y. sadhuram, t.v. ramana murthy, and b.p. rao
31 Tsunamis and Marine Life d.v. subba rao, b. ingole, d. tang, b. satyanarayana, and h. zhao
373
32 Tsunami Impact on Coastal Habitats of India p.n. sridhar, a. surendran, s. jain, and b. veera narayan
393
33
405
Overview and Integration of Part 4 u. aswathanarayana (editor)
PART 5
Quo Vadis
409
34
Protection Measures Against Tsunami-type Hazards for the Coast of Tamil Nadu, India v. sundar
411
35
Protective Role of Coastal Ecosystems in the Context of the Tsunami in Tamil Nadu Coast, India: Implications for Hazard Preparedness a. mascarenhas and s. jayakumar
423
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
xiv
Contents
36
Integrated Preparedness Systems u. aswathanarayana (editor)
37
Social and Political Aspects of Tsunami Response, Recovery, and Preparedness Planning: A Transdisciplinary Approach from Canada c. amaratunga and h. smith fowler
437
445
38 An Ideal Conceptual Tsunami Warning System for the Indian Ocean t.s. murty, n. nirupama, a.d. rao, and i. nistor
455
39
475
Overview and Integration of Part 5 n. nirupama (editor)
Author Index Subject Index
477 487
Figures
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
1.1 1.2 1.3 1.4 2.1 2.2 2.3 2.4 2.5 2.6
2.7
3.1
3.2 3.3
3.4
Rupture areas of great earthquakes of Mw 7.7 or greater and inferred seismic gap areas that could be sites of future tsunamigenic great earthquakes in the Indian Ocean. Source zones of 2004 and 2005 earthquakes in Sumatra–Andaman zone. Rupture areas of past great earthquakes along Sumatra. The southern Sumatra zone is a possible site for future great earthquake. Krakatau volcanic eruption, 1883 (Source: Simkin & Fiske, 1983). Geophysical data coverage over the ECMI (thin solid lines represent cruise tracks over ECMI and Bengal Fan). Bathymetry map of ECMI. Bathymetry sections of ECMI. (a) Magnetic anomaly map of Cauvery basin, with structural interpretation (contour interval: 20 nT). (b) Schematic cross section of basement along Profile A–B of Cauvery Basin. Free air gravity anomaly map of Cauvery offshore basin (contour interval 10 mGal). FZ1 and FZ2 are the inferred faults. Location and focal mechanism of the Pondicherry earthquake are shown (Murty et al., 2002). (a) Bathymetry map of Cauvery offshore basin (F1 and F2 are fault trends inferred from Figure 4(a)). (b) Bathymetry sections of Cauvery offshore basin (horizontal scale: 1 cm = 15 km) (horizontal distance in cms as measured from Figure 4(a)). Bathymetry map of Cauvery offshore basin (Murty et al., 2002), dashed lines indicate ship tracks along which bathymetry and gravity data were acquired. Black dot in the offshore indicates the location of Pondicherry earthquake (Mw 5.5) of 25 September 2001. MBA: Moyar–Bhavani–Attur lineament; PCL: Palghat–Cauvery Lineament. Map of Indian Ocean rim countries affected by the December 26, 2004 Indian Ocean Tsunami due to M9.3 earthquake off the coast of Sumatra. M8.7 earthquake on March 28, 2005 which occurred 250 km south of December 26 event is also shown. (Source: http://www.USGS.gov, adapted and modified.) Map showing different tectonic plates and locations of subduction zones (solid lines) and mid-oceanic ridges (zig-zag lines). Red dots are volcanoes. (Source: www.usgs.gov) Epicenters of the earthquakes of M > 7.0 for the Indian Ocean are shown with different colors of varying depths. Two sources of tsunami generation, Andaman–Sumatra in the east and Makran coast in the west are shown by ellipses. (Source: http://www.USGS.gov, adapted and modified.) Map showing tectonics of the Indo-Australian plate viz-a-viz Burma and Sunda plates. Yellow and Red stars show the epicenter of M9.3 earthquake on
4 5 6 12 20 21 22 25 26
27
29
33 35
36
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
xvi
Figures
December 26, 2004 and M8.7 on March 28, 2005. Solid arrows are direction of the movement of the Indo-Australian plate. Aftershocks are shown in yellow circles. (Source: http://www.USGS.gov) 3.5 (a) Reference map showing the locations of the principal geological features in the Indian Ocean. The red star marks the location of the initiation of rupture of the great Sumatra–Andaman earthquake. Brown lines show active and fossil plate boundaries. Arrows show the relative plate motions. The age of the incoming oceanic plate is shown with colors in millions of years. (b) Distribution of the apparent thermal age which results from the seismic inversion using the thermal parameterization. It is defined as the lithospheric age at which a purely conductive temperature profile would most closely resemble the observed thermal structure (after Ritzwoller et al., 2005). 3.6 Map showing rupture areas of four great earthquakes in the subduction zone from Andaman and Nicobar Islands to Sunda trench. Star shows epicenter of December 26, 2004 earthquake of M9.3 (figure from http://www.drgeorgepc.com). 3.7 (a) Makran subduction zone west of Karachi, Pakistan. (b) Vulnerability of the Indian west coast to the tsunami generated in Makran coast (figures from http://www.drgeorgepc.com). 3.8 Tsunami run-up heights along the east coast of Tamil Nadu. Numbers in the figure are tsunami heights in meters. 3.9 A shore-normal beach profile at Devanaampatnam (11◦ 44.589 N 79◦ 47.289 E). The sea level shown in the figure is the level at 9 a.m. IST on December 26, 2004, the time of tsunami attack. 3.10 Water marks in house in Devanaampatnam (photograph: R.K. Chadha). 3.11 A Bathymetry profile along 13◦ N, south of Pulicat lake. 7.1 Tide gauge record at Hanasaki, Japan, showing a tsunami forerunner (Nakamura and Watanabe, 1961). 7.2 Tide gauge records showing tsunami forerunners at some locations on the Pacific coast of Canada for the 1960 Chilean and the 1964 Alaska earthquake tsunamis (Murty, 1977). 8.1 The four global oceans. 8.2 The area in the Indian Ocean in which the tsunami energy propagated on 26 December 2004. 8.3 The rupture process according to Stein and Okal (2005). 8.4 Spheroidal normal mode 0 S3 . The amplitude of the mode is greatly exaggerated (from Lomnitz and Nilsen-Hofseth, 2005). 8.5 The directions of minimum and maximum energy (from Lomnitz and Nilsen-Hofseth, 2005). 8.6 Simulated snapshot of tsunami consisting of reconstructed peaks and troughs that formed in the Indian Ocean at a moment 1 h, 55 min after the earthquake struck (from Wilson (June 2005) http://www.physicstoday.org). 9.1 Definition sketch for two-layers profile. 9.2 Sketch of initial and boundary (periodic) conditions. 10.1 Spatial grid distribution in the spherical system of coordinates. 10.2 Ocean bathymetry. Computational domain extends from 80◦ S to 69◦ N. 10.3 History of tsunami propagation. Generated by bottom deformation at T = 40 s this tsunami experiences significant transformations and reflections. Black dashed lines denote bathymetry in meters. Numbers for the bathymetry in the figure should be multiplied by 10.
37
38
39 40 41 42 43 44 74 75 82 83 83 85 85 86 92 94 99 101
103
Figures 10.4 10.5 10.6
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
10.7 10.8 10.9 10.10 10.11
10.12 10.13 10.14 10.15 10.16 10.17 10.18 10.19 10.20 10.21 10.22 11.1 11.2 12.1 12.2 12.3 12.4
Tsunami signal propagating from the generation domain into the open ocean. Initial box signal of 20-min period is followed by the signal reflected from the shelf break and signal radiated from the shelf domain. December 26, 2004 Sumatra earthquake uplift as constrained by tsunami travel times. The source function. Maximum uplift is 507 cm and maximum subsidence approximately 474 cm. Coordinates are given in geographical degrees. Point (0,0) is located at 89◦ E and 1◦ N. Distribution of the tsunami amplitude in the Indian Ocean at 2 h 50 min from the tsunami onset. The wave reflected from the India and Sri Lanka propagates back to the source region. Distribution of the tsunami amplitude in the Indian Ocean at hour 4 from the tsunami onset. Along with the reflection shown in Figure 10.7, the reflection from the Maldives also sends energy eastward. Sea level recorded at Cocos Island on December 26, 2004. A birds-eye view of the IOT at time 9 h 25 min from the tsunami onset, looking from Africa toward India and Indonesia. Trapped tsunamis around continents and islands still display a strong signal. Sea level pattern generated by the IOT of December 26, 2004 at 30 h 40 min from the onset. Tsunami signals in the Northern Atlantic and Southern Pacific have been reorganized into coherent waves after passing through the narrows between Africa and South America, and Australia and Antarctica. Maximum modeled tsunami amplitude in the Indian Ocean. Maximum modeled tsunami amplitude in World Ocean. Residual maximum amplitude in World Ocean. Energy flux vectors over the South Pacific ridge at time 26 h 20 min. Colors denote sea level. Dark-brown lines denote the ridge depth – 3000 m depth contour. Southward directed energy flux through the E–W cross-section located in the Indian Ocean along 10◦ S from 80◦ E to 105◦ E. Energy flux through the cross-sections located between Antarctic and major continents. Along 20E from Antarctica to South Africa (AS) (light shading); along 140◦ E, from Antarctic to Australia (AA) (dark shading). Travel time (in hours) for the tsunami of 0.1 cm amplitude. Travel time (in hours) for the tsunami of 0.5 cm amplitude. Observations and computations from the four stations in the Indian Ocean. Ground track of Jason-I and computed tsunami amplitude at 2:55 UT on December 26, 2004 in the Indian Ocean. Computed and observed tsunami amplitude along the Jason-I track. Upper panel: source function given in Figure 10.1. Lower panel: source function orientation and width adjusted. The sea state at 5, 30, 60, 120, 180, 240, 300, 360, 420, 480, 600 min in Indian Ocean (Yalciner et al., 2005). The observed run-up distributions along the east coast of India and comparison with model results (Yalciner et al., 2005). Map of Indonesia (www.worldatlas.com). Andaman and Nicobar Group of Islands (map by ANCOST, NIOT, Chennai). (a)–(d) Real shore profiles of a few Andaman and Nicobar Group of Islands (by ANCOST, NIOT, Chennai). Location map of the study area of the Tamil Nadu coasts (www.sthjournal.org/241/chand.pdf).
xvii
104 105 106 107 107 108 108
109 110 110 111 112 113 114 115 116 117 118 119 127 128 132 133 134 136
xviii
Figures
14.1
Resonance characteristics of a system with a single degree of freedom (Raichlen, 1966). Tsunami travel times in hours (annotation “s” does not stand for second) Yalciner (2005). Maximum wave amplitude for the Global Ocean. Arrival time of first wave of the tsunami for the Global Ocean. Maximum tsunami wave height (cm) in the Indian Ocean. Arrival time of first wave in the Indian Ocean. Observed versus model wave arrival times. Computed and observed inundation in Banda Aceh Indonesia from http://www.wldelft.nl/cons/area/ehy/flood/tsunami.html. Tsunami amplitudes at Banda Aceh (Indonesia), Galle (Sri Lanka), Madras (now Chennai, India). Maximum water elevation. Grids used in the model. Computed maximum water elevation in the Indian Ocean. Geographical map of the Indian Ocean (Source: http://mapsherpa.com/tsunami/). Illustration of the antinode with highest positive amplitude at quarter wavelength of a sine wave. Illustration of the phenomena of Helmholtz resonance in harbours. (a) Tsunami waves reaching Kerala after first diffracting around Sri Lanka and travelling northward and getting reflected from Lakshadweep Islands (for illustrative purposes only, not to scale; basemap adapted from www.mapsofindia.com). (b) Tsunami waves reaching Kerala after being reflected from the coast of Somalia (for illustrative purposes only, not to scale; basemap adapted from www.wikipedia.com). Tsunami tidal interaction (personal communication, Prof. Z. Kowalik, 2006). Amplification of the tsunami at the ocean surface through coupling with internal waves. Tidal amphidromic point in the Arabian Sea. Maximum tsunami amplitudes (m) in Sabang and Banda Aceh area. Maximum tsunami amplitudes (m) in centre of Banda Aceh area. Maximum tsunami amplitudes (m) in West Coast of Banda Aceh area. Maximum tsunami amplitude (m) in Sigli area of Banda Aceh. Maximum tsunami amplitude (m) in Khao Lak (north part) in Thailand. Maximum tsunami amplitude (m) in Phuket beach area in Thailand. Maximum tsunami amplitude (m) in Phi Phi Don area in Thailand. Maximum tsunami amplitudes (m) in Sri Lanka. Maximum tsunami amplitudes (m) in Indian east coast area. Maximum tsunami amplitudes (m) in the Maldives. Location of the epicenter of the main shock of the 26 December 2004 mega-thrust earthquake in Banda-Aceh, Sumatra and the aftershocks. Also shows past seismicity of the region. Tectonics of the region around the epicenter of the devastating earthquake on 26 December 2004, the plate margin, where the India plate is being subducted beneath the Burma plate, along the Sunda trench. This active plate movement generates numerous earthquakes along the entire plate margin. The zone of plate movement stretches up to the Himalayan belt and results in the uplift of the Himalayan range.
15.1
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
15.2 15.3 15.4 15.5 15.6 15.7 15.8 15.9 15.10 15.11 17.1 17.2 17.3 17.4
17.5 17.6 17.7 17.8 17.9 17.10 17.11 17.12 17.13 17.14 17.15 17.16 17.17 19.1 19.2
152 160 161 162 162 163 163 164 167 168 169 170 186 186 187
187 188 189 190 200 201 202 203 204 205 206 207 207 208 216
217
Figures xix
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
19.3
Schematic model showing the generation of pre-earthquake thermal anomaly and detection by thermal remote sensing and meteorological stations in the ground. 19.4 Concentration of positive charges (denoted by “+” sign) on the surface of the butterfly net, especially at the apex (maximum positive curvature). If the net is inverted by pulling up the thread, charges have the tendency to move towards the new outer surface. 19.5 NOAA–AVHRR data derived LST time series maps for the region around the epicentral region of the 26 December 2004 Sumatra earthquake. A thermal anomaly developed before the earthquake and went away along with the event. The LST was seen to be maximum on 25 December 2004, just one day before the earthquake. 19.6 Seawater turbidity observed through NOAA–AVHRR data sets induced by the almost 1300 km fault rupture, which caused the great earthquake on 26 December 2004. 19.7 Turbidity observed through MODIS data sets near coasts in Sumatra induced by the tsunami after the great earthquake on 26 December 2004. 20.1 Standard and extreme ARDC atmospheres (Harkrider, 1964). 21.1 Track of DEMETER orbit on March 23, 2005 when the satellite was above the Sumatra region. Stars indicate the epicenters of the earthquakes. 21.2 From top to bottom the panels successively show the electron density, electron temperature, spectrogram of an electric component between 0 and 2 kHz and earthquakes seen by DEMETER along the orbit. The data are presented as a function of the universal time (UT), The local time (LT), geographic latitude and longitude values are also given. 21.3 Spectrogram of ELF electric waveform above the latitude in which earthquake occurred obtained using Level 1 burst mode data for the orbit shown in Figure 21.1. 21.4 Three hourly Kp values for March 23, 2005. 21.5 Track of DEMETER orbit on March 26, 2005 when the satellite was close to the Sumatra region. 21.6 Data recorded by DEMETER along the orbit shown in Figure 21.5. The top panel shows the electron density, the middle panel shows the ion density and the bottom panel gives the earthquakes “seen” by the satellite. At the bottom, UT, LT, geographic latitude and longitude values are indicated. 21.7 Three hourly Kp values for March 26, 2005. 21.8 Track of DEMETER orbit on July 6, 2005 when the satellite was above the Indonesian region. Stars indicate the epicenters of the earthquakes. 21.9 The top panel gives the spectrogram of an electric component between 0 and 400 Hz and the bottom panel gives the earthquake information. At the bottom, UT, LT, geographic latitude and longitude values are indicated. 21.10 Spectrogram of ULF/ELF electric waveform above the latitude in which earthquake occurred obtained using Level 1 burst mode data for the orbit shown in Figure 21.8. 21.11 Three hourly Kp values for July 6, 2005. 22.1 (a) Top portion of the gauge’s mounting structure, where battery, electronics, solar panel, and cellular modem are fixed (after Prabhudesai et al., 2006). and (b) Illustration of NIO sea-level gauge installed at Verem Jetty, Mandovi estuary, Goa, India (after Prabhudesai et al., 2006). 22.2 Schematic diagram illustrating implementation of realtime coastal sea-level data reception utilizing GPRS technology (after Prabhudesai et al., 2006).
219
220
222 223 224 229 237
238 239 240 240
241 242 242 243 244 244
252 254
xx
Figures
22.3 22.4
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
23.1 23.2 23.3 23.4 23.5 23.6 24.1 24.2 24.3 24.4 24.5 24.6 24.7 24.8 26.1 26.2 26.3 26.4 26.5 26.6 26.7 26.8 26.9 26.10
26.11 26.12 26.13 26.14
Display of predicted fair-weather sea-level, observed real-time sea levels, and residuals from Verem jetty (Mandovi estuary), Goa, India (after Prabhudesai et al., 2006). FTP up-load time during a period of 24 h for a 100 kb test data (after Prabhudesai et al., 2006). Sea level gauge network for the PTWS (McCreery, 2005). Tsunami run-up detectors used by the PTWC (McCreery, 2005). DART systems that were in operation for the Pacific Ocean (Bernard, 2005). The DART system as deployed in the ocean (Bernard, 2005). Schematic diagram of plane-wave incident on a cylindrical island. Input is trace X , response is trace Y . (Reid and Knowles, 1970). A simulated satellite picture of the Indian Ocean tsunami of 26 December 2004 (Wilson (June 2005) http://www.physicstoday.org). 250 locations in 35 countries for which tsunami travel time charts are prepared. Epicenter locations of the past earthquakes that generated tsunamis, which had some impact in the Indian Ocean. Tsunami travel time chart for a location in the Rann of Kutch in India. Tsunami travel time chart for a location off the coast of Sri Lanka. Tsunami travel time chart for a location in the Bay of Bengal. Tsunami travel time chart for a location in the Andaman Sea. Tsunami travel time chart for the city of Cota Raja, Banda Aceh in Indonesia. Tsunami travel time chart for a location in the Indian Ocean. (a) Phuket, Thailand and (b) Khao Lak, Thailand, visited during reconnaissance investigations. (Figures reprinted from Mapsoft world, 2005, Khao Lak Promotions.) Coastal erosion in Rawai Beach on Phuket Island, Thailand. Damage to (a) roof tiles and (b) timber columns, Kata Beach on Phuket Island, Thailand. Damage to first-story masonry walls, Patong Beach on Phuket Island, Thailand. Damage to buildings in Nai Thon Beach on Phuket Island, Thailand. Extensive damage to columns, masonry walls and roofs of low-rise buildings in Khao Lak Beach, Thailand. (a) Damage to the harbor and (b) boat floated inland, just north of Khao Lak Beach, Thailand. Damage to buildings in Phi Phi Island, Thailand. Wave pressure as per: (a) equation (26.1) (Goda, 1995) and (b) equation (26.2) (Hiroi, 1919). Comparisons of lateral forces due to earthquake, wind, and tsunami for an interior frame of a 6-story reinforced concrete building in Vancouver, Canada for 5.0 m tsunami water height, 6.0 m transverse span length and seismic force reduction factor of R = 4.0. Damage to timber frame structures in Thailand. (a) Damaged wall exposing masonry units in Patong Beach; (b) typical 50 mm thick concrete block masonry units; (c) and (d) construction of confined masonry in Khao Lak harbor town. Typical punching failure of masonry walls: (a) Kamala Beach and (b) Phi Phi Island. (a) Column failures in a 2-story reinforced concrete frame units in Khao Lak Beach and (b) Nominal moment-axial force interaction diagram for a 200 mm square column with 0.5% reinforcement and approximate moments imposed due to assumed tsunami pressure.
255 255 260 261 262 263 265 267 274 282 288 288 289 289 290 290 298 298 299 299 300 300 301 301 302
303 304 304 305
305
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Figures xxi 26.15 (a) and (b) Column failures in Khao Lak Beach; (c) and (d) frame damage on Phi Phi Island. 26.16 Non-engineered concrete frames that survived the tsunami in Khao Lak Beach. 26.17 Engineered reinforced concrete frame buildings: (a) and (b) on Phi Phi Island; (c) in Nai Thon Beach; (d) in Khao Lak Beach, that survived the tsunami without structural damage. 26.18 Damage to reinforced concrete frame buildings in Nai Thon Beach. 26.19 Damage to reinforced concrete frame building under construction in Nai Thon Beach. 26.20 (a) and (b) Damage to precast slab strips in Phaton Beach shopping center; (c) Close-up view of a precast slab strip; (d) lifting of precast concrete slab due to water pressure in Nai Thon Beach. 26.21 Damage to precast slab strips of the concrete dock in Khao Lak. 26.22 Soil erosion and related foundation problems: (a) buildings in Kamala Beach and (b) a building in Nai Thon Beach. 26.23 Destruction of residential timber framed construction in Banda Aceh. 26.24 Tsunami damage to non-engineered reinforced concrete buildings in central Banda Aceh. 26.25 Impact loading on columns due to floating debris, Banda Aceh. 26.26 Impact of fishing boats on buildings, Banda Aceh. 26.27 Power generating vessel that floated 3.5 km inland in Banda Aceh. 26.28 Timber framed structures in Banda Aceh. 26.29 Punching failure of masonry infill walls under tsunami pressure, Banda Aceh. 26.30 Damage to non-engineered reinforced concrete framed buildings, Banda Aceh. 26.31 Non-engineered reinforced concrete framed buildings that survived the tsunami. 26.32 Engineered reinforced concrete framed buildings that survived tsunami in downtown Banda Aceh, away from the coastal region. 26.33 A single-story house, displaced by water pressure due to lack of proper anchorage, Banda Aceh. 26.34 Displaced fuel tanks due to tsunami water pressure that were swept by half a kilometer, also destroying houses in their way, Banda Aceh. 26.35 The failure of a steel truss bridge in eastern Banda Aceh. 26.36 (a) and (b) The failure of a precast concrete bridge near downtown Banda Aceh and replacement by a temporary bridge. 26.37 Multi-span reinforced concrete bridges in eastern Banda Aceh, crossing the same river; (a) a bridge close to the ocean that was completely swept off by the tsunami and (b) a bridge approximately 3 km away from the ocean that survived the tsunami. 26.38 (a) A 3-span reinforced concrete bridge in Banda Aceh that was displaced on its abutments due to tsunami and/or seismic ground shakings and (b) close-up view of the separation of girders at a pier. 27.1 Location map. 27.2 Run-up level along the Kerala coast (after Kurian et al., 2006). 27.3 Beach profile stations. 27.4 SLED profiles across two stations off Chavara. 27.5 Bathymetric changes along Arattupuzha–Thangasseri coast. 27.6 Distribution of sand in the inner shelf (a) during 1987 and (b) during 2005. 27.7 Bathymetric changes along the TS canal.
306 307 307 308 309 310 310 311 312 312 313 313 314 314 315 316 317 317 318 318 319 319
319 320 324 326 328 328 329 333 334
xxii 28.1 28.2 28.3
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
28.4
28.5 28.6 29.1 29.2 29.3 29.4 29.5 29.6 29.7 29.8 30.1 30.2
30.3 30.4 30.5 30.6 30.7 30.8 31.1
Figures Indian Remote Sensing Satellite (IRS) imageries of the Kerala coast, captured before and after tsunami. (a) During January 2004 using IRS 1D and (b) December 27, 2004. (a) Selected transects and stations. (b) Isolines of various physical parameters, January 2005: (i) temperature in ◦ C, (ii) pH, (iii) salinity, and (iv) dissolved oxygen. Isolines of various nutrients. (µmol/l). (a) Nitrite – January 2005, (b) nitrite – May 2005, (c) nitrate – January 2005, (d) nitrate – May 2005, (e) phosphate – January 2005, and (f) phosphate – May 2005. Isolines of various biological parameters. Primary productivity in mgC/m3 /h for (a) pre-tsunami, (b) January 2005, and (c) May 2005. Chlorophyll-a in mg/m3 for (d) pre-tsunami, (e) January 2005, and (f) May 2005. Zooplankton biomass in ml/m3 for (g) pre-tsunami, (h) January 2005, and (i) May 2005 periods. Textural variations in sediments off 25 km – Vizhinjam. Latitudinal depth profile along Muttam to Kolachel transect, note the channelized flow at 8.03 latitude. Location map. Coastal geomorphology of the Kanyakumari District. (a) Map showing the inundation between Arokiapuram to Dharmapuram, (b) Map showing the inundation between Rajakkamangalam to Colachel. Impact of tsunami on the beach profile along the study area. Proportion of erosion and accretion of beach volume due to tsunami along the study area. Physical damages along the Kanyakumari District. (a) Impact of tsunami in the pH level along the Kanyakumari District, (b) Impact of tsunami in the TDS level along the Kanyakumari District. Tsunami vulnerability map of the study area. Time series (hourly) data on (a) temperature (◦ C) and (b) salinity (psu). Average profiles of (a) temperature (◦ C), (b) salinity (psu), (c) density (kg/m3 ), (d) sound velocity (m/s) and (e) Brunt Vaisala frequency (N 2 ; cph) based on the above data (solid line). Profiles from Levitus for the month of January are plotted (dotted line). Variations of (a) temperature (◦ C), (b) salinity (psu) and (c) density (kg/m3 ) along the transect off Visakhapatnam harbour on 8 January 2005. Variations of (a) temperature (◦ C), (b) salinity (psu) and (c) density (kg/m3 ) before (20 December 2004) (solid line) and after (7 January 2005) (dotted line) tsunami. Time series data (2 min interval) on temperature (◦ C) direction and speed (m/s) of the current, U and V components at (a) 10 m and (b) 60 m below surface. Spectral analysis of temperature (2 min interval) at (a) 10 m and (b) 60 m depths at the above location. Changes in temperature (◦ C) and salinity (psu) in the top layer as seen from the Argo data before (20 December 2004) (solid line) and after (26 December 2004) (dotted line) tsunami. Vulnerability of the Indian coast for the damages due to tsunamis/storm surges, inferred from the shoreline displacement for 1 m rise in sea level (Shetye et al., 1990). Schematic representation of the duration of a natural hazard and its impact on life and property per unit time (based on Krishna, 2005).
340 341 344
345 347 347 352 353 354 357 357 358 360 361 366
366 367 368 368 369 370 371 374
Figures xxiii
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
31.2
(a) Study area (box-S) showing southeast Asian countries affected by tsunami on 26 December 2004. (b) Study area in detail. The epicentres of earthquakes are marked with star/numbers – 1 for the disaster on 26 December 2004 and 2 for 28 March 2005. Box-T is the region for which time-series data (daily and 8-day) of and SST were examined. Provinces of Indonesia labelled: Aceh, N.S: North Sumatra, S.S.: South Sumatra, W.S.: West Sumatra, Riau and Jambi. 31.3 Cruise track and sample location map of the study area (based on Murthy, 2005). 31.4 Spatial distribution of chlorophyll a (mg/m3 ) (derived from MODIS) in the Indian Ocean during December 2004. Position of the earthquake on 26 December 2004 was indicated with red star in Figure 31.4(d). The circles in Figure 31.4(a,b) show high on northeast of Sumatra Island. Regions X and Y (Fig. 31.4(d)) represent changes in the coastal chlorophyll associated with tsunami waves. The interruption of satellite coverage (white patches over sea surface area) is due to cloud cover. 31.5 Spatial distribution of chlorophyll a (mg/m3 ) during January 2005. Circles 1 and 2 are the phytoplankton blooms. 31.6 Time-series data: (a) Daily concentrations of chlorophyll a derived from MODIS. Line-T passing through the bars is the trend line. Discontinuity of graph could be noticed due to missing values. Numbers 1, 2, 3, and 4 are the peaks of concentration during the events of earthquakes. (b) Comparison of (8-day average) among 3 years. P is the maximum encountered between mid of January and February during 2002–2003 and 2003–2004. (c) and SST during October 2004–May 2005. The dates of earthquakes are indicated with down arrows in all panels. 31.7 SST in the period of tsunami. The position of earthquake was indicated with red star in Figure 31.7(b). Downward arrow (Fig. 31.7(c)) denotes the gradient of lowering temperature from northeast to the southeast coast of India in Bay of Bengal. Variation of temperature close to the epicentre of earthquake was pointed with straight arrows. 31.8 Submerged coral beds, beach and forest area along the eastern coast of Southern Andaman Island near Baratang Island. 31.9 Map showing Malacca in Car Nicobar and settlement areas close to coast. 31.10 Returning wave of tsunami at Chennai, India. 31.11 Schematic representation of impact of tsunami on the various marine biotopes. 32.1 Circled areas indicate the severely affected coastal areas in (1). Andaman–Nicobar Islands and (2) southern east coast of India and north coast of Sri Lanka. 32.2 AWiFs data showing pre- (A) and post-tsunami (B) scenario in Nicobar Islands (a) Katchal, (b) Kamorta and (c) Trankati. 32.3 IRS-LISS III data showing north Chennai coast affected by tsunami waves. 32.4 Quick Bird data showing inundation of tsunami devastated Nagapattanam. 32.5 False color image showing high turbidity in the coastal waters of Andaman Islands on 27th December 2004. 32.6 The pre- and post-tsunami ocean color data showing increase in suspended particulate in coral reef areas of Andaman Islands on 27 December 2004. 32.7 Oceansat-OCM derived suspended sediment concentration along Pulicat Lake, Chennai coast during pre-tsunami (25 December 2004) and post-tsunami (27 and 31 December 2004).
375 375
376 377
378
379 381 382 383 383 394 394 395 395 396 396 397
xxiv
Figures
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
32.8
IRS-LISS III data showing the opening of Pichavaram inlet (2 and 4) and reaching of Vellar and Kollidam mouths (1 and 3) during pre- and post-tsunami. 32.9 The field photography showing damage to the coastal geomorphology and dunes erosion in MGR Tittu in Pichavaram due to tsunami wave runoff. 32.10 (A) Fishermen village houses damaged by tsunami waves in Pichavaram, (B) Tiruchendur, a temple town experienced retreat of sea during tsunami, (C and D) Along the coast of Mannakkudi (Kanyakumari District), the church and the bridges suffered damage where there was no seawall to protect them from tsunami waves. 32.11 Post-tsunami vulnerability ranking of coastal stretch of Tamil Nadu based on the slope, type of coast, presence of barriers and relative elevation, etc. 34.1 System of groins in the Chennai area. 34.2 Satellite imagery showing the shoreline advancement due to groin field in Kanyakumari district. 34.3 Suggested protection measure for the stretch of the coast at Pudukuppam in Parangipettai. 34.4 Proposed groins (thoondil valaivu) at Colachel jetty. 34.5 Proposed layout of groins from Nagore to Keechankuppam. 34.6 Rehabilitation of existing groins at Tranquebar coast. 34.7 Proposed shape of the sand dune at Palayur. 34.8 Proposed coastal protection measure at Palayur. 34.9 Proposed coastal protection measure at Thirumalaivasal. 35.1 Location of 24 stations along Tamil Nadu coast where post-tsunami surveys and beach profiles were carried out. 35.2 Map of Nagore–Velankanni stretch showing some of the coastal geomorphic and vegetal features. 35.3 Landscape changes before (November 1998) and after (April 2005) the tsunami: (a), Human occupation of dunes at Nagore (November/1998). (b) All huts and shacks flattened by the tsunami; only coconut groves survived (April/2005). (c), Crowded improvised structures on the beach opposite the shrine at Velankanni (November/1998). (d) Make-shift structures washed off in totality by violent waves (April/2005). 35.4 Profiles of the beach at five stations along Tamil Nadu coast. The arrow indicates run-up heights at each location. 35.5 Casuarina plantations served as excellent buffers against the tsunami onslaught at Nanjalingampettai (a) and at Karaikal (b); these trees remained intact all along the Tamil Nadu coast (April 2005). 36.1 Risk adequate premiums as a function of the size of the risk community. (Source: Menzinger and Brauner, 2002.) 36.2 Institutional structure for a catastrophe bond (cat bond). (Source: Stipple, 1998.) 37.1 The population health model. 37.2 Suggested schemata for the application of an eco-health approach to development. 37.3 Iterative research strategy for improving human health using a participatory and transdisciplinary approach (Forget, 1997). 38.1 The four global oceans. 38.2 The countries in and around the Indian Ocean. 38.3 250 locations in the Indian Ocean and adjoining region for which tsunami travel time charts have been prepared (Bhaskaran et al., 2005).
399 399
400 401 413 414 416 417 418 419 419 420 421 424 425
428 431 432 440 440 449 451 452 455 456 457
Figures xxv 38.4 38.5 38.6
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
38.7 38.8 38.9 38.10 38.11 38.12 38.13
Tsunami travel time chart for Rann of Kutch, India. Latitude 23.583◦ N; Longitude 68.367◦ E (from Bhaskaran et al., 2005). Tsunami travel time chart for a location off Sri Lanka: Latitude 8.570◦ N; Longitude 81.230◦ E (from Bhaskaran et al., 2005). Irregular triangular grid for a finite element model for a part of the Pacific coast of Canada and USA. Irregular triangular grid for a finite element model of the Queen Charlotte Islands of Canada. The Marching problem (Hyperbolic) (from Crandall, 1956). The Jury problem (Elliptic) (from Crandall, 1956). Method of characteristics (from Crandall, 1956). Schematic illustration of the tsunami numerical modelling concept for the Pacific Ocean. Schematic illustration of the tsunami numerical modelling concept for the Atlantic Ocean. Schematic illustration of the tsunami numerical modelling concept for the Indian Ocean.
458 458 460 460 462 462 465 467 467 468
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tables
1.1 1.2 2.1
List of tsunamis in Sumatra–Java region. List of tsunamis that affected Indian region and vicinity. Shelf/slope characteristics off selected places over the ECMI (Murthy et al., 1993). 3.1 Details of tsunami run-up surveys along the coast of Tamil Nadu. 8.1 Some of the fault parameters (from Kowalik et al., 2005). 10.1 Fault parameters used to generate vertical sea floor movement. 10.2 Observed and calculated travel time. 11.1 Some of the world’s destructive tsunamis of the Pacific and Indian Ocean. 12.1 Maximum water height and maximum inundation distance along the coasts of Indonesia during historical tsunamis. 12.2 Predicted time (t) required to travel from 1 m depth to 0 m depth and tsunami height (Hs ) near coastlines of Indonesia. 12.3 Maximum run-up level distance up to which seawater inundated inland during boxing day tsunami in Andaman and Nicobar Islands. 12.4 Predicted tsunami travel time (t) from 1 m depth to 0 m depth and tsunami height (Hs ) near coastline of Andaman and Nicobar Islands. 12.5 Few examples in Andaman and Nicobar Islands to show the improvement in tsunami height (Hs ) predictions when different slopes on land (Figure 12.3(a)–(d)) are considered. 12.6 Run-up level of sea water during 26 December 2004 Indian Ocean Tsunami at selected locations along Tamil Nadu coasts. 12.7 Prediction of travel time (t) from 1 m depth to 0 m depth and tsunami height near coastline for Tamil Nadu coasts for 26 December 2004 Indian Ocean Tsunami. 12.8 Inundation distance extent along the study area. 12.9 Predicted time (t) required to travel from 1 m depth to 0 m depth and tsunami height (Hs ) near coastline for 26 December 2004 Indian Ocean Tsunami. 12.10 Historical tsunami events and run-up levels in the Pacific and Atlantic oceans. 12.11 Predicted tsunami travel time (t) from 1 m depth to 0 m depth near coastlines and tsunami heights (Hs ) of historical tsunamis of Pacific and Atlantic oceans. 15.1 The fault data used to compute the tsunami source for simulation. 15.2 Fault parameters as used in the DCRC model. 15.3 Model parameters. 15.4 Fault parameters used to generate vertical sea floor movement. 15.5 Observed and calculated travel time. 17.1 Length of pendulum day at different latitudes. 17.2 Listing of observed tsunami arrival times and run-up. 19.1 Time of acquisition of NOAA–AVHRR GAC data used to prepare LST time series maps to study pre-earthquake thermal anomaly.
8 14 23 43 84 105 116 125 132 133 135 136 137 138 138 139 139 140 141 160 166 168 170 171 189 191 221
xxviii Tables 19.2 19.3 21.1
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
21.2 23.1 24.1 24.2 24.3 27.1 27.2 28.1 28.2 28.3 29.1 29.2 29.3 29.4 31.1 31.2 31.3 31.4 34.1 34.2 35.1 35.2 36.1 36.2 38.1
Time of acquisition of NOAA–AVHRR GAC data, used for analysis of seismically induced turbidity in the seawater by the mega-thrust earthquake of 26 December 2004. Time of acquisition of terra-MODIS data, used for analysis of tsunami induced turbidity in the seawater by the mega-thrust earthquake of 26 December 2004. Time, locations, depth and region of the earthquakes that occurred in the Sumatra region (from the web server http://www.iris.edu/seismon). Time, locations, depth and region of the earthquakes that occurred in the Indonesian region (from the web server http://www.iris.edu/seismon). Use of the data from DART systems in tsunami warning (Gonzalez et al., 2005). Number of locations in each country for which tsunami travel time charts are prepared. Locations in each country for which tsunami travel time charts have been prepared. Geographical coordinates and the source region of the epicenters used for the study. All longitudes are east and negative latitudes denote southern hemisphere. Run-up level and maximum inundation time along Kerala coast (after Kurian et al., 2006). Volume changes at different stations adjoining the Kayamkulam inlet (after Kurian et al., 2006). Comparison of physico-chemical parameters. Comparison of biological parameters. Granulometric data on sediments. Inundation extent and run-up level along the study area. Amount of accretion/erosion of beach volume (m3 ) due to the tsunami. Loss of life and properties due to the tsunami. Criteria adopted for the preparation of vulnerability map. Satellite based data on inundation of seawater in Nicobar Islands during tsunami. Impact of tsunami on mangrove stands of Andamans (based on Roy and Krishnan, 2005). Summary of observations made by dives on coral reefs. Impact of tsunami on aquaculture in peninsular India. Rate of erosion along the Tamil Nadu coast (Public Work Department, Tamil Nadu, 2002). Area of beach in-between groins 5 and 6. State of coastal landforms along Nagore–Velankanni sea front of Tamil Nadu in November 1998, and landscape changes and impacts recorded after the tsunami of December 2004. Monetary loss (crops, property and infrastructure) incurred and deaths reported as a consequence of some of the extreme events along the Indian coasts during the last 53 years. Populations affected by the tsunami. Science-based and people-based preparedness systems. Tsunami characteristics of the four oceans.
221 221 237 242 264 274 275 282 330 331 343 346 346 355 355 359 361 382 386 387 388 412 414 426 429 437 443 466
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Foreword
The tsunami of 26 of December of 2004 in the Indian Ocean, with its sequel of death and destruction, showed us nature at its horrendous worst. It will take years, if not decades, for the region to recover fully from the loss of livelihoods and the negative effects on development efforts. But the tragedy also gave us the opportunity to see human nature at it best. Immediately after the disaster, the Secretary General of the UN, Kofi Annan, issued an appeal to the solidarity of the world. The response to this call was unprecedented in the record of humanitarian missions of the UN. Nations, International Organizations, private industries, individual citizens, rich and poor – all did contribute to this magnanimous response. We have the knowledge; we have the science and the technology to mitigate the impact of this and other natural disasters. If a tsunami warning system, like the one existing in the Pacific Ocean since 1965,would had been in place, much of that damage would have been prevented. This is the hard lesson of the Tsunami of 2004. The Science of tsunami is still growing, as all science does. The learning never ends or stops. It is the role and mission of scientists, to constantly push the outer boundary of knowledge. But at any given time, science, what we know from theory and experience, can always be applied for the benefit of humankind. The application itself may not constitute science sensu stricto; it may be something else – it is designing a tool to do a job. The more we know, the better will be the tool, but the will to build the tool must be there. For that we need scientists, engineers and technicians that humbly assume this less glamorous task of putting science to work. The architecture of the international tsunami warning system now being put in place in the Indian Ocean under the leadership of the Intergovernmental Oceanographic Commission of UNESCO, is composed of two different networks: the upstream detection network of instruments, seismographs, sea-level gauges and deep ocean pressures sensors, and the downstream network of national tsunami centres, in charge of delivering the warnings to the people at risk with at least one national centre in each participating nation. In a minimal configuration, the national tsunami centre must have the operational capability of receiving warnings 24 hours a day, 7 days a week, and of disseminating these warnings both to the responsible authorities and to the general public. National tsunami centres must also be capable of defining national preparedness procedures and of putting in place national education and awareness plans. A timely, cent-per-cent accurate and precise warning will not provide any protection if the information doesn’t reach in time the people at risk and the people do not know how to respond to the emergency. Early warning is thus as much an issue of “soft” organizational technology, communication and community based systems, as it is of “hard” science and technology, numerical modelling and instrumental networks. In our experience, the real effectiveness of such a system is achieved when national centres move away from this minimal configuration, and start to develop their own national detection networks and risk-assessment protocols. These developments take time and require sustained efforts at both the national and international levels. One of the editors of this book, Tad S. Murty, published in 1977 a pioneering book entitled “Seismic Sea Waves: Tsunamis”, that has been extensively used in many parts of the world as a
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
xxx
Foreword
training tool. What has happened in the Indian Ocean in recent months furnishes many examples of actions taken at the community level, with the support of a wide variety of national and international organizations, many of them NGOs. But we must recognize that building national preparedness is the most difficult part of establishing early warning systems. In this new book “The Indian Ocean tsunami”, co-edited with U. Aswathanarayana and N. Nirupama, these scholars are giving us further demonstration of their deep-rooted concern on the importance of using science for the public good. The many contributions collected in this volume, give a thorough account of the different aspects of Indian Ocean Tsunami of December 2004. The wealth of information contained in these pages will certainly contribute to improve the design of the end-to-end systems needed for providing true protection from tsunami hazard to the populations living around the Indian Ocean rim. Through the acknowledgement to all those contributing to make this book, we also wish to thank all the scientists, technicians, professionals and engineers that during the last 18 months have volunteered their work to this noble cause. Patricio A. Bernal Executive Secretary of IOC Assistant Director-General of UNESCO Paris, July 2006.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Prologue
The 2004 Indian Ocean Tsunami was a great catastrophe. The tsunami directly affected the coastlines of the Indian Ocean, and the event was recorded by coastal tide gauges around the world. The loss of life, property and livelihoods was horrific. But, the response to the disaster has been dramatic, both in terms of support from individuals and nations from around the world and by scientists around the Indian Ocean and elsewhere. In the immediate aftermath of the tsunami, scientists worked together to find support and launch field surveys around the Indian Ocean. Scientists made field observations and surveyed run-up and inundation to provide important constraints for numerical modeling of the tsunami, then underway. Engineers visited sites of destruction along the coastlines of many countries and observed the destruction of man-made structures brought by a major tsunami. And, unfortunately many individuals lost their lives, particularly women and children, and others were permanently affected by the trauma associated with the loss of lives, injuries, loss of homes and property and the loss of their livelihood. This volume is the product of the energy and talent of many scientists within the region and outside the region who are dedicated to achieving a better understanding the phenomena of tsunami. Expertise involves specialists in the areas of: seismology, oceanography, tsunami science, structural geology, tectonics, sedimentology, geomorphology, hydrology, and many related fields of study. Their aim is the same: to better understand how tsunami are generated, how they propagate across the ocean, how the tsunami waves change when they reach shallow water and how they flood onto the coastline and inundate the land. These studies aim at increasing our understanding of tsunami, improving our capability to predict tsunami, identify ways to mitigate damage, and ultimately to save lives. It is very rewarding to see Indian Scientists writing about the Indian Ocean Tsunami. We hope the collaborations between scientists will continue in the future. The reports in this volume are integrated and reviewed by the editors. All tsunami scientists will benefit from the work described in this volume. I thank all of the contributors to this volume and the editors for a job well done. Your work will save lives during the next tsunami. Aloha, Dr. Barbara H. Keating University of Hawaii, Marine Geologist and President of the Tsunami Society June 5, 2006
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Preface
The purpose of the volume is to develop methodologies for the prediction of, and preparedness for, the tsunamis per se, on the basis of the knowledge obtained from a study of the Indian Ocean Tsunami (also known as the Asian Tsunami or the Sumatra Tsunami) of December 26, 2004. On the basis of the analysis of seismological data, hydroacoustic signals, geodynamic environment of genesis, magnetic images of the crust of the Sumatra region, Global Position System (GPS) analyses, etc., it has been found that the Sumatra earthquake which had a magnitude Mw of 9.3, energy of 1.1 × 1017 Nm, occurred at a depth of 20–30 km. close to Indonesian forearc. The earthquake rupture had a maximum length of 1200 km along the interface between the IndoAustralian and Burmese plates. There was a 20 m displacement of the fault plane, and the sea floor thrusted up several metres. The earthquake appears to have produced a recognizable pole shift, and a small change in the length of the day, and oblateness of the earth. Jason I altimetry satellite could detect the tsunami wave. The tsunami affected 12 countries from Indonesia to Somalia, and killed 176,260 people. The number of people missing is 49,682, and the number of persons displaced is 1,726,270 (Science, August 12, 2005). The tsunami destroyed billions of dollars worth of property, and caused horrendous human suffering. Indonesia suffered the greatest damage as the epicenter of the earthquake, and area of initiation of the tsunami are located there. There is a scientific explanation for the linear path of the tsunami – why it affected Sri Lanka and the Tamilnadu coasts, but not Bangladesh and Orissa coasts. Singh (Nature, 2005) showed that as a consequence of the existence of a lithosphere-scale boundary around the Simeulue Island which continues upto the east of Nicobar Island, the December 26, 2004 earthquake rupture which seems to have been initiated west of this boundary, did not cross the boundary to the east, but got propagated northwards upto the Andaman Islands. The same boundary also explains why the after shocks of Mw 9.3 earthquake of December 26, 2004 and Mw 8.6 earthquake of March 28, 2005 did not overlap. The design of a cost-effective strategy for the warning and preparedness for tsunamis has to take into account two attributes of the tsunami, namely, its genesis and its impact. Tsunami is triggered by earthquakes, volcanism, submarine landslides, etc., and hence it can be conveniently dovetailed with the existing administrative structures for earthquake disaster management. A tsunami is akin to a tidal wave in its impact on the coasts, and can hence be treated as an add-on to tidal wave warning systems. The volume seeks to provide the knowledge base needed for the purpose. In order to predict the future occurrence of tsunami, and to design warning and preparedness systems for the purpose, we need to understand why the tsunami got generated where, when and how it got generated, and why and how it could cause so much devastation as it did. The volume is structured to seek answers to these questions: Part 1: Geostructural environment of tsunami genesis – the sites wherefrom the future tsunami could be generated; Part 2: Modeling of the tsunami generation and propagation – how will such tsunamis be expected to move, Part 3: Tsunami detection and monitoring systems – how to use the precursors to enhance the advance warning time, how to detect the tsunami when it occurs, how to communicate the warning to the
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
xxxiv
Preface
populations likely to be affected, and how to get it implemented; Part 4: Biophysical and socioeconomic dimensions of tsunami damage – how the variation in the severity of tsunami damage at various sites is determined by the geophysical, ecological and socioeconomic ambiences of the sites and Part 5: Quo Vadis – Where do we go from here. Preparedness systems based on dual-use technologies. The Editors are grateful to Dr. Patricio Bernal, Executive Secretary, Intergovernmental Oceanographic Commission andAssistant Director General of Unesco, Paris, for his perceptive Foreword, and to Dr. Barbara Keating, President of the Tsunami Society, Honolulu, for her appreciative Prologue. The knowledge base contained in the volume is relevant not only to the Indian Ocean countries, but also globally. It will be useful to university students, professionals and administrators concerned with seismology, ocean science, meteorology, disaster management, coastal management, etc. The impetus for the volume came from the Brain Storming Session (regarding the Indian Ocean Tsunami of December 26, 2004), New Delhi, India, January 21 and 22, 2005, organized by the Department of Science and Technology and the Department of Ocean Development, Government of India, under the aegis of the Indian National Science Academy. One of us (TSM) is grateful to the Government of Canada for associating him with the various authorities concerned with the Indian Ocean Tsunami. Tad S. Murty U. Aswathanarayana N. Nirupama
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Part 1
Geostructural environment of tsunami genesis
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 1
A Historical Account of the Earthquakes and Tsunamis in the Indian Ocean
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
B.K. Rastogi National Geophysical Research Institute, Hyderabad, India
1.1
INTRODUCTION
A catalog of tsunamis in the Indian Ocean during the period 326 BC to 2005 AD has been compiled. Possible source zones of large tsunamigenic earthquakes in the Indian Ocean have been identified. Repeat periods of strong earthquakes in these zones are also assessed. Tsunamis are not as common in the Indian Ocean as in the Pacific. As compared to average eight tsunamis per year in the Pacific, Indian Ocean has no more than one in 3 years or so. Eighty percent of the tsunamis of the Indian Ocean originate in the Sunda arc covering Java and Sumatra. A catalog of tsunamis presented here includes about 70 tsunamis from Sunda arc and about 20 tsunamis from the rest of the Indian Ocean. The same belt extends northward to Andaman–Nicobar Islands where a few tsunamis have originated. Further north, Bangladesh– Myanmar coast has produced some well-documented tsunamis. Makran coast in the northwest is known to have generated at least one major tsunami. Karachi–Kutch coast region is a possible source zone. Other regions like Chagos Ridge can give rise to local tsunamis. Tsunamis are mostly caused by thrust-type subduction zone earthquakes which are sometimes associated with landslides. The seismic gap areas along the subduction zones are possible sites of future great earthquakes. Along the Sunda arc, great earthquakes of M8.5 or greater can be repeated every two centuries at a site but smaller tsunamigenic earthquakes can be repeated every few decades. Along Sunda arc volcanic eruptions have also given rise to large tsunamis. There appears to have been a hiatus in tsunami generation in this region, with a significant gap in events occurring from around 1909 through 1967 (Tsunami Laboratory, Novosibirsk, Russia). Tsunamis from the Java region are not described in detail as their effects are restricted to Indonesia. 1.2 TSUNAMIGENIC EARTHQUAKE SOURCE ZONES IN THE INDIAN OCEAN Thrust-type earthquakes occurring along subduction zones that cause vertical movement of ocean floor tend to be tsunamigenic. During rupture of a subduction megathrust, for example along the Sumatra arc, the portion of southeast Asia that overlies the megathrust jumps westward (toward the trench) and upward by several meters. This raises the overlying ocean, so that there is briefly a “hill” of water overlying the rupture. The flow of water downward from this hill triggers a series of broad ocean waves that are capable of traversing the entire Bay of Bengal. When the waves reach shallow water they slow down and increase greatly in height – from a few meters to some tens of meters – and thus are capable of inundating low-lying coastal areas. The maximum tsunami run-up is generally twice the vertical movement of the ocean floor and its power depends on the rupture area. Landslide of the ocean floor associated with earthquakes adds to the power of tsunami. 3
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
4
B.K. Rastogi
Figure 1.1.
Rupture areas of great earthquakes of Mw 7.7 or greater and inferred seismic gap areas that could be sites of future tsunamigenic great earthquakes in the Indian Ocean.
In the near future, earthquakes along southern Sumatra, Makran coast, Indus Delta, Kutch– Saurashtra coast, Bangladesh and southern Myanmar might cause tsunamis which can affect India (Figures 1.1–1.3). The seismic gap areas along subduction zones like Andaman–Sumatra and Makran can be assessed as possible future source zones of tsunami-generating earthquakes in the Indian Ocean and the repeat periods of great earthquakes can be assessed from past seismicity. Along the Andaman–Sumatra trench, the convergence rate is 40–50 mm/year yielding return periods of 150–200 years for great to giant earthquakes of M8.5 or greater. Major tsunamigenic earthquakes of M < 8.0 have been repeated more frequently at intervals of over a few decades. Occurrence of M8.5 earthquake of 2005 at the rupture zone of M8.5 earthquake of 1861 matches with the estimated recurrence rate. However, some great earthquakes have occurred more frequently: rupture zone of 1833 M8.7 earthquake overlapped 1797 M8.2 rupture zone (i.e. within 36 years). Smaller but tsunamigenic earthquakes, of M7.5–8.0 have been repeated more frequently at intervals of over a few decades like 1907 (Ms7.6) and 1935 (Mw 7.7) major earthquakes that occurred near the 1861 (Mw 8.5) source zone. The Sumatran subduction zone is one of the most active plate tectonic margins in the world, accommodating over 50 mm/year of oblique northward convergence between the south Asian and India–Australian plates, which arcs 5500 km from Myanmar past Sumatra and Java toward Australia. The plates meet 5 km beneath the sea at the Sumatran trench, on the floor of the Indian Ocean. The trench runs roughly parallel to the western coast of Sumatra, about 200 km offshore. At the trench, the Indian–Australian plate gets subducted; that is, it is plunging into the earth’s interior and being overridden by southeast Asia. The contact between the two plates is a “megathrust”. The two plates do not glide smoothly past each other along the megathrust
5
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
A historical account of the earthquakes and tsunamis in the Indian Ocean
Figure 1.2.
Source zones of 2004 and 2005 earthquakes in Sumatra–Andaman zone.
but move in “stick-slip” fashion. This means that the megathrust remains locked for centuries, and then slips suddenly a few meters, generating a large earthquake. Some coastal areas east of the megathrust sink by a meter or so, leading to permanent swamping of previously dry, habitable ground. Islands above the megathrust rise 1–3 m, so that shallow coral reefs emerge from the sea.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
6
B.K. Rastogi
Figure 1.3.
Rupture areas of past great earthquakes along Sumatra. The southern Sumatra zone is a possible site for future great earthquake.
Sunda arc extends further north to the Andaman–Nicobar group of islands which is also seismically active zone and generates frequent large earthquakes. Large earthquakes occurred in 1847 (Mw > 7.5), 1868, 1881 (Mw 7.9) and 1941 (M7.7). Future locales are seismic gap areas that have remained unruptured in the past few decades. The 2004 Sumatra earthquake of M9.3 occurred in one such gap. The 28 March 2005 earthquake of M8.7 has occurred in a gap area south of the 2004 rupture identified by Kerry Sieh of Caltech. The 2005 earthquake was probably triggered by the 2004 Sumatra earthquake. Further south along Sumatra trench large and great earthquakes can be expected within a few decades. Mw 7.8 earthquake of 2000 has occurred along the southernmost tip of Sumatra. Northern Sumatra and Andaman regions are assessed to be probably free from great earthquakes for a few decades due to occurrence of 2004 Mw 9.3 and 2005 Mw 8.7 earthquakes. Southern Sumatra has potential for a great earthquake. However, the effect of tsunami due to this in India and Sri Lanka may be a limited one as the path of tsunami will be oblique to the rupture zone.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
A historical account of the earthquakes and tsunamis in the Indian Ocean
7
The Makran subduction zone of southern Pakistan and Iran is seismically less active, but has produced great earthquakes. The 28 November 1945 (Mw 8.0) earthquake generated the last major tsunami in the Arabian Sea. More than 4000 people were killed on the Makran coast by both the earthquake and the tsunami. The tsunami caused damage in Mumbai with 2-m run-up and affected Karwar (Karnataka). This earthquake occurred in the eastern part of the Makran zone, two sides of which remain potential zones for great earthquakes. Indus Delta and probably the coasts of Kutch and Saurashtra are also potential zones for great earthquakes and tsunami. Tsunami was generated by an earthquake in 1762 in Myanmar and possibly in 1874 by an earthquake near Bangladesh. Some earthquakes in future also in these regions can possibly generate tsunamis. Normal fault-type earthquakes can also generate moderate tsunami. Strike-slip earthquakes that cause horizontal movement of ocean floor are not tsunamigenic but oblique-slip/dip-slip component in them can generate weak tsunamis. Spreading zones like Carlsberg ridge, NinetyEast ridge, etc. are sites of such earthquakes. The Chagos ridge east of Carlsberg ridge had given rise to a local tsunami due to thrust component of motion for a major earthquake of Mw 7.7 earthquake of 30 November 1983 near Diego Garcia.
1.3
EARTHQUAKES AND TSUNAMIS FROM SUNDA ARC REGION
Newcomb and McCann (1987) compiled historic records of earthquakes and tsunamis from Sunda arc region. An updated list is given in Table 1.1. The Sumatra part of the Sunda arc had been much more active than Java part. Detailed description of some of the significant earthquakes and the tsunamis caused by them are given below. 1.3.1
Earthquakes in Sumatra
11 December 1681
“Strong earthquake” shook the Sumatra mountains near Mentawai Archipelago and a seaquake was observed. 3 November 1756 Many houses collapsed in several towns of Sumatra near to Engano Island. No tsunami was reported. No date, 1770 Severe damage in the same general area as the 1756 event, but a tsunami was reported. 10–11 February 1797, A large earthquake and tsunami was observed in ports on the coast of Mw 8.2 the mainland and on the Batu Island. Waves of great force hit the area near Padang (0.99◦ S 100.37◦ E), the town was inundated and more than 300 fatalities occurred (Heck, 1947). 18 March 1818 A very strong shock associated with both tsunami and seaquake near to Engano Island. 24 November 1833 The great earthquake of M > 8.7 had maximum intensities and generated a tsunami over 550 km along the south central coast of Sumatra that also caused much damage to the coast. Numerous deaths occurred in western Sumatra. This earthquake ruptured the plate margin from the southern island of Enggano to Batu. 5–6 January 1843, The earthquake caused severe damage, liquefaction and many fatalities Mw 7.2 in Nias Island. A tremendous tsunami wiped out towns on the east coast of Nias and mainland. The damage and associated tsunami were much localized. The village of Barus (2◦ N 98.38◦ E) and Palan Nias (Nias Island 1.1◦ N 97.55◦ E) reported large waves on 2 days. 11 November 1852 Earthquake near Nias generated seaquake.
Year
416.09.10 1681.12.11 1768.06.22 1770 1797.02.10 1799 1815.04.10 1815.11.22 1816.04.29 1818.03.18 1818.11.08 1820.12.29 1823.09.09 1833.01.29 1833.11.24
Sep. 1837 1843.01.05 1843.01.06 1852.11.11 1856.07.25 1857.05.13 1859.10.20 1861.02.16 1861.03.09 1861.04.26 1861.06.05 1861.06.17 1861.09.25 1864 1883.08.26 1883.08.27
1 2 3 3 4 5 6 7 8 9 10 11 12 13 14
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Banda Ache Southwest Sumatra Southwest Sumatra Sibolga, Sumatra Java-Flores sea Bali Sea Southern Java sea Southwest Sumatra Southwest Sumatra Southwest Sumatra Java Southwest Sumatra Southwest Sumatra Sumatra Krakatau Krakatau (Volcano)
Java-S Sumatra Bali Sea Southwest Sumatra Southwest Sumatra Southeast Sumatra Java-Flores Sea Bali Sea Penang Island Bengkulu, Sumatra Bali Sea Flores Sea Java Bengkulu, Sumatra Southwest Sumatra
Location
−6.10 −6.06
105.423 105.25
−3.5
102.2 5.5 1.5 1.05 1.7 −8.5 −8 −9 −1 0.3 1 −6.3 1 −1.5
−7 −5 −1 −2.983 −8.2 −8 5.383 −3.767 −7 −7 −6.5
115 102 99 104.75 118 115.2 100.25 102.267 117 119 108.5
96 98 97.33 98.8 116 115.5 111 97.5 99.37 97.5 107.3 97.5 100
−10
Lat.
120
Long.
List of tsunamis in Sumatra–Java region.
S.N.
Table 1.1.
1 1
Ms 6.8 Ms 6.5
Ms 8.5 Ms 7 Ms 7
6 6
1 1 1 1 1 1 1
Ms 6.8 Ms 7
4 1
1
6 1 1 1 1 1 6 3 1 3 1 1 1
Ca.
Ms 7.2 Ms 7.2
Mw 8.7
Ms 7 Ms 8.5 Ms 7.5 Ms 6.8
Ms 7
Ms 7.5 Ms 7 Ms 8
Mag.
3 4
1 2 4 2 4 4 4 2 3 3
2 4
4
2 4 4 3 4 2 4 3 2 3 2 4 2
Pro.
(1) (3)
0.5 3.0
(1) (8) 35 (67)
1.0 4.5
(2) (1) (9) (4) (1) 1.5
3.0 2.0 1.5
(3)
2.5
(1)
(1) (1)
(1) (1)
Max. run-up (run-ups)
1.5
0.5 3.0
I
NOAA/NESDIS Newcomb and McCann (1987) NOAA/NESDIS NGDC/NOAA Berninghausen (1966) Berninghausen (1966) NOAA/NESDIS NOAA/NESDIS NGDC/NOAA Berninghausen (1966) NOAA/NESDIS NOAA/NESDIS NOAA/NESDIS Berninghausen (1966) NGDC/NOAA, Newcomb and McCann (1987) NGDC/NOAA Berninghausen (1966), Heck (1947) Berninghausen (1966), Heck (1947) NGDC/NOAA NOAA/NESDIS NOAA/NESDIS Berninghausen (1966) Berninghausen (1966) NGDC/NOAA NGDC/NOAA NOAA/NESDIS NOAA/NESDIS Berninghausen (1966) Berninghausen (1966) Berninghausen (1966) Berninghausen (1966)
References
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012 8 B.K. Rastogi
Feb. 1884 1885.07.29 1889.08.16 1892.05.17 1896.10.10 1904.07.04 1907.01.04
1908.02.06 1909.06.03 1914.06.25
1917.01.21 1921.09.11
1922.07.08 1926.06.28 1928.03.26 1930.03.17 1930.06.19 1930.07.19 1931.09.25 1935.12.28 1936.08.23 1948.06.02 1949.05.09 1955.05.17 1957.09.26 1958.04.22 1963.12.16 1964.04.02
31 32 33 34 35 36 37
38 39 40
41 42
43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58
Lhoknga, Ache Southwest Sumatra Krakatau Java-south Java-south Southern Java Sea Southwest Sumatra Southwest Sumatra Malay Peninsula Malay Peninsula Malay Peninsula Malay Peninsula Southern Java Sea Southwest Sumatra Java Off northwest coast of Indonesia
Southwest Sumatra Sumatra West coast of South Sumatra Bali Sea Southern Java Sea
Krakatau Ajerbangis Java, Indonesia Malay Peninsula Southwest Sumatra Sumatra Southwest Sumatra
95.233 99.5 105.423 105.4 105.3 114.3 102.7 98.25 95 94 95 94 107.3 104 105.4 95.7
115.4 111 5.467 −1.5 −6.102 −6.1 −5.6 −9.3 −5 .001 6 5.5 5 6.5 −8.2 −4.5 −6.2 5.9
−8 −11
−5 −2 −4.5
2
94.5 100 101 102.5
−6.10 0.2 −6 2.5 −3.5
105.423 99.383 106 99.5 102.5
Ms 6 Ms 6.5 Ms 7.5 Ms 8.1 Ms 7.3 Ms 6.5 Ms 6.7 Ms 7.2 Ms 5.5 Ms 6.5 6.5 Ms 7.0
Ms 6.7
Ms 6.5 Ms 7.5
Ms 7.5 Ms 7.7 Ms 8.1
Ms 7.6
Ms 6.8 Ms 6 Ms 7.5 Ms 6.8
3 4 1 0 1 1 3 2 3 1 2 2 2 2 3 2 2 3
1 1 6 6 1 1 1 1 1 1 1 1 1 1 1 8
4 2 0
4
2 2 3 3 2
1 1
1 1 1
1
1 1 1 1 1
1.4 (1) 1.4
1.0 1.0
0.7 1 0.7 0.7
0.7
0.7 0.1 31.4
2 0.2
2.8 (7)
4 (4) 1 (1)
2.0
1.0
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
(Continued )
NGDC/NOAA NGDC/NOAA/Newcomb and McCann (1987) NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA NOAA/NESDIS
NGDC/NOAA/Newcomb and McCann (1987) NGDC/NOAA NGDC/NOAA NGDC/NOAA
Murty et al. (1999) NGDC/NOAA NGDC/NOAA NGDC/NOAA NGDC/NOAA
A historical account of the earthquakes and tsunamis in the Indian Ocean 9
1964.04.02 1967.04.12 1977.08.19 1982.02.24 1984
1985.04.13 1994.02.15 1994.06.02 2000.06.04
2000.06.18 2004.12.26
2005.03.28
2005.04.10
59 60 61 62 63
64 65 66 67
68 69
70
71
Malay Peninsula Malay Peninsula Sunda Islands Java trench Off west coast of Sumatra Bali Island, Indonesia Southern Sumatra Java, Indonesia Off west coast of Sumatra South Indian Ocean Off west coast of Sumatra Off west coast of Sumatra Kepulauanmentavia
Location
99.607
97.013
97.45 95.947
114.2 104.3 112.8 102.09
95.7 97.3 118.4 97.7 97.955
Long.
−1.64
2.074
−13.8 3.307
−9.2 −5 −10.5 −4.72
5.9 5.5 −11 4.37 0.18
Lat.
Ms 6.7
Mw 8.7
Ms 7.8 Mw 9.3
Ms 6.2 Ms 7.0 Ms 7.2 Ms 7.8
Ms 7.0 Ms 7.5 Ms 8 Ms 5.4 7.2
Mag.
1
1
1 1
1 1 1
1 1 1 1
Ca.
4
4
4 4
4
2
3 3 4 4
Pro.
3.0
1.5
I
NOAA/NESDIS NGDC/NOAA NOAA/NESDIS NOAA/NESDIS
13 (15) (1) 0.3 24 (302 4 (2) 1 (1)
NOAA/NESDIS NGDC/NOAA NOAA/NESDIS NGDC/NOAA Engdahl et al. (1998)
References
NGDC/NOAA NGDC/NOAA NGDC/NOAA USGS/NEIC(PDE)
2
Max. run-up (run-ups)
Cause Code: Cause code indicates the cause or source of the tsunamis. Valid values: 1 to 12 1 = earthquake 2 = questionable earthquake 3 = earthquake and landslide 4 = earthquake and volcano 5 = earthquake, volcano and landslide 6 = volcano 7 = volcano and earthquake 8 = volcano and landslide 9 = volcano, earthquake, and landslide
I is tsunami intensity, maximum run-up is in meters, reported number of run-ups are given within brackets. The data are taken from National Geophysical Data Center (NGDC); National Oceanic and Atmospheric Administration (NOAA) and National Environmental Satellite, Data, and Information Service (NESDIS). A “−1” is used as a flag (missing) value in some fields. The cause and probability of the tsunamis are shown by “Ca.” and “Pro.” respectively. The cause and probability of the tsunamis are given by following codes.
Year
(Continued)
S.N.
Table 1.1.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012 10 B.K. Rastogi
where “h” is the maximum run-up height of the wave.
I = log2 (21/2 ∗ h)
Some other formulae are also in use. Tsunami Intensity: Tsunami intensity scales have been suggested based on its effect and damage caused by it. There are many formulae for intensity based on tsunami run-ups. Tsunami intensity is defined by Soloviev and Go (1974) as
10 = landslide 11 = meteorological 12 = explosion Event Probability: Probability of actual tsunami occurrence is indicated by a numerical rating of the validity of the reports of that event. Valid values: 0 to 4 4 = definite tsunami 3 = probable tsunami 2 = questionable tsunami 1 = very doubtful tsunami 0 = erroneous entry Tsunami Magnitude: Tsunami magnitude, Mt is defined in terms of tsunami-wave amplitude by Iida et al. (1967) as: Mt = log2 Hmax
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
A historical account of the earthquakes and tsunamis in the Indian Ocean 11
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
12
B.K. Rastogi
Figure 1.4.
Krakatau volcanic eruption, 1883 (Source: Simkin & Fiske, 1983).
16 February 1861 A great earthquake of M8.5 ruptured a major segment of the plate boundary in northern Sumatra. The tsunami that was generated extended over 500 km along the arc. Tsunami destroyed southern towns of Batu Island, and a town on the southwest side of Nias experienced a tsunami of height 7 m. The earthquake and tsunami caused thousands of fatalities along the west coast of Sumatra. Two aftershocks on 9 March and 26 April 1861 also caused tsunamis. There was no major shock for almost 50 years. Historic records show that the strongest tsunami was associated with the volcanic eruption of Krakatau in Indonesia on 27 August 1883. The 35 m-high tsunami took a toll of 36,000 lives in western Java and southern Sumatra. The island volcano of Krakatau exploded with devastating fury, blowing its underground magma chamber partly empty so much so that much overlying land and seabed collapsed into it. Figure 1.4 shows the parts of the island that fell into the sea. A series of large tsunami waves was generated from the explosion, some reaching a height of over 35 m above sea level. Tsunami waves were observed throughout the Indian Ocean, the Pacific Ocean, the American west coast, South America and even as far away as the English Channel. On the facing coasts of Java and Sumatra the sea flood went many kilometers inland and caused such vast loss of life that one area was never resettled and is now the Ujung Kulon nature reserve. At its peak, the island of Rakata, which the volcano of Krakatau had formed, had reached a height of 790 m above sea level. Subsequent local tsunamis in the Sunda Strait were generated by the 1927 and 1928 eruptions of the new volcano of Anak Krakatau (Child of Krakatau) that formed in the area. Although large tsunamis were generated from these recent events, the heights of the waves attenuated rapidly away from the source region, because their periods and wavelengths were very short. There was no report of damage from these more recent tsunamis in the Sunda Strait (George, 2003).
A historical account of the earthquakes and tsunamis in the Indian Ocean
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
1.3.2
13
Krakatau explosive eruption (1883)
According to ancient Japanese scriptures, the first known supercolossal eruption of Krakatau occurred in the year 416 AD – some gave the date as 535 AD. The energy of this eruption is estimated to have been about 400 Mt of TNT, or the equivalent of 20,000 Hiroshima bombs. This violent early eruption destroyed the volcano, which collapsed and created a 7-km-wide submarine caldera. The remnants of this earlier violent volcanic explosion were the three islands of Krakatau, Verlaten and Lang (Rakata, Panjang and Sertung). Undoubtedly the 416 AD eruption/ explosion/collapse must have generated a series of catastrophic tsunamis, damage from which must have been much greater than those generated in 1883. However, there are no records documenting the size of these early tsunamis or the destruction they caused. Subsequent to the 416 AD eruption and prior to 1883, three volcanic cones of Krakatau and at least one older caldera had combined again to form the island of Rakata. The volcanic cones on the island were aligned in a north–south direction. The northernmost was called Poeboewetan and the southernmost was called Rakata. Overall approximate dimensions of the island were 5 km × 9 km (George, 2003). 4 January 1907, This event caused tsunamis that devastated Simeuleu, Nias and Batu Ms 7.6 Islands of Sumatra and extended over 950 km as measured by tide gauges. 25 June 1914, M7.6 This earthquake destroyed buildings in southern Sumatra. No tsunami was reported. 1935, Mw 7.7 Tsunami in southwest Sumatra. The 2004 Sumatra–Andaman earthquake of M9.3 generated 30-m-high tsunami when upward slip of the ocean floor was up to 15 m along a 1300-km long and 160–240-km-wide rupture. It was the deadliest tsunami killing about 300,000 people in 13 countries situated all around the Indian Ocean. The tsunami power was enhanced by large landslides over the rupture zone of 2004 earthquake. The earthquake had created large thrust ridges, about 1500 m high, which collapsed in places to produce large landslides, several kilometers across. The force of displaced water was such that blocks of rocks, massing millions of tons apiece, were dragged as much as 10 km. An oceanic trench several kilometers wide was also formed. The M8.7 great Sumatra earthquake of 28 March 2005 with an upward movement of 2 m of seafloor in an area of 400 km × 100 km generated a locally damaging 4-m-high tsunami that struck nearby islands and coastal Sumatra and was recorded by tidal stations in the Indian Ocean (asc.India.org). The earthquake and tsunami killed 665 people. The tsunami struck Nias Island with wave heights of 4–5 m. A 3–4 m wave struck the islands of Banyak and Simeulue, and the Singkil District of Sumatra. According to the Pacific Tsunami Warning Center (PTWC), tide gauges in the Indian Ocean recorded minor wave activity in the Australian Cocos Island (10–22 cm), the Maldives (10 cm) and Sri Lanka (25–30 cm). 1.4 TSUNAMIS THAT AFFECTED THE INDIAN REGION AND VICINITY Though rare, tsunamis have hit India earlier. The tsunamis in the Indian region and vicinity are listed in Table 1.2. The oldest record of tsunami is available from November 326 BC earthquake near the Indus Delta/Kutch region that set off massive sea waves in the Arabian Sea. Alexander the Great was returning to Greece after his conquest and wanted to go back by a sea route. But a tsunami due to an earthquake of large magnitude destroyed the mighty Macedonian fleet (Lisitzin, 1974). Poompuhar is a town in the southern part of India in the state of Tamil Nadu. It was a flourishing ancient town known as Kaveripattinam that was washed away in what is now recognized as an ancient tsunami in about 500 AD. This time matches with the Krakatau explosion.
326 BC About 500 AD 900 AD 1008 1762.04.12 1819.06.16 1842.11.11 1845.06.19 1847.10.31
1868.08.19 1874 1881.12.31
January 1882 1883.08.27 1884 1935.05.31 1935.11.25 1941.06.26 1945.11.27 1983.11.30 2004.12.26
1 2 3 4 5 6 7 8 9
10 11 12
13 14 15 16 17 18 19 20 21
Sri Lanka Krakatau (Volcanic Eruption) West of Bay of Bengal Andaman–Nicobar Andaman–Nicobar Andaman Islands Makran Coast Chagos ridge Off west coast of Sumatra and Andaman–Nicobar
Andaman Islands Sunderbans (Bangladesh) West of Car Nicobar
Indus Delta/Kutch region Poompuhar, Tamil Nadu Nagapattinam, Tamil Nadu Iranian Coast Bay of Bengal (Bangladesh) Kutch Northern Bay of Bengal Kutch Little Nicobar Island
Location
Note: For footnotes refer Table 1.1.
Date
8.34 −6.06 5.5 12.1 25.2 −6.85 3.307
94 92.5 63.5 72.11 95.947
11.67 22 8.52
11.12 10.46 25 22 71.9 21.5 68.37 7.333
Lat.
81.14E 105.25
92.73 89 92.43
79.52 79.53 60 92 26.6 90 23.6 93.667
Long.
List of tsunamis that affected Indian region and vicinity.
S.N.
Table 1.2.
Mw 7.5 Ms 6.5 Mw 7.7 Mw 8.0 Mw 7.7 Mw 9.3
Mw 7.9
Mw 7.5–7.9
Mw 7.8
Eq. Mag
1 1 1 1 1 3
1 6
1 1 1
4 2 4 4 4 4
3 4
4 2 4
4 4 4 4 4 3 4 3 3
1 1 1 1 1 1 1
Pro.
Cau.
3.0
4.5
I
1.25 11 1.5 (2) 5
(1)
2
1.2
4
(3)
>2 (1)
(run ups)
Max. run-up
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Lisitzin (1974) Wikipedia Kalki Krishnamurthy Murty et al. (1999) Mathur (1988) Macmurdo (1821) Oldham (1883) Nelson (1846) Berninghausen (1966), Heck (1947) NGDC/NOAA Mihir Guha, Free Journal Berninghausen (1966), Ortiz and Bilham (2003) Berninghausen (1966) Berninghausen (1966) Murty et al. (1999) NGDC/NOAA NGDC/NOAA Bilham et al. 2005 Murty et al. (1999) NGDC/NOAA NGDC/NOAA
References
14 B.K. Rastogi
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
A historical account of the earthquakes and tsunamis in the Indian Ocean
15
There is mention in the scriptures of tsunami effect at Nagapattinam in 900 AD that destroyed a Budhist monastery. According to literature available in the library of Thondaiman Kingdom in Pudukottai, Tamil Nadu, it was during the reign of Chola that waves had washed away the monastery and several temples, and killed hundreds of people. There is evidence of this in Kalki Krishnamurthy’s book “Ponniyin Selvan,” The Pinacle of Sacrifice. In the chapter “The Sea Rises”, the author explains how the sea had risen very high and a black mountain of water surged forward. The sea inundated warehouses and sheds, and began to flow into the streets. Ships and boats seemed suspended in mid-air, precariously poised on the water peaks. The book also describes how an elephant was washed away by the gushing water. Tsunami has been observed in the north Indian Ocean on the Iranian coast from a local earthquake between 1 April and 9 May 1008 (Murty et al., 1999). An earthquake occurred during 1524 AD off the coast of Dabhol, Maharashtra, and a resulting large tsunami caused considerable alarm to the Portuguese fleet that was assembled in the area (Bendick and Bilham, 1999). A tsunami is known to have occurred in the Bay of Bengal on 2 April 1762, caused by an earthquake in Bangladesh–Myanmar border region. The epicenter is believed to be 40 km southeast of Chittagong, or 61 km north of Cox’s Bazaar, or 257 km southeast of Dhaka, Bangladesh. The shock caused severe damage at Chittagong and other areas on the eastern seaboard of the Bay of Bengal. The Arakan coast got elevated for a length of more than 160 km. The quake also caused a tsunami in the Bay of Bengal. The water in the Hoogly River in Kolkata rose by 2 m. The rise in the water level at Dhaka was so sudden that hundreds of boats capsized and many people were drowned. This is the earliest well-documented tsunami in the Bay of Bengal (Mathur, 1998). 16 June 1819 in India, Kutch (Mw 7.8) there was a Severe earthquake with large changes in the elevation of the land. The town of Sindri (26.6◦ N 71.9◦ E) and adjoining country were inundated by a tremendous rush of water from the ocean. There was submergence, with the ground apparently sinking by about 5 m (Macmurdo,1821) An earthquake on 11 November 1842 near the northern end of Bay of Bengal caused a tsunami by which waters of the distributaries of the Ganges Delta were agitated. Boats were tossed about as if by waves in a squall of wind (Oldham, 1883). 19 June 1845 in India, Kutch “The sea rolled up the Koree (Kori creek, 23.6◦ N 68.37◦ E) (the east) mouth of the Indus overflowing the country as far westward as the Goongra river, northward to the vicinity of Veyre, and eastward to the Sindree Lake” (Nelson, 1846). On 31 October 1847 the small island of Kondul (7◦ 13 N 93◦ 42 E) near Little Nicobar was inundated (Heck, 1947; Berninghausen, 1966) by an earthquake whose Mw magnitude could have been greater than 7.5 (Bilham et al., 2005). Mihir Guha (http://www.freejournal.net), Former Director General of the India Meteorological Department, reported that a tsunami struck Sunderbans (Bangladesh) in May 1874, killing several hundred-thousand people. It was the result of an earthquake in Bhola district. Earthquake and tsunami both played havoc in vast areas of Sunderbans, 24-Parganas, Midnapore, Barishal, Khulna and Bhola. Even Kolkata felt its impact. It was the same year that the meteorological center in Alipore was set up. However, no written record of such an earthquake or tsunami is available. Other minor tsunamis of height up to 2 m hit the east coast of India in 1842 and 1861 (from Sumatra), 1881 (from Car Nicobar), 1883 (Krakatau), 1907 (Sumatra) and 1941 (Andaman). The 1881 Andaman earthquake of Mw 7.9 caused 1.2-m-high tsunami. Indonesian earthquake of 1907 registered about 1-m-high tsunami in India. Chennai Port Trust recorded a 2-m-high tsunami due to the eruption of the Krakatau volcano in Indonesia on 27 August 1883. Andaman earthquake of Mw 7.7 in 1941 registered a 1.5-m-high tsunami. Some of these tsunamis are described below. An earthquake of magnitude Mw 7.9 occurred at Car Nicobar Island on 31 December 1881. A tsunami was generated by this earthquake in the Bay of Bengal. Though the run-ups and waves
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
16
B.K. Rastogi
heights were not large, its effects were observed in the Andaman–Nicobar Islands and were recorded on the east coast of India. In the Nicobar Islands, the waves were less than 75-cm high (Ortiz and Bilham, 2003). A 1-m-high wave was recorded (Berninghausen, 1966) at Port Blair on South Andaman Island. On the east coast of India, the tsunami first arrived at Nagapattinam at around 10:15 a.m. local time (LT) in the form of 1.2-m-high waves (Ortiz and Bilham, 2003). Tidal gauges at other locations recorded minor variations from normal tidal changes. The tsunami then struck the rest of the Tamil Nadu coast, first hitting Chennai (Ortiz and Bilham, 2003) at 10:20 LT and then progressing north toward Vishakhapatnam in Andhra Pradesh at 10:43 LT. Waves arrived (Ortiz and Bilham, 2003) at False Point on the Mahanadi delta in Orissa at 11:15 LT and at Pamban in the Gulf of Mannar at 11:32 LT. Waves less than 0.3-m high were recorded later in the day in West Bengal by tidal gauges at Dublat at the mouth of the Hoogly river at 13:00 LT and then in Diamond Harbour at 15:10 LT. Waves attributed (Berninghausen, 1966) to this tsunami were also observed at Batticaloa and Trincomalee on the east coast of Sri Lanka. No tsunami (Ortiz and Bilham, 2003) was reported from tidal gauges in Myanmar. A tsunami was noticed at Dublat (mouth of Hoogly River) near Kolkata due to earthquake in the western part of the Bay of Bengal in 1884 (Murty et al., 1999) that reached up to Port Blair. On 26 June 1941, Andaman had an earthquake that had a moment magnitude Mw 7.7 and was located at 12.1◦ N and 92.5◦ E (Bilham et al., 2005). A tsunami was triggered by this earthquake in the Bay of Bengal. Height of the tsunami was reported to be of the order of 0.75–1.25 m. At the time no tidal gauge was in operation. Mathematical calculations suggest that the height could be of the order of 1 m. This tsunami was witnessed along the eastern coast of India. It is believed that nearly 5000 people were killed by the tsunami on the east coast of India. Local newspapers are believed to have mistaken the deaths and damage to a storm surge; however, a search of meteorological records does not show any storm surge on that day on the Coromandel Coast (Murty, 1984). National dailies like the Times of India, which reported the quake’s shaking effects, did not mention any deaths, either as a result of a storm surge or a tsunami. The deadliest tsunami prior to 2004 in south Asia was in 28 November 1945 that originated off the Makran coast of Pakistan in the Arabian Sea, and caused deaths as far as Mumbai. More than 4000 people were killed on the Makran coast by both the earthquake and the tsunami. The earthquake was also characterized by the eruption of a mud volcano, a few kilometers off the Makran coast, which are common features in western Pakistan and Myanmar. It led to the formation of four small islands. A large volume of gas that erupted from one of the islands sent flames leaping “hundreds of meters” into the sky (Mathur, 1998). The most significant aspect of this earthquake was the tsunamis that it triggered. The tsunami reached a height of 17 m in some Makran ports and caused great damage to the entire coastal region. A good number of people were washed away. The tsunami was also recorded at Muscat and Gwadar. The tsunami had a height of 11.0–11.5 m in Kutch, Gujarat (Pendse, 1945). At 8:15 a.m., it was observed on Salsette Island (i.e. Mumbai) (Newspaper archives, Mumbai). It was recorded in Bombay Harbour, Versova (Andheri), Haji Ali (Mahalaxmi), Juhu (Ville Parle) and Danda (Khar). At Versova (Andheri, Mumbai), five persons who were fishing were washed away. At Haji Ali (Mahalaxmi, Mumbai), six persons were swept into the sea. At Danda and Juhu, several fishing boats were torn off their moorings. The tsunami did not do any damage to Bombay Harbour. Most persons who witnessed the tsunami said that it rose like the tide coming in, but much more rapidly. The height of the tsunami in Mumbai was 2 m. A total of fifteen persons were washed away in Mumbai. The Mw 7.7, 30 November 1983 earthquake in Chagos Archipelago, was one of the strongest earthquakes ever recorded in the Indian Ocean. It occurred at 17:46 p.m. UTC. The earthquake caused some damage (NEIC) to buildings and piers on Diego Garcia, which is part of the Chagos Archipelago. The 1983 earthquake triggered a tsunami in the region. In the lagoon, on Diego Garcia, there was a 1.5-m rise in wave height and there was some significant wave damage near the southeastern tip of the island. A 40-cm wave was also recorded at Victoria, Seychelles. There was a large zone of discolored seawater observed 60–70 km north–northwest of Diego Garcia.
A historical account of the earthquakes and tsunamis in the Indian Ocean
17
Moment-tensor solution indicated normal faulting along an east–west plane at a depth of 10 km with source duration of 34 s.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
1.5
CONCLUSIONS
• It is assessed that Sumatra and Andaman regions will probably not generate great earthquakes for a few decades due to occurrence of 2004 Mw 9.3 and 2005 Mw 8.7 earthquakes. Southern Sumatra has potential for a great earthquake. However, the effect of tsunami due to this in India and Sri Lanka may be a limited one as the path of tsunami will be oblique to the rupture zone. • The Makran subduction zone of southern Pakistan is seismically less active, but has produced great earthquakes. The 28 November 1945 (Mw 8.0) earthquake generated the last major tsunami in the Arabian Sea. This earthquake occurred in the eastern part of the Makran zone, two sides of which remain potential zones for great earthquakes that can generate tsunamis in future. • Indus Delta and possibly the coasts of Kutch and Saurashtra are also potential zones for great earthquakes and tsunami. • Tsunami was generated by an earthquake in 1762 in Myanmar and in 1874 by an earthquake near Bangladesh. In future also, some earthquakes in these regions can possibly generate tsunamis. • The Carlsberg ridge, Chagos ridge and Ninety-East ridge can give rise to local tsunami due to normal and thrust component of motion for major earthquakes as happened due to Mw 7.7 earthquake of 30 November 1983 near Diego Garcia along the Chagos ridge. • Eighty percent of the tsunamis in the Indian Ocean are from Sunda arc region where on average tsunamis are generated once in 3 years. In the rest of the Indian Ocean tsunamis can be generated once in 10 years or so. REFERENCES Bendick, R. and Bilham, R. (1999). A Search for Buckling of the SW Indian Coast related to Himalayan Collision. In: A. Macfarlane, R.B. Sorkhabi, and J. Quade, (eds.), Himalaya and Tibet: Mountain Roots to Mountain Tops: Geological Society of American Special paper 328. pp. 313–322. Berninghausen, W.H. (1966). Tsunamis and Seismic Seiches Reported from Regions Adjacent to the Indian Ocean. Bull. Seism. Soc. Am., 56(1), 69–74. Bilham, R., Engdahl, R., Feld, N., and Sayabala, S.P. (2005). Partial and Complete rupture of the IndoAndaman plate Boundary 1847–2004. Seism. Res. Lett., 76(3), 299–311. Engdahl, E.R., van der Hilst, R.D., and Buland, R.P. (1998) Global teleseismic earthquake relocation with improved travel times and procedures for depth determination. Bull. Seis. Soc. Am., 88, 722–743. George, P.-C. (2003). Near and Far-Field Effects of Tsunamis Generated By the Paroxysmal Eruptions, Explosions, Caldera Collapses and Massive Slope Failures of The Krakatau Volcano in Indonesia on August 26–27, 1883. Sci. Tsunami Hazards, 21(4), 191–222. Heck, N.H. (1947). List of Seismic Sea Waves. Bull. Seism. Soc. Am., 37, 269–286. Lisitzin, E. (1974). Sea Level Changes, Elsevier Oceanographic Series, No.8, New York, 273 pp. Macmurdo, C. (1821). Account of the Earthquake which Occurred in India in June 1819. Edinburgh Phil. J., 4, 106–109. Mathur, S.M. (1998). Physical Geology of India. National Book Trust of India, New Delhi. Murty, T.S., Bapat, A., and Prasad, V. (1999). Tsunamis on the Coastlines of India. Sci. Tsunami Hazards, 17(3), 167–172. Murty, T.S. (1984). Storm Surges – Meteorological Ocean Tides. Bull. Fish. Res. Board Can., Ottawa. Nelson, C. (1846). Notice of an Earthquake and a Probable Subsidence of the Land in the District of Cutch, near the Mouth of Koree, or the Eastern Branch of the Indus in June 1845. Quart. J. Geol. Soc. London, 2, 103.
18
B.K. Rastogi
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Newcomb, K.R. and McCann, W.R. (1987). Seismic History and Tectonics of the Sunda Arc. J. Geophys. Res., 92(B1), 421–439. Oldham, T.A. (1883). Catalogue of Indian earthquakes. Geo. Surv. Ind. Mem, 19(Pt 3), 163–215. Ortiz, M. and Bilham, R. (2003). Source Area and Rupture Parametres of the 31 December 1881 Mw = 7.9 Car Nicobar Earthquake Estimated from Tsunamis Recorded in the Bay of Bengal. J. Geophys. Res. Solid Earth, 108(4), ESE 11, 1–16. Pendse, C.G. (1945). The Mekran earthquake of the 28th November 1945. India Met. Deptt. Scientific Notes, 10, 141–145. Simkin, T. and Fiske, R.S. (1983). Krakatau 1883: The Volcanic Eruption and its Effects. Smithsonian Institution Press, Washington, DC. 464 pp. Soloviev, S.L., and Go, C.N. (1974) Catalog of tsunamis in the western coast of Pacific Ocean (in Russian), Nauka Publishing Co., Moscow.
CHAPTER 2
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Impact of Coastal Morphology, Structure and Seismicity on the Tsunami Surge K.S.R. Murthy, V. Subrahmanyam, G.P.S. Murty and K. Mohana Rao National Institute of Oceanography, Regional Centre, Lawsons Bay, Visakhapatnam, Andhra Pradesh, India
2.1
INTRODUCTION
The Sumatra earthquake (Mw9.3) with its epicenter located southwest of Sumatra island (3.7◦ N, 95◦ E) occurred at the interface of the Indian and Burmese plates, where the former subducts beneath the Burmese plate all along the Andaman–Sumatra–Sunda Arc (McCloskey et al., 2005; Sieh, 2005; Gupta, 2005; Satish Singh, 2005). It is the Fifth largest earthquake since 1990 and the largest since the 1964 Alaska earthquake. In this region the Indian plate moves towards northeast at the rate of 6 cm/year relative to the Burmese plate which results in an oblique convergence at the Sunda trench. The oblique motion is partitioned into thrust faulting and strike-slip faulting which occurs at the plate interface. The 26 December earthquake was the result of such thrust faulting (Radhakrishna, 2005). The earthquake with a focal depth of 30 km has affected a length of nearly 1300 km along the interface, with the rupture mainly propagating towards the north of the epicenter within an average width of 100 km. The average displacement of the fault plane is 15 m. The rupture appears to have occurred in two stages; in the first stage it propagated rapidly to about 400 km followed by a slow rupture of about 500 km. The total duration of the rupture is around 800 s (13 min). Subsequent observations indicate that this catastrophic event made the Earth wobble on its axis. An anticlockwise rotation of the islands of the Andaman and Nicobar region by about 2–3 m has been reported. The southern part of these islands has undergone submergence of about 2 m, whereas the Northern Andaman area has experienced an uplift of about 1 m (Subramanian, 2005; Murthy, 2005a,b). As many as 2000 aftershocks have been reported in this area even after 7 months of the main event, with some of them having magnitudes greater than 7.0 (e.g., Mw 7.3, 26 December 2004; Mw 8.4, 28 March 2005; and Mw 7.2, 24 July 2005). This earthquake appears to have disturbed a water column of nearly 100 × 70 × 1.5 km3 volume (Radhakrishna, 2005), resulting in an unique Indian Ocean Tsunami that has engulfed the coastal area of Indian Ocean rim countries including Indonesia, Thailand, India, Sri Lanka and as far west as Somalia in EastAfrica. The inundation on the coastal stretches is as much as 2 km in some cases, the worst affected being the Andaman and Nicobar Islands and the southern parts of east coast of India. Tidal measurements at Port Blair, Chennai and Visakhapatnam indicate surge heights of the order of 1.5–5.5 m above mean sea level, with maximum height near Nagapattinam and Karaikal (Chadha et al., 2005). The tsunami waves reached the Nicobar Islands, within few minutes, the Andaman Islands within half-an hour and the Indian coast in about 2–2.5 h (Subramanian, 2005; Chadha et al., 2005). Sea level transgression was of the order of 1100–1200 m in case of Andaman coast, 350–1100 m in case of Nicobar and nearly 200–800 m in case of Tamil Nadu coast of India. The twin events of the earthquake and tsunami not only stress the importance of an integrated warning system for the Indian Ocean rim countries but also the need for new disaster management 19
20
K.S.R. Murthy et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
plans to mitigate such hazards. As far as the Indian coastal regions are concerned, detailed geophysical surveys on the coastal morphology, structure and coastal seismicity are very essential in order to understand the relationship between tsunami run–up heights, inundation extent and the shelf/slope morphology. Studies on coastal structure and seismicity are useful in estimating the seismic hazard risk of the coastal region as well as in understanding the influence of tectonics on the tsunami initiation and surge. In this chapter, we present a detailed analysis on the morphology, structure and seismicity of the Cauvery offshore basin, which includes the tsunami-affected Nagapattinam-Cuddalore shelf of Tamil Nadu margin. 2.2
COASTAL MORPHOLOGY AND STRUCTURE IN RELATION TO TSUNAMI SURGE – A CASE STUDY
The east coast of India is more prone for natural hazards like cyclones, storm surges and now the new hazard in the form of tsunami, in comparison to the west coast. The Eastern Continental
Figure 2.1.
Geophysical data coverage over the ECMI (thin solid lines represent cruise tracks over ECMI and Bengal Fan).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Impact of coastal morphology, structure and seismicity on the tsunami surge
21
Margin of India (ECMI) is a passive margin evolved during the process of breakup of Eastern Gondwanaland in the late Cretaceous (Murthy et al., 1995). In the pre-breakup scenario, the present Krishna–Godavari Basin was conjugate with the Enderbyland of east Antarctica (Murthy et al., 1997). Most of the Indian rivers, having an eastward slope, join the Bay of Bengal, thereby resulting in a mosaic of basinal and non-basinal morphology. The shallow bays associated with the basinal areas are more affected by the crossing of cyclones and storm surges, due to the wider shelf with gentle slope. One of the important parameters to be considered in the context of cyclones/storm surges/tsunami is the seabed morphology, including the shelf/slope characteristics of the margin. In this connection, the Regional Centre, National Institute of Oceanography (NIO), Visakhapatnam, India has collected bathymetry, magnetic and gravity data over the ECMI from Karaikal in the south to Paradip in the north (Figure 2.1). Morphology, stratigraphy, structure and tectonics of ECMI, including the offshore river basins like the Cauvery, Krishna–Godavari and Mahanadi were analyzed from this data by Murthy et al. (1993), Murthy et al. (1995), Subrahmanyam et al. (1995), Murthy et al. (1997) and Murty et al. (2002). In this chapter, we make use of this data to analyze the factors that facilitated the high tsunami surge in case of the Nagapattinam–Cuddalore shelf of Tamil Nadu margin.
Figure 2.2.
Bathymetry map of ECMI.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
22
K.S.R. Murthy et al.
2.3
FACTORS GUIDING THE HIGH TSUNAMI SURGE AT NAGAPATTINAM–CUDDALORE SHELF
It is evident from the bathymetric maps and sections of the eastern continental margin that the shelf is in general steep and narrow in the south (except off Chennai) whereas it is relatively wider and gentle in the north (Figures 2.2 and 2.3 and Table 2.1). However, some of the offshore river basins are associated with shallow bays with a concave coastline (e.g., Cauvery and Krishna– Godavari). There is a significant change in the direction of the coastline at 14◦ N, south of which the coastline is oriented in north–south direction, whereas north of it is northeast–southwest in orientation. The southern part of the eastern continental shelf of India from Karaikal in the south (11◦ N) and Visakhapatnam in the north (approximately 18◦ N) was affected by the tsunami surge caused by the Sumatra earthquake (Figure 2.1). However, the Tamil Nadu shelf, in particular Nagapattinam– Cuddalore part (shown as shaded zone in Figures 2.1–2.6) was the worst affected by the tsunami surge of 26 December 2004. The observed run-up heights in this part are as much as 5.2 m (e.g., Nagapattinam) with inundation to a distance of nearly 800 m into the interior (Chadha et al., 2005). This part of the shelf represents the offshore Cauvery Basin. Qualitative inferences drawn
Figure 2.3.
Bathymetry sections of ECMI.
Impact of coastal morphology, structure and seismicity on the tsunami surge
23
by Raval (2005) and Subrahmanyam et al. (2005) suggest that some of the main reasons for the high Tsunami surge in case of Nagapattinam–Cuddalore shelf of Tamil Nadu margin are:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
1 its relative proximity to the source of the tsunami, 2 the configuration of the coastline and the bathymetry of the shelf and 3 the tectonics associated with the offshore Cauvery basin. Earlier analysis from magnetic data of the Cauvery Basin (Subrahmanyam et al., 1995) suggest that the basin is fault controlled, with two major east–west faults (shown as F1 and F2, in Figure 2.4(a)) extending from the land into the offshore. The northern fault is located north off Pondicherry, whereas the southern fault is located north off Vedaranyam. Magnetic and gravity data indicate shallow basement on either side of the main basin, with a down throw underlying major part of the basin (Schematic cross section shown in Figure 2.4(b)). Dyke intrusions, with Table 2.1.
Shelf/slope characteristics off selected places over the ECMI (Murthy et al., 1993).
Location (See Figure 2.1)
Water depth at shelf break (m)
Shelf edge distance from coast (km)
Shelf gradient (ratio)
Slope gradient (ratio)
Depth at which marginal high (M.H.) is recorded (m)
Paradip
100
68
1 : 320
1 : 50
Not clear
Puri
130
56
1 : 300
1 : 43
Not clear
Chilka lake
220
52
1 : 200
1 : 28
800?
South of Chilka lake
220
47
1 : 200
1 : 35
Not clear
South of Gopalpur
220
51
1 : 200
1 : 16
1500–2000
Kalingapatnam
120
44
1 : 280
1 : 16
1800–2000
South of Kalingapatnam Visakhapatnam
130
53
1 : 345
1 : 15
2000–2100
200
54
1 : 250
1 : 12
1900–2100
North of Kakinada
200
57
1 : 225
1 : 12
2400–2600
Krishna river
70
35
1 : 300
1 : 25
Nizampatnam
70
57
1 : 300
1 : 25
Madras
200
55
1 : 200
1:8
Uncertain due to slumping over the slope Uncertain due to slumping over the slope 2700
North of Pondicherry Karaikal
90
51
1 : 400
1:6
3000
70
29
1 : 200
1 : 19
3000
Nagapattinam
70
47
1 : 340
1 : 17
3000
Remarks (only relative terms) Wide shelf, gentle slope Wide shelf, gentle slope Wide shelf, gentle slope Wide shelf, gentle slope Wide shelf, steep slope Narrow shelf, steep slope Wide shelf, steep slope Wide shelf, steep slope Wide shelf, steep slope Narrow shelf, gentle slope Wide shelf, gentle slope Wide shelf, very steep slope Wider shelf, very steep slope Narrow shelf, steep slope Narrow shelf, steep slope
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
24
K.S.R. Murthy et al.
north–south trends were inferred north and south of these two east-west fault trends. In the central part of the basin, a minor basement rise representing probable offshore extension of the Kumbakonam Ridge was also delineated (shown as KR in Figure 2.4(a)). Thus the Nagapattinam– Cuddalore part of Tamil Nadu margin is a structurally controlled basin flanked by two major fault lineations. The free-air gravity anomaly map (Figure 2.5) is characterized by a significant north–south trending linear gravity low with major discontinuities, FZ1 and FZ2 along 12◦ 15 N and 11◦ 45 N respectively, indicating major fault zones (Murty et al., 2002). The discontinuity observed in the gravity data (FZ1) correlates well with the major discontinuity observed in the magnetic data (F1). Similar trend has also been evidenced from the bathymetry, which may indicate a major structural trend. The east–west trending lineation F2, (Figure 2.4(a)) which correlates well with the landward Palghat–Cauvery Lineament (PCL) can be considered as its offshore extension. The PCL has been termed as a shear zone. These studies have also revealed a good correlation between land–ocean tectonics and coastal seismicity and the epicenter of a moderate earthquake of magnitude 5.6 that occurred on 26 September 2001 was located over the fault trend FZ1, which was inferred as the offshore extension of the Moyar–Bhavani–Attur (MBA) lineament (Figure 2.5 and Murty et al., 2002) Detailed bathymetry map and sections of the Nagapattinam–Cuddalore shelf (from 10.5◦ N to about 12◦ N, Figures 2.6(a) and (b)) indicate that one of the main reasons for the higher run-up heights and inundation in Nagapattinam–Cuddalore coast could be the concave shape of the shelf with a gentle slope, which might have accelerated the tsunami surge to flush through at a rapid force. Bathymetry sections off Pondicherry (CB3) and Cuddalore (CB4) indicate a gentle continental shelf and slope up to about 3000 m water depth representing the concave nature of the shelf. The sections off Vedaranyam (CB9) in the south and those in the north of Pondicherry (CB1 and CB2) indicate a wider shelf with a steeper slope (Figure 2.6(b)), representing the southern and northern boundaries of the concave shelf. The area within these boundaries is more affected by the tsunami surge. Earliest bathymetry observations over the offshore Cauvery basin revealed the presence of submarine canyons off Cuddalore and Pondicherry (Varadachari et al., 1968). Subsequent geophysical studies also suggested that major valleys off Pondicherry are formed due to the existence of mega lineaments (Rao et al., 1992). The lineament pattern played a major role in shaping the continental slope morphology, besides erosional and depositional processes (Figure 2.5 of Rao et al., 1992). The high run-up heights and inundation in case of Nagapattinam and Cuddalore shelf are therefore the result of a combination of a structurally controlled basin with a favorable seabed morphology. 2.4
COASTAL SEISMICITY
The effect of the Sumatra earthquake is relatively insignificant on the east coast of India, though its southern part, namely the Tamil Nadu coast is severely affected more by the tsunami surge. The Andaman and Nicobar Islands however were affected both by the earthquake and the tsunami. Nevertheless it is important to note that during the months of December 2004 to July 2005, the coastal areas all along the east coast of India have experienced tremors of few seconds duration both due to the main event of 26 December 2004, and by the aftershocks of 28 March 2005 (Mw 8.3) and 24 July 2005 (Mw 7.3). This is a new development and it is quite likely that the aftershocks are likely to continue in the Andaman and Nicobar region with relatively high frequency and with increased amplitude (>5.0). This implies that the coastal regions have to take note of this new seismic hazard. Hitherto the observed seismicity in the coastal regions of the Stable Continental Region (SCR) is mainly due to the reactivation of weak zones due to
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
(a)
(b)
Figure 2.4.
(a) Magnetic anomaly map of Cauvery basin, with structural interpretation (contour interval: 20 nT). (b) Schematic cross section of basement along Profile A–B of Cauvery Basin.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
26
K.S.R. Murthy et al.
Figure 2.5.
Free air gravity anomaly map of Cauvery offshore basin (contour interval 10 mGal). FZ1 and FZ2 are the inferred faults. Location and focal mechanism of the Pondicherry earthquake are shown (Murty et al., 2002).
the stresses developed as a result of the northward movement of the Indian plate. However there is now a new possibility of reactivation of weak zones of the coastal areas, due to the aftershock, of high amplitude, occurring continuously at the eastern end, that is along the Andaman and Nicobar arc. Under these circumstances, it is very essential to carryout geophysical studies in the coastal region in order to identify the land–ocean tectonic lineaments and their correlation with earlier reported seismicity so that seismic zonation maps can be generated for the Coastal regions. Some attempts were made in this direction particularly over the Tamil Nadu and Andhra Pradesh margin. Geophysical data comprising of bathymetry, magnetic, gravity and in some cases shallow seismic data were collected on the continental shelf region of Tamil Nadu and Andhra Pradesh margin. These data sets were utilized to delineate the offshore extension of the Coastal lineaments. Correlation some of these lineaments with earlier reported seismicity
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Impact of coastal morphology, structure and seismicity on the tsunami surge
Figure 2.6.
27
(a) Bathymetry map of Cauvery offshore basin (F1 and F2 are fault trends inferred from Figure 4(a)). (b) Bathymetry sections of Cauvery offshore basin (horizontal scale: 1 cm = 15 km) (horizontal distance in cms as measured from Figure 4(a)).
indicate neotectonic activity associated with these lineaments. As a case study the Cauvery basin study is presented below. 2.5
EVIDENCE OF FAULT REACTIVATION OFF PONDICHERRY COAST FROM MARINE GEOPHYSICAL DATA
Present study is mainly focused on the northern part of the Cauvery offshore basin, between latitudes 10◦ –14◦ N and longitudes 80◦ –82◦ E (Figure 2.7), over which the bathymetry, gravity and magnetic data were collected (Murty et al., 2002). The bathymetry map (Figure 2.7) in general, shows a linear north–south trend, except between latitudes 10◦ 45 N and 12◦ 15 N, where the contours trend coastward indicating a major fault zone. Similar trend is also evidenced by the geophysical anomalies. The free-air gravity anomaly map (Figure 2.5) is characterized by a significant north–south trending linear gravity low with major discontinuities, FZ1 and FZ2 along 12◦ 15 N and 11◦ 45 N respectively, indicating major fault zones. Total field magnetic anomaly map (Figure 2.4) also shows two major discontinuities, F1 and F2 along 12◦ 15 N (off Pondicherry), and 10◦ 45 N (south of Karaikal), respectively (Subrahmanyam et al., 1995). The discontinuity observed in the gravity data (FZ1) correlates well with the major discontinuity observed in the magnetic data (F1). Similar trend has also been evidenced from the bathymetry,
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012 28 K.S.R. Murthy et al.
Figure 2.6.
(Continued)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Impact of coastal morphology, structure and seismicity on the tsunami surge
Figure 2.7.
29
Bathymetry map of Cauvery offshore basin (Murty et al., 2002), dashed lines indicate ship tracks along which bathymetry and gravity data were acquired. Black dot in the offshore indicates the location of Pondicherry earthquake (Mw 5.5) of 25 September 2001. MBA: Moyar–Bhavani–Attur lineament; PCL: Palghat–Cauvery Lineament.
which may indicate a major structural trend. F1, trending eastnortheast–westsouthwest, which was interpreted earlier as an offshore extension of northeast–southwest trending Kumbum– Pondicherry lineament on land. The east–west trending lineation F2, which correlates well with the landward PCL can be considered as its offshore extension. Another major discontinuity, having eastnortheast–westsouthwest trend (FZ2), as observed in the gravity data, may be spatially correlated to the east–west trending MBA lineament on land. This lineament was interpreted as a shear zone and also been suggested as a suture zone and a steeply dipping thrust fault. It can be suggested that FZ2 may be the offshore extension of the MBA lineament (see Subrahmanyam et al., 1995 for references). An earthquake of M5.5, that occurred on 25 September 2001 off Pondicherry, can be considered as a fairly larger event for the south Indian shield so far. Located at 11.95◦ N and 80.23◦ E, the epicenter falls over the continental slope, about 40 km off Pondicherry at 1900 m water depth with a focal depth of 10 km. The epicenter of the earthquake falls over the reported fault zone FZ2, which has been interpreted as an offshore extension of the MBA lineament. The focal
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
30
K.S.R. Murthy et al.
mechanism solution of the earthquake (Figure 2.5) suggests thrust faulting and indicates small strike-slip component with left lateral motion along a northeast striking nodal plane (Rastogi, 2001). The other nodal plane in the solution strikes in the east–west direction and is preferred to be the fault plane on the basis of the seismicity trend in this region along 12◦ N latitude. Chandra (1977) has reviewed the seismicity of Peninsular India and obtained the focal mechanism solutions for some of the prominent earthquakes, including the Bhadrachalam (Mw 5.7, 1969) and Ongole (Mw 5.4, 1967) earthquakes. Some similarities can be drawn between these earthquakes and the Pondicherry earthquake, as they all fall in the vicinity of the 80◦ E longitude, along which the coastline trends nearly north–south in the south Indian shield. The fault-plane solutions suggest that they are of thrust type. Interestingly, the Ongole earthquake also suggests faulting along an east–west trending fault, which is very similar to the Pondicherry earthquake. Based on the qualitative analysis of geophysical data, it is conjectured that the Pondicherry earthquake might have occurred due to the release of compressional stresses along the fault zone FZ2, which may be the offshore extension of east–west trending MBA lineament. This further indicates the recent reactivation of the Precambrian shear zones in the offshore regions of Tamil Nadu, particularly the MBA lineament. Similar studies have been carried out on the land–ocean tectonics and neotectonics associated with some of these lineaments over the Palar basin, Visakhapatnam and Vizianagaram shelf of Andhra Pradesh (Subrahmanyam et al., 1999; Murthy et al., 2001). The results suggest reactivation of some coastal lineaments and thus form part of the seismicity associated with SCR regions, as discussed already. In the context of the seismicity experienced over the coastal regions of east coast of India in the post-Sumatra earthquake scenario, these studies assume significant importance and must be continued for the entire east coast.
ACKNOWLEDGEMENTS The authors are thankful to Dr. Satish R. Shetye, Director, N.I.O, Goa for his encouragement. Thanks are also due to Ms. T. Sridevi, Ms. Sunita Rani Panda, Ms. A. Anuradha and Ms. B. Adilakshmi and for their help in the preparation of the manuscript. REFERENCES Chadha, R.K., Latha, G., Harry, Y., Peterson, C., and Toshctama, K. (2005). The Tsunami of the Great Sumatra Earthquake of M.9.0 on 26, December, 2004 – Impact on the east coast of India. Curr. Sci., 88, 1297–1300. Chandra, U. (1977). Earthquakes of peninsular India – A seismotectonic study. Bull. Seism. Soc. of Am., 67, 1387–1413. Gupta, H.K. (2005). Early warning system for oceanographic disasters in Indian Ocean (tsunami and storm surges): The Indian initiative. J. Geol. Soc. India, 65, 639–646. McCloskey, J., Nalbant, S.S., and Steacy, S. (2005). Scientists issue Indonesia earthquake warning. Nature, 434, 582. Murthy, K.S.R. (1995). Some geodynamic complexities related to the evolution of Bengal Fan and the neotectonic activity of the South Indian Shield. Curr. Sci., 73, 10. Murthy, K.S.R. (2005a). First oceanographic expedition to survey the impact of the Sumatra earthquake and the Tsunami at 26th December 2004. Curr. Sci., 88, 1038–1039. Murthy, K.S.R. (2005b). Oceanographic expedition to study the Post-Tsunami impact in the Bay of Bengal and Andaman and Nicobar Islands, paper presented at the International Symposium on “External Flooding Hazards at Nuclear Power Plant Sites” (IAEA/NPCIL) held at Kalpakkam, Tamilnadu, India, 29th August, 2005.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Impact of coastal morphology, structure and seismicity on the tsunami surge
31
Murthy, K.S.R., Rao, T.C.S., Subrahmanyam, A.S., Malleswara, Rao M.M., and Lakshminarayana, S. (1993). Structural lineaments from magnetic anomaly map of eastern continental margin of India and Bengal Fan. Mar. Geol., 114, 169–183. Murthy, K.S.R., Subrahmanyam, A.S., Lakshminarayana, S., Chandrasekhar, D.V., and Rao, T.C.S. (1995). Some geodynamic aspects of Krishna–Godavari basin, east coast of India. Conti. Shelf Res., 15, 779–788. Murthy, K.S.R., Malleswara Rao, M.M., Venkateswarlu, K., Subrahmanyam, A.S., Lakshminarayana, S., and Rao, T.C.S. (1997). Marine magnetic anomalies as a link between granulite belts of east coast of India and Enderbyland of Antarctica. J. Geol. Soc. India, 49, 153–158. Murthy, K.S.R., Subrahmanyam, A.S., Murty, G.P.S., and Sarma, K.V.L.N.S. (2001). Reactivation of some land–ocean tectonic lineaments of the eastern continental shelf under the compressional stress regime in the South Indian Shield – A Geophysical study, 38th Annual Convention of IGU, Visakhapatnam, pp. 47–48. Murty, G.P.S., Subramanyam, A.S., Murthy, K.S.R., and Sarma, K.V.L.N.S. (2002). Evidence of Fault Reactivation off Pondicherry Coast from Marine Geophysical Data. Curr. Sci., 83, 1446–1449. Radhakrishna, B.P. (2005). Devastating Tsunami strikes coastline of India on 26th December 2004. J. Geol. Soc. India, 65, 129–134. Rao, L.H.J., Rao, T.S., Reddy, D.R.S., Biswas, N.R., Mohapatra, G.P., and Murty, P.S.N. (1992) Morphology and sedimentation of continental slope, rise and abyssal plain of western part of Bay of Bengal. Geol. Surv. of India, Special Publication, 29, 209–217. Rastogi, B.K. (2001). A note on the Focal Mechanism of Pondicherry earthquake. EQ News, Biannu. News Lett., Department of Science and Technology (DST), New Delhi, 2, 3. Raval, U. (2005). Some factors responsible for the devastation in Nagapattinam region due to Tsunami of 26th December 2004. J. Geol. Soc. India, 65, 647–649. Satish Singh (2005). Sumatra earthquake indicate why the rupture propagated Northwards. EOS, 86, 48. Sieh. K. (2005). Aceh-Andaman earthquake: What happened and what next? Nature, 434, 573–574. Subrahmanyam, A.S., Lakshminarayana, S., Chandrasekhar, D.V., Murthy, K.S.R., and Rao, T.C.S. (1995). Offshore Structural trends from magnetic data over Cauvery Basin, east coast of India. J.Geol. Soc.India, 46, 269–273. Subrahmanyam, A.S., Venkateswarlu, K., Murthy, K.S.R., Malleswara Rao, M.M., Mohan Rao, K., and Raju, Y.S.N. (1999). Neotectonism–An offshore evidence from eastern continental off Visakhapatnam, Curr. Sci., 76, 1251–1254. Subrahmanyam, C., Gireesh, R., and Gahalaut, V. (2005). Continental slope characteristics along the Tsunami – affected areas of eastern offshore of India and Sri Lanka, Jour Geol. Soc. India, 65, 778–780. Subramanian, B.R. (2005) – Restricted circulation. Preliminary report on studies of seismic pattern, tidal pattern and submergence to help locate resettlement areas in Andaman and Nicobar Islands, submitted to Department of Science and Technology, New Delhi. Varadachari, V.V.R., Nair, R.R., and Murthy, P.S.N. (1968). Submarine canyons off the Coramandel coast. Bull. Nat. Inst. Science, 38, 457–462.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 3
Tsunamigenic Sources in the Indian Ocean: Factors and Impact on the Indian Landmass
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
R.K. Chadha National Geophysical Research Institute, Hyderabad, Andhra Pradesh, India
3.1
INTRODUCTION
The Indian Ocean Tsunami of December 26, 2004 was the deadliest ocean related disaster in living memory, claiming about 300,000 human lives. The event was associated with the world’s second largest earthquake of M9.3 which occurred off the coast of Sumatra in Sunda trench. Initially, the magnitude of this earthquake was estimated to be 9.0. This became an extraordinary event due to its magnitude and rarity of this phenomenon in the Indian Ocean region relative to Pacific Ocean, where tsunami are very frequent. The tsunami from Sumatra propagated throughout the oceans on the earth with devastating effects on the Indian Ocean rim countries like Indonesia, Thailand, Malaysia, Myanmar, Bangladesh, India, Sri lanka, Maldives and Africa (Figure 3.1).
Figure 3.1.
Map of Indian Ocean rim countries affected by the December 26, 2004 Indian Ocean Tsunami due to M9.3 earthquake off the coast of Sumatra. M8.7 earthquake on March 28, 2005 which occurred 250 km south of December 26 event is also shown. (Source: http://www.USGS.gov, adapted and modified.)
33
34
R.K. Chadha
Another great earthquake of M8.7 occurred on March 28, 2005 about 150 km further southeast of the December 2004 event, with its hypocenter below Nias Island in the Sunda trench. This earthquake set off alarms leading to tsunami warnings along the Indian east coast and also in parts of Indonesia and Thailand. The tsunami warnings were called off within few hours, as this earthquake did not generate any consequential tsunami. Although this earthquake was 6 times smaller than the December 26, 2004 earthquake, it was still strong enough to generate large tsunami. However, there was only a single report of small tsunami of few tens of centimeters height from Cocos Islands.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
3.2
CAUSES OF TSUNAMI
Tsunamis are water waves generated by the disturbance caused by submarine earthquakes, landslides, explosive volcanism and meteorite impact with the ocean. Among these, submarine earthquakes are the major cause for tsunami generation. More than 90% of the tsunamis generated in the Pacific Ocean over the last 200 years were caused due to earthquakes. But, not all earthquakes generate tsunamis. Major factors for tsunami generation are: (i) magnitude, (ii) depth and (iii) the nature of earthquake faulting and rupture of ocean floor. It has been generally observed that the tsunami caused by earthquakes of M7.5–7.8 are local in nature and will not cause great damage at distant regions. However, this type of events can cause secondary effects like triggering submarine landslides or slumps. Most of the earthquakes of magnitude greater than 7.9 cause both destructive local tsunamis near the epicenter and also significant sea level changes leading to high run-up heights causing severe damages at great distances. The second most important factor is the depth of the earthquake. Shallow marine events that deform the seafloor generate tsunami rather than a deeper event. Lastly, earthquakes with thrust or normal fault mechanism are more likely to generate a tsunami than a strike-slip fault where there is only horizontal displacement. Submarine landslides can also generate tsunamis in some cases, provided volume of material moved is substantial and move at a great speed. The characteristics of these tsunamis are different from those of earthquake generated, which displaces seabed. In case of earthquakes the maximum energy is focused perpendicular to the strike of the fault and decreases in intensity along the strike of the fault. The landslide-generated tsunamis are more focused. The slide moves in a down slope direction and the wave propagates both upslope and parallel to the slide. While earthquakegenerated waves are very symmetrical close to their source, landslide generated ones have a shape that is best characterized by N-shaped waves. The wave train is led by a very low crested wave followed by a trough up to 3 times greater in amplitude. The second wave in the wave train has the same amplitude as the trough, but over time, it decays into three or four waves with decreasing wave periods. The initial inequality between the crest of the first wave and the succeeding trough enables landslide-generated tsunami to acquire greater run-up heights than those induced by earthquakes (Bryant, 2001). Submarine eruptions within 500 m of the ocean surface can disturb the water column enough to generate a surface tsunami wave. Below this depth, the weight and volume of the water suppress surface wave formation. Tsunami from this cause rarely propagates more than 150 km from the site of eruption. More significant are submarine explosions that occur when ocean water comes in contact with the magma chamber. This water is converted instantly to steam and in the process produces a violent explosion. Krakatau volcanic explosion during 1883 produced a tsunami of 40 m height killing 36,000 people mostly on the islands of Java and Sumatra. The impact of the meteorite into an ocean can also cause tsunami, but more dangerous will be the splash that travels at high velocity for hundreds of kilometers and fall back to the surface of the earth over equivalent distances. But this is a very rare phenomenon.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunamigenic sources in the Indian Ocean
Figure 3.2.
35
Map showing different tectonic plates and locations of subduction zones (solid lines) and mid-oceanic ridges (zig-zag lines). Red dots are volcanoes. (Source: www.usgs.gov)
3.3 TECTONIC ENVIRONMENT OF TSUNAMI Typically, large earthquakes in the subduction zones or trenches are the main source of tsunami generation through out the world oceans. The tsunami can be generated either by earthquake faulting which deforms the ocean floor or by triggering a huge submarine landslide. The Pacific Ocean is encircled through out with such trenches due to the interaction of the Pacific plate with other plates like, Nazca, Cocos, Indo-Australian, Eurasian, Kamchatka, Philippines, North American and South American plates. Most of the world’s great earthquakes have occurred in these subduction zones, which are seismically very active and often referred to as “Ring of Fire” (Figure 3.2). The Hawaiian Hot Spot is another source of tsunami generation in the Pacific Ocean. During 1900 to 2001, 796 tsunamis were observed or recorded in the Pacific Ocean according to the Tsunami Laboratory in Novosibirsk, Russia. Of them, 117 caused casualties and damage, mostly near the source, and 9 caused widespread destruction throughout the Pacific. The greatest number of tsunamis during any 1 year was 19 in 1938, but all were minor and caused no damage. There was no single year of the period that was free of tsunami. Seventeen percent of the total tsunamis were generated in or near Japan. The distribution of tsunami generation in other areas is as follows: South America, 15% ; New Guinea Solomon Islands, 13%; Indonesia, 11%; Kuril Islands and Kamchatka, 10%; Mexico and Central America, 10%; Philippines, 9%; New Zealand and Tonga, 7%; Alaska and west coasts of Canada and the United States, 7% ; and Hawaii, 3% (www.prh.noaa.gov/ptwc/abouttsunamis.htm). Compared to this, there are a very few tsunami reported in the Indian Ocean during the last 200 years, viz., (i) 1883 due to Krakatau volcanic explosion, (ii) December 31, 1881 due to Mw 7.9 earthquake in the Nicobar Islands, (iii) June 26, 1941 due to Mw 8.1 earthquake in the Andaman Islands, (iv) November 27, 1945 due to M8.3 earthquake in the Makran coast, Arabian Sea and (v) December 26, 2004 due to M9.3 earthquake off the coast of Sumatra.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
36
R.K. Chadha
Figure 3.3.
3.4
Epicenters of the earthquakes of M > 7.0 for the Indian Ocean are shown with different colors of varying depths. Two sources of tsunami generation, Andaman–Sumatra in the east and Makran coast in the west are shown by ellipses. (Source: http://www.USGS.gov, adapted and modified.)
INDIAN OCEAN TSUNAMI SOURCES
Indian Ocean exhibits an environment dominated by the presence of mid-oceanic ridges where Indo-Australian and African plates are moving away from the Antarctic plate along these ridges. The earthquakes associated with these ridges are mostly occurring on strike-slip faults where the dominant movement is horizontal along transform faults and hence, not a source for tsunami generation. Due to the movement of the Indian plate in north northeast direction, the subduction zones are confined in the north along the Himalayan region due to the continent–continent collision between Indian and Eurasian plates, in the east along the Andaman–Sumatra Sunda trench where the Indian plate is subducting below the Burmese plate, and a small subduction zone in the west along Makran coast, near Karachi, Pakistan. Thus, the Andaman–Sumatra and Makran subduction zones are the two main sources in the Indian Ocean (Figure 3.3) where earthquakes of magnitudes between 7.9 and above can occur giving rise to tsunamis, which can affect the east and the west coasts of India. 3.4.1 Andaman–Sumatra Tsunami source The M9.3 earthquake off the coast of Sumatra triggered the tsunami (Figure 3.4). The earthquake occurred due to the thrusting of the Burmese plate over the Indian plate. The fracture propagated unidirectionally from Sumatra, toward north along the arcuate plate boundary paralleling the Andaman and Nicobar Islands, at a velocity of about 2.4 km/s for the first 600 km and then it slowed down to about 1.5 km/s (de Groot-Hedin, 2005). The total length of the fracture as seen from the aftershocks distribution is about 1250 km. McCloskey et al. (2005) estimated a maximum displacement of the order of about 20 m, with most of the slip being concentrated in the first 500 km from the epicenter. The 10–20 m of the
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunamigenic sources in the Indian Ocean
Figure 3.4.
37
Map showing tectonics of the Indo-Australian plate viz-a-viz Burma and Sunda plates. Yellow and Red stars show the epicenter of M9.3 earthquake on December 26, 2004 and M8.7 on March 28, 2005. Solid arrows are direction of the movement of the Indo-Australian plate. Aftershocks are shown in yellow circles. (Source: http://www.USGS.gov)
vertical displacement of the ocean floor, up to a distance of about 600 km in the first 3–4 min, caused enormous displacement of water in the Indian Ocean, creating large-scale tsunami which then traveled to several thousands of kilometers in the world oceans. As the tsunami source was large and elongated in the northwardly direction from Sumatra, the east coast of India and Sri Lanka were the worst affected in the west. On March 28, 2005, another great earthquake of M8.7 occurred 150 km further southeast of the December 26, 2004 event. This event also occurred on a thrust fault in Sunda trench at a depth of 30 km. The fracture propagation for this event was also unidirectional but in the southeast direction from the epicenter for a length of about 300 km as seen from the aftershock zone (Figure 3.4). From the modeling of earthquake ruptures of these two earthquakes (Lay et al., 2005; Ammon et al., 2005) and their directions, Singh (2005), showed that a lithosphere-scale boundary around Simeulue Island could have acted as a barrier for rupture propagation on either side of this boundary. Earlier, Ritzwoller et al. (2005) have shown that December 26, 2004 earthquake got initiated where the incoming Indian plate lithosphere is warmest and the dip of the WadatiBenioff zone is least steep along the subduction zone extending from the Andaman trench to the Java trench. Anomalously high temperatures are observed in the supra-slab mantle wedge in the
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
38
R.K. Chadha
Figure 3.5.
(a) Reference map showing the locations of the principal geological features in the Indian Ocean. The red star marks the location of the initiation of rupture of the great Sumatra– Andaman earthquake. Brown lines show active and fossil plate boundaries. Arrows show the relative plate motions. The age of the incoming oceanic plate is shown with colors in millions of years. (b) Distribution of the apparent thermal age which results from the seismic inversion using the thermal parameterization. It is defined as the lithospheric age at which a purely conductive temperature profile would most closely resemble the observed thermal structure (after Ritzwoller et al., 2005).
Andaman back-arc. The subducting slab is observed along the entire plate boundary to a depth of at least 200 km. These factors contribute to the location of the initiation of rupture, the strength of seismic coupling, the differential rupture speed between the northern and southern segments of the earthquake, and the cause of convergence in the Andaman segment (Figure 3.5(a) & (b)). The subduction zone boundary from Andaman and Nicobar Islands to Sunda trench in the Indian Ocean has experienced several great earthquakes in past which either generated or had the potential to generate tsunami (Ortiz and Bilham, 2003). Figure 3.6 shows rupture areas of four great earthquakes since 1800. From the rupture areas and the damages, these earthquakes could be of M ≥ 8.0. It is seen from the figure that the December 26, 2004 event occurred in the gap between 1861 and 1881 earthquakes in the Sumatra–Nicobar region and ruptured a fault length of about 1250 km from Sumatra to Andaman including areas ruptured in earlier earthquakes during 1881 and 1941. The March 28, 2005 episode, however, ruptured a 300 km fault segment in southeast direction agreeing with typical fault lengths associated with earthquakes with M8.0 or greater. It is now observed that earthquakes with M ≥ 9.0 will have greater fault lengths like, 1960 Chile earthquake of M9.5 with 1000 km, 1964 Alaska earthquake of M9.1 with 700 km and now 2004 Sumatra earthquake of M9.3 with about 1300 km. 3.4.2
Makran Tsunami source
The convergence of the Indian plate with the Arabian and Iranian microplates of the Eurasian tectonic blocks has created an active east–west subduction zone along the Makran coast in southern
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunamigenic sources in the Indian Ocean
Figure 3.6.
39
Map showing rupture areas of four great earthquakes in the subduction zone from Andaman and Nicobar Islands to Sunda trench. Star shows epicenter of December 26, 2004 earthquake of M9.3 (figure from http://www.drgeorgepc.com).
Pakistan. Although large earthquakes along the Makran subduction zone are infrequent, the November 28, 1945 earthquake generated a tsunami which claimed more than 4000 human lives in southern Pakistan, and affected western coast of India, Iran, Oman and possibly elsewhere in other Indian Ocean islands (Figure 3.7). 3.5
IMPACT ON THE INDIAN LANDMASS
The impact of the tsunami was immediate and highest in the Andaman and Nicobar Islands as they were lying on the tsunami source. Small islands were totally inundated and in one case an island was cut up into two parts due to tsunami water, which overran the entire island from one end to the other. A maximum tsunami run-up height of 7.0 m was observed at Malacca in Car Nicobar Islands. The maximum tsunami run-up height is defined as the vertical water-surface elevation reached by the tsunami above sea level. It is measured using the standard surveying techniques and instrumentation.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
40
R.K. Chadha
Figure 3.7.
(a) Makran subduction zone west of Karachi, Pakistan. (b) Vulnerability of the Indian west coast to the tsunami generated in Makran coast (figures from http://www.drgeorgepc.com).
The most severe distant effects were observed along the western coast of India, more than 2000 km from the tsunami source. These effects were studied in detail by several workers to understand the impact along the east coast of India (Chadha et al., 2005; Yeh et al., 2005; Peterson et al., 2005). These studies describe the measurements of maximum tsunami run-up heights and inundation distances, flow patterns of tsunami runup and rundown, recording of eyewitness accounts, examination of sediment deposits, observations of structural damages, etc. Because of the distant tsunami, no observation of subsidence, uplift, and landslides was made, although geomorphological changes due to tsunami were examined. The tsunami took about 150 min to reach the east coast of India. The worst affected was the coastline along Tamil Nadu coast from Chennai in the north to Nagapattinam in the south. Relatively, the Andhra Pradesh coast suffered less. The tsunami claimed 106 human lives, with Krishna and Prakasam districts recording 27 and 35 deaths, respectively. Other affected districts were Guntur, Nellore, west Godavari, east Godavari and Vishakapatnam. The tsunami is reported to have encroached 500 m to 2 km inland at various places owing to the flatness of several beaches. Tide gauge recorder at Vishakapatnam port in Andhra Pradesh showed tsunami heights to be about 1.4 m at 09:05 h (IST). Although, eyewitnesses reported tsunami heights up to 5 m, the surveys showed the maximum tsunami run-up height to have reached about 2.5 m along Andhra Pradesh coastline with higher splashes at a few places. However, there was a general agreement amongst
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunamigenic sources in the Indian Ocean
41
Figure 3.8. Tsunami run-up heights along the east coast of Tamil Nadu. Numbers in the figure are tsunami heights in meters.
the people that four waves hit the coast in which the second wave was the strongest claiming most of the lives. Several people reported receding of the sea up to 500 m prior to the arrival of tsunami waves in the region. The 350-km-long stretch of the shoreline of Tamil Nadu state from Pulicat (13.3◦ latitude) in the north to Vedaranniyam (10.3◦ latitude) in the south was affected to varying degrees. The tsunami run-up heights varied between 2.5 and 5.2 m in these regions after applying tidal corrections from tide tables published by Survey of India. Maximum surge elevations were also measured and found to vary between 3.8 and 6.0 m (Mean tidal level). Shore-normal inundation profiles studied at 11 locations to estimate the run-up heights of the tsunami along this section are shown in Figure 3.8. Survey localities were selected on the basis of the reports of maximum damage and loss of lives. The profiles within the survey localities were selected on the basis of representative high-water marks and line-of-sight traverses to beach swash zones. High-water marks were measured from the highest elevations of several different indicators. These indicators include: (i) mud lines on standing structures, i.e., maximum still-water elevation, (ii) physical damage to standing structures, i.e., maximum surge elevation and (iii) flotsam debris on tree branches, roofs and ground slopes, i.e., maximum splash elevations and/or maximum inundation distances. Maximum still-water elevations were preferentially taken from interior wall mud line, which should minimize effects from turbulent flow around structures. Exterior wall mud lines were used where horizontal mud lines could be correlated between buildings. The vertical distance between the highest horizontal mud line and ground elevation, i.e., tripod footings surface was measured to the nearest centimeter with a tape. Maximum surge elevations were estimated from features reflecting apparent large debris damage at elevations above the mud lines. These features included displaced roof tiles, broken masonry, fresh gouges in plaster and heavy woody debris left in broken or bent tree branches. Maximum splash elevations were established from light flotsam hanging in limbs of standing vegetation and/or draped on standing structures such as railings and roofs. Horizontal sighting distances were generally less than 100 m between level and stadia rod. Elevations were measured to the nearest centimeter. Total profile errors of ±0.1 m
42
R.K. Chadha 4
Elevation (m)
Tsunami inundation 3 2 1
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Sea level 0 50
0
50
100
150
Inshore distance from the shore (m)
Figure 3.9. A shore-normal beach profile at Devanaampatnam (11◦ 44.589 N 79◦ 47.289 E). The sea level shown in the figure is the level at 9 a.m. IST on December 26, 2004, the time of tsunami attack.
are assumed for the single sighting, single direction, leveling surveys. End points of the profiles were approximately located by 12-channel GPS using the WGS84 datum. Three levels of flow competence were established from maximum landward transport of gravel, sand and flotsam debris. The gravel-size material included large (>2-cm diameter) shells, brick and mortar fragments and road gravel. General lack of gravel source material likely precluded evidence of high-flow competence in some areas. Beach sand (0.06 to 2 mm grain diameter) was present in almost all the profiles. Maximum inundation positions were generally evident from semicontinuous lines of debris that crossed streets, vacant lots, fields and wetland surfaces. Large inundation distances in some locations were related to localized flow through dune-ridge gaps or tidal inlets. Overland flow direction was measured from several different features. These features included vegetation, flop-overs and debris shields wrapped around trunks and sand ripples and linear-scours. Figure 3.9 shows a shore-normal beach profile at Devanaampatnam. The true inundation was measured at the mud line inside the house in Devanaampatnam shown in Figure 3.10 (Yeh et al., 2005). Three zones of flow competence were established from the maximum transport distances of gravel, sand and flotsam in the 11 profiles surveyed. Gravel transport ranged from 30 to 60-m distance from the swash zone in Pattinapakam, Periakalapet, Devanaampatnam and Tarangambadi profiles. The gravel-size clasts were largely derived from tsunami damaged brick walls, foundations and roofing tiles in the region. Maximum sand transport ranged from 90 to 430-m distance from the swash zone in most of the profiles (Table 3.1). Beach width in most of the profiles varies between 30 and 80 m, except at Parangipettai and Nagapattinam where it was about 300 m. With the exception of these profiles, the average sand transport distance is about 100 m beyond the beach backshore. Tsunami sand deposits ranged from coarse upper (700–1000 µ) to very fine upper (88–125 µ) in grain size, based on comparisons with grain-size cards. Sand sheet thickness ranged from several tens of centimeters near beach backshore to 1 cm thickness at the distal end of sand transport. With increasing distance landward, the mean grain size of the sand sheets appeared to decrease. The graded sequence from coarse to fine upwards in each of 2–3 sand layers was observed at the 80 m position at Devanaampatnam. Maximum inundation distances along the profiles were established on the basis of most landward distribution of flotsam in debris lines or of anomalous articles, e.g., clothing mats, fishing floats etc. Maximum inundation ranged between 140 and 800 m from the swash zone. Based on local topography, flow direction indicators and the orientation of debris lines it was apparent that maximum landward
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunamigenic sources in the Indian Ocean
43
Figure 3.10. Water marks in house in Devanaampatnam (photograph: R.K. Chadha). Table 3.1.
Details of tsunami run-up surveys along the coast of Tamil Nadu.
Sl. No.
Location
1
Pulicat
2
Pattinapakam
3
Kovalam
4
Kalpakkam
5
Periakalapet
6
Puttupatnam
7
Devanaampatnam
8
Parangipettai
9
Tarangambadi
10
Nagapattinam
11
Vedaranniyam
Latitude ◦ N/ Longitude ◦ E
Run-up elevation (m)
Lateral inundation (m)
Maximum sand distance (m)
13◦ 23.040 / 80◦ 19.984 13◦ 01.263 / 80◦ 16.722 12◦ 47.455 / 80◦ 15.003 12◦ 30.378 / 80◦ 09.688 12◦ 01.544 / 79◦ 51.888 11◦ 51.618 / 79◦ 48.926 11◦ 44.576 / 79◦ 47.230 11◦ 30.965 / 79◦ 45.947 11◦ 01.620 / 79◦ 51.350 10◦ 45.785 / 79◦ 50.928 10◦ 23.597 / 79◦ 52.014
3.2
160
90
2.7
145
120
4.3
180
120
4.1
360
190
3.9
170
130
2.6
–
–
2.5
340
180
2.8
700
400
4.4
400
150
5.2
800
430
3.6
–
–
R.K. Chadha
inundation occurred by lateral flow at Pulicat, Devanaampatnam, Parangipettai and Tarangambadi. Lateral flows filled interdune-ridge valleys that were landward of shore-parallel dune ridges at Devanaampatnam and Parangipettai. The interdune-ridge valleys at the landward ends of these two profiles were connected to tidal inlet channels. Lateral flow also filled shallow valleys in Pulicat and Tarangambadi where breaches in shore-parallel dune-ridges allowed tsunami to inundate back-ridge areas. Flow features were recorded in most of the profiles that include vegetation flop-over, orientated beams, debris shields around tree trunks and sand ripples. The mean bearing of measured flow direction was observed to be 250◦ from true north. The data suggest an oblique angle of tsunami wave attack, particularly in profiles between 11.5◦ and 12◦ latitude where the shoreline trends north northeast. The tsunami wave attack was observed to be of the order of 30–40◦ from shore normal, in the study area. It is remarkable that the run-up heights are fairly uniform (2.5–5.2 m) along the 600-km long coast of Andhra Pradesh and Tamil Nadu states of the Indian east coast. The energetic and uniform tsunami run-up distribution along the very long coastal stretch must be attributed to its very long tsunami source coinciding with the initial rupture of the fault plane of the December 26, 2004 earthquake. Variations in the run-up heights must be caused by the bathymetry and coastal topography. According to the General Bathymetric Chart of the oceans (GEBCO), depth of the Bay of Bengal is fairly uniform with very gradual inclination toward north. Ninetyeast Ridge is the only major disturbance in the abyssal plain but it is running in the north–south direction, perpendicular to the tsunami propagation and parallel to the Indian coast; hence the wave refraction must be minimal. The topography along the coast is fairly straight without significant features of headlands, sounds and coves. Furthermore, the incident tsunami was very long – approximately 430 km – hence detailed bathymetry in small scale could not cause significant local amplification. The continental shelf along the south-east coast is narrow (ranging less than 20–50 km long), and the continental slope is steep. The exception is the south of 10◦ 30 N, near Palk Straight facing Sri Lanka, where the breadth of continental shelf becomes as wide as 100 km. Figure 3.11 shows a typical bathymetry profile taken from the Marine Chart along 13◦ N (DMA63270), south of Pulicat; the continental shelf is about 40-km wide and the continental slope is approximately 1/10. 0 500
Bathymetry profile along 13°N
1000 Depth (m)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
44
1500 2000 2500 3000 3500 4000
0
20
40
60 80 Distance offshore (km)
Figure 3.11. A Bathymetry profile along 13◦ N, south of Pulicat lake.
100
120
140
Tsunamigenic sources in the Indian Ocean 3.6
45
FACTORS OF TSUNAMI IMPACT ON INDIAN COAST
Several factors that are responsible for severe impact on Indian coast line have become obvious from the earthquakes of December 26, 2004 and March 28, 2005. The most critical factors are as follows.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
3.6.1
Strike of the fault
The strike of the earthquake fault is one of the most important factors in energy distribution in case of a tsunami. The maximum energy is focused perpendicular to the strike of the fault and decreases in intensity toward the strike direction. The M9.3 earthquake on December 26, 2004 occurred on a mega-thrust with 10–20 m of vertical slip on the fault. The rupture propagated from Sumatra in northwest direction up to Nicobar Islands and then turned toward Andaman Islands in the north. Due to this large earthquake source, the tsunami propagated with maximum energy in the perpendicular direction, toward Sri Lanka, east coast of India, Maldives and African countries that are located in this path. Although, there were few deaths reported from Bangladesh, the impact of tsunami was minimal along the northern coastal regions of India in Orissa and West Bengal where tsunami of few centimeters height were reported. 3.6.2
Energy dissipation
During the propagation of tsunami in the open ocean, the energy gets dissipated whenever there is an obstruction. This is one of the reasons that the coastal regions in the Palk Bay, between India and Sri Lanka viz., Rameswaram, Ramananthapuram and Tuticorin Port along the Tamil Nadu coast did not suffer damage because of the natural protection provided by the presence of Sri Lanka Island in front of them which took the frontal attack of the tsunami. But due to the wrap-up effect, which is caused by focusing of energy into the marginal areas of a bay, the Kanyakumari region, situated at the tip of the Indian landmass and some parts of Kerala on the western coast of India were also affected. The damage was more in some coastal regions of Kerala due to the reflection of tsunami from Maldives in the Indian Ocean. Similarly, the western coast of Sri Lanka suffered less damage compared to the east and southern coast. 3.6.3
Coastal topography
The observations made during tsunami run-up surveys clearly show the role of coastal topography on the impact of a tsunami attack. The run-up heights were found to be in the range from 2.5 to 5.2 along the east coast of India. Most of the loss of life and damage to property was in the first 100 m from the shore where several settlements were washed away. Small differences in local run-up and coastal topography resulted in large differences in tsunami inundation and associated loss of life and damage within the Tamil Nadu coastal areas. However, the combination of local high runup, low topography and dense development apparently accounted for the large loss of lives and property. The surge water elevations, together with surge water depths appear to be other important parameters during a tsunami attack. Another important coastal feature like low valleys behind shore-parallel dune ridges claimed several lives due to lateral flows from tidal inlets or from breaches in the dune ridge. 3.7
SUMMARY
The December 26, 2004 Great Indian Ocean Tsunami is undoubtedly a remarkable event because of its size and the aerial extent of damage caused in several of the Indian Ocean countries. This
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
46
R.K. Chadha
event has also provided an impetus to explore both basic and applied research in tsunami science and engineering fields in the Indian Ocean, which will lead to better preparedness for the future disaster. Although the causes of tsunami are many, more than 90% of them are generated due to large earthquakes associated with thrust faulting in subduction zones. Unlike the Pacific Ocean where the tsunami sources are several in view of the subduction zones which surround the entire Pacific plate, the Indian Ocean has mainly two such sources, viz., Andaman–Sumatra and Makran subduction zones in the east and west, respectively. Both these sources are distant sources of tsunami generation and hence, the chances of India being affected by the local tsunami are remote. The danger from local tsunami generation is mostly limited to the Andaman and Nicobar Islands. The entire east coast of India is vulnerable to varying degree of tsunami threat from large earthquakes occurring in the Andaman–Sumatra subduction zone. The three primary conditions for earthquakes to generate tsunami are: (i) the magnitude should be ≥7.9, (ii) the nature of faulting should be either thrust or normal and (iii) the earthquake should be shallow enough to cause vertical uplift in the ocean floor. The December 26, 2004 earthquake off the coast of Sumatra fulfills all these three conditions, as the magnitude was M9.3, the faulting was a thrust type, the depth was 30 km or less which deformed the ocean floor by 10–20 m. Although the earthquake of M8.7 below the Nias Island in Sumatra, on March 28, 2005 was strong enough, it did not deform the ocean floor. Further, the focus of the earthquake was below the Nias Island and any displacement of ocean water would not have been significant to create large tsunami, even if there is some deformation of the ocean floor. However, there were reports of few centimeters of tsunami recorded at Cocos Islands. Even if a tsunami was generated due to this earthquake, it must have propagated in the open ocean, as the strike of the earthquake fault is west northwest. This also brings the important factor of focusing of energy in perpendicular direction of the strike of the fault. Similarly, an earthquake of M8.1 which occurred on December 24, 2004, south of Australia, did not generate any tsunami as this earthquake occurred on a strike-slip fault associated with the spreading mid-oceanic ridge. From the Indian point of view, if tsunamis are generated due to the earthquakes occurring on the Andaman–Nicobar section of the subduction zone, which has a north northwest to north– south trend, the impact on the east coast of India will be much severe due to directivity of energy and also lesser distance to the Indian coastline. Conversely, if tsunamis are generated due to earthquakes further south along the Sumatra–Java axis along the Sunda trench, it will not have any damaging effect on the Indian coast, as the strike of the fault in this region changes to west northwest to east–west. On the western side the Indian coast is likely to be affected due to large earthquakes in the Makran subduction zone, southern Pakistan. There are reports of tsunami affecting coastal regions up to Goa on the west coast of India, during the earthquake of November 28, 1945 earthquake in Makran subduction zone. In the recent times, there have been some suggestions about the breakup of Indo-Australian plate into two along an east– west nascent boundary developing in the Indian Ocean (Orman et al., 1995). There have been few earthquakes recorded from this region. At present the tsunami potential of this region is unknown. Earthquakes generating tsunami have occurred in the past, it has occurred now and will continue to occur. To mitigate this hazard, efforts in three directions are needed. On one hand, work has to be done in terms of developing a Tsunami Early Warning System for the Indian Ocean based on online monitoring of damaging earthquakes with all its parameters and tsunami propagation modeling in the Indian Ocean. On the other hand, tsunami hazard maps have to be prepared showing the possible inundation areas in case of a tsunami attack. Lastly, educating people and disseminating information about the impending disaster to the people likely to be affected is another important aspect which should be looked into from the mitigation point of view.
Tsunamigenic sources in the Indian Ocean
47
ACKNOWLEDGEMENT I am thankful to Profs. Tad Murthy and U.Aswathanarayana for inviting me to contribute this paper to their book on tsunami and also making suggestions during the preparation of this manuscript. I am grateful to the Director, National Geophysical Research Institute, Hyderabad, India for fruitful discussions. The information provided by Dr. George Pararas-Carayannis is gratefully acknowledged.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
REFERENCES Ammon, C.J., Chen, J., Thio, H.-K., Robinson, D., Ni, S., Hjorleifsdottir, V., Kanamori, H., Lay, T., Das, S., Helmberger, D., Ichinose, G., Polet, J., and Wald, D. (2005). Rupture process of the 2004 Sumatra–Andaman earthquake. Science, 308, 1133–1139. Bryant, E. (2001). Tsunami – The Underrated Hazard, Cambridge University Press, UK, 320 pp. Chadha, R.K., Latha, G., Yeh, H., Peterson, C., and Katada, T. (2005). The tsunami of the great Sumatra earthquake of M 9.0 on 26 December 2004 – Impact on the east coast of India. Curr. Sci., 88(8), 1297–1300. de Groot-Hedin, C.D. (2005). Estimation of the rupture length and velocity of the Great Sumatra earthquake of December 26, 2004 using hydroacoustic signals. Geophys. Res. Lett., 32, L11303, doi:10.1029/2005GL022695. Lay, T., Kanamori, H., Ammon, J., Nettles, M., Ward, S.N., Aster, R.C., Beck, S.L., Bilek, S.L., Brudzinski, M.R., Butler, R., DeShon, H.R., Ekstrom, G., Satake, K., and Sipkin, S. (2005). The great Sumatra– Andaman earthquake of 26 December 2004. Science, 308, 1127–1132. McCloskey, J., Nalbant, S.S., and Steacy, S. (2005). Earthquake risk from so-seismic stress. Nature, 434, 291. Orman, J., Van Cochran, J.R., Weissel, J.K., and Jestin, F. (1995). Distribution of shortening between the Indian and Australian plates in the central Indian Ocean. Earth Planet. Sci. Lett., 133, 35–46. Ortiz, M., and Bilham, R. (2003). Source area and rupture parameters of the 31 December 1881 Mw 7.9 Car Nicobar earthquake estimated from tsunami recorded in the Bay of Bengal. J. Geophys. Res., 108(B4), 2215, doi:10:1029/2002JB001941. Peterson, C., Chadha, R.K., Cruikshank, K.M., Francis, M., Latha, G., Katada, T., Singh, J.P., and Yeh, H. (2005). Preliminary comparison of December 26, 2004 tsunami records from southeast Indian and southwest Thailand to paleotsunami records of overtopping height and inundation distance from the Central Cascadia margin, USA. Communicated to the 8th NCEE Conference, San Francisco, USA, April 2006. Ritzwoller, M.H., Shapiro, N.M., and Engdahl, E.R. (2005). Structural context of the great Sumatra– Andaman Island earthquake (personal communication). Singh, S.C. (2005). Sumatra Earthquake Research indicates why rupture propagated northward. EOS. Trans. Am. Geophys. Un., 86(48), 497, 502. Yeh, H., Peterson, C., Francis, M., Latha, G., Chadha, R.K., Katada, T., Singh, J.P., and Raghuram, G. (2005). Tsunami survey along the south-east Indian coast, SPECTRA, Earthquake Engineering Research Institute, OR, USA (in press). Yeh, H., Francis, M., Peterson, C.D., Katada, T., Latha, G., Chadha, R.K., Singh, J.P. and Raghuram, G. (2006). Effects of the 2004 Great Sumatra Tsunami: Southeast Indian coast, Jour. Am. Soc. Civil. Engg., (In Press).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 4
Paleo-Tsunami and Storm Surge Deposits
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
K. Arun Kumar, H. Achyuthan, and N. Shankar Department of Geology, Anna University, Chennai 600 025, Tamil Nadu, India
4.1
INTRODUCTION
Tsunamis are generated by seafloor displacement that occurs during earthquakes, collapse of sub-aerial landmasses into lakes or reservoirs, landslides, and volcanic activity independent of earthquakes. An earthquake recurrence interval is an important factor in seismic hazard assessment, but they are often poorly determined. In areas with low deformation rates, mainly intraplate regions, the recurrence and intervals of strong earthquakes usually exceed the time span covered by historical records, so geological records serve the purpose. Estuarine environment can preserve tsunami records that can be extended beyond the historical record in several ways. Generally, only large tsunamis (>5 m) leave visible deposits, with the identification of smaller events requiring higher-resolution studies. Such deposits are not found in all coastal sites. However they are vital for not only extending the magnitude and frequency record back in time, but also for verification and augmentation of model data and for iteration with model data to ensure realistic inundation scenarios. The records and evidences are difficult to achieve due to the anthropogenic activities in the coastal belts. Rapid industrialization coupled with the population growth makes it difficult to retrieve the evidences of historic tsunamis.
4.2
METHODS FOR IDENTIFICATION OF PALEO-TSUNAMI AND STORM SURGE DEPOSITS
Tsunami deposits are typically thin and fine landward and must contain marine and brackish fossils (Bobrowski et al., 1999). The identification and analysis of tsunamigenic deposits provide insight into their characteristics, including the recurrence interval of the catastrophic events near the shore. The low-elevation lakes less than 5 m above the mean sea level situated close to the sea shore are recent target to study paleo-tsunami deposits (Hutchinson et al., 2000). Tsunami deposits commonly exhibit evidence of rapid deposition, such as grading or massive structure. Tsunami deposits are not uniquely identifiable, and other kinds of deposits share some of their characteristics, but in general will not share all. Storm deposits most closely resemble tsunami deposits, but storm waves will not penetrate the distances of a long wave such as a tsunami. Tsunami deposits will tend to show less contemporaneous reworking than storm deposits. Moreover, in the case of Kamchatka, cyclones are weaker than in Japan, for example, where tsunami deposits have been described to the exclusion of storms at elevations of less than 3 m (Minoura et al., 1994). Compared to tsunami and storm deposits, aeolian sands are typically very well sorted, very fine sand, and form thicker, wedge-shaped layers. Silt and very fine aeolian sand are also disseminated in the peat. Flood deposits are typically browner and muddier, and fluvial sediment less mature than on the beach. Colluvium is poorly sorted, with angular grains. The sites chosen for paleo-tsunami study should not be susceptible to river flooding. 49
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
50
K. Arun Kumar et al.
A tsunami deposit is usually identified by the sedimentary context (e.g. deposited on soil associated with coseismic subsidence), larger grain size than surrounding sediments indicating higher-energy depositional conditions, spatial distribution of the deposit, and by ruling out other high-energy depositional modes (e.g. storm surges or floods). For example at Cascadia, paleotsunami deposits were identified as being anomalous sand layers in low-energy marsh or lacustrine environments (Peters et al., 2001). Additional information that indicates a seaward source of sediments, such as microfossils (Hemphil-Haley, 1995) or geochemical signature (Schlichting, 2000), are also useful for determining that a deposit was formed by a tsunami. The sedimentation is often such that the thicker deposits with larger grain sizes indicate faster flows. A deposit is formed by spatial gradients in transport (more coming into an area than leaving it), by change in storage of sediment in suspension in the water column, or by a combination of these processes. The variation (both horizontal and vertical) in grain size in the deposit may be used to constrain the relative contributions of transport gradients and sediment storage in the water column to forming the deposit. For example, when sediment settles from suspension (change in storage in the water column) the deposit will have more particles with higher settling velocities near the bottom and more particles with lower settling velocities near the top. When density of particles is similar, larger particles have higher settling velocities. The resulting deposit will have larger particles near the bottom creating a normal grading (Jaffe and Gelfenbaum, in preparation). When deposits are formed by spatial gradients in transport, the bed may or may not be normally graded, depending on sediment source, the time history of sediment transport, and the spatial gradients in transport of each particle size. Foraminiferal assemblages are the best way to study the inundation caused by the tsunamis. When the sediment is characterized as a tsunami deposit the number of broken shell fragments is much lower than the number of foraminiferal assemblages; particularly the deep-sea forams are in abundance. To differentiate a storm surge from the tsunami deposit foraminiferal analyses are the best. A storm surge consists of foraminifera, which are characteristic to the beach environment, but a tsunami deposit contains more of deep-sea foraminifers. In a lacustrine environment, plant detritus of diverse sizes and reworked submarine shelf or intertidal material can also be encountered within the sand sheet and or towards the top. Cataloging and assessing tsunami records are important for long-term tsunami prediction and for tsunamihazard mapping. Historical records of tsunamis are too short to develop a predictive chronology of events using only historical data. The way to obtain long-term data is to study paleo-tsunami, that is, to identify, map, and date prehistoric and historical tsunami deposits. These deposits provide a proxy record of large earthquakes. Paleo-tsunami sediments can be chronologically dated using radiocarbon, optically stimulated luminescence, and thermoluminescence dating method and these are being carried out in several sites along the east coast of India. This chapter deals with several occurrences of past tsunami that occurred in the Indian Ocean and elsewhere. Scientists have known that for some 50 million years, the Indian subcontinent has been pushing northward into Eurasia, forcefully raising the Tibetan Plateau and the Himalayan Mountains. The new research suggests that starting about 8 million years ago, the accumulated mass became so great that the Indo-Australian Plate buckled and broke under the stress. The world’s largest recorded earthquakes were all megathrust events and occur where one tectonic plate subducts beneath another. These include: the magnitude 9.5, 1960 Chile earthquake; the magnitude 9.2, 1964 Prince William Sound, Alaska earthquake; the magnitude 9.1, 1957 Andrean of Alaska earthquake, and the magnitude 9.0, 1952 Kamchatka earthquake. Megathrust earthquakes often generate large tsunamis that cause damage over a much wider area than is directly affected by ground shaking near the earthquake’s rupture. In the Pacific Ocean where the majority of these waves have been generated, the historical record shows tremendous destruction. There is also a history of tsunami destruction in Alaska, in the Hawaiian Islands in South America and elsewhere in the Pacific, although the historic records
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Paleo-tsunami and storm surge deposits
51
for these areas do not go back sufficiently in time. Historical records also document considerable loss of life and destruction of property on the western shores of the North and South Atlantic, the coastal regions of northwestern Europe, and in the seismically active regions around the eastern Caribbean. Fortunately tsunami in the Indian Ocean, Atlantic, and the Caribbean do not occur as frequently as in the Pacific. Destructive tsunamis have also occurred in the Indian Ocean and in the Mediterranean Sea. The most notable tsunami in the region of the Indian Ocean was that associated with the violent explosion of the volcanic island of Krakatoa in August 1883 that it forced much of the seabed below to collapse. A 30-m (100 ft) tsunami resulting from this explosion killed 36,500 people in Java and Sumatra. Japan is very vulnerable to the tsunami hazard. In Japan, which has one of the most populated coastal regions in the world and a long history of earthquake activity, tsunamis have destroyed entire coastal populations. All the major Japanese islands have been struck by devastating tsunamis. A total of 68 destructive tsunami have struck Japan between AD 684 and 1984 with thousands of lives lost and with the destruction of hundreds of villages. In this century alone, at least six major destructive tsunamis have hit Japan. On March 3, 1933 a tsunami in the Sanriku area reached a height of about 30 m and killed over 3000 people, injured hundreds more, and destroyed approximately 9000 homes and 8000 boats. Other similarly destructive tsunami occurred in 1944, 1946, 1960, and in 1983. The 1983 event, although not very destructive in terms of lives lost and property damage, occurred in the Sea of Japan in an area not known before for seismic or tsunami activity. In the Hawaiian Islands, tsunamis have struck repeatedly, causing great loss of life and immense damage to property. Most noteworthy of the recent Hawaiian tsunami is that of April 1, 1946 which inundated and destroyed the city of Hilo, killing 159 people. Other recent tsunamis that have hit Hawaii occurred in 1952, 1957, 1960, 1964, and 1975. A large earthquake in the Moro Gulf in the Philippines on August 16, 1976 generated one of the most devastating recent tsunami. The tsunami waves killed over 8000 people in Mindanao, leaving 10,000 injured and 90,000 more homeless. In August 1977 a large earthquake in the Lesser Sunda Islands, Indonesia, generated a destructive tsunami, which killed hundreds of people on Lombok and Sumbawa Islands along the eastern side of the Indian Ocean. Another devastating tsunami occurred on December 12, 1979 in the southwest corner of Colombia destroying several fishing villages, taking the lives of hundreds of people and creating economic chaos in an already economically depressed region of that country. Many more events have occurred in the last 20 years. 4.3 TSUNAMI CAUSED DUE TO FALL OF VOLCANO INTO THE SEA On the early morning of March 13, 1888, about 5 km3 of the Ritter Island volcano fell violently into the sea northeast of Papua New Guinea. This event, the largest lateral collapse of a volcanic island in historical time, flung devastating tsunamis tens of meters high onto adjacent shores (Cooke, 1981). During the recent research cruise, the debris avalanche and associated debris flows off Ritter were studied using a hull-mounted Em120 multibeam system, which gave both bathymetry and acoustic reflectivity, or “backscatter” data. The survey also deployed an mr1 towed vehicle to collect higher-quality backscatter images, an underway 3.5-kHz sonar, and an Sio EdgeTech chirp 2–6 kHz seismic system. The latter two provided sub-bottom sediment profiles to depths of tens of meters, and thus thickness estimates for large areas of the debris deposits. 4.4
OTHER VOLCANO COLLAPSES IN THE BISMARCK SEA
Collapse is known to be a major process of volcano denudation in the Hawaiian (Moore et al., 1989) and Canary Islands (Urgeles et al., 1997) and in the Lesser Antilles island arc (Deplus et al.,
52
K. Arun Kumar et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
2001). In addition to the well documented collapse of Ritter Island, 10–12 additional debris avalanche deposits related to collapse were discovered throughout the Bismarck volcanic arc. The older debris avalanches are exposed off Tolokiwa and Sakar islands near Ritter, Garove, and Unea islands in the Witu group (north of west New Britain), Dakataua volcano (the northernmost of a north–south volcanic peninsula on central New Britain), Crown, Long, and Karkar volcanoes north of Madang, and Manam and Bam volcanoes off northwestern Papua New Guinea. The Tolokiwa deposit is especially impressive, with large blocks (hundreds of meters long) individually imaged as far as 20 km from its north coast. The debris avalanche off the breached caldera Garove island, expansive enough for the Kilo Moana to sail in, extends 20 km from the breach along the south side of the caldera. A prominent debris field can be seen to a distance of 15 km north of Crown island. Several of the volcanoes show multiple collapse events. 4.5
PALEO-TSUNAMI DEPOSITS FORMED DUE TO VOLCANIC ERUPTIONS
The geological record suggests that megatsunamis are rare, but due to their size and power, can produce immensely devastating effects. However as with Lituya Bay, this is often localized. A megatsunami that is known to have a widespread impact which reshaped an entire coastline occurred approximately 4000 years ago on Reunion Island, to the east of Madagascar. Extensive geological investigations indicate that the risk of a re-occurrence is minimal. There are indications that a giant tsunami was generated by the bolide impact that created the Chesapeake Bay impact crater, a shallow-water near-shore impact off the eastern North American coastline about 35.5 million years ago, in the late Eocene Epoch. Around Kamchatka more than 24 Holocene key marker tephra layers from 11 different volcanic centers have been identified (Braitseva et al., 1997). Ages of prehistoric marker tephra have been determined by multiple radiocarbon dates of enclosing strata, calibrated to calendar ages (Braitseva et al., 1997). These dated tephra layers have provided a record of the most voluminous explosive events. Each marker tephra has been traced for tens to hundreds of kilometers away from the volcanic source and characterized by stratigraphic position, area of dispersal, radiocarbon age, typical grain-size distribution, and chemical and mineral composition. They occur as patchy deposits and do not occur over the entire inundated surface. In Cascadia, Darienzo and Peterson (1990) have dated a series of tsunami deposits and determined the recurrence interval for subduction zone earthquakes and associated tsunamis is from 200 to 600 years. Darienzo and Peterson (1990) provided evidence for paleo-tsunami deposition across a series of salt marshes along the northern Oregon coastline. They described a series of sediment sheets occasionally containing clay/silt units, marine brackish diatoms, and generally massive structure. These authors argued that the sands were transported and deposited out of turbulent suspension rather than due to small-scale currents that produce ripples and dunes on the seabed. However they also noted that the lateral extent of the sediment sheets indicated that the tsunami surges were capable of transporting fine sands over distances greater than 0.75 km despite being associated with bottom shear stresses that were insufficient to disturb or remove the stems of plants rooted in the underlying marsh surfaces. Detailed information from coastal Washington State indicated the former occurrence of a large tsunami that accompanied an episode of coseismic coastal submergence during a large earthquake took place ca. 300 years ago (Atwater and Yamaguchi, 1991). Sedimentary evidence for this tsunami is widespread throughout the Pacific west coast (Clague, 1997). Around 6100 BC, Storegga slides occurred under the water near the edge of Norway’s continental shelf. An area roughly the size of Iceland shifted causing a megatsunami in the North Atlantic Ocean. Harbitz (1992) attempted to develop a numerical model of the second Storegga submarine slide. He noted that the scale of tsunami run-up along the Scottish and Norwegian
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Paleo-tsunami and storm surge deposits
53
coastlines very much depended upon the average landslide velocity that was introduced into the model. He also noted that an average slide velocity of 20 m/s resulted in run-up values on to adjacent coastlines of between 1 and 2 m. By contrast a modeled landslide velocity of 50 m/s resulted in run-up values between 5 and 14 m, values significantly in excess of the estimates for adjacent coastlines based on geological data. Atwater and Moore (1992) described stratigraphical evidence from the Puget Sound, Washington, for a paleo-tsunami that flooded coastal areas ca. 1000 years ago. At Cultus Bay a sand sheet between 5 and 15 cm in thickness containing marine foraminifera is enclosed within peat. The deposits typically have a medium grain size of 0.1 mm and exhibit a progressive fining inland. Shi et al. (1995) demonstrated that the Flores Tsunami of December 12, 1992 was associated with the deposition of extensive sheets of sediments up to 1 m thick, and these are continued landward by discontinued sediment accumulations. The highest of these sediment accumulations always occurs below the upper limit of tsunami run-up. 4.6
EARTHQUAKE-TRIGGERED TSUNAMI
In the historical past earthquakes have triggered the occurrence of devastating tsunami. Clague and Bobrowski (1994) described the tidal marshes near Tofino and Ucleuelet Vancouver Island British Columbia salt marshes overlain by sand sheets containing marine foraminifera and vascular plant fossils, demonstrating rapid submergence prior to burial by marine sands. Similar sheets of sand attributed to the 1964 great Alaska earthquake and tsunami have been described by Clague et al. (1994) for Port Alberni, British Columbia. Recent studies of coastal sediments deposited by paleo-tsunamis have shown that tsunami sediments deposition is frequently associated with the deposition of sediment sheets that rise in altitude inland as tapering sediment wedges (Dawson, 1994). Around 1650 BC, at Santorini a Greek volcanic island eruption caused a tsunami, estimated to be between 100 and 150 m high and devastated the island of Crete 75 km away. Santorini is believed to be the cause of the Great Flood recorded in Jewish, Christian, and Islamic historical texts. The violent eruption and explosion of the volcano of Santorini, in the 15th century BC generated a tremendous tsunami, which destroyed most of the coastal Minoan settlements on the Aegean Sea islands acting as the trigger for the decline of the advanced Minoan civilization. A massive tsunami caused by an earthquake along a 1000-mile fault hit the coastal areas of northern California, Oregon, Washington, and British Columbia, on January 26, 1700. The tsunami also caused flooding and damage in Japan. Brian Atwater of the US Geological Survey made many discoveries exposing the history of the land and the coastal peoples of the Northwest. Layers of beach sand enabled him to pinpoint the exact date of the 1700 tsunami. The 1700 Cascadia tsunami can be identified with confidence from a sheet of sand that tapers landward, contains marine fossils, extends kilometers inland from the limit of sand deposition by storm surges, and coincides stratigraphically with evidence for abrupt tectonic subsidence and seismic shaking (Atwater et al., 2005). Tsunami geology, which began with surveys of the 1946 Aleutian tsunami in Hawaii (Shepard et al., 1950) and the 1960 Chile tsunami in Japan (Kitamura et al., 1961), now encompasses a broad range of stratigraphic and geomorphic evidence, and it includes several published comparisons between tsunami and storm deposits. The Sunday earthquake in 1775 at Lisbon, Portugal, that devastated Lisbon sent many people fleeing from churches to the coastlines to avoid falling debris. The tsunami that followed killed tens of thousands of people. In Algarve, Portugal, tsunami deposits were produced during the great Lisbon earthquake of November 1, 1755 AD. At Boca do Rio tsunami deposits occur as a continuous sheet of sediments inland from the coast but farther inland are replaced by
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
54
K. Arun Kumar et al.
discontinuous and eventually sporadic sediment sheets until a point is reached when there was no sedimentary trace of the tsunami despite historical observations that tsunami flooding took place to considerably higher altitudes (Hindson et al., 1996). An earthquake that measured 7.2 on the Richter scale occurred beneath the ocean on the Grand Banks underwater plateaus southeast of Newfoundland in 1929. The tsunami reached heights of over 7 m and hit the southern coast of Newfoundland. On April 1, 1946 an earthquake triggered a tsunami near the Aleutian Islands of Alaska. The magnitude of this earthquake was 7.8. The height of the tsunami is not known but in the Aleutian Islands it had killed 165 people and caused over $26 million in damage. A Pacific-wide tsunami was also created as a result of this earthquake. The tsunami traveled through the Pacific Ocean and struck Hawaii and the French Marquesas Islands. In the Marquesas Islands, the local people knew the dangers of a tsunami and some of the warning signs. Survivors of the 1946 tsunami or “taitoko” as it is called on the islands, recall being warned by their elders to flee for higher ground. Waters ran up into low-lying areas of this small group of islands during this tsunami at a depth of 20 m in the low-lying regions. The largest recorded earthquake of the 20th century occurred on May 22, 1960 off the coast of south-central Chile. It was measured at a magnitude of 9.5 and generated a Pacific-wide tsunami similar to the tsunami of 1946. The death toll in Chile was estimated at 2300 people. In Hilo, Hawaii the destructive waves took the lives of 61 people. The waves also reached Japan, damaging coastlines and the fishing industry. An earthquake that measured 9.2 generated tsunamis and struck Alaska, British Columbia, California, and Pacific Northwest towns in 1964 on Good Friday and hence also called as Good Friday Tsunami. Waves reached a height of nearly 6 m and struck as far away as Crescent City, California. A 7.9 magnitude earthquake occurred off the Pacific coast of Colombia and Ecuador on December 12, 1979. This tsunami killed an estimated 400 people and left 798 wounded. It is a common misconception that tsunamis only occur in oceanic areas. The 1999 tsunami that struck parts of western Turkey originated in the Sea of Marmara at Izmit Bay, part of the Turkish Straits. The earthquake event known as Kocaeli was located on the Northern Anatolian Fault, sending water from the sea towards Turkey. Areas incurring the largest damage were Golcuk, where water run-up reached a height of 4 m. The cities of Degirmendere and Karamursel also experienced heavy damage due to flooding.
4.7
PALEO-TSUNAMI AND STORM SURGES IN INDIAN OCEAN
Tsunamis are a common phenomenon in the Pacific and the Atlantic but they are not quite frequent in the Indian Ocean. In earliest known tsunami occurred in the Bay of Bengal in 1762, caused by an earthquake on Myanmar’s, Arakan Coast. This tsunami event was experienced in multiple regions throughout the world. Tsunami has also been triggered due to debris flow and avalanches in higher latitudinal regions. Though tsunamis cause a havoc of destruction they have also shaped the coastal geomorphology and created an esthetic landscape. REFERENCES Atwater, B.F. and Moore, A.L. (1992). A tsunami about 1000 years ago in Puget Sound, Washington. Science, 241, 1614–1617. Atwater, B.F. and Yamaguchi, D.K. (1991). Sudden, probably co-seismic submergence of Holocene trees and grass in coastal Washington State. Geology, 19, 706–709.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Paleo-tsunami and storm surge deposits
55
Atwater, B.F., Musumi-Rokkaku, S., Satake, K., Tsuji, Y., Ueda, K., and Yamaguchi, D.K. (2005). The orphan tsunami of 1700: Japanese clues to a parent earthquake in North America, U.S. Geol. Surv. Prof. Paper., 1707, 144. Bobrowski, P.T., Clauge, J., Hutchinson, J., and Lopez, G.I. (1999). Earthquake induced land subsidence and sedimentation on the west coast of Canada. In: Proceedings and Abstracts, XV International INQUA Congress, Durban, South Africa, August 3–11, 1999, pp. 25–26. Braitseva, O.A., Ponomareva, V.V., Sulerzhitsky, L.D., Melekestsev, I.V., and Bailey, J. (1997). Holocene key-marker tephra layers in Kamchatka, Russia. Quatern. Res., 47, 125–149. Clague, J.J. (1997). Evidence of large earthquakes at the Cascadia subduction zone. Rev. Geophy., 35, 439–460. Clague, J.J., Bobrowski, P.T., and Hamilton, T.S. (1994). A sand sheet deposited by the 1964 Alaska Tsunami at Port Alberni British Columbia. Estuar. Coast. Shelf Sci., 38, 413–421. Cooke, R.J.S. (1981). Eruptive history of the volcano at Ritter Island. In: R. W. Johnson, (ed.), Cooke-Ravian Volume of Volcanological Papers, Mem. 10, Geological Survey of Papua New Guinea, Port Moresby, pp. 115–123. Darienzo, M.E. and Peterson, C.D. (1990). Episode tectonic subsidence of late Holocene salt marshes, northern Oregon, central Cascadia margin. Tectonics, 9, 1–22. Dawson, A.G. (1994). A geomorphological effects of tsunami run-up and backwash. Geomorphology, 10, 83–94. Deplus, C., Le Friant, A., Boudon, G., Komorowsk, J.C., Villemant, B., Harford, C., Segoufin, J., and Cheminee J.L. (2001). Submarine evidence for large-scale debris avalanches in the Lesser Antilles Arc. Earth Planet. Sci. Lett., 192, 145–157. Harbitz, C.B. (1992). Models simulation of tsunami generated by the Storegga Slide. Mar. Geol., 105, pp. 1–21. Hemphil-Haley, E. (1995). Diatom evidence for earthquake-induced subsidence and tsunami 300 yr ago in southern coastal Washington. Geol. Soc. Am. Bull., 107, 367–378. Hindson, R., Andrade, C., and Dawson, A. (1996). Sedimentary processes associated with the tsunami generated by the 1755 Lisbon earthquake on the Algarve coast, Portugal, Phys. Chem. Earth, 21, 57–63. Hutchinson, I., Guilbault, J.P., Clauge, J.J., and Bobrowski, P.T. (2000). Tsunamis and the tectonic deformation at the northern Cascadia margin: A 3000 year record from Deserted Lake, Vancouver Island, British Columbia, Canada. Holocene, 10, 249–439. Kitamura, N., Kotaka, T., and Kataoka, J. (1961). Ofunato-Shizugawa chiku (region between Ofunato and Shizugawa), In: E. Kon’no (ed.), Geological Observations of the Sanriku Coastal Region Damaged by Tsunami due to the Chile Earthquake in 1960, Contribution Institute of Geology Paleontology, Tohoku University, 52, 28–40. Minoura, K., Nakaya, S., and Uchida, M. (1994). Tsunami deposits in a lacustrine sequence of the Sanriku coast, northeast Japan. Sedimen. Geol., 89(1/2), 25–31. Moore, J.G., Clague, D.A., Holcomb, R.T., Lipman, P.W., Normark, W.R., and Torresan, M.E. (1989). Prodigious submarine landslides on the Hawaiian Ridge. J. Geophy. Res., 94(B12), 465–484. Peters, B., Jaffe, B.E., Peterson, C., Gelfenbaum, G., and Kelsey, H. (2001). An overview of tsunami deposits along the Cascadia margin. Proceedings of the International Tsunami Symposium, pp. 479–490. Schlichting, R.B. (2000). Establishing the inundation distance and overtopping height of paleotsunami from the late-Holocene geologic record at open-coastal wetland sites, central Cascadia margin. MS Thesis, Portland State University, Portland, OR, US, pp. 166. Shepard, F.P., Macdonald, G.A., and Cox, D.C. (1950). The tsunami of April 1, 1946, Bull. Scripps Inst. Oceanogr., 5, 391–528. Shi, S., Dawson, A.G., and Smith, D.E. (1995). Coastal sedimentation associated with the December 12th 1992 Tsunami in Flores, Indonesia. In: K. Satake and K. Imamura. (eds.), Recent Tsunamis, Pure and Applied Geophysics, 144, 525–536. Urgeles, R., Canals, M., Baraza, J., Alonso, B., and Masson, D. (1997). The most recent megalandslides of the Canary Islands: El Golfo debris avalanche and Canary debris flow, west El Hierro Island. J. Geophy. Res., 102(B9), 20305–20323.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 5
Overview and Integration of Part 1
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
U. Aswathanarayana Mahadevan International Centre for Water Resources Management, Hyderabad, Andhra Pradesh, India
5.1
GEOSTRUCTURAL ENVIRONMENT OF TSUNAMI GENESIS
Raison d’ etre: Part 1 deals with the Geostructural Environment of Tsunami Genesis. It is based on the dictum, “The Past is the key to the Future”. By tracing the geostructural environments in which the tsunamis were generated in the past, and their periodicity, we identify the sites wherefrom and when they could possibly get generated in the future. The vulnerability of a given coastal site to tsunami damage is dependent on the slope and morphology of the coast. In order for an earthquake to generate tsunami, the magnitude should be ≥7.9, the nature of faulting should be either thrust or normal, and the earthquake should be shallow enough to cause vertical uplift in the ocean floor. Paleo-tsunami records are useful for extending the magnitude and frequency record of the tsunamis back in time. Rastogi (Chapter 1) compiled a catalogue of tsunamis that occurred in the Indian Ocean during the period 326 bc–2005 ad. In order to identify possible sites of future great earthquakes, the sites of large megathrust earthquakes (>Mw 7.5) are plotted on the subduction zones, along with information on the rupture zones of the great earthquakes. Thrust type earthquakes occurring along subduction zones that cause vertical movement of ocean floor tend to be tsunamigenic. As compared to average eight tsunamis per year in the Pacific, Indian Ocean has no more than one in 3 years or so. Eighty percent of the tsunamis of the Indian Ocean originate in the Sunda arc covering Java and Sumatra. The seismic gap areas along subduction zones like Andaman–Sumatra and Makran can be assessed as possible future source zones of tsunami generating earthquakes in the Indian Ocean and the repeat periods of great earthquakes can be assessed from past seismicity. Along the Andaman–Sumatra trench convergence rate is 40–50 mm/year, yielding return periods of 150–200 year for great to giant earthquakes of magnitude 8.5 or greater. Major tsunamigenic earthquakes of magnitude <8.0 have been repeated more frequently at intervals of over a few decades. Normal fault type earthquakes can also generate moderate tsunami. Strike-slip earthquakes that cause horizontal movement of ocean floor are not tsunamigenic but oblique-slip/dipslip component in them can generate weak tsunamis. Spreading zones like Carlsberg ridge, Ninety-East ridge, etc. are sites of such earthquakes. It is assessed that Sumatra and Andaman regions will probably not generate great earthquakes for a few decades due to occurrence of 2004 Mw 9.3 and 2005 Mw 8.7 earthquakes. Southern Sumatra has potential for a great earthquake. However, the effect of tsunami due to it in India and Sri Lanka may be a limited one as the path of tsunami will be oblique to the rupture zone. In the near future, earthquakes along southern Sumatra, Makran coast, Indus Delta, Kutch– Saurashtra coast, Bangladesh and southern Myanmar might cause tsunamis which can affect India. Murthy et al. (Chapter 2) made use of the bathymetry, magnetic and gravity data over the Eastern Continental Margin of India (ECMI) from Karaikal in the south to Paradip in the north, 57
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
58
U. Aswathanarayana
to account for wide variations in the damage inflicted by the tsunami at different sites along the east coast of India. The ECMI is a passive margin evolved during the process of break-up of Eastern Gondwanaland in the late Cretaceous. Most of the Indian rivers, having an eastward slope, join the Bay of Bengal, thereby resulting in a mosaic of basinal and non-basinal morphology. The shallow bays associated with the basinal areas are more affected by the crossing of cyclones and storm surges, due to the wider shelf with gentle slope. Detailed bathymetry map and sections of the Nagapattinam–Cuddalore shelf (from 10.5◦ N to about 12◦ N) indicate that one of the main reasons for the higher run-up heights and inundation in Nagapattinam–Cuddalore coast could be the concave shape of the shelf with a gentle slope, which might have accelerated the tsunami surge to flush through at a rapid force. Bathymetry sections off Pondicherry and Cuddalore indicate a gentle continental shelf and slope up to about 3000 m water depth representing the concave nature of the shelf. The sections off Vedaranyam in the south and those in the north of Pondicherry indicate a wider shelf with a steeper slope, representing the southern and northern boundaries of the concave shelf. The area within these boundaries is more affected by the tsunami surge. Earliest bathymetry observations over the offshore Cauvery basin revealed the presence of submarine canyons off Cuddalore and Pondicherry. Subsequent geophysical studies also suggested that major valleys off Pondicherry are formed due to the existence of mega lineaments. The lineament pattern played a major role in shaping the continental slope morphology, besides erosional and depositional processes. The high run-up heights and inundation in case of Nagapattinam and Cuddalore shelf are therefore the result of a combination of a structurally controlled basin with a favourable seabed morphology. It is important to note that during the months of December 2004–July 2005, the coastal areas all along the east coast of India have experienced tremors of few seconds duration both due to the main event of 26 December 2004, and by the aftershocks of 28 March 2005 (Mw 8.3) and 24 July 2005 (Mw 7.3). This is a new development and it is quite likely that the aftershocks are likely to continue in the Andaman and Nicobar region with relatively high frequency and with increased amplitude (>5.0). This implies that the coastal regions have to take note of this new seismic hazard. Hitherto the observed seismicity in the coastal regions of the Stable Continental Region (SCR) is mainly due to the reactivation of weak zones due to the stresses developed as a result of the northward movement of the Indian plate. However there is now a new possibility of reactivation of weak zones of the coastal areas, due to the aftershock, of high amplitude, occurring continuously at the eastern end, i.e., along the Andaman and Nicobar arc. Under these circumstances, it is very essential to carryout geophysical studies in the coastal region in order to identify the land–ocean tectonic lineaments and their correlation with earlier reported seismicity so that seismic zonation maps can be generated for the Coastal regions. According to Chadha et al. (Chapter 3), the entire east coast of India is vulnerable to varying degree of tsunami threat from large earthquakes occurring in the Andaman–Sumatra subduction zone. The three primary conditions for an earthquake to generate a tsunami are: (i) the magnitude should be ≥7.9, (ii) the nature of faulting should be either thrust or normal, and (iii) the earthquake should be shallow enough to cause vertical uplift in the ocean floor. The 26 December 2004 earthquake off the coast of Sumatra fulfills all these three conditions, as the magnitude was M9.3, the faulting was a thrust type, the depth was 30 km or less which deformed the ocean floor by 10–20 m. Although the 28 March 2005 earthquake of M8.7 below the Nias Island in Sumatra, was strong enough, it did not deform the ocean floor. Further, the focus of the earthquake was below the Nias Island and any displacement of ocean water would not have been significant to create large tsunami, even if there was some deformation of the ocean floor. This also brings in the important factor of focusing of energy in a direction perpendicular to the strike of the fault. Similarly, an earthquake of M8.1 which occurred on 24 December 2004, south of Australia, did
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Overview and integration of Part 1
59
not generate any tsunami as this earthquake occurred on a strike-slip fault associated with the spreading mid-oceanic ridge. From the Indian point of view, if tsunamis are generated due to the earthquakes occurring on the Andaman–Nicobar section of the subduction zone, which has a north northwest to north–south trend, the impact on the east coast of India will be much severe due to directivity of energy and also lesser distance to the Indian coastline. Conversely, if tsunamis are generated due to earthquakes further south along the Sumatra–Java axis along the Sunda trench, it will not have any damaging effect on the Indian coast, as the strike of the fault in this region changes to west northwest to east–west. On the western side the Indian coast is likely to be affected due to large earthquakes in the Makran subduction zone in southern Pakistan. There are reports of tsunami affecting coastal regions up to Goa on the west coast of India, during the earthquake of 28 November 1945 in Makran subduction zone. In the recent times, there have been some suggestions about the breakup of Indo-Australian plate into two along an east–west nascent boundary developing in the Indian Ocean, but there have been few earthquakes recorded from this region. At present the tsunami potential of this region is unknown. According to Arun Kumar et al. (Chapter 4), paleo-tsunami records are useful not only for extending the magnitude and frequency record of the tsunamis back in time, but also for verification and augmentation of model data and for iteration with model data to ensure realistic inundation scenarios. Rapid industrialization coupled with the population growth in the coastal areas some times makes it difficult to retrieve the evidences of historic tsunamis. Tsunami deposits are typically thin and fine landward and must contain marine and brackish water fossils. They commonly exhibit evidence of rapid deposition, such as grading or massive structure. Tsunami deposits are not uniquely identifiable, and other kinds of deposits share some of their characteristics, but in general will not share all. Storm deposits most closely resemble tsunami deposits, but storm waves will not penetrate the distances of a long wave such as a tsunami. Tsunami deposits will tend to show less contemporaneous reworking than storm deposits. A tsunami deposit is usually identified by the sedimentary context (e.g. deposited on soil associated with coseismic subsidence), larger grain size than surrounding sediments indicating higher-energy depositional conditions, spatial distribution of the deposit, and by ruling out other high-energy depositional modes (e.g. storm surges or floods). Thicker deposits with larger grain sizes indicate faster flows. A deposit is formed by spatial gradients in transport (more coming into an area than leaving it), by change in storage of sediment in suspension in the water column, or by a combination of these processes. The variation (both horizontal and vertical) in grain size in the deposit may be used to constrain the relative contributions of transport gradients and sediment storage in the water column to forming the deposit. Foraminiferal assemblages are the best way to study the inundation caused by the tsunamis. When the sediment is characterized as a tsunami deposit the number of broken shell fragments is much lower than the number of foraminiferal assemblages; particularly the deep-sea forams are in abundance. To differentiate a storm surge from the tsunami deposit foraminiferal analyses are most useful. A storm surge consists of foraminifera, which are characteristic to the beach environment, but a tsunami deposit contains more of deep-sea foraminifers. In a lacustrine environment, plant detritus of diverse sizes and reworked submarine shelf or intertidal material can also be encountered within the sand sheet and or towards the top. Cataloguing and assessing tsunami records are important for long-term tsunami prediction and for tsunami-hazard mapping. Historical records of tsunamis are too short to develop a predictive chronology of events using only historical data. The way to obtain long-term data is to study paleo-tsunami, i.e., to identify, map and date prehistoric and historical tsunami deposits. These deposits provide a proxy record of large earthquakes. Paleo-tsunami sediments can be chronologically dated using radiocarbon, optically stimulated luminescence and thermoluminescence dating methods and these are being carried out for several sites along the east coast of India.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Part 2
Modelling of Tsunami Generation and Propagation
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 6
A Review of Classical Concepts on Phase and Amplitude Dispersion: Application to Tsunamis
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
N. Nirupama Atkinson School of Administrative Studies, York University, Toronto, Canada T.S. Murty and I. Nistor Department of Civil Engineering, University of Ottawa, Ottawa, Canada A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology, New Delhi, India
6.1
INTRODUCTION
In the analytical and numerical modelling of tsunami generation and propagation some of the classical concepts, such as, phase (frequency) and amplitude (non-linear effects) dispersion are one of the main considerations. It is beneficial to review here some of the mathematical concepts underlying these important processes. The phase and amplitude dispersions, while both are important in tsunami modelling, play their role at different stages of the tsunami event. The phase dispersion mainly happens during the propagation of the tsunami, whereas, the amplitude dispersion is relevant in the coastal amplification of the tsunami. Even if there is only one wave, when the tsunami is generated, the ocean bathymetry and phase dispersion, among other things, contribute to the separation and spreading of the tsunami into a multi-wave event. The amplitude dispersion is due to the non-linear interaction of the tsunami as it enters shallow water and interacts with the coastal topography. It is generally known (Murty, 1977) that the non-linear effects amplifies the currents several times more than the amplification of the water level. The Indian Ocean Tsunami of 26 December 2004 exhibited both phase and amplitude dispersion effects (Kowalik, 2005a, b; Murty et al., 2005a–e; Murty et al., 2006; Nirupama et al., 2005; Nirupama et al., 2006).
6.2
DISPERSION AND THE URSELL PARAMETER
Amplitude and phase dispersion relations relevant for long gravity waves, and the theory of generation of gravity waves by deformation at the bottom (e.g. earthquake), or deformation at or near the surface (e.g. explosions), will be discussed. Under the assumption that depth is small compared to a horizontal length scale, there are three regions of approximation for the long-wave theory (Chen et al., 1975): (a) linear equations, (b) finite-amplitude equations, and (c) Boussinesq or Kortweg–de Vries (KdV) type equations. Three characteristics lengths determine which equation is most appropriate: water depth, D, wavelength, λ, and wave amplitude, η. 63
64
N. Nirupama et al.
Three non-dimensional parameters can be defined: ε≡
η ε D2 ηλ2 ; µ≡ 2; U ≡ = 3 D λ µ D
(6.1)
U is generally referred to as the “Ursell parameter” and expresses the relative significance of amplitude and phase dispersion. In the linear periodic wave theory (see Lamb, 1945) the frequency, ω, is given by:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
ω2 = gk tanh (KD)
(6.2)
where g is gravity, K is wave number, and the phase velocity, C, is given by: C=
ω K
(6.3)
For very long waves, tanh (KD) can be approximated by the leading term in its expansion. Then from equations (6.2) and (6.3): C=
2π2 D2 gD 1 − 3 λ2
(6.4)
where wave number, K, is 2π/λ. From equation (6.4) it can be seen that long waves travel with a speed mainly determined by water depth but subject to a small negative correction proportional to µ. Two wave components with a slightly different value of µ will tend to separate as they progress; then µ is a measure of “frequency dispersion.” To understand the second type of dispersion, consider the formula for the celerity of a solitary wave (see equation (6.39)): ε C∼ (6.5) = gD 1 + 2 √ The celerity is approximately gD but is subject to a small positive correction proportional to the relative amplitude. Thus, ε is a measure of the amplitude dispersion. One can distinguish among the following three regimes of U : ⎧ 1 Amplitude dispersion can be ignored. Linear long-wave theory is valid. ⎪ ⎪ ⎪ ⎪ ⎪ 0(1) Both amplitude and phase dispersions are important. The Boussinesq ⎪ ⎨ equations (to be introduced later) are appropriate. Under certain conditions U these equations reduce to the KdV equations. ⎪ ⎪ ⎪ ⎪ ⎪ 1 Amplitude dispersion dominates. Finite-amplitude, non-linear, long-wave ⎪ ⎩ theory is appropriate. In tsunami studies, both linear and non-linear long-wave equations have been utilized. However, for tsunami travel over the continental shelf, neither the linear nor non-linear cases might be relevant; indeed, one might have to use the intermediate type, e.g. Boussinesq-type equations. LeMéhauté (1969) pointed out that the Ursell parameter is not wholly satisfactory in delineating the different regimes. He agrees that when U 1, the linear small-amplitude wave theory applies.
A review of classical concepts on phase and amplitude dispersion
65
However, for very long waves in shallow water (flood waves, bore, near-shore tsunami) the value of U (supposed to be 1) depends on the interpretation given to λ. (For very long waves, the concept of wavelength loses its meaning because the wavelength of a solitary wave is ∞, but the flow curvature under the crest is that of a cnoidal wave for which a finite wavelength can be defined.) The relative amplitude, η/D, is then more relevant than U for interpreting the importance of non-linear terms.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
6.3
MATHEMATICAL DEVELOPMENT OF THE URSELL PARAMETER
Following Broer (1964), the Ursell parameter will be formally developed, and the problem of treatment of interaction of non-linearity (or amplitude dispersion) and dispersion (i.e. phase dispersion) in wave propagation will be discussed. The Boussinesq equation (6.6) approximately describes the unidirectional propagation of finite-amplitude waves on a water layer of uniform depth, when the ratio of depth of wavelength, although small, is not negligible as in the theory of tides. The classical form of this equation is: ∂2 h ∂2 h 3 ∂2 1 3 ∂4 h 2 − gD = (h − D) + g gD 4 ∂t 2 ∂x2 2 ∂x2 3 ∂x
(6.6)
where t is time and x is the horizontal direction of wave propagation. In this equation, the first term on the right is the non-linear term due to finite wave amplitude, and the second term represents the dispersion due to finite depth to wavelength ratio. Here h(x, t) is the local wave height above the horizontal bottom. Assuming irrotational flow, the velocity components, u and w, are given in terms of the velocity potential, (x, z, t), by: u=
∂ ; ∂x
w=
∂ ∂z
(6.7)
The potential, , satisfies Laplace’s equation: ∂2 ∂2 + 2 =0 ∂x2 ∂z
(6.8)
The boundary condition at the bottom is no flow normal to it, i.e. ∂ = 0 for z = 0 ∂z
(6.9)
There are two surface boundary conditions. The kinematic condition is (e.g. Lamb, 1945): ∂h ∂ ∂h ∂ + = ∂t ∂x ∂x ∂z
for z = h(x, t)
The dynamic condition is:
∂ 1 ∂ 2 ∂ 2 gh + = gD + + ∂t 2 ∂x ∂z
(6.10)
for z = h
(6.11)
66
N. Nirupama et al.
Assuming that is analytical in x and z (i.e. no singularities), equations (6.8) and (6.9) are satisfied by writing:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
(x, z, t) = φ(x, t) −
z 2 ∂2 φ z 4 ∂4 φ + − ··· 2! ∂x2 4! ∂x4
(6.12)
where φ(x, t) is the potential at the bottom. From equations (6.10) and (6.11) after using equation (6.12), two simultaneous equations for φ(x, t) and η(x, t) are obtained. These differential equations will be of infinite degree in η and infinite order in ∂/∂x. To obtain an equation similar to equation (6.6), one has to truncate equations (6.10) and (6.11). However, before this truncation, it is convenient to work with dimensionless variables. Choose dimensionless variables (denoted by prime) such that: Lt x ≡ Lx ; z = Dz ; t = √ ; φ = φ εL gD gD
(6.13)
Also write h = D(1 + εη )
(6.14)
where ε is the relative wave amplitude defined in equation (6.1) and η (x, t ) is so chosen that its maximum value is unity for some initial or boundary value. This means the dimensionless slope ∂η /∂x is of the order of unity provided L is chosen appropriately. Suppose the dominant wavelength, λ, is chosen for L, and noting the definition of µ from equation (6.1) and from equations (6.10), (6.11), and (6.12), retaining terms of order zero and one only, then: 2
∂η ∂2 φ ∂ φ ∂η ∂φ 1 ∂4 φ + 2 = −ε η 2 + + µ 4 ∂t ∂x ∂x ∂x ∂x 6 ∂x
(6.15)
and η+
2 ∂φ 1 ∂3 φ 1 ∂φ + µ 2 =− ε ∂t 2 ∂x 2 ∂x ∂t
(6.16)
From equations (6.15) and (6.16) the solutions will depend on the ratio ε/µ, the Ursell parameter defined earlier in equation (6.1).
6.4
FREQUENCY AND AMPLITUDE DISPERSION
Lighthill (1958) appears to be the first to coin the words “frequency dispersion” and “amplitude dispersion.” Other authors used the terms “dispersion” to refer to “frequency dispersion” and “non-linear effects” to refer to amplitude dispersion. Frequency dispersion means wave components of different frequencies propagate with different velocities whereas amplitude dispersion refers to the situation where greater values of surface elevation propagate with greater velocities to cause steepening of the waves. Situations when phase and amplitude dispersions tend to balance each other will be discussed.
A review of classical concepts on phase and amplitude dispersion
67
First, consider the case when the terms with ε are ignored in equations (6.15) and (6.16). The equations then become linear and on eliminating η between them, one obtains: ∂2 φ ∂2 φ 1 ∂4 φ 1 ∂3 φ − = − µ 2 µ ∂x2 ∂t 2 6 ∂x4 2 ∂x ∂t
(6.17)
express φ = ei(Kx−ωt)
(6.18)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
√ where i = −1, the following dispersion equation is obtained: 1 + 16 µK 2 ω2 1 ∼ = = 1 − µK 2 K2 3 1 + 12 µK 2
(6.19)
To this order, the exact dispersion relation is in the units used so far. (gravity is contained in the non-dimensionalization of t): K √ ω2 = √ tanh (K µ) µ
(6.20)
Next, in equations (6.15) and (6.16) ignore the terms with µ and define: u=
∂φ ∂x
(6.21)
the velocity at the bottom in the x direction. Then equations (6.15) and (6.16) become: ∂η ∂ [(1 + εη)u] = 0 + ∂t ∂x ∂η ∂u ∂u + + εu =0 ∂t ∂t ∂x
(6.22) (6.23)
In these equations there is no restriction on ε because if the expansions leading to equations (6.15) and (6.16) are continued, the terms with ε2 and higher powers occur always with µ, and these terms drop out when terms with µ are ignored. When terms with both ε and µ are ignored in equations (6.15) and (6.16), linear equations without dispersion are obtained. For waves travelling to the right, the solutions is: η = u = η(x − t)
(6.24)
Hence: ∂η ∂η ∂u ∂u + = + =0 ∂x ∂t ∂x ∂t 6.5
(6.25)
REDUCED FORM OF THE BOUSSINESQ EQUATION
In this non-dimensional notation, the Boussinesq equation (6.6) becomes: ∂2 η ∂2 η 3 ∂2 1 ∂4 η − 2 = ε 2 (η)2 + µ 4 2 ∂t ∂x 2 ∂x 3 ∂x
(6.26)
68
N. Nirupama et al.
However, the left side of equation (6.26) can be deduced from equations (6.15) and (6.16) and this gives: 2 2
∂2 η ∂2 η u ∂ ∂2 1 ∂4 η − = ε − (ηµ) − µ ∂t 2 ∂x2 ∂x2 2 ∂x · ∂t 3 ∂x3 ∂t
(6.27)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
It is clear that equation (6.27) is not the same as equation (6.26) but will reduce to equation (6.26) provided equations (6.24) and (6.25) are valid. Actually, due to the approximate nature of equations (6.15) and (6.16), all one has to assume instead of equations (6.24) and (6.25) are η − O(ε, µ) and ∂ ∂ + = O(ε, µ) ∂x ∂t
(6.28)
The significance is that solutions of equations (6.15) and (6.16) (where the main part is waves travelling to the right, assuming these exist at least during some time interval) could be found from the simpler equation (6.26) under the present approximation. For waves propagating to the left, the signs in equation (6.28) can be changed; then also equation (6.26) can be obtained from equation (6.27). Thus, provided equation (6.28) holds and assuming equation (6.21) throughout, equations (6.15) and (6.16) become: ∂η ∂u ∂ 1 ∂3 η + = −ε (ηu) + µ 3 ∂t ∂x ∂x 6 ∂x 2 ∂η ∂u ∂ u 1 ∂3 u + = −ε + µ 2 ∂x ∂t ∂x 2 2 ∂x ∂t
(6.29) (6.30)
From equation (6.28) one can write: ∂η ∂η ∂u ∂u + − − =0 ∂t ∂x ∂t ∂x
(6.31)
Take equation (6.29) + equation (6.30) − equation (6.31) and use equation (6.28) for the right side to give: ∂η ∂η 3 ∂ 1 ∂3 η + + ε (η)2 + µ 3 = 0 ∂t ∂x 4 ∂x 6 ∂x
(6.32)
This is the reduced equation (from Boussinesq’s equation) that can be used to understand the interaction between amplitude and phase dispersions.
6.6
CNOIDAL WAVES
Solitary Wave: To examine the properties of the reduced equation following Broer (1964), drop the numerical factors in equation (6.32) and write: ∂η ∂η ∂η ∂3 η + + εη + µ 3 = 0 ∂t ∂x ∂x ∂x
(6.33)
A review of classical concepts on phase and amplitude dispersion
69
This equation is suitable for initial value problems in which: η(x, 0) = F(x)
(6.34)
is given. Both amplitude and phase dispersions will tend to distort the wave forms; however, there might be situations when both effects cancel each other for special wave forms. In this case, the solution is simplified:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
η = F(x − at)
(6.35)
where a is the speed of propagation of the waves. From equations (6.33) and (6.35) after introducing t = 0 and integrating with respect to x one gets: 1 ∂2 F (1 − a)F + εF 2 + µ 2 = 0 2 ∂x
(6.36)
Multiply this with ∂F/∂x and integrating again leads to: 2 1 1 1 ∂F (1 − a)F 2 + εF 3 + µ =b 2 6 2 ∂x
(6.37)
where b is a constant of integration. Equation (6.37) can be solved in terms of the elliptic functions. As these are represented by “cn”, the name “cnoidal waves” was coined to refer to solutions of equation (6.37). For b = 0, the simple solution is: F=
p cosh2 (qx)
(6.38)
where 3(a − 1) p≡ ε
and
1 q≡ 2
a−1 µ
This is the solution for the so-called “solitary wave” and the Ursell parameter becomes ε/(3µ/2) and is independent of a and p. The speed of the solitary wave from equation (6.36) is: 1 µ ∂2 η a = 1 + εη + 2 η ∂x2
(6.39)
If the third term is ignored, and it is taken into consideration that equation (6.39) is in nondimensional units, then equation (6.39) reduces to equation (6.5). Before going into some detail of cnoidal waves, solitary waves, the Stokes finite-amplitude waves (to be introduced later), it is worthwhile to indicate the relevance to tsunamis of the various waves and approximations to the theories discussed so far. In deep water, especially in the near field of tsunami generation, the linear theory (phase dispersion alone is relevant) is probably adequate; on the continental shelf both phase and amplitude dispersions will be important, thus, cnoidal waves and solitary waves are relevant. In the very shallow coastal areas (bays, harbours, inlets) the amplitude dispersion dominates.
70
N. Nirupama et al.
Both amplitude and phase dispersions will tend to cause a gradual distortion of the waves. The nature of the distortion produced by amplitude and phase dispersions need not be the same. The reduced form of Boussinesq equation (6.33) will be used to examine this problem. Consider a frame of reference that moves with unit speed (in non-dimensional units). For this, define: S =x−t
(6.40)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Then equation (6.33) becomes: ∂η ∂η ∂3 η + εη +µ 3 =0 ∂t ∂S ∂S
(6.41)
The effects of amplitude and phase dispersions will be estimated qualitatively by integrating equation (6.41) with respect to S. However, solutions will be restricted to those where η is either periodic or of the nature of a solitary wave, so that η and its derivatives tend to zero for large |S|. In both cases, the integrated terms can be discarded after integrating by parts. Integration of equation (6.41) gives: d ηdS = 0 (6.42) dt Multiplying equation (6.41) by η, η2 , ηn gives the following relations after integration: d 1 2 η dS = 0 dt 2 d dt
1 3 η dS = −µ 3
∂η ∂S
3
dS
(6.43) (6.44)
and the general relation is: d dt
η
n+1
1 dS = − n(n + 1)(n − 1)µ 2
∂η ∂S
3
ηn−2 dS
(6.45)
Differentiate equation (6.41) with respect to S, multiply by ∂η/∂S and integrate to give: d dt
∂η ∂S
2
dS = −ε
∂η ∂S
3
dS
(6.46)
The integrals on the right sides of equations (6.44)–(6.46) can be considered as expressing the asymmetry of the waves. For waves with steep fronts the integrals will be negative. Thus, these equations show that phase dispersion will tend to heighten the crests and flatten the troughs. Equation (6.46) shows that the amplitude dispersion will increase the averaged square of the slope of the waves. 6.7
SUMMARY
There are two types of dispersion associated with tsunami waves: phase dispersion mainly happens during the propagation of the tsunami over the ocean, while amplitude dispersion occurs due to
A review of classical concepts on phase and amplitude dispersion
71
non-linear interactions in shallow water near the coastline. The Ursell parameter involving the tsunami amplitude, its wavelength, and the water depth determines which type of equations should be used in analytical–numerical modelling of tsunamis. Here the classical concepts on these two types of dispersions as well as cnoidal waves and reduced forms of the Boussinesq equation and their mathematical development are briefly reviewed.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
REFERENCES Broer, L.J.F. (1964). On the interaction of nonlinearity and dispersion in wave propagation I. Boussinesq’s equation. Appl. Sci. Res. Sect. B, 2, 273–285. Chen, M., Diroky, D., and Hwang, L.S. (1975). Nearfield Tsunami Behaviour. US National, Science Foundation. By Tetra Tech. Inc., Pasadena, CA. 49 pp. Kowalik, Z., Knight, W., Logan, T., and Whitmore, P. (2005a). Numerical modeling of the global tsunami: Indonesian tsunami of 26 December 2004. Sci. Tsunami Hazards, 23(1), 40–56. Kowalik, Z., Knight, W., Logan, T., and Whitmore, P. (2005b). The tsunami of 26 December 2004: numerical modeling and energy considerations. In: G.A. Papadopoulos and K. Satake (eds.), Proceedings of International Tsunami Symposium. Chania, Greece, 27–29 June, pp. 140–150. Lamb, H. (1945). Hydrodynamics, 6th edn. Dover Publishing, Inc., New York, NY, 738 p. LeMéhauté, B. (1969). On the nonsaturated breaker theory and the wave run-up. In: 8th Conference of Coastal Engineering, American Society Civil Engineering, Mexico City, Mexico, pp. 77–92. Lighthill, M.J. (1958). River Waves, Proceeding 1st Symposium Naval Hydrodynamics, Washington, DC, 1956, US National Research Council, pp. 17–44. Murty, T.S. (1977). Seismic seawaves-tsunamis, Bulletin 198, Fisheries Research Board of Canada, Ottawa, 337 pages (in English and Russisan). Murty, T.S., Nirupama, N., and Rao, A.D. (2005a). Why the earthquakes of 26th December 2004 and the 27th March 2005 differed so drastically in their tsunami-genic potential. Newslett. Voice Pacific, 21(2), 2–4. Murty, T.S., Rao, A.D., and Nirupama, N. (2005b). Inconsistencies in travel times and amplitudes of the 26 December 2004 Tsunami. J. Mar. Med., 7(1), 4–11. Murty, T.S., Nirupama, N., Nistor, I., and Hamdi, S. (2005c). Why the Atlantic generally cannot generate trans-oceanic tsunamis. ISET J. Earthq. Technol., 42(4), 227–236. Murty, T.S., Nirupama, N., Nistor, I., and Hamdi, S. (2005d). Far field characteristics of the Tsunami of 26 December 2004. ISET J. Earthq. Technol., 42(4), 213–217. Murty, T.S., Nirupama, N., Nistor, I. and Rao, A.D. (2005e). Conceptual differences between the Pacific, Atlantic and Arctic tsunami warning systems for Canada. Sci. Tsunami Hazards, 23(3), 39–51. Murty, T.S., Rao, A.D., Nirupama, N., and Nistor, I. (2006). Numerical modelling concepts for the tsunami warning systems. Curr. Sci., 90(8), 1073–1081. Nirupama, N., Murty, T.S., Rao, A.D., and Nistor, I. (2005). Numerical tsunami models for the Indian Ocean countries and states. Indian Ocean Survey, 2(1), 1–14. Nirupama, N., Murty, T.S., Nistor, I., and Rao, A.D. (2006). The energetics of the tsunami of 26 December 2004 in the Indian Ocean: a brief review. Mar. Geod., 29 (1), 39–48.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 7
A Partial Explanation of the Initial Withdrawal of the Ocean during a Tsunami
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
N. Nirupama Atkinson School of Administrative Studies, York University, Toronto, Canada T.S. Murty Department of Civil Engineering, University of Ottawa, Ottawa, Canada A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology, New Delhi, India I. Nistor Department of Civil Engineering, University of Ottawa, Ottawa, Canada
7.1
INTRODUCTION
One of the interesting phenomenon that occur at some locations during tsunami events is the withdrawal of the ocean, before the main tsunami waves arrive at that location, which is generally referred to as the initial withdrawal of the ocean (IWO). However, always IWO neither occurs at the same location for every tsunami nor at every location for the same tsunami. Sometimes IWO leads to tragic circumstances because curious people go to the ocean to watch the ocean bottom, which under normal circumstances is never exposed. While it is not a random occurrence, intuition tells us that the local ocean bathymetry and coastal topography, as well as the tsunami wave characteristics must have something to do with IWO. However, these factors alone may not be able to completely account for IWO. Until now there has been no completely satisfactory explanation for IWO. Here we attempt to provide at least a partial explanation. Tadepalli and Synolakis (1994) suggested that the so-called N-waves, which are waves with a leading depression can account for the IWO process. However, they say that their theory is relevant only if the epicenter of the earthquake is about 100 km from the coastline where the IWO occurs. Hence, it is difficult to envisage whether the N-wave phenomena can account for IWO on far off coast such as Tamil Nadu in India and Sri Lanka since the distances involved are more than an order of magnitude greater than 100 km.
7.2
SOLITARY WAVES AND ROLE OF VISCOSITY IN TSUNAMI PROPAGATION, FORERUNNER
It is well known that solitary waves are single crests propagating over the ocean surface. Single troughs will quickly fill up and are not stable to have equilibrium configurations or waveforms 73
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
74
N. Nirupama et al.
Figure 7.1. Tide gauge record at Hanasaki, Japan, showing a tsunami forerunner (Nakamura and Watanabe, 1961).
(Murty, 1977). Cherkesov (1966) showed that inclusion of viscosity in the calculations of the form of the free surface produced a depression wave that arrived before the leading wave of the tsunami. Although the effect of viscosity on the main tsunami wave itself is rather small (it might reduce its amplitude by 2% at most), if viscosity is ignored, then there will be no depression wave before the main wave arrives. Tsunami forerunners can occur even before the initial withdrawal (Figure 7.1) and at other times, there are no forerunners at all (Figure 7.2). Nakamura and Watanabe (1961) developed a simple theory to explain the occurrence of the forerunner and showed that, for large angles of incidence, the forerunner cannot be well developed. Also, for a forerunner to be produced, the tsunami period has to be considerably greater than the seiche period of the bay in which the tide gauge is located. Nakamura and Watanabe (1961) explained that the absence of forerunner at the Japanese coast associated with the tsunamis generated by the Kamchatka earthquake of 1957 and Aleutian earthquake of 1946 was due to oblique incidence. They gave the same reason for the absence of the forerunner from the 1960 Chilean tsunami at stations other than Japanese. Munk (1947) treated the problem of increase in the period of waves traveling over large distances and applied his general treatment to tsunamis, and seismic surface waves. Although this work may not properly belong to this section, it was included because Munk’s theory examined, as a by-product, the period of forerunners. The forerunners Munk dealt with were those observed at Pendeen, England, and Woods Hole, MA. Nevertheless, his theory is sufficiently general to be of relevance to the tsunami forerunner problem. He applied his theory to three tsunamis, the Chilean tsunami of November 10, 1922, the Kamchatka tsunami of April 13, 1923, and the Aleutian tsunami of April 1, 1946. He found good agreement between the calculated and observed increase in period of the tsunami during propagation. The results showed that the period of the tsunami increases with the increase of travel distance but decreases with time at a given station.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
A partial explanation of the initial withdrawal of the ocean during a tsunami
75
Figure 7.2. Tide gauge records showing tsunami forerunners at some locations on the Pacific coast of Canada for the 1960 Chilean and the 1964 Alaska earthquake tsunamis (Murty, 1977).
Munk’s theory showed that the period of the forerunner is proportional to I , where I is an integral defined below and inversely proportional to the square root of time 2π2 1 I= 3 D 2 dx (7.1) g2 Where g is gravity, D(x) is the depth of the ocean, and x is the direction of tsunami travel. 7.3 THE THEORY OF SPIELVOGEL Spielvogel (1976) showed theoretically that in certain situation, the run-up on a beach was caused by a leading negative wave, followed by a positive wave. Following Speilvogel, we will present his theory here.
76
N. Nirupama et al.
Consider a two-dimensional flow on a sloping beach. Let V *, η*, X *, and t* be velocity, surface elevation, horizontal coordinate, and time (* denotes dimensional quantity). Introduce non-dimensional quantities through: V ∗ ≡ V0 V
x∗ ≡ Lx
η∗ ≡ βLη
T ∗ ≡ Tt
(7.2)
Where L is a typical length, g is gravity, β is the inclination of the beach, and:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
T ≡ V0 ≡
L βg
(7.3)
g Lβ
Spielvogel transforms the problem from the x, t plane to the σ, λ plane defined by: V ≡
1 ∂ φ(σ, λ) σ ∂σ
(7.4)
x=
1 ∂ V2 σ2 φ− − 4 ∂λ 2 16
(7.5)
η=
1 ∂ V2 φ− 4 ∂λ 2
(7.6)
t=
λ −V 2
(7.7)
where σ ≥ 0. The significance of this transformation is that σ = 0 gives the instantaneous shoreline and λ = 0 gives the initial time. The momentum equation valid for shallow water in the σ, λ plane becomes:
∂2 ∂2 3 ∂ − − 2 2 ∂λ ∂σ σ ∂σ
V =0
(7.8)
or alternatively 1 ∂ ∂ ∂2 σ − 2φ = 0 σ ∂σ ∂σ ∂λ
(7.9)
The following Jacobian has to be nonzero in the proper domain for transformation back to the x, t plane: Carrier and Greenspan (1958) gave the following solution for (7.9) φ=−
0
∞
1 J0 (τσ) sin (τλ)I (τ)dτ τ
(7.10)
A partial explanation of the initial withdrawal of the ocean during a tsunami
77
with I (τ) =
∞
4σJ1 (στ)(η0 )σ dσ
0
(7.11)
where (η0 )σ = η(σ, 0)
(7.12)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
gives the initial shape. The following formulae could be derived from above: η1 (σ, λ) = −
1 4
∞
0
J0 (τσ) cos (τλ)I (τ)dτ
(7.13)
and V (σ, λ) =
0
∞
1 J1 (τσ) sin (τλ)I (τ)dτ σ
(7.14)
Spielvogel assumed run-ups of the form: η˜ (x, 0) = A e16p(x−A)
(7.15)
where A is positive and is the amplitude of run-up and is small enough to satisfy (7.9). Here p is also a positive quantity and is a measure of the width of the run-up and has to satisfy (7.9). To specify the initial condition as a function of σ, invert: σ 2 = 16[A e16p(x−A) − x]
(7.16)
For x(σ). However, instead of this, one can start with the following initial shape: 2
η0 = η(σ, 0) = A e−pσ ,
A>0
(7.17)
η σ2 1 0 = η0 + n 16 16p A
(7.18)
Thus 2
x0 = x(σ, 0) = A e−pσ − which gives: η0 = A e16p(x−η0 )
as the initial condition. The fact that pA is small implies that η0 obeys (7.15) except in a small area near σ = 0. Then: x∼−
σ2 16
η0 ∼ A e−px
(7.19) (7.20)
78
N. Nirupama et al.
and 1
0 < (η0 )x =
1+
e(pσ 2 /16pA)
<1
(7.21)
If in Equation (7.16) Ap 1, then: (η0 )x ≈ 16Ap = ηx |σ=0
(7.22)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Under this initial condition, (7.11) could be evaluated to give: I (τ) = −
2A −τ 2 /4p e p
(7.23)
Then the solutions are: 2A ∞ −τ 2 /4p V =− τe J1 (τσ) sin (τλ) dτ σp 0 A ∞ −τ 2 /4p η1 = τe J0 (τσ) cos (τλ) dτ 2p 0 η = η1 − t=
V2 2
(7.24) (7.25) (7.26)
λ −V 2
(7.27)
For numerical purposes it is convenient to write the solution in the following form: η1 = A
∞ ∞ (−4pλ2 )s (−pσ 2 )k (s + k)! 2s!k!k!
k=0 s=0
(7.28)
and V = −8pAλ
∞ ∞ (−4pλ2 )s (−pσ 2 )k (sk + 1)! (2s + 1)!(k)!(k + 1)! k=0 f =0
Both are convergent everywhere in principle. But in practice, their use is restricted because of round-off errors to the case of small pσ 2 and pλ2 . Speilvogel represented the solution in another manner after defining: √ πi erf (−iα) 2α 0 ∞ ∞ (−4α2 )n n! α2n 2 = e−α (2n + 1)! (2n + 1)!n! n=0 n=0 g(α2 ) =
1
2 (1−y 2 )
e−α
dy =
(7.29)
A partial explanation of the initial withdrawal of the ocean during a tsunami
79
and noting that: lim α2 g(α2 ) =
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
α2 →∞
1 2
The solutions can be expressed as follows: A π η1 = [1 − (α2+ )g(α2+ ) − (α2− )g(α2− )] dφ π 0 4A π V = sin φ[(α2− )g(α2− ) − (α2+ )g(α2+ )] dφ πσ 0
(7.30)
(7.31) (7.32)
Here: (α2+ ) = p(σ sin φ + λ)2
(7.33)
(α2− ) = p(σ sin φ − λ)2
(7.34)
The behavior on the shoreline could be inferred from equation (7.30). Then: η1 (λ, 0) = A[1 − 2λ2 pg(pλ2 )]
(7.35)
V (λ, 0) = −4Aλp[1 + (1 − 2λ2 p)g(pλ2 )]
(7.36)
The last two equations show that the exponential run-up at the shoreline is caused by a leading negative wave, followed by a positive wave.
7.4
SUMMARY AND CONCLUSIONS
One of the unsolved issues in tsunami research is the dynamics of tsunami forerunners in general, and IWO, in particular. Surprisingly, very little work has been done on this important topic in recent years and all the relevant references are dated a few decades back. The closest to providing any explanation for IWO is a somewhat qualitative paper by Cherkesov (1996) and a mathematical theory by Spielvogel (1976), which at least provides a partial explanation. The main difficulty with tsunami forerunners and IWO is the fact that these do not occur universally, that is, no locations show these for all tsunamis and no tsunami show these for all locations. REFERENCES Carrier, G.F. and Greenspan, H.P. (1958). Water waves of finite amplitude on a sloping Beach. J. Fluid Mech., 4, 97–109. Cherkesov, L.V. (1966). The influence of viscosity on the propagation of tsunami type waves, atmos. Ocean Physics, 2, 793–797. Munk, W.H. (1947). Increase in the period of waves traveling over large distances: with applications to tsunamis, swell, and seismic surface waves. Trans. Am. Geophys. Union, 28, 198–217. Murty, T.S. (1977). Seismic sea waves-tsunamis, Bulletin No. 198. Can. Bull. Fish. Aquat. Sci., Ottawa, Canada, 337.
80
N. Nirupama et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Nakamura, K. and Watanabe, H. (1961). Tsunami forerunner observed in case of the Chile tsunami of 1960, Report on Chileon Tsunami, Field Investigation Committee for the Chilean Tsunami, Earthquake Research Institute, Tokyo, pp. 82–99. Spielvogel, L.Q. (1976). Run-up of single waves on a sloping beach, Proceedings of IUGG Symposium Tsunamis and Tsunami Research, January 29–February 1, 1974, Wellington, N.Z.R. Soc. N.Z. Bull. 15, pp. 113–119. Tadepalli, S. and Synolakis, C.E. (1994). The runup of N-waves on sloping beaches. Proc. Roy Soc. Lond. A, 445, 99–112.
CHAPTER 8
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
The Energetics of the Tsunami of 26 December 2004 in the Indian Ocean: A Brief Review N. Nirupama Emergency Management, Atkinson School of Administrative Studies, York University, Toronto, Canada T.S. Murty & I. Nistor Department of Civil Engineering, University of Ottawa, Ottawa, Canada A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology, New Delhi, India
8.1
INTRODUCTION
The tsunami of 26 December 2004 in the Indian Ocean is by far the most destructive tsunami ever in historical time and indeed one of the worst natural disasters in human history. This one single tsunami killed more people than all the Pacific Ocean Tsunamis combined for the 19th and 20th centuries. This is extraordinary, considering the fact that tsunamis are rare in the Indian Ocean, while they occur frequently in the Pacific Ocean. Three factors have been cited as the main reasons for the great loss of life associated with this tsunami: lack of an early warning system for tsunamis in the Indian Ocean, lack of public awareness, and high population density on the coastlines of south and southeast Asia. Undoubtedly these three factors contributed significantly to the heavy loss of life. However, there is much more to this tsunami, than just socio-economic factors. There are several physical oceanographic processes that acting together made this tsunami so violent (Murty et al., 2005a–h). This is the second global tsunami in historical time, the first one being the tsunami from the eruption of the volcano Krakatoa in the Sunda Strait on 27 August 1883. This is the first global tsunami since modern seismographic, oceanographic, and communication networks are in place. For the present purpose, a global tsunami is defined as one that propagated at least into three of the four global oceans. Of the four global oceans, the Pacific, Atlantic, and Indian oceans are connected in the south, through the so-called Southern Ocean (Figure 8.1). The Pacific and Atlantic oceans are connected to the Arctic Ocean in the north. The Arctic and Indian oceans are not connected, at least, directly. The Indian Ocean Tsunami propagated into the Pacific and Atlantic oceans. Since a vast amount of data (both seismic as well as oceanographic) are available for this tsunami (and the earthquake), it is possible to estimate the energy of the earthquake and the tsunami. However, it should be noted that, the estimates provided here suffer an error of at least 10–20%. 81
82
N. Nirupama et al.
Arctic Ocean
Atlantic Ocean
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Pacific Ocean
Indian Ocean
Figure 8.1. The four global oceans.
8.2 THE RUPTURE PROCESS At 00 hours 58 min 47 s GMT on 26 December 2004 the rupture started and continued for at least 7 min (Park et al., 2005). The rupture extended toward northwest along the Sunda trench for 1200 km to the Andaman Islands. The strong directionality toward the northwest is particularly striking. The total area ruptured is estimated to be about 250,000 km2 (Wilson, 2005). According to Wilson (2005), the rupture started off the coast off Sumatra and progressed slowly for 1 min with little displacement or slip, and then grew in intensity with the slip reaching a maximum at the northern end of the Sumatra Island (Figure 8.2). Then it expanded toward northwest past the Nicobar Islands and turned clockwise toward the Andaman Islands, with a long tail (Figure 8.3). The epicentre was at 3.298◦ N latitude and 95.779◦ E longitude and was some 257 km southwest of Banda Aceh. The focal depth was 30 km. Table 8.1 lists some of the fault parameters. It has been estimated that the vertical deformation has ranged from 0 to 20 m with an average value of about 10 m. Following the earthquake, some of the smaller islands southwest of Sumatra, may have moved toward southwest by up to 20 m. The northern tip of Sumatra may also have moved toward southwest by up to 36 m. It took only 4 min for the P-wave to arrive at the south-eastern tip of Sri Lanka, 12 min to arrive in Antarctica and Europe in 21 min. The P-wave arrived at every location on the globe (US Geological Survey, web site). The ground moved up and down by 9 cm in Sri Lanka in an oscillation following the earthquake.
8.3 THE NORMAL MODES OF THE EARTH Berman (2005) mentions that certain areas where the tsunami caused great destruction did not experience large amplitude surface waves associated with the earthquake. For this reason, he refers to an alternate mechanism, put forward by Lomnitz and Nilsen-Hofseth (2005) invoking the normal modes of oscillation of the earth. The reason the tsunami was very destructive at certain
The energetics of tsunami: a brief review
Bangaladesh
Pakistan
MIDDLE EAST
India Arabian sea
Myanmar
Tamil Nadu
Lakshadweepa Islands Kerala
Bay of Bengal
Thailand
Andaman-Nicobar
Sri Lanka
Banda Aceh
Phuket Is Malaysia
Maldives Somalia AFRICA
Indonesia Seychelles
1000 mi 1000 km Australia
Madagascar AFRICA Countries most affected by the Tsunami
Figure 8.2. The area in the Indian Ocean in which the tsunami energy propagated on 26 December 2004.
INDIAN PLATE
BURMA MICROPLATE
SLOW SLIP 2/3 FAULT AREA AFT ERS KZ
COMBINED: Mw 9.3
HOC
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
83
ONE
SU
FAST SLIP 1/3 FAULT AREA Mw 9.0
400 km
Figure 8.3. The rupture process according to Stein and Okal (2005).
M
AT
RA
84
N. Nirupama et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Table 8.1.
Some of the fault parameters (from Kowalik et al., 2005).
Earthquake parameter
Southern fault segment
Northern fault segment
Strike Dip Slip Length Depth (SW corner) SW corner latitude SW corner longitude Moment Rigidity
335◦ 8◦ 110◦ 300 km 8 km 3.0◦ N 94.4◦ E 3.2 × 1029 dyn/cm 4.2 × 1011 dyn/cm2
350◦ 8◦ 90◦ 700 km 8 km 5.6◦ N 93.3◦ E 7.6 × 1029 dyn/cm 4.2 × 1011 dyn/cm2
locations can be completely and satisfactorily explained using certain physical oceanographic processes alone, and there need not be any correlation with earthquake surface waves at that location (Murty et al., 2005a–h). However, the normal mode approach needs consideration based upon its own merits. The traditional thinking about tsunami generation involves a substantial displacement of the ocean bottom in the vertical direction during the earthquake. The vertical deformation of the ocean bottom creates a huge mound of water at the ocean surface, which then spreads across the ocean as a tsunami. The dynamic process presented by Lomnitz and Nilsen-Hofseth (2005) is completely different. According to them, normal modes of the earth are either standing waves of the toroidal (Love Waves) or the spheroidal (Rayleigh Waves). The tsunami of 26 December 2004 in the Indian Ocean could have been either fully or partly generated by low order spheroidal modes of the earth, such as 0 S2 , 0 S3 , and 0 S4 that may have excited a wave guide, which is a oceanic layer that confines and guides a propagating wave. If an impulsive upheaval of the ocean floor by 10 m is assumed as in traditional thinking, the ratio R of the kinetic to the potential energy is: R=
V2 2gh
(8.1)
where V is the velocity, h is the elevation of the bulge, and g is acceleration due to gravity. Near peak ground accelerations in large subduction earthquakes is about 0.25 g. This low value means, R could be about 0.001 or even as small as 10−6 . Hence, according to Lomnitz and Nilsen-Hofseth (2005), the assumption of an impulsive source could overestimate the kinetic energy by a factor of a million. Lomnitz and Nilsen-Hofseth (2005) asked the question: could this tsunami be generated by the spheroidal normal modes of the earth. The earthquake can excite normal modes with wavelengths comparable to the circumference of the earth. Toroidal modes do not alter the shape of the earth, but spheroidal modes do. They excite oscillations of the ocean floor as well the ionosphere. From the Jason satellite, this tsunami was detected and a dominant period of 37 min was noticed. This matches the period of the spheroidal mode 0 S3 . Figure 8.4 shows schematically this spheroidal mode and Figure 8.5 shows the directions of maximum and minimum energy. The City of Galle in Sri Lanka was badly damaged by the tsunami, while Cocos Island, which is roughly at the same great circle distance (1754 km) from the epicentre, recorded only a maximum tsunami amplitude of 42 cm. This drastic difference in the tsunami behaviour can be easily understood from Figure 8.5.
The energetics of tsunami: a brief review
85
Epicenter
Spheroidal normal mode
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
0S 3
Figure 8.4.
Spheroidal normal mode 0 S3 . The amplitude of the mode is greatly exaggerated (from Lomnitz and Nilsen-Hofseth, 2005).
Figure 8.5. The directions of minimum and maximum energy (from Lomnitz and Nilsen-Hofseth, 2005).
8.4
SATELLITE DETECTION OF THE TSUNAMI
This is the first time in history, a tsunami has been detected by a satellite. Even though at present satellite coverage is hopelessly inadequate for real-time detection of tsunamis, just by
86
N. Nirupama et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
chance, this tsunami was detected by a satellite. About 2 h after the earthquake, the low-orbiting altimetry satellite Jason-I was tracking the ocean surface some 1500 km from Sri Lanka and headed northwest toward the Bay of Bengal (Wilson, 2005). During 10 min, it took the satellite to reach the coast of Myanmar, the satellite passed over the front of the tsunami wave over a swath some 5 km wide (Figure 8.6). NOAA web site provides the following information. Data was reviewed from four earth orbiting radar satellites. At 2 h after the earthquake, the tsunami amplitude was 60 cm, by 3 h and 15 min, the amplitude decreased to 40 cm, by 8 h and 50 min, the tsunami waves spread over most of the Indian Ocean and had amplitudes of only 5–10 cm, which is the limit of the satellite resolution. Even at this time, in the Bay of Bengal the tsunami had amplitude of about 25 cm. The four satellites that detected this tsunami are: (1) (2) (3) (4)
The TOPEX/POSEIDON, JASON-I, The European Space Agency’s ENVISAT, US navy’s GEOSAT follow on.
The first two are jointly operated by NASA and the French Space Agency CNES.
Figure 8.6.
Simulated snapshot of tsunami consisting of reconstructed peaks and troughs that formed in the Indian Ocean at a moment 1 h, 55 min after the earthquake struck (from Wilson (June 2005) http://www.physicstoday.org).
The energetics of tsunami: a brief review
87
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
8.5 THE ENERGY OF THE EARTHQUAKE According to the USGS, the strongest earthquake in historical time was the Chilean earthquake of 22 May 1960, which had a moment magnitude of 9.5. The second strongest is the 26 December 2004 earthquake in the Indian Ocean with a moment magnitude of 9.3, followed by the Alaska earthquake of 28 March 1964 at 9.2, the 9 March 1957 earthquake in the Andreanof Islands at 9.1, and the Kamchatka earthquake of 4 November 1952 at 9.0. According to the USGS, the energy released by the earthquake was 2.0 EJ, which according to them is equivalent to the total amount of energy needed to boil 150 l (40 US Gallons) of water for all the 6 billion human inhabitants of this planet. The strain energy released by this earthquake is equal to all the energy released by earthquakes on the globe for the period 1976–1990. Wikipedia encyclopaedia gives the energy release in the earthquake as 4.3 EJ (4.3 × 1018 J). This is equal to 100 gigatons of TNT, or the total energy used in the USA in 6 months. One can also understand the power of the earthquake in terms of the normal mode oscillations of the earth. If the earth’s surface oscillated by as much as 25–30 cm as was reported, it is equivalent to the combined tidal forces of the moon and the sun on the earth. There is even a speculation that the earth’s rotation was altered following the quake. The length of the day was shortened by 2.68 ms (or about one-billionth of the length of the day). It also caused the earth to wobble on its axis by up to 2.5 cm in the direction of 145◦ E longitude. Because of the tidal effects of the moon, any rotational changes will disappear quickly over time.
8.6
ENERGY OF THE TSUNAMI
We estimated that the total amount of ocean water that temporarily invaded land during this tsunami event was about 8000 km3 . After a few hours, almost all of this ocean water flowed back into the Indian Ocean. To put this in terms to grasp, consider a lake 10 km long, 10 km wide, and 10 m deep. This is 1 km3 , thus about 8000 such lakes of water invaded the land. This is about half of the total water content in the atmosphere at any given time. If all the atmospheric water suddenly falls to the ground all at once (actually it does not, atmospheric water precipitates on the average of every 9 days and thus 40 times in the year. Evaporation, mainly from the oceans, replenishes the atmospheric water), it will cover the entire earth to a thickness of 2.5 cm. Hence it is easy to see that the tsunami waves which were responsible for transporting an amount of 8000 km3 of water onto a coastal strip of land of width varying from a few meters to up to 3 km, could have attained amplitudes of 10 m or greater. We estimated that the total energy of this tsunami was about 4.5 megatons of TNT. For comparison purposes, the entire munitions used in the Second World War was about 2 megatons. Again to put this in easy to comprehend terms, 1 megaton of TNT will completely occupy a standard US freight train some 300 miles (480 km) in length. It may be noted that 1 megaton of TNT is equivalent to 4 PJ (petajoules). According to Kowalik et al. (2005), the total potential energy related to the bottom deformation given in Figure 8.3 which is transferred to the sea level oscillations is calculated as: Ep = 0.5
ρgζ 2 R20 δφ δλ
(8.2)
For a detailed discussion of this equation, see Kowalik et al. (2005). Calculation over the area of deformation sets the potential energy to 5.39 × 103 TJ (terrajoule).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
88
N. Nirupama et al.
8.7
SUMMARY AND CONCLUSIONS
The earthquake of 26 December 2004 in the Indian Ocean is the second strongest on record, with a moment magnitude of 9.3. The only earthquake stronger than this in historical time was the Chilean earthquake of 22 May 1960 with a moment magnitude of 9.5. The tsunami generated by this earthquake is the most destructive tsunami on record and is among the top natural disaster ever in historical time. This tsunami alone caused a loss of life that exceeded the combined cumulative loss of life from all the Pacific Ocean Tsunamis for the 19th and 20th centuries. This is astonishing, considering the fact that tsunamis are quite rare in the Indian Ocean, whereas they are frequent in the Pacific Ocean. On 27 March 2005, another earthquake occurred in the general area, but somewhat south of the epicenter of the 26 December 2004 earthquake. A small tsunami was generated with amplitudes up to 30 cm. This tsunami is truly a global tsunami in the sense that it not only propagated throughout the Indian Ocean in which it originated, but also into the Pacific and Atlantic oceans with discernable amplitudes, if not destructive. The only other global tsunami was the one on 27 August 1883 from the eruption of the volcano Krakatoa in the Sunda Strait, but this event occurred before modern seismographic and sea level monitoring instrumentation is put in place. This very unique and extensive dataset of the tsunami of 26 December 2004 in the Indian Ocean is very valuable in calibrating and fine tuning tsunami generation, propagation and coastal inundation models, which then can be adapted to the other oceans, with some modifications. It may be sometime before a comparable dataset for a global tsunami will be available.
REFERENCES Berman, A.E. (2005). Letters from Jakarta: Indian Ocean Nations Select a tsunami warning system. Houston Geol. Soc. Bull., April, 9–19 and 91. Kowalik, Z., Knight, W., Logan, T. and Whitmore, P. (2005). Numerical modeling of the global tsunami: Indonesian Tsunami of 26 December 2004. Sci. Tsunami Hazards, 23(1), 40–56. Lomnitz, C. and Nilsen-Hofseth, S. (2005). The Indian Ocean disaster: tsunami physics and early warning dilemmas. EOS, 86(7), 65–70. Murty, T.S., Nirupama, N. and Rao, A.D. (2005a). Tsunami warning system for the Indian Ocean. Proceedings of the Brain Storming Session, Department of Science and Technology, New Delhi, India, January 21–22. Murty, T.S., Nirupama, N. and Rao, A.D. (2005b). Historical tsunami data for the Indian Ocean with special reference to India. Proceedings of the Brain Storming Session, Department of Science and Technology, New Delhi, India, January 21–22. Murty, T.S., Nirupama, N. and Rao, A.D. (2005c). Why the earthquakes of 26th December 2004 and the 27th March 2005 differed so drastically in their tsunamigenic potential. Newslett. Voice Pacific, 21(2), 2–4. Murty, T.S., Nirupama, N., Nistor, I. and Hamdi, S. (2005d). Far field characteristics of the tsunami of 26 December 2004. ISET J. Earthq. Technol., 42(4), 213–217. Murty, T.S., Nirupama, N., Nistor, I. and Hamdi, S. (2005e). Why the atlantic generally cannot generate trans-oceanic tsunamis. ISET J. Earthq. Technol., 42(4), 227–236. Murty, T.S., Rao, A.D. and Nirupama, N. (2005f ). Inconsistencies in travel times and amplitudes of the 26th December 2004 tsunami. J. Mar. Med. Soc., 7(1), 7–14. Murty, T.S., Rao, A.D., Nirupama, N. and Nistor, I. (2005g). Numerical modelling concepts for the tsunami warning systems. Curr. Sci., 90(8), 1073–1081. Murty, T.S., Rao, A.D., Nirupama, N. and Nistor, I. (2005h). Tsunami warning systems for the hyperbolic (Pacific), parabolic (Atlantic) and elliptic (Indian) oceans. J. Indian Geophys. Union (submitted). Park, J., Anderson, K., Aster, R., Lay, T. and Simpson, D. (2005). Global seismographic network records the great Sumatra–Andaman earthquake. EOS, 86(6), 57–64.
The energetics of tsunami: a brief review
89
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Radhakrishna, B.P. (2005). O, horrible! most horrible! (devastating tsunami strikes coastline of India on 26 December 2004). J. Geol. Soc. India, 65, 129–134. Stein, S. and Okal, E. (2005). Ultra-Long Period Seismic Moment of the Great December 26, 2004 Sumatra Earthquake and Implications for the Slip Process, posted at http://www.earth. northwestern.edu/people/seth/research/sumatra2.html Wilson, M. (2005). Modelling the Sumatra–Andaman earthquake reveals a complex, non-uniform rupture. Phys. Today, June, 19–21.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 9
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Possible Amplification of Tsunami Through Coupling with Internal Waves N. Nirupama Emergency Management, Atkinson School of Administrative Studies, York University, Toronto, Canada T.S. Murty and I. Nistor Department of Civil Engineering, University of Ottawa, Ottawa, Canada A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology, New Delhi, India
9.1
INTRODUCTION
The tsunami of 26 December 2004 in the Indian Ocean has been analyzed and studied in detail by several authors (Kowalik et al., 2005a, b; Murty et al., 2005a–e; Murty et al., 2006; Nirupama et al., 2005a–b). About a dozen different physical oceanographic processes need to be invoked to explain the detailed physical characteristics of this tsunami. One of these processes is the possible amplification of the tsunami through coupling with internal waves. This coupling with internal waves contributed at least 1 m to the overall tsunami amplitude at some locations on the Tamil Nadu coast of India, as well as in Sri Lanka. An analytical–numerical model originally developed by Imamura and Imteaz (1995) has been adapted to check if this level of amplification is possible through internal wave coupling. Since the Bay of Bengal is subjected to strong stratification, it is quite conceivable that generation of internal waves at the density interfaces is quite routine.
9.2 THE ANALYTICAL MODEL OF IMAMURA AND IMTEAZ A mathematical model for two-layered flow in a wide channel with non-horizontal bottom was studied assuming a hydrostatic pressure distribution, negligible friction and negligible interfacial mixing which are, however, important in a underwater landslides. Also uniform density and velocity distributions in each layer were assumed. Considering the two-dimensional case, Figure 9.1, Euler equations of mass and momentum continuities are integrated in each layer, with the kinetic and dynamic conditions at the free surface and interface. The details of the derivation are described in Imteaz (1993). Governing equations for an upper layer are given by: ∂η1 − η2 ∂M1 + =0 ∂t ∂x
(9.1) 91
92
N. Nirupama et al. z Top surface
h1
0 r1
h1 −h1
h2
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
h2
Interface
−(h1 + h2)
Figure 9.1.
X
r2
Definition sketch for two-layers profile.
∂M1 + ∂t
∂
M12 D1
∂x
+ gD1
∂η1 =0 ∂x
(9.2)
and those for a lower layer are: ∂η2 ∂M2 + =0 ∂t ∂x 2 M ∂ D22 ∂M2 ∂η1 ∂h1 ∂η2 ∂η2 ∂h1 + + gD2 α + − + − =0 ∂t ∂x ∂x ∂x ∂x ∂x ∂x
(9.3)
(9.4)
where η is water surface elevation, D = h + η the total depth, h the still water depth, M the discharge, ρ the density of fluid, α = ρρ12 and subscripts 1 and 2 indicate the upper and lower layers, respectively. Generally it is difficult to handle the derived momentum equation, because some non-linear terms are included in the momentum equations. As long as the amplitude of waves are small compared with the still water depth, such non-linear terms are neglected. Then, the linear governing equations can be used in the case of small amplitude waves. Linearized equations for an upper layer are obtained as follows: ∂(η1 − η2 ) ∂M1 + =0 ∂t ∂x
(9.5)
∂M1 ∂η1 + gh1 =0 ∂t ∂x
(9.6)
and for a lower layer: ∂η2 ∂M2 + =0 ∂t ∂x
(9.7)
Possible amplification of tsunami through coupling with internal waves
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
∂M2 ∂η1 ∂h1 ∂h1 ∂h1 ∂η2 ∂η2 + gh2 α + − − + + gη2 (α − 1) =0 ∂t ∂x ∂x ∂x ∂x ∂x ∂x
93 (9.8)
The effects of a lower layer on an upper one are found in the mass conservation equation in a lower layer, and those of an upper layer on a lower one are in the momentum equation in a lower layer. The both effects of interaction between the two layers play an important role on wave propagation and the amplification of a top surface or an interface. The solution of the non-linear governing equations cannot be obtained analytically without some assumption or simplification. So linearized governing equations assuming a flat bottom have been solved analytically to test the validity of the numerical model. By simple differential operation and substitution, four linearized equations are transformed into two equations. Upper and lower layer equations are respectively obtained as: ∂ 2 η1 ∂2 η1 ∂ 2 η2 − gh (1 + αβ) − gh (1 − α) =0 1 2 ∂t 2 ∂x2 ∂x2
(9.9)
and (1 + αβ)
∂ 2 η2 ∂ 2 η2 ∂ 2 η1 − gh (1 − α) − αβ =0 2 ∂t 2 ∂x2 ∂t 2
(9.10)
where β is h2 /h1 . The last terms as an external force in the above two equations are added into the simple wave equation. Let us derive the solution using the Fourier transform. If we consider the progressive wave at the interface, η2 can be expressed by, η2 =
+∞
−∞
η˜ (k) eik(x−c2 t) dk
(9.11)
where η˜is the amplitude of the Fourier series for the initial condition, c2 = gh2 (1 − α)/(1 + αβ) and k is the wave number. Then, the water surface in an upper layer can be assumed as, +∞ η1 = a(k, t)˜η(k) eik(x−c1 t) dk (9.12) −∞
where c1 = gh1 (1 + αβ) and α(k, t) is a function of time and wave number to be solved. It is a reasonable assumption for initial boundary condition that at t = 0, η1 = 0 and ∂η1 /∂t = ∂η2 /∂t. Initial and boundary conditions are illustrated in Figure 9.2. Substituting equations (9.11) and (9.12) into equation (9.9) we get the following ordinary differential equation for a:
+∞
−∞
d2 a da 2 ik(c1 −c2 )t η˜ eik(x−c1 t) dk = 0 − 2ikc k a e + gh 1 2 dt 2 dt
(9.13)
Integration of equation (9.13) with respect to time with the initial condition that ∂η1 /∂t = ∂η2 /∂t yields, da = ikac1 − ikc2 dt
(9.14)
94
N. Nirupama et al. u/s BC
h1 = 0
d/s BC
X
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
h2 = a sin k(xc2t)
Figure 9.2.
Sketch of initial and boundary (periodic) conditions.
In addition, using the first initial condition that at t = 0, a = 0, we can obtain the solution of a:
gh2 (1 − α) eik(c1 −c2 )t c2 gh2 eik(c1 −c2 )t 1 c2 a=− + e2ikc1 t − − c1 − c 2 c1 + c 2 2c1 2c1 2c1 (c1 + c2 ) 2c1
(9.15)
Finally, substitution of equation (9.15) into equation (9.12) yields the solution for the wave profile in the upper layer as follows:
+∞
+∞ gh2 (1 − α) gh2 (1 − α) c2 ik(x−c2 t) η1 = − η˜ e dk + η˜ eik(x−c1 t) dk + 2c1 (c1 − c2 ) 2c1 −∞ c12 − c22 −∞
+∞ gh2 (1 − α) c2 + η˜ eik(x+c1 t) dk − (9.16) 2c1 (c1 + c2 ) 2c1 −∞
9.3
DISCUSSION OF RESULTS
Imamura and Imteaz (1995) defined the following two parameters: α=
density of upper layer ; density of lower layer
β=
depth of the lower layer depth of the top layer
In their numerical solutions, they gave results for a range of values for α and β and showed that the ratio η/h2 could achieve values of up to 0.04. Here η is the surface wave (tsunami) amplitude and h2 is the depth of the lower layer. Since they took h2 as 25 m, this means, η could be about 1 m. 9.4
SUMMARY
One of the physical oceanographic processes which contributes to the amplification of tsunamis is the coupling of surface waves (tsunami) to internal waves generated at density interfaces. Making use of an analytical–numerical model developed by Imamura and Imteaz (1995), it has been shown that the internal wave coupling could add at least 1 m to the overall tsunami amplitude.
Possible amplification of tsunami through coupling with internal waves
95
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
REFERENCES Kowalik, Z., Knight, W., Logan, T., and Whitmore, P. (2005a). Numerical modeling of the global tsunami: Indonesian Tsunami of 26 December 2004. Sci. Tsunami Hazards, 23(1), 40–56. Kowalik, Z., Knight, W., Logan, T., and Whitmore, P. (2005b). The tsunami of 26 December 2004: Numerical modeling and energy considerations. In: Proceedings of International Tsunami Symposium. G.A. Papadopoulos and K. Satake, (eds.), Chania, Greece, 27–29 June, pp. 140–150. Imamura, F., and Imteaz, M.M.A. (1995). Longwaves in two layers: governing equations and numerical model. Sci. Tsunami Hazards, 13(1), 3–24. Imteaz, M.M.A. (1993). Numerical model for the long waves in the two-layers. MS Thesis, Asian Institute of Technology, Bangkok, 86 pp. Murty, T.S., Rao, A.D., and Nirupama, N. (2005a). Inconsistencies in travel times and amplitudes of the 26 December 2004 tsunami. J. Marine Med., 7(1), 4–11. Murty, T.S., Nirupama, N., Nistor, I., and Rao, A.D. (2005b). Conceptual differences between the Pacific, Atlantic and Arctic tsunami warning systems for Canada. Sci. Tsunami Hazards, 23(3), 39–51. Murty T.S., Nirupama, N., Nistor, I., and Hamdi, S. (2005c). Far field characteristics of the tsunami of 26 December 2004. ISET J. Earthq. Technol., 42(4), 213–217. Murty T.S., Nirupama, N., Nistor, I., and Hamdi, S. (2005d). Why the Atlantic generally cannot generate trans-oceanic tsunamis. ISET J. Earthq. Technol., 42(4), 227–236. Murty T.S., Nirupama, N., and Rao, A.D. (2005e). Why the earthquakes of 26th December 2004 and the 27th March 2005 differed so drastically in their tsunami-genic potential, Newsletter: Voice Pacific, 21(2), 2–4. Murty T.S., Rao, A.D., Nirupama, N., and Nistor, I. (2006). Numerical modelling concepts for the tsunami warning systems. Curr. Sci., 90(8), 1073–1081. Nirupama, N., Murty, T.S., Nistor, I., and Rao, A.D. (2005a). The energetics of the tsunami of 26 December 2004 in the Indian Ocean: a brief review. Marine Geod., 29(1), 39–48. Nirupama, N., Murty, T.S., Rao, A.D., and Nistor, I. (2005b). Numerical tsunami models for the Indian Ocean countries and states. Indian Ocean Surv., 2(1), 1–14.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 10
Numerical Modeling of the Indian Ocean Tsunami
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Z. Kowalik Institute of Marine Science, University of Alaska, Fairbanks, AK, USA W. Knight NOAA/NWS/West Coast and Alaska Tsunami Warning Center, Palmer, Alaska, USA T. Logan Arctic Region Supercomputing Center, University of Alaska, Fairbanks, AK, USA P. Whitmore NOAA/NWS/West Coast and Alaska Tsunami Warning Center, Palmer, Alaska, USA
10.1
INTRODUCTION
Recently, Kowalik et al. (2005a, b) applied a global tsunami model (GTM) to the Indian, Atlantic, and Pacific oceans. The model domain covered the entire World Ocean extending from 80◦ S to 69◦ N. The model spatial resolution was 1 min and its domain included approximately 200 million grid points. In order to carry out this simulation, a parallel version of the model code was developed and run on a supercomputer. The 1 min spatial resolution (Kowalik et al., 2005a, b) produced very small numerical dispersion even when tsunamis traveled large distances. The preliminary model calibration against the December 26, 2004 Indian Ocean Tsunami (IOT) showed the following: 1 The chart of the maximum tsunami amplitude showed enhancement of the tsunami amplitude over the major oceanic ridges. Oceanic ridges acted as ducts for tsunami waves, transferring energy for many thousand kilometers without noticeable dissipation. 2 Travel time computation based on Fermat’s principle may lead to errors in the prediction of tsunami arrival time in case of tsunamis which propagate to other oceans. 3 In passing through straits between continents, a tsunami signal is reorganized from noisy into coherent wave motion. 4 The Coriolis force plays an important role in modifying the global tsunami propagation pattern. The model for the IOT showed that the tsunami encompassed the entire World Ocean. In the Indian Ocean the tsunami properties were related to the source function, that is to the magnitude and direction of the bottom displacement. In the southern, Pacific, and especially Atlantic Ocean, tsunami propagation was marked by wave energy ducting over oceanic ridges. Travel times obtained from computations as arrival of the first significant wave show a clear and consistent pattern only in the region of high amplitudes, and in the simply connected domains such as in the case of the tsunami which traveled from Indonesia, around New Zealand, and into the Pacific Ocean. The path through the deep ocean to western NorthAmerica carried miniscule energy, while the stronger signal traveled a much longer distance via South Pacific ridges. Travel time for these 97
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
98
Z. Kowalik et al.
amplified energy fluxes was much longer than the arrival of the first wave. Such wave behavior indicates that the computations of tsunami travel time based on the ray theory of wave propagation and on Fermat’s principle of least time might lead to large errors in predicting travel time for tsunamis large enough to propagate into contiguous oceans. The tsunami characterized by large energy fluxes, while traveling between Australia and Antarctic, was reorganized into coherent wave-like forms. The sources for these larger fluxes were multiple reflections from the Seychelles, Maldives and a slower, reflected signal from the Bay of Bengal. The energy flux into the Atlantic Ocean showed a unidirectional pattern since the energy was pumped into this domain through the directional properties of the source function. The energy flow into the Pacific Ocean was approximately 75% of the total flow to the Atlantic Ocean. In many locations along the Pacific and Atlantic coasts, the first arriving signal, or forerunner, had lower amplitude than the main signal, which often is much delayed. Understanding this temporal distribution of tsunami characteristics is important for the application to tsunami warning and prediction procedures and technology.
10.2
BASIC EQUATIONS AND TOOLS
To study tsunamis the equations of motion and continuity are formulated in the spherical polar coordinates. Longitude, latitude, and distance from the Earth’s center are defined as λ, φ, and R. If the origin of the system is located on the ocean surface, it is more suitable to introduce a vertical coordinate z = R − R0 . Here R0 is the radius of Earth and is equal to 6370 km. Because Earth is not exactly spherical, the equations given below will better describe the large scale motion relative to the geopotential and not to the spherical surfaces. For further discussion of this problem see Gill (1982). The vertically averaged equations of motion and continuity in the spherical system are: ∂u u ∂u v ∂u u g ∂ζ τb + + − 2 + v sin φ = − − λ ∂t R0 cos φ ∂λ R0 ∂φ R0 cos φ R0 cos φ ∂λ ρ0 D
(10.1)
τφb u ∂v v ∂v u ∂v g ∂ζ + + + 2 + − u sin φ = − ∂t R0 cos φ ∂λ R0 ∂φ R0 cos φ R0 ∂φ ρ0 D
(10.2)
∂ζ ∂η 1 ∂uD 1 ∂ − + + (Dv cos φ) = 0 ∂t ∂t R0 cos φ ∂λ R0 cos φ ∂φ
(10.3)
In the above equations, u is the velocity in the λ (E–W) direction, v denotes the velocity in the φ (N–S) direction, and ζ is the sea level, η is the bottom displacement, t is the time, g is Earth’s gravity acceleration (g = 981 cm s2 ), ρ is sea water density, H is the average (undisturbed) water depth, and D is the total depth D = H + ζ − η. The Coriolis parameter will be taken as f = 2 sin φ. It is a function of the Earth’s angular velocity = 7.29 × 10−5 s−1 and the latitude φ. The components of the bottom friction force are non-linear functions of velocity: τλb = ru (u2 + v2 ) and τφb = rv (u2 + v2 ) To simplify the bottom friction terms in equations (10.1) and (10.2) the following notation is introduced: τλb ru (u2 + v2 ) = = Rx u (10.4a) ρ0 D ρ0 D
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Numerical modeling of the Indian Ocean tsunami
Figure 10.1.
99
Spatial grid distribution in the spherical system of coordinates.
rv (u2 + v2 ) = = Ry v ρ0 D ρ0 D τφb
(10.4b)
The dimensionless bottom friction coefficient r is taken as 3.3 × 10−3 . In order to identify important steps in the construction of a global numerical code, we will discuss basic numerical formulas for the spherical coordinate system. The computation will be done in a space staggered grid (C grid) given in Figure 10.1. The u velocity grid points denoted as horizontal bars are offset from the v velocity grid points (vertical bars). Sea level grid points are denoted by crosses. The grid size (space step) along the E–W direction is hλ = R0 cos φλ. Index j = 1, … stands for the space stepping along the parallels of latitude, thus the distance along the parallels is expressed as jhλ . As the parallels of latitudes become very small circles near the poles, this geographical region must to be either excluded from consideration or introduced into the computation through a different map projection. In this study, we exclude the poles from the computational domain. The space step along the N–S direction is hφ = R0 φ. Index k stands for the space stepping along the meridians of longitude. Locations of the grid points on the sphere are given by their j and k coordinates. The u, v, and ζ points are organized into triplets as shown by the triangles in Figure 10.1. The water depths are defined at the sea level points. To resolve some terms in the equations of motion the v velocity is needed at the u locations and vice versa. For this reason the circles are introduced to explain how the averaged values are constructed. The four values of v given inside the circles, when averaged will define the averaged v velocity at the u point location. This point location is given by uj,k . The averaged v velocity at this location is vu = 0.25(vj,k−1 + vj,k + vj−1,k + vj−1,k−1 ). In a similar way the average u velocity (four values of u given inside circles) at the vj,k point is uv = 0.25(uj+1,k + uj+1,k+1 + uj,k+1 + uj,k ). The solution of equations (10.1)–(10.3) is usually advanced in time by the two-time-level numerical scheme (Kowalik and Murty, 1993a; Imamura, 1996). For the spatial derivatives the second order of approximation is constructed. ζ upm T m T tan φk u,m m gT m m m m (ζj,k − ζj−1,k ) + Tf vu,m + v uj,k − TRm (u − uj−1,k ) x,j,k uj,k − R0 hλ j,k hλ T vu,m um T m T vu,m p n m m m m m − n (uj+1,k − uj,k )− (uj,k − uj,k−1 )− (uj,k+1 − uj,k ) (10.5) hλ hφ hφ
m+1 m uj,k = uj,k −
100
Z. Kowalik et al.
m+1 m vj,k = vj,k −
−
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
−
T tan φkv v,m v,m+1 gT m m m (ζj,k+1 − ζj,k ) − Tf uv,m+1 − u u −TRm y,j,k vj,k hφ R0
T uv,m p hλ
m m (vj,k − vj−1,k )−
Tvpm m T uv,m n m m m (vj+1,k − vj,k )− (v − vj,k−1 ) hλ hφ j,k
Tvnm m m (v − vj,k ) hφ j,k+1
m+1 m ζj,k = ζj,k −
(10.6)
T T (fluxλ,j+1,k − fluxλ,j,k ) − (fluxφ,j,k − fluxφ,j,k−1 ) ζ hλ cos φk hφ
m + ηm+1 j,k − ηj,k
(10.7)
In the numerical approach we aimed to construct a higher order of approximation in space as did Lynett et al. (2002). For this purpose we modified the upwind/downwind flux code proposed by Mader (2004). For the large scale computations the upwind/downwind scheme is essential as it displays strong stability (Imamura, 1996). We have improved the original code by an additional interpolation between the grid points based on the method of characteristics and the resultant code given by equations (10.8) and (10.9) is close to the third order of approximation in space. m+1 m m+1 m fluxλ,j,k = upm+1 (ζp,λ − ηm (ζn,λ − ηj,k ) + uj,k j−1,k ) + un m ζp,λ
m ζn,λ
= 0.5 + = 0.5 +
upm+1
T hλ
unm+1
T hλ
m ζj−1,k
m ζj−1,k
m+1 m+1 upm+1 = 0.5(uj,k + |uj,k |)
m+1 T ζm + 0.5 − up hλ j,k m+1 T ζm + 0.5 − un hλ j,k
(Hj,k + Hj−1,k ) 2
m+1 m+1 and unm+1 = 0.5(uj,k − |uj,k |)
(10.8a) (10.8b) (10.8c) (10.8d)
m+1 (Hj,k + Hj,k+1 ) m m+1 m fluxφ,j,k = cos φkv vpm+1 (ζp,φ − ηm ) + v (ζ − η ) + v (10.9a) j,k+1 j,k n n,φ j,k 2 m m+1 T m m+1 T ζ + 0.5 − vp ζm ζp,φ = 0.5 + vp (10.9b) hφ j,k hφ j,k+1 T T m m ζn,φ ζj,k ζm = 0.5 + vnm+1 + 0.5 − vnm+1 hφ hφ j,k+1
(10.9c)
m+1 m+1 vpm+1 = 0.5(vj,k + |vj,k |)
(10.9d)
m+1 m+1 and vnm+1 = 0.5(vj,k − |vj,k |)
In the above code the index m stands for the time stepping and the time step is T .
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Numerical modeling of the Indian Ocean tsunami
Figure 10.2.
10.3
101
Ocean bathymetry. Computational domain extends from 80◦ S to 69◦ N.
DOMAIN, BOUNDARY CONDITIONS AND NUMERICAL GRID
The integration domain is shown in Figure 10.2. It extends from 80◦ S to 69◦ N. The boundaries include both wet and dry points. At coastal (dry points) the normal velocity is set to zero. At the wet boundary points (along 69◦ N) the radiation condition, established by Reid and Bodine (1968) is used. The entire globe is cut along 20◦ E longitude, requiring a cyclic boundary condition for sea level and the E–W velocity on this meridian. It appears at the first glance that the above boundary conditions are sufficient to derive a solution. Introductory numerical experiments show that even with the relatively large space step of 1 min, new dry and wet points may be generated due to run-up or run-down. A numerical scheme for the wetting and drying needs to be introduced. The total depth (H + ζ − η) is usually taken as the parameter to be tested for the presence of the wet or dry points (Flather and Heaps, 1975; Goto and Shuto, 1983; Kowalik and Murty, 1993b; Imamura, 1996; Titov and Synolakis, 1998). Wet and dry points are identified by setting the average (undisturbed) ocean depth as positive (wet points) and elevations (dry points) as the negative values. The total depth in the dry grid points is taken as zero D = H + ζ = 0. A simple run-up condition is used. The following steps are taken when the dry point (jwet + 1) is located to the right of the wet point jwet . IF(ζ m ( jwet ) > −H ( jwet + 1))
then ujmwet +1 = ujmwet
If wetting is possible (as stated by the above condition) the velocity from the wet point is extrapolated to the right (dry point), but sea level is calculated through the equation of continuity. The model has been calibrated and tested through a comparison with analytical solutions as well as with laboratory experiments. Initially, a series of comparisons was made against benchmark
102
Z. Kowalik et al.
problems suggested at the Third International Workshop on Long-wave Run-up Models held at Catalina Island in 2004, namely:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
1 A wave run-up on uniformly sloping beach. An analytical problem introduced by Carrier et al. (2003). 2 A run-up due to Gaussian shape landslide which initially is located at the shoreline. An analytical solution was proposed by Liu et al. (2003). 3 A maximum run-up on Okushiri Island during tsunami 1993 reproduced in the large wave flume by Matsuyama and Tanaka (2001). In order to further test how the model’s numerical approximation works, a case of a wave running up the beach without friction (which was solved analytically by Carrier and Greenspan, 1958) was also simulated. The solution derived by Thacker (1981) for a two-dimensional oscillation in a parabolic basin was useful in the model testing and in predicting the extent of inundation. A stringent condition on the run-up computations is set by a dam-break problem (Shigematsu et al., 2004), which was also used for testing and validation of the model. Comparison of the model against analytical solutions and testing cases proposed by the International Tsunami Community and against experimental data showed very good agreement. But in some cases, as one would expect a priori, the analytical and numerical solutions converge only when the numerical model uses very fine spatial and temporal stepping (resolution). This situation is described in detail by Horrillo et al. (2005) particularly for a case representing the dam-break problem formulated by Shigematsu et al. (2004). The spatial grid step of numerical computation is 1 (R0 φ = 1.852 km), and it changes with latitude as R0 φ cos φ. Numerical stability requires that this step be smaller than the distance √ T gH . The deepest point in the World Ocean (h 11000 m) is located close to 11◦ N and therefore, the time step of numerical integration is less than 7.9 s. This step was diminished to 2 s as the run-up scheme requires smaller time stepping. The total number of the grid points was close to 200 million, therefore the simple time stepping solution, even on a supercomputer may take several weeks. To reduce the computational time the entire domain was split along meridians into 60 subdomains to apply 60 processors. With this parallelization, 50 h of tsunami propagation was reproduced in 9 h on a CRAY X1. Consideration of a small spatial step is important as the shortperiod waves can be obliterated during large distances of propagation when using large spatial steps. Taking the average depth of the World Ocean as 4000 m, a wave with 10-min period has a wavelength close to 120 km. Such a wave length is discretized by the 1.852 km grid into about 64 mesh lengths. The amplitude of a sinusoidal wave propagating over 10000 km distance will diminish only about 2%, but some shorter dispersive waves will be generated as well (Kowalik, 2003). 10.4 A SIMPLE MODEL OF TSUNAMI GENERATION BY THE BOTTOM DISPLACEMENT The IOT observations indicated strong amplifications of the tsunami in the near-shore regions due to bottom shoaling. Additionally, observations described numerous reflections and long ringing of tsunami oscillations in the coastal regions (Merrifield et al., 2005; Rabinovich, 2005), suggesting either the local resonance or the local trapping of tsunami energy. To elucidate how a tsunami generated by the bottom displacement interacts with the shelf/shelf break bathymetry to generate a complex signal which travels into open ocean domain, we consider a channel of the 1000 km long and 3 km deep connected to the shelf/shelf break domain (see Figure 10.3). We start the computation by considering a tsunami generated by uniform bottom uplift at the source region located between 200 and 400 km (Figure 10.3). In Figure 10.3 a 2 m tsunami wave mirrors the uniform bottom uplift occurring at T = 40 s from the beginning of the uplift process.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Numerical modeling of the Indian Ocean tsunami
Figure 10.3.
103
History of tsunami propagation. Generated by bottom deformation at T = 40 s this tsunami experiences significant transformations and reflections. Black dashed lines denote bathymetry in meters. Numbers for the bathymetry in the figure should be multiplied by 10.
Later, this water elevation separates into two waves of 1m height each traveling in opposite directions (T = 16.7 min). The wave traveling toward the open boundary exits the domain without reflecting (radiation condition), and the wave traveling toward the shelf propagates without change of amplitude. This is because the bottom friction at 3 km depth is negligible (T = 39 min). At T = 57.6 min, the tsunami impinges on the shelf break resulting in the tsunami splitting into two waves (T = 1 h 16 min) due to partial reflection: a backward traveling wave with amplitude of ∼0.5 m, and a forward traveling wave toward the very shallow domain (T = 1 h 27 min). While the wave reflected from the shelf break travels without change of amplitude, the wave on the shelf is amplified and becomes shorter. The maximum amplitude attained is approximately 7.2 m (not shown). Figure 10.3 (T = 3 h) shows the time when the wave reflected from the shelf break left the domain and the wave over the shelf oscillates with an amplitude diminishing toward the open ocean. These trapped and partially leaky oscillations continue for many hours, slowly losing energy due to waves radiating into the open ocean and due to the frictional dissipation at the bottom. This behavior is also described in Figure 10.4, where temporal changes of the sea level are given in proximity to the open boundary. The initial box signal of about 20-min period is followed by the wave reflected from the shelf break and the semi-periodic waves radiated
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
104
Z. Kowalik et al.
Figure 10.4. Tsunami signal propagating from the generation domain into the open ocean. Initial box signal of 20-min period is followed by the signal reflected from the shelf break and signal radiated from the shelf domain.
back from the shelf/shelf break domain. The open boundary signal which is radiated into the open ocean is therefore the result of superposition of the primary wave and secondary waves. The period of the primary wave is defined by the width of the bottom deformation and the ocean depth (initial wave generated by earthquake), while the periods of the secondary waves are defined by reflection and generation of the new modes of oscillation through an interaction of the tsunami waves with the shelf/shelf break geometry. The evidence for tsunamis transformation and trapping have been presented both, theoretically (Nekrasov, 1970; Abe and Ishii, 1980) and in observations (Loomis, 1966; Yanuma and Tsuji, 1998; Mofjeld et al., 1999). 10.5
SOURCE FUNCTION
The generation mechanism to model the IOT is mainly the static sea floor uplift caused by abrupt slip at the India/Burma plate interface. Permanent, vertical sea floor displacement is computed using the static dislocation formulae from Okada (1985). Inputs to these formulae are fault plane location, depth, strike, dip, slip, length, and width as well as seismic moment and rigidity. The earthquake’s total rupture extent can be estimated by several approaches. Finite fault seismic data inversion is one method, which yielded fault lengths on the order of 350–650 km (e.g. Ji, 2004; Yagi, 2005). Another traditional method to delineate earthquake fault zones is by plotting the aftershocks which occur in the first 24 h following the main shock. The aftershocks are expected to cluster within the slip zone. This approach led to an estimate of 1200 km for the fault length (NEIC, 2004). In this study, the fault extent is constrained by observed tsunami travel times to the northwest, east, and south of the slip zone. Figure 10.5 displays the tsunami arrival time constraints on the fault zone. Tsunami arrival times at Paradip–India (SOI, 2005), Ko Tarutao–Thailand (Iwasaki, 2005), and Cocos Island (Merrifield et al., 2005) tide gauges are plotted in reverse. That is, the observed travel time contour is plotted with the tide gauge location as the origin point. This method indicated a fault zone approximately 1000 km by 200 km. The epicenter location lay on the southern end of the fault zone. To accommodate trench curvature, the fault plane is broken into two segments. Fault parameters for the two segments are listed in Table 10.1. Strike, dip, and slip are based on the definitions from Aki and Richards (1980). Strike is determined by the trench orientation. Dip is taken from the
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Numerical modeling of the Indian Ocean tsunami
Figure 10.5.
105
December 26, 2004 Sumatra earthquake uplift as constrained by tsunami travel times.
Table 10.1.
Fault parameters used to generate vertical sea floor movement.
Earthquake parameter
Southern fault segment
Northern fault segment
Strike Dip Slip Length Depth (SW corner) SW corner latitude SW corner longitude Moment Rigidity
335◦ 8◦ 110◦ 300 km 8 km 3.0N 94.4E 3.2 × 1029 dyn cm 4.2 × 1011 dyn cm−2
350◦ 8◦ 90◦ 700 km 8 km 5.6N 93.3E 7.6 × 1029 dyn cm 4.2 × 1011 dyn cm2
Harvard CMT solution (HRV, 2005). The slip for the southern segment is based on the Harvard CMT solution while the slip for the northern segment is set at 90◦ based on observed tsunami first motions on Indian tide gauges (NIO, 2005). Depth is based on the finite fault inversion of Ji (2004). The total moment released (derived by assuming an average slip of 13 m and rigidity of 4.2 × 1011 dyn cm−2 ) in the two segments equals 1.08 × 1030 dyn cm (Mw = 9.3), which is in good agreement with 1.3 × 1030 dyn cm proposed by Stein and Okal (2005) based on normal mode analysis. The 3-D rendering of the source function is shown in Figure 10.6.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
106
Z. Kowalik et al.
Figure 10.6. The source function. Maximum uplift is 507 cm and maximum subsidence approximately 474 cm. Coordinates are given in geographical degrees. Point (0,0) is located at 89◦ E and 1◦ N.
The total potential energy related to the bottom deformation given in Figure10.6 which is transferred to the sea level oscillations is calculated as: Ep = 0.5 ∫ ∫ ρgζ 2 R20 cos(φ)δφδλ
(10.10)
Calculation over the area of deformation sets the potential energy to 5.39 × 103 TJ (terra joule). 10.6
IMPORTANT SNAPSHOTS OF IOT DEVELOPMENT
The IOT of December 26, 2004 was quite different from the large scale tsunamis of the Pacific and Atlantic oceans. Post tsunami analysis suggests that boundary reflections significantly influenced tsunami levels in the Indian Ocean (Murty et al., 2006). Reflections from land, focusing and trapping of tsunami energy by the long island chains, and amplification and reorganization of the tsunami signal in the straits between continents all played an essential role in the IOT propagation. Figure 10.7 depicts multiple reflections from the Indian Peninsula and Sri Lanka as a reverse wave traveled back toward Indonesia. The initial reflection occurred about 2 h after the earthquake onset. We will demonstrate later that this reflection sends more energy southward than the initial wave generated by the earthquake. The tsunami interaction with the semi-transparent Maldives islands was different from that with the India and Sri Lanka coasts (Figure 10.8). While propagating toward Africa, the tsunami impinged on the Maldives Island chain. Only a portion of the incoming tsunami energy crossed this chain. Some energy was trapped around the islands and the rest was reflected backwards. As Figure 10.8 depicts, the forward signal is changed through refocusing. An island (or a few islands) splits the tsunami into two parts. These two parts coalesce behind the island, often generating local amplification of the tsunami (see wave front south-west of the Maldives). The trapped signal around the islands tends to interact with the local bathymetry generating quite large amplitudes which are slowly dissipated through bottom friction and radiation into the open ocean. This behavior is similar to the process discussed earlier and demonstrated in Figure 10.4. The role of multiple reflections in the sea level variations during IOT is well depicted by the sea level recorded on the Cocos Island (Figure 10.9). Even though Cocos Island (12◦ 7 S 96◦
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Numerical modeling of the Indian Ocean tsunami
107
Figure 10.7.
Distribution of the tsunami amplitude in the Indian Ocean at 2 h 50 min from the tsunami onset. The wave reflected from India and Sri Lanka propagates back to the source region.
Figure 10.8.
Distribution of the tsunami amplitude in the Indian Ocean at hour 4 from the tsunami onset. Along with the reflection shown in Figure 10.7, the reflection from the Maldives also sends energy eastward.
53 E) is not within the field of view of the figures (southern most latitude in Figure 10.8 is 10◦ S), it is close enough to draw valid conclusions on the arrival time of reflected waves. There is very good correspondence between predicted reflections from Sri Lanka and the Maldives and features recorded in the mareogram. The reflected wave from Sri Lanka arrived about 3 h after direct wave and the Maldives reflected wave arrived close to 5 h after the direct wave. Figure 10.10 gives a different perspective on the tsunami. Here, one is looking at the IOT from the African continent toward Indonesia. The time of this snapshot is about 9 h 25 min from the
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
108
Z. Kowalik et al.
Figure 10.9.
Sea level recorded at Cocos Island on December 26, 2004.
Figure 10.10. A birds-eye view of the IOT at time 9 h 25 min from the tsunami onset, looking from Africa toward India and Indonesia. Trapped tsunamis around continents and islands still display a strong signal.
onset of the earthquake. Although by this time the tsunami has essentially dissipated in the open ocean domain, a strong tsunami signal still persists along the continents and islands demonstrating trapping of tsunami energy by local shelf bathymetry. The IOT in the global ocean displays many interesting features. One of them observed through the animation technique is tsunami transformation when it travels through the narrows between oceans. The passage between Antarctica and Australia/New Zealand plays a noticeable role in tsunami amplification. As the passage is wide on the Indian Ocean side and constrained on the Pacific side, the eastward moving signal is amplified and also reorganized into periodic wave-like structures. A similar reorganization of a quasi-turbulent signal into oscillatory wave pattern can be observed in the passage between South America and Africa for the tsunami propagating from the Southern into the Northern Atlantic (Figure 10.11).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Numerical modeling of the Indian Ocean tsunami
Figure 10.11.
10.7
109
Sea level pattern generated by the IOT of December 26, 2004 at 30 h 40 min from the onset. Tsunami signals in the Northern Atlantic and Southern Pacific have been reorganized into coherent waves after passing through the narrows between Africa and South America, and Australia and Antarctica.
GLOBAL DISTRIBUTION OF MAXIMUM AMPLITUDE
Model computations using the source discussed in Section 4 were made for the first 50 h of propagation by such time the tsunami signal had traveled over the entire World Ocean. During this computation the maximum tsunami amplitude in every grid point was recorded. The plot of maximum amplitude in the proximity of the generation domain is given in Figure 10.12, and the corresponding plot for the World Ocean is given in Figure 10.13. The strong directional signal generated by the elongated source dominated the Indian Ocean domain. The main energy lobe was directed toward Sri Lanka and the secondary lobe pointed toward South Africa, sending a strong signal into the Atlantic Ocean. The computation indicated that the maximum amplitude was 15.5 m in proximity to the fault, 9.3 m at the shore of Thailand, 8.1 m at Sri Lanka, and 3.3 m at the coast of East Africa. This figure also depicts the amplitude enhancement in shallow water and especially in proximity to peninsulas and islands due to energy concentration through the refraction process. The large domain of the Arabian Sea is located in the shadow of the main energy beam. Both computation and observation demonstrated significant increase of the tsunami amplitude up to 1.5 m at the coast of Oman as recorded by the tide gauge in Salalah. This global maximum amplitude distribution (Figure 10.13) shows that the IOT encompassed all over the World Ocean. Although the source directed most of the wave energy towards South Africa, a strong signal was also directed towards the Antarctica. It is obvious that the ocean bathymetry affected the tsunami propagation. For example, tsunamis tended to propagate toward Antarctica along the oceanic ridges, and subsequently continued to transfer higher energy along the South Pacific ridge toward South and Central America. This mode of propagation resulted in the tsunami amplitude increase up to 65 cm along the Pacific coast of South America. A similar mode of energy transfer is observed in the Atlantic, where the Mid-Atlantic ridge channeled the tsunami to produce 30 cm wave amplitude as far north as Nova Scotia. An especially large energy flux was ducted from the South Atlantic Ridge toward Brazil and Argentina. The filaments of energy trapped along the South Pacific ridges are most spectacular as they ducted tsunami
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
110
Z. Kowalik et al.
Figure 10.12.
Maximum modeled tsunami amplitude in the Indian Ocean.
Figure 10.13.
Maximum modeled tsunami amplitude in World Ocean.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Numerical modeling of the Indian Ocean tsunami
111
energy for many thousands of kilometers. A simple explanation of the energy trapping using the continuity equation leads us to conclude that the amplitude should increase over the ridges due to the relatively shallower depth there. At the same time the role of bottom friction over the 3 km deep ridge is negligible and, therefore, the tsunami can travel long distance without much energy losses. The trapping of this energy is probably related to the long wave trapped along the ridge (Mei, 1989). The cross-ridge trapping length, which is responsible for energy concentration, is approximately defined by the tsunami wavelength. As the IOT carried a wide spectrum of waves with periods from 20 to 50 min, the wavelength for the mid-ocean ridge is in the range of 100–600 km. A simple explanatory model for the long wave trapping may be based on tsunami wave speed over and off the ridge. As the waves over ridge are slower and the waves off ridge are faster, the joint tsunami wave front is curved in such a way that the energy is fluxed toward the ridges (see sea level pattern in Figure 10.11 over the South Pacific ridge). However, the above explanation neglects the influence of the Coriolis force on tsunami propagation. Tsunamis are typically computed without Coriolis force because their periods are much smaller than the inertial period. As propagation proceeds over long distances the compounding effect of Coriolis force may sum up and significantly increase. In Figure 10.14 the residual maximum amplitude is given as the difference between two computed distributions, one with and one without Coriolis force. The difference given in Figure 10.14 shows locations where Coriolis force dominated. The amplitudes are not very large and according to expectation the influence was increasing toward the south since the Coriolis effect increases poleward from equator. Consistent change is observed along the South Pacific Oceanic Ridge. Residuals due to Coriolis force were close to 1 cm and since the total amplitude along this ridge according to Figure 10.13 is approximately 4 cm, we conclude that Coriolis force plays a role in the energy trapping along the oceanic ridges (see also trapping in the South Atlantic). A simple model for energy trapping due to the Coriolis force is a Kelvin wave propagating along the depth discontinuity
Figure 10.14.
Residual maximum amplitude in World Ocean.
112
Z. Kowalik et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
(Longuet-Higgins, 1969). The across-discontinuity trapping distance is defined by the Rossby radius of deformation (Gill, 1982). This distance is a function of depth and latitude and for ranges of depth 1–4 km and for latitude of 40–60◦ the Rossby radius ranges from 1000–2000 km. As this length is much larger than the tsunami wavelength we can conclude that Coriolis force is less effective in the concentrating tsunami energy along the oceanic ridges. To demonstrate the pattern of the energy trapped over the various bathymetric features we considered the energy flux. In the rectangular system of coordinates, with the x coordinate along E–W direction and y along N–S direction, the u component of velocity along x direction can be combined with the sea level (ζ) to define E–W component of the energy flux vector (Kowalik and Murty, 1993a): Ex = ρgHuζ
(10.11)
Similarly, the N–S component of the energy flux vector is defined (with v, the velocity component along the y direction): Ey = ρgHvζ
(10.12)
Where ρ is the sea water density, g = 9.81 m s−2 is the Earth’s gravity acceleration and H is the ocean depth. The energy flux vector for the progressive wave is always propagating into the same direction as the sea level and velocity and its direction is perpendicular to the wave front. To preserve direction of the energy flux in the progressive wave the velocity and sea level elevation remain in phase (Henry and Foreman, 2001). Energy flux units which are expressed as Joule/(s cm) is an energy flux per unit width and per unit time. To derive the total energy flux the above expressions should be multiplied by the length of a cross-section and integrated over the time period. In Figure 10.15 the energy flux vectors are shown in the south-western part of the Pacific Ocean. The larger tsunami amplitudes are located above the oceanic ridge and the energy flux is directed along the ridge. This small group of higher amplitude waves does not belong to the
Figure 10.15.
Energy flux vectors over the South Pacific ridge at time 26 h 20 min. Colors denote sea level. Dark-brown lines denote the ridge depth – 3000 m depth contour.
Numerical modeling of the Indian Ocean tsunami
113
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
first tsunami signal to arrive in this region. Its average wavelength was about 1350 km, as the depth of the ridge is close to 3 km the wave period was about 2 h. This is a somewhat long period for a free tsunami wave, but the longer period suggested possible influence of the Coriolis force. The wave pattern also showed that the waves over ridges were slower than the off-ridge waves, suggesting trapping due to refraction and focusing of off-ridge energy toward the ridge. Nonetheless, we cannot exclude the possibility of a resonance interaction of the tsunami wave and ridge bathymetry since Snodgrass et al. (1962) demonstrated the presence of discrete spectra in waves trapped over depth discontinuities. Mei (1989) showed that over a stepped bottom ridge the discrete spectra exist as well. If an incident wave can excite these trapped modes, an amplification of the tsunami signal due to resonance will follow. 10.8 TIME-DEPENDENT PROPAGATION Although the maximum amplitude defines tsunami distribution in the World Ocean, it does not reveal the temporal development of tsunamis. To improve understanding of the large scale temporal processes we used the temporal change of the tsunami energy fluxes passing through various cross-sections. The first cross-section considered is in the Indian Ocean, from 80◦ E to 105◦ E along 10◦ S (see Figure 10.12). The southward directed energy flux shown in Figure 10.16 is responsible for the tsunami signal propagating into the Pacific. The first maximum in this figure has been associated with the direct wave passing through the latitude 10◦ S at 2 h after the initial source motion. The next, even bigger energy influx arrived 2 h later, and is caused by the reflection from Sri Lanka and the east coast of India. The reflection from the Maldives Islands generates a signal which passed the cross-section at about 6.5 h from the initial disturbance. This cross-section is located quite close to the Bay of Bengal and therefore a large portion of the Maldives-reflected signal omitted this route. Since the Bay of Bengal acted as a parabolic mirror, it reflected many signals of smaller amplitude southward. The conclusion from the above experiment is that the reflected signal may send more energy south than the direct signal. With the major maxima in the southward directed signal identified, the task to associate them with the signal propagating into the Pacific Ocean remains. For this purpose an energy flux is considered through the three cross-sections located between Antarctica and the major continents. The cross-section (light shading in Figure 10.17) along the longitude 20◦ E from AS shows the time variation of the energy flux between Indian and Atlantic oceans. This flux remains negative for the entire period of 50 h, thus confirming that the inflow is directed into the Atlantic Ocean.
Figure 10.16.
Southward directed energy flux through the E–W cross-section located in the Indian Ocean along 10◦ S from 80◦ E to 105◦ E.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
114
Z. Kowalik et al.
Figure 10.17.
Energy flux through the cross-sections located between Antarctic and major continents. Along 20E fromAntarctica to SouthAfrica (AS) (light shading); along 140◦ E, fromAntarctic to Australia (AA) (dark shading).
On the other hand the energy flow through a cross-section along 140◦ E (dark shading in Figure 10.17) from AA is at all times positive (from the Indian to Pacific oceans). The flux through the cross-section located between South America and Antarctica at 70W, reveals a small in-flow from the Atlantic into the Pacific. Figure 10.17 clearly demonstrates that the magnitude and the time variability of energy fluxes through cross-sections AA and AS are quite different in character. The flux passing AS has a large value and the maximum energy in-flow to the Atlantic is located close to the initial wave front, even though the first arriving signal is not related to the maximum energy. The energy flow into the Atlantic is a result of the source orientation as shown in Figure 10.12. The main energy maximum is slower to arrive than the initial signal because, as Figure 10.13 depicts, the maximum energy directed toward South Africa cross-section is located along the oceanic ridge. Due to smaller depth over this ridge the more energetic signal propagates slower. The energy flow through AA demonstrates that tsunami arrives about 10 h from the onset of the earthquake; it initially has small amplitude which slowly increases in time from 18 to 21 h to achieve a few maxima. To understand the origin of this complicated temporal pattern of energy flux through AA we turn to Figure 10.16 and analyze the southward energy flux from the source area. The first signal arriving at the southern boundary in Figure 10.16 also crosses AA as the initial signal, since the travel time for this signal is close to 10 h. The second signal arriving 2 h later is caused by the reflection from the Sri Lanka and Indian coasts. The maxima in Figure 10.17 occurring from 18 to 21 h are related to the energy flux arriving by the various routes from the Indian Ocean. The arrival time of these signals depends on the depth and on the traveled distance. Therefore, it is useful to notice that the route from the generation domain to South America via passages between Australia and Antarctic is the great circle of a sphere. Signals which travel from the generation area to the section AA through the deep ocean travel faster, in about 10 h, but transfer less energy. The slower signal travels along oceanic ridges and transfers more energy as confirmed in Figure 10.15 by the energy flux vectors. This appears to be only a part of the story. Tracking (through animations) the signal shown in Figure 10.17 backwards (in time), the
Numerical modeling of the Indian Ocean tsunami
115
tsunami which passes between Australia and Antarctica, as depicted in cross-section AA (Figure 10.17), then loses its identity going back into the southern Indian Ocean. The large in-flow of energy shown in Figure 10.17, is related to the reflected signals off the Seychelles, Maldives and Africa, and to a slowly traveling reflection which originated in the Bay of Bengal. To compare the total energy flux entering Pacific and Atlantic oceans over the first 50 h of process, the energy fluxes given in Figure 10.17 have been integrated in time. The total energy flowing into the Pacific Ocean is approximately 75% of the total flow to the Atlantic Ocean.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
10.9 TRAVEL TIME Tsunami travel time from the source region to the given location is an important parameter in the tsunami prediction and warning. The Indonesian tsunami arrival times have been determined for many locations (Merrifield et al., 2005; Rabinovich, 2005; http://www-sci.pac.dfo mpo.gc.ca/osap/projects/tsunami/tsunamiasiax_e.htm; http://ilikai.soest.hawaii.edu/uhslc/iotd/; http://www.nio.org/jsp/tsunami.jsp). This set of data presents a possibility for the ocean-wide comparison of the data and our model. The first numerical experiment delineates the tsunami arrival time at every grid points for a signal of 0.1 cm amplitude. The computed tsunami travel time chart is depicted in Figure 10.18. The chart shows that even at such small limiting amplitudes the tsunami signal arriving at Alaska and North America did not pass through the Indonesian Straits but rather around Australia and New Zealand. The next numerical experiment computes isolines of arrival time for the tsunami signal of 0.5 cm amplitude (Figure 10.19). In the vast regions of Northern and Central Pacific this figure does not show a consistent arrival time. We may conclude that the main premise used to construct these figures, namely that the first train of tsunami waves is associated with the largest wave, does not hold true. We were able to construct isolines of arrival time in the regions of larger amplitudes, that is in the Indian Ocean, in the South Pacific (especially along the South Pacific ridge) and in the South
Figure 10.18. Travel time (in hours) for the tsunami of 0.1 cm amplitude.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
116
Z. Kowalik et al.
Figure 10.19. Travel time (in hours) for the tsunami of 0.5 cm amplitude. Table 10.2.
Observed and calculated travel time.
Station location
Travel time observed
Travel time for 0.1 cm amplitude
Travel time for 5 cm amplitude
Chennai (80◦ .17E, 13◦ .04N) Male (73◦ .52E, 4◦ .18N) Hanimadhoo (73◦ .17E, 6◦ .77N) Diego Garcia (72◦ .40E, 7◦ .28S) Hillarys (115◦ .73E, 31◦ .82S) Salalah (54◦ .00E, 16◦ .93N) Pt. La Rue (55◦ .53E, 4◦ .57S) Lamu (40◦ .90E, 2◦ .27S) Zanzibar (39◦ .18E, 6◦ .15S) Portland (141◦ .60E. 38◦ .33S) Richard’s Bay (32◦ .08E, 28◦ .80S) Port Elizabeth (25◦ .63E, 33◦ .97S) Jackson Bay (168◦ .62E, 43◦ .98S) Arraial de Cabo (42◦ .02W, 22◦ .97S) Arica (70◦ .21W, 18◦ .22S) Char. Amalie (64◦ .55W. 18◦ .20N) San Diego (117◦ .12W, 32◦ .45N) Halifax (63◦ .59W, 44◦ .66N) Atl.City (74◦ .25W, 39◦ .21N) Toffino (125◦ .55W. 49◦ .09N) Adak (176◦ .65W, 51◦ .87N)
2 h 36 min 3 h 25 min 3 h 41 min 3 h 55 min 6 h 41 min 7 h 17 min 7 h 25 min 9 h 9 min 9 h 49 min 10 h 39 min 11 h 13 min 12 h 28 min 18 h 18 min 21 h 56 min 26 h 36 min 28 h 42 min 31 h 25 min 31 h 30 min 31 h 48 min 32 h 1 min 35 h
2 h 18 min 3 h 12 min 3 h 24 min 3 h 40 min 6 h 24 min 7 h 6 min 7 h 24 min 8 h 30 min 10 h 24 min 9 h 48 min 11 h 00 min 12 h 00 min 12 h 30 min 20 h 54 min 26 h 6 min 27 h 45 min 29 h 0 min 30 h 6 min 30 h 45 min 29 h 0 min 27 h
2 h 20 min 3 h 18 min 3 h 30 min 3 h 40 min 6 h 36 min 7 h 6 min 7 h 24 min 8 h 30 min 10 h 36 min 10 h 18 min 11 h 12 min 12 h 6 min 19 h 30 min 21 h 30 min 29 h 20 min 33 h 30 min 35 h 30 min 32 h 6 min 33 h 30 min 38 h 30 min 40 h
Atlantic. By checking results of computations at the coastal locations we found that a tsunami of 0.5 cm amplitude arrived at every location in the Pacific Ocean. This wave did not arrive at western North America by refracting around New Zealand; it traveled closer to South America via energy ducts located over South Pacific ridges. This is a long travel time compared to the travel time depicted in Figure 10.18. In Table 10.2 the observed arrival time is compared with the computed arrival time of 0.1–5 cm tsunami amplitude. The observations define travel times uniquely when amplitude of the signal
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Numerical modeling of the Indian Ocean tsunami
117
is above the noise level. The mixed signal of meteorological and tsunami origin is difficult to differentiate. We took, somewhat arbitrarily, the amplitude of 5 cm as a signal strong enough to be seen above the meteorological noise. As can be seen from Figures 10.16 and 10.17 in many locations, and as close to the source as New Zealand, the first waves to arrive were quite small and they slowly increased in amplitude. For example, the observed arrival time at Jackson Bay, NZ was 18 h 18 min, while according to the travel time computed by the first perturbation of 0.1cm at this location, the arrival time for the first wave was 12 h 30 min. The stations located in the Northern Pacific showed the largest differences between the calculated and observed travel time. This is caused either by small tsunami signal-to-noise ratio, or by multiple paths between the source and gauge locations. In the latter, especially important is an interaction of the higher energy tsunami signals which travel slowly over the oceanic ridges and the lower energy signals which travel faster over the deep oceanic regions.
10.10
OBSERVATIONS VERSUS COMPUTATIONS
Although the 1-min computational mesh resolves many coastal and bathymetric features, nonetheless it is too large to resolve the local dynamics such as run-up. In Figure 10.20, the sea level at four stations have been chosen from the Indian Ocean area and compared with observations described by Merrifield et al. (2005).
Figure 10.20.
Observations and computations for four stations in the Indian Ocean.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
118
Z. Kowalik et al.
Figure 10.21.
Ground track of Jason-I and computed tsunami amplitude at 2:55 UT on December 26, 2004 in the Indian Ocean.
The stations are located on Maldives Island (Male), on the Seychelles Islands (Pt LaRue), on the African coast of Kenya (Lamu), and on the Arabian Peninsula coast of Oman (Salalah). The model reproduces quite well the maximum amplitude and the temporal behavior of the tsunami, indicating that with higher resolution bathymetry an even better comparison can be achieved. As luck would have it, the Jason-I altimetry satellite traversed the Indian Ocean about 2 h after the event origin time. It crossed the equator on a NNE path at 02:55 UTC on December 26, 2004 (Figure 10.21). The non-validated altimetry data was downloaded from the JPL Physical Oceanography web site at ftp://podaac.jpl.nasa.gov/pub/sea_surface_height/jason/j1nrtssha/data/ Altitude is sampled approximately once per second and is corrected for tides. In order to remove any background, the raw data from the “tsunami pass” was corrected by subtracting out an average of 4 “non-tsunami” passes. Jason-I repeats its track about once every ten days, so four repeat paths were averaged to generate the background. The corrected signal, was then smoothed by removing fluctuations with wavelengths shorter than 20 km. The smoothed signal with background removed was compared against the model data. Note that Jason-I was above the tsunami for about 10 min. During this time the tsunami was in motion, so the comparison is made to dynamic model predictions and not against a static snapshot. The model wave heights at the moment of equatorial crossing of Jason-I are shown in Figure 10.21. Jason-I crosses the leading edge wave at a point on the wave front where the amplitude is rapidly increasing toward the NW. The comparison between data and model is therefore sensitive to small variations in source details. The final comparison is shown in Figure 10.22 (upper panel). The leading edge wave location is predicted accurately by the model, even if the amplitude is not. The modeled leading wave with the amplitude similar to the recorded by satellite was in the satellite footprint 15 min earlier (see, Figure 10.21). A closer amplitude/period match was obtained by rotating the source strike counterclockwise slightly, and by reducing the fault width from 200–125 km. The comparison, given in Figure 10.22 (lower panel) shows that the model as driven by the adjusted source function reproduces more accurately the leading wave recorded by Jason-I.
Numerical modeling of the Indian Ocean tsunami
119
Ssh on Jason track 80
Model Jason
60
Ssh (cm)
40 20 0 20
60 20
0
10
10
20
Latitude (deg) Ssh on Jason track 80 Model Jason
60 40 Ssh (cm)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
40
20 0
20 40 60 80
100 20
Figure 10.22.
10.11
10
0 Latitude (deg)
10
20
Computed and observed tsunami amplitude along the Jason-I track. Upper panel: source function given in Figure 10.1. Lower panel: source function orientation and width adjusted.
DISCUSSIONS AND CONCLUSIONS
The main purpose of the present chapter was to derive a global picture of the IOT of December 26, 2004 based on a computer model. By comparison to the global observation we have identified major patterns of tsunami propagation. The new physics observed through application of the GTM underscores the importance of using a GTM in tsunami investigations. The IOT of December 26, 2004 was quite different from large scale tsunamis in the Pacific and Atlantic oceans. Post IOT analysis depicted numerous reflections and quite long ringing of the tsunami oscillations in the coastal regions suggesting local resonance and local trapping of tsunami energy. Computed distributions of the maximum amplitude compare well with observations analyzed by Merrifield et al. (2005) and by Rabinovich (2005). The observed and computed temporal variation of the tsunami at gauge stations in the Indian Ocean display quite similar amplitude and variability, although the model resolution of about 2 km still needs to be improved for proper run-up calculations. The comparison against satellite data shows that improvements in the source function are needed. The source location was constrained by tsunami travel times to tide
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
120
Z. Kowalik et al.
gauge locations and by earthquake parameters computed from seismic data inversions. Further investigation through comparison to the magnitude and location of the satellite signal (Gover, 2005) should also improve the source parameters. As the source function is one of the major building blocks of the tsunami model and, as some new insight on the source function pattern and tsunami generation has been suggested by Lay et al. (2005) and Hirata et al. (2005), it will be important to improve the GTM model using this new information. The model computations reveal peculiarities of tsunami signal when it travels from the source over entire World Ocean. The most interesting is ducting the tsunami energy along oceanic ridges, (Kowalik et al., 2005; Titov et al., 2005), which is so clearly shown in Figures 10.12 and 10.13. To demonstrate the pattern of energy trapping over the ridges, the energy flux function is used. The energy flux vectors show magnification over the South Pacific ridge and their distribution suggest trapping due to refraction and focusing the off-ridge energy toward the ridge. Further investigations of energy fluxes show more complicated temporal and spatial patterns in tsunami propagation. The primary signal traveling toward the Pacific Ocean depicted low energy level, therefore it was not well observed at sea level gauges. The more energetic signal arrived with some delay. The investigation of the energy flux along the South Pacific ridge reveals tsunami of approximately 2 h period. Tracking (through animations) tsunamis passing between Australia and Antarctica we have found that the tsunami moving from west to east is amplified and is also reorganized into periodic wave-like structures. Similar reorganization of the tsunami occurs between South America and Africa for the tsunami propagating from the Southern into the Northern Atlantic.
ACKNOWLEDGEMENTS We wish to express our gratitude to Juan Horrillo, Institute of Marine Science, University of Alaska, Fairbanks for testing our model using VOF approach and offering suggestions on the model improvements. We are grateful to Roger Edberg, Arctic Region Supercomputing Center, for constructing high-quality animations which allowed us to grasp the nature of the global tsunami propagation. We are also indebted to Professor Sathy Naidu, Institute of Marine Science, University of Alaska, Fairbanks, who made comments toward improving manuscript.
REFERENCES Abe, K., and Ishii, H. (1980). Propagation of tsunami on a linear slope between two flat regions. Part II reflection and transmission, J. Phys. Earth, 28, 543–552. Aki, K., and Richards, P.G. (1980). Quantitative Seismology Theory and Methods, Vol. 2. W.H. Freeman and Co., San Francisco, p. 557. Carrier, G.F., and Greenspan, H.P. (1958). Water waves of finite amplitudes on a sloping beach. J. Fluid Mech., 4, 97–109. Carrier, G. F., Wu, T.T., and Yeh, H. (2003). Tsunami run-up and draw-down on a plane beach. J. Fluid Mech., 475, 79–99. Flather, R.A., and Heaps, N.S. (1975). Tidal computations for Morecambe Bay. Geophys. J. Roy. Astr. Soc., 42 489–517. Gill, A.E. (1982). Atmosphere-Ocean Dynamics. Academic Press, New York, p. 662. Goto, C., and Shuto, N. (1983). Numerical simulation of tsunami propagations and run-up. In: (eds.), K. Iida and T. Iwasaki. Tsunamis–Their Science and Engineering, Terrapub, Tokyo, pp. 439–451. Gover, J. (2005). Jason 1 Detects the 26 December 2004 Tsunami. EOS, Trans. AGU, 86, 37–38. Henry, R.F., and Foreman, M.G.G. (2001). A representation of tidal currents based on energy flux. Mar. Geod., 2493, 139–152.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Numerical modeling of the Indian Ocean tsunami
121
Hirata, K., Satake, K., Tanioka, Y., Kuragano, T., Hasegawa, Y., Hayashi, Y., and Hamada, N. (2005). The Indian Ocean tsunami: tsunami source model from satellite altimetry. In: G.A. Papadopoulos and K. Satake, (eds.), Proceedings of the International Tsunami Symposium, Chania, Greece, 2005, pp. 72–76. Horrillo, J.J., Kowalik, Z., and Knight, W. (2005). Full navier-stokes approximation in tsunami investigation. In: G.A. Papadopoulos and K. Satake, (eds.), Proceedings of the International Tsunami Symposium, Chania, Greece, 2005, pp. 77–88. HRV Harvard CMT Catalog (2005). Harvard Seismology: Centroid-Moment Tensor Project, posted at http://www.seismology.harvard.edu/CMTsearch.html Imamura, F. (1996). Review of tsunami simulation with a finite difference method. In: H.Yeah, P. Liu and C. Synolakis (eds.), Long-Wave Runup Models World Scientific, Singapore, pp. 25–42. Iwasaki, S.I. (2005). Posting of Thailand Tide Gauge Data to Tsunami Bulletin Board, also posted at http://www.navy.mi.th/hydro/tsunami.htm Ji, C. (2004). Preliminary Result of the 04/12/26 (Mw 9.0), OFF W COAST of Northern Sumatra Earthquake, posted at http://www.gps.caltech. edu/%7Ejichen/Earthquake/2004/aceh/aceh.html Kowalik, Z. (2003). Basic relations between tsunami calculation and their physics II. Sci. Tsunami Hazards, 21(3), 154–173. Kowalik, Z., and Murty, T.S. (1993a). Numerical Modeling of Ocean Dynamics. World Scientific, Singapore, 481pp. Kowalik, Z., and Murty, T.S. (1993b). Numerical simulation of two-dimensional tsunami runup. Mar. Geod., 16, 87–100. Kowalik, Z., Knight, W., Logan, T., and Whitmore, P. (2005a). Numerical modeling of the global tsunami: indonesian tsunami of 26 December 2004. Sci. Tsunami Hazards, 23(1), 40–56. Kowalik, Z., Knight, W., Logan, T., and Whitmore, P. (2005b). The tsunami of 26 December 2004: numerical modeling and energy considerations. In: G.A. Papadopoulos and K. Satake, Chania, (eds.) Proceedings of the International Tsunami Symposium, Greece, 27–29 June, 2005, pp. 140–150. Lay, T., Kanamori, H., Ammon, C.J., Nettles, M., Ward, S.N., Aster, R.C., Beck, S.L., Bilek, S.L., Brudzinski, M.R., Butler, R., DeShon, H.R., Ekstrom, G., Satake, K., and Sipkin, S. (2005). The Great Sumatra–Andaman Earthquake of 26 December 2004. Science, 308, 1127–1139. Liu, P.L.F., Lynett, P., and Synolakis, C.E. (2003). Analytical solution for forced long waves on a sloping beach. J. Fluid Mech., 478, 101–109. Longuet-Higgins, M.S. (1969). On the transport of mass by time-varying ocean currents. Deep-Sea Res., 16, 431–447. Loomis, H.G. (1966). Spectral analysis of tsunami records from sations in the Hawaiian islands. Bull. Seis. Soc. Amer., 56, 697–713. Lynett, P.J., Wu, T.R., and LIU, P.L.F. (2002). Modeling wave run-up with depth-integrated equations. Coast. Eng., 46(2), 89–107. Mader, C.L. (2004). Numerical Modeling of Water Waves, CRC Press, pp. 274. Matsuyama, M. and Tanaka, H. (2001). An experimental study of the highest run-up height in the 1993 Hokkaido Nansei-oki Earthquake Tsunami. International Tsunami Symposium, Seattle, Washington, USA, pp. 879–889. Mei, C.C. (1989) The Applied Dynamics of Ocean Surface Wave. World Scientific, Singapore, pp. 740. Merrifield, M.A., Firing, Y.L., Aarup, T., Agricole, W., Brundrit, G., Chang-seng, D., Farre, R., Kilonsky, B., Knight, W., Kong, L., Magori, C., Manurung, P., Mccreery, C., Mitchell, W., Pillay, S., Schindele, F., Shillington, F., Testut, L., Wijeratne, E.M.S., Caldwell, P., Jardin, J., Nakahara, S., Porter, F.-Y., and Turetsky, N. (2005). Tide Gauge Observations of the Indian Ocean Tsunami, December 26, 2004, Geophys. Res. Letters, 32, L09603, doi:10.1029/2005GL022610. Mofjeld, H.O., Titov, V.V., Gonzalez, F.I., and Newman, J.C. (1999). Tsunami wave scattering in the north pacific. IUGG 99 Abstracts, Week B, July 26–30, B.1326. Murty, T.S., Rao, A.D. Nirupama, N., and Nistor, I. (2006). Tsunami warning systems for the hyperbolic (Pacific), parabolic (Atlantic) and Eliptic (Indian) oceans. J. Indian Geophys. Union, 10(2), 69–78. NEIC – US National Earthquake Information Center (2004). Magnitude 9.0 off the West Coast of Northern Sumatra. Sunday, December 26, 2004 at 00:58:53 UTC Preliminary Earthquake Report, posted at http://neic.usgs.gov/neis/bulletin/neic_slav_ts.html
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
122
Z. Kowalik et al.
Nekrasov, A.V. (1970). Transformation of tsunami on the continental shelf. In: W.M. Adams. (eds.) Tsunami in the Pacific Ocean. East-West Center Press, Honolulu, 1970. 337–350. NIO – National Institute of Oceanography of India (2005). 26 December 2004 Tsunami, posted at http://www.nio.org/jsp/tsunami.jsp Okada, Y. (1985). Surface deformation due to shear and tensile faults in a half-space. B. of the Seismol. Soc. Am., 75, 1135–1154. Rabinovich, A.B. (2005). Web Compilation of Tsunami Amplitudes and Arrival Times. http://wwwsci.pac.dfo-mpo.gc. ca/osap/projects/tsunami/tsunamiasia_e.htm Reid, R.O., and Bodine R.O. (1968). Numerical model for storm surges in Galveston Bay. J. Waterway Harbor Div., 94(WWI), 33–57. Shigematsu, T., Liu, P. L.-F., and Oda, K. (2004). Numerical modeling of the initial stage of dam-break waves. J. Hydraul. Res., 42(2), 183–195. Snodgrass, F.E., Munk, W.H., and Miller, G.R. (1962). California’s continental borderland. Part I. Background spectra. J. Mar. Res., 20, 3–30. SOI – Survey of India (2005). Preliminary Report of Tsunami Observations, posted at http://www.surveyofindia.gov.in/tsunami4.htm Stein, S., and Okal, E. (2005). Ultra-Long Period Seismic Moment of the Great December 26, 2004 Sumatra Earthquake and Implications for the Slip Process, posted at http://www.earth.northwestern.edu/people/seth/research/sumatra2. html Thacker, W.C. (1981). Some exact solutions to the nonlinear shallow—water wave equations. J. Fluid Mech., 107, 499–508. Titov, V., Rabinovich, A.B., Mofjeld, H.O., Thomson, R.E., and González, F.I. (2005). The Global Reach of the 26 December 2004 Sumatra Tsunami. Science, 309, 2045–2048. Titov, V.V., and Synolakis, C.E. (1998). Numerical modeling of tidal wave run-up. J. Waterway Port Coast. Ocean Eng., 124(4), 157–171. Yagi, Y. (2005). Preliminary Results of Rupture Process for 2004 off Coast of Northern Sumatra. Giant Earthqauke (ver. 1), posted at http://iisee.kenken.go.jp/staff/yagi/eq/Sumatra2004/Sumatra2004. html Yanuma, T. and Tsuji, Y. (1998). Observation of edge waves trapped on the continental shelf in the vicinity of makurazaki harbor, Kyushu, Japan. J. Oceanog., 54, 9–18.
CHAPTER 11
Modeling Techniques for Understanding the Indian Ocean Tsunami Propagation
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
V.P. Dimri and K. Srivastava National Geophysical Research Institute, Hyderabad, Andhra Pradesh, India
11.1
INTRODUCTION
The most destructive tsunami experienced by humanity has now been attributed to the great Indian Ocean Tsunami with death toll around 300,000. This earthquake that triggered the tsunami has changed the scientific considerations and the understanding of hazard globally. Researchers are now geared up in quantifying and understanding all different aspects for the evaluation and assessment of this natural hazard as this event has all the observations recorded and documented. The devastating mega thrust earthquake occurred in a tectonically active region where the Indian plate is subducting beneath the Burmese platelet in the Sunda trench. The focal depth has been estimated to be about 30 km and the length of the fault rupture has been inferred to be about 1300 km. Numerous studies have been carried out to constrain the rupture parameters however, one notable study was using far-field GPS observations where an average of 11 m of reverse slip has been estimated for the southern part of the rupture zone whilst it is about 10 m in the northern part (Catherine et al., 2005). The earthquake triggered giant tsunami propagated throughout the Indian Ocean. In the larger oceans such as Pacific reflections of the direct tsunami waves are not significant. However, in Indian Ocean, reflections played a very significant role and Kerala was affected by reflections from the Lakshadweep Islands. The sustained high water level in the Andaman and Nicobar Islands could be also due to the reflection of direct tsunami waves and trapping of wave energy. The location of the earthquake was such that it was land locked more or less from the three sides and instead of tsunami energy being spread to higher northern latitudes, the Bay of Bengal and the Arabian Sea got the brunt of it. Ocean depth gradients that give rise to convergences and divergences of tsunami wave energy (i.e. constructive and destructive interference) were responsible for devastating waves at Sri Lanka. The tsunami traveled and arrived in north of Sumatra within half an hour after the earthquake and a few hours later they arrived in Thailand, Sri Lanka, India and Maldives and after about 10 hours the tsunami reached the east coast of Africa. The tsunami waves arrived at different times at different locations on the Indian coast (Nair et al., 2005). They reached the Andaman and Nicobar Islands after an hour of the earthquake i.e around 07.25 h. They then reached Chennai at 08.45 h, Velankani coast at 09.05 h and in Kanyakumari around 11.45 h. In the west coast near Kayamkulam lagoonal mouth they reached around 12.30 h. Heavy loss of life and property at Nagapattinam (Tamil Nadu) could be because of resonance of tsunami waves with natural frequency of the coast. Subrahmanyam et al. (2005) have shown that the tsunami traveled with great speed across the Bay of Bengal and approached the continental slope and moved along the shoreline and surged through the bathymetric window between Nagapattinam and Cuddalore. They bought out the importance of bathymetry on tsunami amplification along the continental slopes. The wave heights observed at some locations are: along the north west coast of Sumatra 123
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
124 V.P. Dimri and K. Srivastava is around 10–15 m, whilst in Sri Lanka’s east coast was found to be around 5–10 m, in Andaman Islands it is >5 m and in Thailand, Phuket Island it is 3–5 m. Chadha et al. (2005) measured the tsunami run-up heights from water marks on houses and ocean debris transported on land along the east coast of India and found it to vary from 2.5 m to 5.2 m with the maximum height measured at Nagapattinam. The tsunami of the December 26, 2004 in the Indian perspective has been complied in book edited by Ramasamy and Kumanan (2005). Ramalingeswara Rao (2005) has discussed about the possible reasons for the triggering of the tsunami. Mahadevan et al. (2005) have simulated the tsunami waves using Delft 3D-Flow model which has been developed by WL/Delft Hydraulics of the Netherlands. The effect of tsunami on the ground water flow in coastal zones of Tamil Nadu was studied by Elango and Sivakumar (2005). Solving the ground water flow through porous media under non-equilibrium and anisotropic conditions they showed that the saline water recharges the coastal aquifer and the salinity of the ground water increased by about eight times which was later expected to decrease as the ground water flows towards the sea. 11.2
SOME GREAT AND DESTRUCTIVE TSUNAMIS OF THE WORLD
According to the US National Oceanic and Atmospheric Administration (NOAA) the Pacific is by far the most active tsunami zone; tsunamis have been generated in other bodies of water, including the Caribbean and Mediterranean Seas, and the Indian and Atlantic oceans. Some great and destructive tsunamis of the world have been shown in Table 11.1 using the information from Dudley and Lee (1988), Lander and Patricia (1989), Choi et al. (2003), Yalciner et al. (2005) with details about the magnitude, wave heights and the affected places. 11.3
MATHEMATICAL MODELS
11.3.1 Tsunami magnitude scales The tsunami magnitude scales have been discussed in Satake (2002). The tsunami magnitude scale is expressed as the Imamura-Iida scale m, I is the tsunami intensity. This scale was originally expressed by Idia et al. (1967) as, m = log2 h
(11.1)
where h is the maximum run-up height in m. This scale was later modified by Hatori (1979) to include the far-field data and gave the magnitude as, m=3+
log{(h/0.5)(R/1000)1/2 } √ log 5
where h is in m and R is in km. Tsunami intensity scale is defined by Soloviev (1970) as: √ ¯ i = log2 ( 2h)
(11.2)
(11.3)
Later, Chubarov and Gusiakov (1985) defined the tsunami intensity on the Soloviev-Imamura scale as, I = 3.55Mw − 27.1
(11.4)
where Mw is the moment magnitude of the earthquake. This relation holds good for steep dip-slip or low angle thrust fault mechanism. However for large strike slip faults tend to generate tsunamis of decreasing wave heights.
March 9, 1957; Aleutian
May 2, 1960; Chilean March 28, 1964; Alaska
September 2, 1992; Nicaragua November 27, 1999; Sydney, Australia October 4, 1994; Russia-Kuril Islands, Shikotan November 15, 1994; Phillipines-Mindora October 9, 1995; Mexico-Manzanilo February 2, 1996; Mexico-Manzanilo July 17, 1998; Papua New Guinea November 26, 1999; Vanuatu June 23, 2001; Peru-Southern January 2, 2002; Vanuatu
3
4 5
6 7 8 9 10 11 12 13 14 15
8 7.7 7.8 7.3
November 27, 1945; Makaran Coast, Pakistan December 12, 1992; Indonesia – Flores Island June 2, 1993; Indonesia, Java May 3, 2000; Indonesia – Sulawesi Island
December 26, 2004; Indian Ocean
4 5 6 7
8
9.3
7.9 – 7.7
Indian Ocean earthquakes, volcanic eruption and tsunami 1 December 31, 1881; Car Nicobar 2 August 27, 1883; Karkatau volcanic eruption 3 June 26, 1941; Andaman
7.6 7.1 8.1 7.1 8 7.8 7.0 7.4 8.4 7.4
9.5 8.4
8.3
8.2
November 4, 1952; Kamchatka
2
6
11–11.5 26.2 1–14 6
1 42 Not available
10 – 10 5–8 1–5.7 0.6 10 2–3 3–4.5 3
2–6
4–16
1–15
3.5
Magnitude Wave height of of earthquake tsunami (m) 7.8
Earthquake and tsunami
Affected places
Kutch and Mumbai The islands of the Pacific nation of Vanuatu Indonesia Sulawesi and the neighboring offshore smaller islands in Indonesia From Indonesia in the east, to the coast of Africa, some 7000 km (4000 miles) away
All along east coast of India Indonesia All along east coast of India
Near Unimak Island in Alaska’s Aleutian Island Chain, Hawaiian The coast of Kamchatka Peninsula, the Kuril Islands and other areas of Russia’s Far East South of the Andreanof Islands, in the Aleutian Islands of Alaska The coast of south central Chile Alaska, Vancouver Island (British Columbia), the states of Washington, California and Hawaii, in the USA Nicaragua The islands of the Pacific nation of Vanuatu The islands of the Pacific nation of Vanuatu Mindora, Verde and Baco Island Mexico Mexico Northern coast of Papua New Guinea The islands of the southwest Pacific North of town of Ocona in Southern Peru The islands of the southwest Pacific
Some of the world’s destructive tsunamis of the Pacific and Indian Ocean.
Pacific ocean earthquakes and tsunamis 1 April 1, 1946; Aleutian
S.no.
Table 11.1.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
300,000
Heavy loss 690 423 46
Not available 36,000 3000
70 8 11 78 40 12 2182 10 96 10
2300 120
Nil
Nil
165
Death toll
Modeling techniques for understanding the Indian Ocean Tsunami propagation 125
126 V.P. Dimri and K. Srivastava 11.3.2 Tsunami propagation
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
The non-linear form of long wave equations to describe the tsunami wave propagation is given below. The governing equation are expressed as; ∂η ∂[u(h + η)] ∂[v(h + η)] + + =0 ∂t ∂x ∂y
(11.5)
∂u ∂u ∂u ∂η τx +u +v +g + =0 ∂t ∂x ∂y ∂x ρ
(11.6)
∂v ∂v ∂v ∂η τy +u +v +g + =0 ∂t ∂x ∂y ∂y ρ
(11.7)
where η is the water elevation, u and v are components of the horizontal velocities, τx and τy are the bottom shear stress components, x and y are horizontal coordinates, t is time, h(x,y) is unperturbed depth, g is the gravitational acceleration. Using the total depth D = h + η, the discharge fluxes M and N in x, y directions respectively can be described as; M = u(h + η) = uD;
N = v(h + η) = vD
and
(11.8)
Equations (11.5)–(11.7) can be written in the following form for the discharge fluxes M and N : ∂η ∂M ∂N + + =0 ∂t ∂x ∂y ∂M ∂ + ∂t ∂x ∂N ∂ + ∂t ∂x
M2 D
MN D
+
∂ ∂y
+
∂ ∂y
(11.9)
MN D N2 D
+ gD
+ gD
∂η gn2 + 7/3 M M 2 + N 2 = 0 ∂x D
∂η gn2 + 7/3 N M 2 + N 2 = 0 ∂y D
(11.10) (11.11)
This formulation was used to develop the TSUNAMI-N2 model first by Professor Fumihiko Imamura in Disaster Control Research Center in Tohoku University (Japan). The TSUNAMI-N2 is one of the key tools in studying the propagation and coastal amplification of tsunamis in relation to different initial conditions. It solves equations (11.9)–(11.11) using a leap-frog scheme in finite difference technique for the basins of irregular shape and topography (Yalciner et al., 2003). This program is used to compute the water surface fluctuations and velocities at all locations, even at shallow land regions. However, there are limitations on the grid size (Imamura, 1996). 11.3.3
Discussions and approaches
Using the TSUNAMI-N2 program, the December 26, 2004 earthquake was modeled for wave propagation and the estimated results are shown in Figure 11.1 (Yalciner et al., 2005). The initial wave was simulated using the source parameters of this earthquake (Yalciner et al., 2005) and state of the sea at 5, 30, 60, 120, 180, 240, 300, 360, 420, 480, 600 min in Indian Ocean is shown in Figure 11.1. The observed and modeled run-up distributions along the east coast of India are taken from Yalciner et al. (2005) and are presented in Figure 11.2.
Figure 11.1. The sea state at 5, 30, 60, 120, 180, 240, 300, 360, 420, 480, 600 min in Indian Ocean (Yalciner et al., 2005).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Modeling techniques for understanding the Indian Ocean Tsunami propagation 127
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
128 V.P. Dimri and K. Srivastava
Figure 11.2. The observed run-up distributions along the east coast of India and comparison with model results (Yalciner et al., 2005).
The wave equations have been extensively used in different fields and analytically have been solved for various initial and boundary conditions (Witham, 1974). Tsunami wave propagation modeling can be projected on to an easier hyperbolic approach for the larger oceans like Pacific and Atlantic, but in the case of Indian Ocean one has to use the more difficult elliptic modeling. This essentially means that the coast line geometry and topography has to be mapped in much greater detail for the Indian Ocean. While modeling refraction (change of velocity from deep to shallow water), reflection and diffraction of the waves the vertical obstacles like sea mount or ridges that obstruct the path needs to be taken into account. Also the boundary conditions for closed domain need to be specified. The advantage of the FEM is precise error control of the approximate solution of the partial differential equation and another advantage is the diversity of possibilities to discretize the numerical domain (2D: quadrilaterals, triangles; 3D: tetrahedral, prisms, pyramids) and to adapt complex geometries. Having obtained a numerical solution for propagation, several simulations can be performed by varying parameters of uneven bathymetry. This can be efficiently done with fractal approach. There is a solution for accurate mapping of topography in terms of Voronoi tessellation, which can generate realistic and uneven fractal surfaces as desired with the help of a few parameters known as Voronoi centers (Dimri and Srivastava, 2005). The placement of these centers decides about the structure of final uneven surfaces and can be used to improve girding and mapping technique for uneven bathymetry. They have generalized the notion of Voronoi tessellation by using Lp distances instead of the least square distances so that Voronoi domains are not necessarily of polygonal shape. Analytically the wave has been solved and used extensively in other branches of science (Witham, 1974). The linear and non-linear partial differential equations are now being solved using the Adomians decomposition method. This is a relatively new method and the solution to the problem is built using a series solution. This method is now being used to solve deterministic, stochastic, linear or non-linear equations in various branches of science and engineering (Adomian, 1994; Wazwaz and Gorguis, 2004). The method has been shown to be systematic, robust,
Modeling techniques for understanding the Indian Ocean Tsunami propagation
129
and sometimes capable of handling large variances in the controlling parameters. The convergence of the decomposition series is very rapid and only a few terms in the series are required for an accurate solution. The forcing function and boundary conditions both have randomness and variability. Thus tsunami fields needs to be characterized by its statistical characteristics such as mean, variances and probability density functions. Theory of stochastic partial differential equations which have been developed extensively in statistical and theoretical physics literature can be used effectively to characterize these effects.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
11.4
CONCLUSIONS
There is need to develop high-resolution models for tsunami wave propagation in the Indian Ocean using denser bathymetry and irregular coastal topography by employing fractals and finite element method for obtaining run-up heights as well as inundation regions for different earthquake sources and magnitudes which are likely to occur in the Bay of Bengal, Indian Ocean and Arabian Sea. It would be desirable to develop analytical and numerical models of tsunami propagation for earthquake sources in Andaman and Nicobar and off the coast of Makaran as these have generated tsunamis in the past. Quantifying and understanding of Indian Ocean Tsunami will provide us useful knowledge for better evaluation of the tsunami hazards along the east and west coast of India in order to mitigate the suffering of the population living along these vast coastlines. ACKNOWLEDGEMENT The authors wish to thank Prof. Yalciner, Middle East Technical University, Civil Engineering Department, Ocean Engineering Research Center, Ankara, Turkey, for sharing his modeling results and permitting us to use them. REFERENCES Adomian, G. (1994). Solving Frontier Problems of Physics – the Decomposition Method. Kluwer Academic, Boston. Catherine, J.K., Gahalaut, V.K., and Sahu, V.K. (2005). Constraints on rupture of Dec 26, 2004 Sumatra Earthquake from Far Field GPS Observations. Earth Plan. Sci. Lett., 273, 673–679. Chadha, R.K., Latha, G., Harry, Y., Peterson, C., and Katada, T. (2005). The Tsunami of the Great Sumatra Earthquake of Magnitude 9.0 on 26 Dec 2004, Impact on east coast of India. Curr. Sci., 88(8), 1297–1301. Choi, B.H., Pelinovski, E., Kim, K.O., and Lee, J.S. (2003). Simulation of the Trans Oceanic Tsunami Propagation due to the 1883 Karkatau Volcanic Eruption. Nat. Hazards Earth Syst. Sci., 3, 321–332. Chubarov, L.B., and Gusiakov, V.K. (1985). Tsunamis and earthquake mechanism in the island arc regions. Sci. Tsunami Hazards, 3(1), 3–21. Dimri, V.P. and Srivastava, R.P. (2005). Fractal modeling of complex subsurface geological structures. In: V.P. Dimri (ed.), Fractal Behavior of the Earth System. The Netherlands, Springer, 208p. Dudley, W.C. and Lee, M. (1988). Tsunami! University of Hawaii Press, Honolulu, Hawaii. Elango, L., and Sivakumar, C. (2005). Numerical modeling of effect of tsunami on ground water flow and solute transport. In: S.M. Ramasamy and C.J. Kumanan (eds.), Tsunami: The Indian Context, pp. 221–229, Allied Publishers Ltd, Chennai, India. Lander, J.F. and Lockridge, P.A. (1989). United States Tsunamis. Publication, 41–42. U.S. Department of Commerce. Hatori, T. (1979). Relation between tsunami magnitude and wave energy. Bull. Earthquake Res. Inst. Univ. Tokyo, 54, 531–541. Idia, K., Cox, D.C., and Paras-Carayannis, G. (1967). Preliminary catalog of tsunamis occurring in the Pacific Ocean. Data report No 5, HIG 67-10, University of Hawaii, Honolulu.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
130 V.P. Dimri and K. Srivastava Imamura, F. (1996). Review of tsunami simulation with a finite difference method. In: H. Yeh, P. Liu, and C. Synolakis (eds.), Long-Wave Runup Models. World Scientific (ISBN 981-02-2909-7), Singapore, pp. 25–42. Mahadevan, R., Chandramohan, P., and Van Holland, G. (2005). Hydrodynamics of Tsunamis. In: S.M. Ramasamy and C.J. Kumanan (eds.), Tsunami: The Indian Context, pp. 69–77. Nair, M.M., Nagarajan, K., Srinivasan, R., and Kanishkan B. (2005). Indian Ocean Tsunami 2004 – An Indian Perspective. In: S.M. Ramasamy and C.J. Kumanan (eds.), Tsunami: The Indian Context, pp. 99–109, Allied Publishers Ltd, Chennai, India. Ramasamy, S.M. and Kumanan, C.J. (2005). Tsunami: The Indian Context. Allied Publishers Ltd, Chennai, India. Ramalingeswara Rao, B. (2005). Tsunami Triggering Mechanism on Indian Coast with Reference to 26 Dec 2004. In: S.M. Ramasamy and C.J. Kumanan (eds.), Tsunami: The Indian Context, pp. 43–50, Allied Publishers Ltd, Chennai, India. Satake, K. (2002). Tsunamis. In: International handbook of Earthquake and Engineering and Seismology, Vol. 81(A). Academic Press, New York, pp. 437–451. Soloviev, S.L. (1970). Recurrence of Tsunamis in the Pacific. In: W.M. Adams (ed.), Tsunamis in the Pacific Ocean. Honolulu, East-West Center Press, pp. 149–164. Subrahmanyam, C.S., Girish, R., and Gahalaut V. (2005). Continental slope characteristics along the tsunami affected areas of eastern offshore of India and Sri Lanka. J. Geol. Soc. India., 65, 778–780. Wazwaz, A.M. and Gorguis, A. (2004). Exact solutions of heat-like and wave-like equations with variable coefficients (with). Appl. Math. Comput., 149(1), 15–29. Witham, G.B. (1974). Linear and Non Linear Waves. Wiley-Interscience, New York. Yalciner, A.C., Pelinovsky, E., Synolakis, C., and Okal, E. (2003). In: A.C. Yalçıner, E. Pelinovsky, C. Synolakis, E. Okal (eds.), NATO SCIENCE SERIES Submarine Landslides and Tsunamis. Kluwer Publishers, The Netherlands, 329p. Yalciner, A.C., Karakus, H., Ozer, C., and Ozyurt, G. (2005). Short course on “Understanding the generation, propagation near and far field impacts of Tsunamis and planning strategies to prepare for future events”, Kuala Lumpur.
CHAPTER 12
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Validation of Tsunami Beach Run-up Height Predictive Model Based on Work–Energy Theorem G. Muraleedharan and A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology Delhi, New Delhi, India T.S. Murty Department of Civil Engineering, University of Ottawa, Ottawa, Canada M. Sinha Centre for Atmospheric Sciences, Indian Institute of Technology Delhi, New Delhi, India
12.1
INTRODUCTION
Natural hazards have always been a challenge to the modelers and tsunamis are in the top list. Even though there are many numerical models for understanding the propagation of tsunamis such as MOST (Method Of Splitting Tsunamis) its behavior in the coastal waters is highly complicated. Hence prediction of a destructive tsunami is highly difficult. It is to be noted that it is not the prediction of a tsunami event but the run-up heights on beaches is more important. The same tsunami will behave in a very different way due to the bottom topography and greater momentum of the terminal speed. A tsunami which is harmless in one coastal location will be very destructive in another location. Hence a clear estimation of the beach run-up heights at each and every coastal locations which are vulnerable to tsunami attacks is highly recommended. The importance of the predictive models developed (Muraleedharan et al., 2006) based on work–energy theorem for estimating the time required by the tsunami to travel from 1 m depth to 0 m depth and thereby the beach run-up heights is to be seen from this point of view. 12.2
MATERIALS AND METHODS
The run-up heights and the maximum inundation distance along a few coasts of Indian Ocean due to 26 December 2004 Indian Ocean Tsunami are mainly considered in this work. A few historical information provided by National Geophysical Data Centre (NGDC) of tsunamis that had occurred along the coasts of Indonesia is also considered in this study (Figure 12.1 and Table 12.1). The scientists of Andaman and Nicobar Centre for Ocean Science and Technology (ANCOST) of National Institute of Ocean Technology (NIOT) Chennai, conducted run-up measurements of the 26 December 2004 Indian Ocean Tsunami from 18 January to 5 February 2005 (Figure 12.2). Elevations at clearly visible sea water mark on building/structures were taken as the run-up levels for measurements. Table 12.2 gives the details of the measured run-up levels, which 131
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
132
G. Muraleedharan et al.
Figure 12.1. Table 12.1.
Map of Indonesia (www.worldatlas.com). Maximum water height and maximum inundation distance along the coasts of Indonesia during historical tsunamis.
Tsunami source date
Run-up location
Year
Month
Day
Country
Name
1992 1992 1994 1994 1994
12 12 6 6 6
12 12 2 2 2
Indonesia Indonesia Indonesia Indonesia Indonesia
Flores Riangkroko Lampon Pancer Rajekwesi
Latitude (S)
Longitude (E)
Maximum water height
−8.500 −8.150 −8.620 −8.589 −8.560
121.000 122.800 114.090 114.005 113.940
25.00 26.20 11.00 9.50 7.00
Maximum inundation distance 300.00 600.00 1000.00 300.00 100.00
have been corrected to mean sea level (approximately 0.8 m added to Mean Sea Level (MSL) to accommodate the land subsidence occurred during earthquake). Beach profiles for some coastal locations of the Andaman and Nicobar Islands are given in Figure 12.3(a)–(d). The significant beach angle (θs – average of 1/3 highest slopes (angles) on land) is suggested for a complicated beach profile. Details of beach run-up heights and sea water inundation inland of north Chennai coast due to Boxing Day tsunami are provided by Department of Ocean Development (Table 12.3). Another study area (Figure 12.4) is the southern and western part of the Tamil Nadu State, India (8◦ 04 to 8◦ 17 N:77◦ 32 to 77◦ 54 E). It is bounded by Indian Ocean in the south, Arabian Sea in the west and Bay of Bengal in the east but the main part of the coast is in the Arabian Sea. The study area is marked with marine terrace, sand dunes, beaches, mangroves, uplands, etc. The continental shelf along the study area extends far away from the shoreline (Chandrasekar et al., 2006). The run-up level of sea water due to 26 December 2004 Indian Ocean Tsunami in this area is given in Table 12.4.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Validation of tsunami beach run-up height
Figure 12.2. Andaman and Nicobar Group of Islands (map by ANCOST, NIOT, Chennai). Table 12.2.
Location Flores Riangkroko Lampon Pancer Rajekwesi
Predicted time (t) required to travel from 1 m depth to 0 m depth and tsunami height (Hs ) near coastlines of Indonesia. Maximum water height (m)
Maximum inundation distance (m)
Beach slope (tan θ)
t (s) (assumed θ = θ1 )
Predicted tsunami height (Hs ) (m)
25.00 26.20 11.00 9.50 7.00
300.00 600.00 1000.00 300.00 100.00
0.0836 0.0437 0.011 0.0317 0.0702
9.43 18.77 81.10 26.41 11.36
10.12 10.19 03.93 03.62 02.80
133
134
G. Muraleedharan et al. Transect near Malacca (Car Nicobar) 8 7
Elevation (m)
6
Maximum Water Level Observed during Tsunami
5 4 3
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
2 Beach
1 0 0
200
400
600
(a)
1000
1200
1400
Distance (m) Profile of coastal land and run-up levels at Campbell Bay (Great Nicobar)
3.5
Maximum Water Level Observed during Tsunami
3 Elevation (m)
800
2.5 2 1.5 1 0.5 0
50
0.5
100
150
200
250
300
Distance (m)
(b)
Elevation (m)
Transect near Hut Bay (Little Andaman)
(c)
Figure 12.3.
12 11 10 9 8 7 6 5 4 3 2 1 0
Maximum Water Level Observed during Tsunami Crop land Fishing village
0
200
400
600
800
1000
1200
1400
1600
Distance (m)
(a)–(d) Real shore profiles of a few Andaman and Nicobar Group of Islands (by ANCOST, NIOT, Chennai).
Validation of tsunami beach run-up height
135
Transect near Chidiyatopu Maximum Water Level observed during Tsunami 5 Elevation (m)
4 3 2 1
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
0 10
10
30
Figure 12.3.
50
70
90
110
130
150
Distance (m)
(d)
(Continued).
Table 12.3.
Maximum run-up level distance up to which seawater inundated inland during boxing day tsunami in Andaman and Nicobar Islands.
Location South Andaman (Port Blair) JNRM College, Aberdeen Bamboo Flat New Wandoor Wandoor Chidiyatopu Chouldari Sippighat (Creek) North Andaman Diglipur Rangat Little Andaman Hut Bay Car Nicobar Malacca Air base Great Nicobar Campell Bay (central) Campell Bay (south)
Maximum run-up level (m)
Distance up to which seawater inundated inland (m)
2.9 3.5 3.7 3.9 4.5 2.0 2.0
130 250 215 215 130 250 2000
1.5 1.5
100 200
5.0
1200
7.0 7.0
1000 1100
3.0 6.0
300 50
A few historical tsunami events in the Pacific and Atlantic oceans provided by the NGDC is also considered in this study (Table 12.5). This study is the validation of the expression derived from the work–energy theorem for estimation of beach run-up heights and the time required to travel from 1 m depth to 0 m depth by tsunamis (Muraleedharan et al., 2006) given below as: Hr =
g Hs t tan θ d1
(12.1)
136
G. Muraleedharan et al. 8°30
r a barani R.
Index Map
Namb iya rR
R.
Se
a
P a la l yar R.
Panniy ar
Inayam Colachel Midalam Vaniakudy Mandakadu Kadiapattinam Muttom Rajakkamangalam Pallam
Kuttapuli
Manakudy
n
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
A r a bia
Study Area
R. an m
Taingapatnam Tamilnadu
Investigated Beach
77°00
Vattakottai Lakshmipuram Chinnamuttom Kanyakumari Cape Camorin
Indian Ocean
Navaladi
Idhinthakarai
Scale – 1:20,000 l
V a lliy ar R.
Ha nu
Ovari Perumanal
.
Th
ami
Tamilnadu
N
ga
India
Be B a y of
n
8°00
77°30
Figure 12.4.
Location map of the study area of the Tamil Nadu coasts (www.sthjournal.org/241/ chand.pdf ).
Table 12.4.
Predicted tsunami travel time (t) from 1 m depth to 0 m depth and tsunami height (Hs ) near coastline of Andaman and Nicobar Islands.
Location South Andaman (Port Blair) JNRM College, Aberdeen Bamboo Flat New Wandoor Wandoor Chidiyatopu Chouldari Sippighat (Creek) North Andaman Diglipur Rangat Little Andaman Hut Bay Car Nicobar Malacca Air base Great Nicobar Campell Bay (central) Campell Bay (south)
Maximum run-up level (m)
Distance up to which seawater inundated inland (m)
Beach slope (tan θ)
t (s) (assumed as θ = θ 1 )
Predicted tsunami height (Hs ) near coastline (m)
2.9 3.5 3.7 3.9 4.5 2.0 2.0
130 250 215 215 130 250 2000
0.02 0.01 0.02 0.02 0.03 0.008 0.001
38.31 62.79 50.44 47.70 24.03 113.68 1031.61
1.08 1.27 1.36 1.44 1.73 0.70 0.62
1.5 1.5
100 200
0.02 0.008
58.36 121.75
0.55 0.52
5.0
1200
0.004
227.12
1.69
7.0 7.0
1000 1100
0.007 0.006
130.98 144.93
2.44 2.42
3.0 6.0
300 50
0.01 0.12
89.72 6.38
1.07 2.48
Validation of tsunami beach run-up height Table 12.5.
Few examples in Andaman and Nicobar Islands to show the improvement in tsunami height (Hs ) predictions when different slopes on land (Figure 12.3(a)–(d)) are considered. Distance up to Beach slope – Predicted Maximum which seawater | tan θ s | t (s) tsunami height run-up inundated (significant beach (off shore (Hs ) near level (m) inland (m) angle – θs ) slope θ1 ) coastline (m)
Location
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
137
South Andaman (Port Blair) Chidiyatopu
4.9
130
Little Andaman Hut Bay
5.0
Car Nicobar Malacca Great Nicobar Campell Bay (central)
0.74 (θs = 36.5◦ )
0.65 (θ1 = 46◦ )
3.25
1200
0.02 (θs = −1.17◦ )
42.08 (θ1 = 1.2◦ ) assumed
1.86
6.8
1000
0.167 (θs = 9.5◦ )
4.28 (θ1 = 10◦ )
3.03
3.0
300
1.96 (θs = 63◦ )
0.33 (θ1 = 63◦ )
1.48
where g is acceleration due to gravity, Hr is beach run-up height by the tsunami, d1 (>0) is water depth (= 1 m), Hs is tsunami height at 0 m depth and t is time required by the tsunami wave to travel from d1 m (here it is equal to 1 m) to 0 m depth which is functionally related with offshore angle θ1 (Muraleedharan et al., 2006) as: t = 0.6791(tan θ1 )−1.0606
(12.2)
tan θ = beach slope (= average of 1/3 highest slopes on land for a complicated beach configuration). If the beach profile is not known, then the beach angle (θ) is calculated as: θ = sin−1(maximum water height/maximum inundation distance)
(12.3)
The offshore slope (tan θ1 ) is considered to be the same as beach slope (tan θ) for unknown offshore slopes. Tsunami heights near coastlines are estimated for various coastal locations along the rim of the Indian Ocean using expression (12.1) as:
Hs =
12.3
Hr
d1 g
t tan θ
RESULTS AND DISCUSSIONS
The tsunami travel time from 1 m depth to 0 m depth (t) and tsunami height (Hs ) near coastlines for a few coastal locations along the rim of the Indian Ocean are predicted. Since the horizontal distance from 1 m depth to 0 m depth will be more for small offshore slopes compared to large offshore slopes, the time required by the wave to travel from 1 m depth to 0 m depth will be more
138
G. Muraleedharan et al. Table 12.6.
Run-up level of sea water during 26 December 2004 Indian Ocean Tsunami at selected locations along Tamil Nadu coasts.
Location
Maximum run-up level (m)
Distance of seawater inundation inland (m)
3.9 2.8 1.8 1.4 3.5
750 200 190 45 80
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Nagapattinam (Light House transect) Chennai (Besant Nagar) Chennai (Kattupalli) Chennai (Kalanji) Sathan Kuppam
Table 12.7.
Prediction of travel time (t) from 1 m depth to 0 m depth and tsunami height near coastline for Tamil Nadu coasts for 26 December 2004 Indian Ocean Tsunami.
Location Nagapattinam (Light House transect) Chennai (Besant Nagar) Chennai (Kattupalli) Chennai (Kalanji) Sathan Kuppam
Maximum run-up level (m)
Distance of seawater inundation inland (m)
Beach slope (tan θ)
t (s) (assumed as θ = θ1 )
Predicted tsunami height (Hs ) near coastline (m)
3.9
750
0.0052
179.56
1.33
2.8 1.8 1.4 3.5
200 190 45 80
0.0140 0.0095 0.0311 0.0438
62.79 95.03 26.91 18.74
1.02 0.64 0.53 1.36
in the first case. Muraleedharan et al. (2006) have shown using equation (12.1) that it is true for the numerical experiment conducted by Marchuk and Anisimov in 2001 for run-up height estimations for different beach angles. Also tsunami run-up heights will be more for small beach slopes than for large beach slopes (Marchuk and Anisimov, 2001). Another important phenomena is that a tsunami wave height near coastline can have a run-up height more than double its height due to bottom topography and greater momentum of the terminal velocity. These findings are reaffirming here for real shore profiles. The time required to travel from 1 m depth to 0 m depth (t) and the tsunami height at 0 m depth (Hs ) are predicted for the historical tsunami events (Table 12.1) for the Indonesian coasts and are given in Table 12.6. Similar predictions are carried out for the 26 December 2004 Indian Ocean Tsunami for the most affected Andaman and Nicobar Group of Islands and for Tamil Nadu coasts (Tables 12.7–12.10). Real shore profiles for a few coasts in Andaman and Nicobar Islands are given in Figure 12.3(a)–(d). Significant beach angles are computed and time (t) and tsunami heights (Hs ) are predicted (Table 12.8). Nearshore tsunami heights are predicted for historical tsunami events in the Pacific and Atlantic oceans (Table 12.11). It is very interesting to see that in all these studies the time (t) is more for small offshore slopes (here the offshore slopes are considered to be equal to the respective beach slopes) compared to large offshore slopes and small tsunami heights can have large run-up heights for small beach slopes.
Validation of tsunami beach run-up height Table 12.8.
Inundation distance extent along the study area.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Location
Longitude (E)
Latitude (N)
Elevation (m)
Inundation distance (m)
77.49 77.45 77.39 77.37 77.34 77.34 77.34 77.33 77.29 77.25 77.22 77.19 77.18 77.16 77.15 77.14 77.12 77.09 77.1
8.17 8.14 8.09 8.08 8.07 8.07 8.06 8.04 8.05 8.05 8.06 8.07 8.08 8.09 8.1 8.11 8.12 8.13 8.14
19 18 17 16 15 16 17 21 09 14 16 11 14 16 12 16 17 15 14
150 175 200 200 300 250 350 300 600 400 150 200 100 100 750 100 300 130 200
Ovari Idinthakarai Perumanal Navaladi Vattakottai Lakshmipuram Chinna muttam Kanyakumari Keelamanakudi Pallam Rajakkamangalam Muttom Kadiapatanam Mandakadu Colachel Vaniakudy Midalam Enayam Taingapatnam
Table 12.9.
Predicted time (t) required to travel from 1 m depth to 0 m depth and tsunami height (Hs ) near coastline for 26 December 2004 Indian Ocean Tsunami.
Location Ovari Idinthakarai Perumanal Navaladi Vattakottai Lakshmipuram Chinna muttam Kanyakumari Keelamanakudi Pallam Rajakkamangalam Muttom Kadiapatanam Mandakadu Colachel Vaniakudy Midalam Enayam Taingapatnam
Elevation (m)
Inundation distance (m)
Beach slope (tan θ)
t (s) (assumed θ = θ1 )
19 18 17 16 15 16 17 21 09 14 16 11 14 16 12 16 17 15 14
150 175 200 200 300 250 350 300 600 400 150 200 100 100 750 100 300 130 200
0.1277 0.1034 0.085 0.0803 0.0500 0.0641 0.0486 0.0702 0.0150 0.0350 0.1073 0.0551 0.14 0.16 0.0160 0.16 0.0567 0.1162 0.07
6.02 7.53 9.24 9.85 16.26 12.50 16.77 11.36 58.36 23.75 7.24 14.69 5.41 4.68 54.50 4.68 14.23 6.66 11.36
Predicted tsunami height (Hs ) near coastline (m) 7.89 6.97 6.88 6.46 5.88 6.37 6.65 8.41 3.28 5.37 6.57 4.34 5.84 6.73 4.39 6.73 6.72 6.19 5.61
139
Month
2 2 11
11 11 11 3 1 9 10 10 12 4 3 11 11 9 9 4 9 9 6
1835 1835 1867
1867 1867 1867 1868 1878 1899 1918 1918 1944 1946 1964 1969 1975 1985 1985 1991 1992 1994 2001
18 18 18 17 20 10 11 11 7 1 28 22 29 19 21 22 2 19 23
20 20 18
Day
USA USA USA USA Japan USA USA Russia USA Mexico Mexico Costa Rica Nicaragua Papua New Guinea Peru
Chile Chile Saint Vincent and the Grenadines USA Territory British Virgin Islands USA
Country
Virgin Islands: Frederiksted Tortola Island: Road Town Puerto Rico: Arroyo Bequia Island: Admiralty Bay Waialua, Oahu, HI Yakutat Bay, West Shore, AK Puerto Rico: Caja De Muertos Puerto Rico: Punta Borinquen Temma Pakala Point, HI Klawock, AK Ok’khovaya River Punaluu Bay, Hawaii, HI Playa Azul Zihuatanejo Cahuito-Puerto Viejo Masachapa Simpson Harbour, New Britain Camana
Maule River Talcahuano Bequia Island: Admiralty Bay
Location
Historical tsunami events and run-up levels in the Pacific and Atlantic oceans.
Year
Table 12.10.
−64.883 −64.616 −66.050 −61.250 −139.840 −66.533 −67.169 135.933 −133.083 162.800 −155.500 −102.350 −101.550 −82.770 −86.520 152.170 −72.720
59.730 17.867 18.484 33.633 55.550 57.000 19.140 17.980 17.650 9.650 11.780 −4.210 −16.630
−72.417 −73.130 −61.250
Longitude
17.717 18.414 17.983 13.000
−35.317 −36.740 13.280
Latitude
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
7.60 1.50 0.90 0.90 3.00 9.00 1.50 4.50 5.00 10.00 4.60 15.00 7.60 2.50 2.50 2.00 6.00 1.20 8.00
3.50 9.00 1.80
Maximum water height
76.00 9.00 40.00 21.00 150.00 500.00 15.00 100.00 200.00 227.00 15.20 500.00 137.00 150.00 200.00 300.00 100.00 200.00 1000.00
140.00 4000.00 146.00
Maximum inundation distance
140 G. Muraleedharan et al.
USA USA USA USA Japan USA USA Russia USA Mexico Mexico Costa Rica Nicaragua Papua New Guinea Peru
Chile Chile Saint Vincent and the Grenadines USA Territory British Virgin Islands USA
1835 1835 1867
1867 1867 1867 1868 1878 1899 1918 1918 1944 1946 1964 1969 1975 1985 1985 1991 1992 1994 2001
Country
Virgin Islands: Frederiksted Tortola Island: Road Town Puerto Rico: Arroyo Bequia Island: Admiralty Bay Waialua, Oahu, HI Yakutat Bay, West Shore, AK Puerto Rico: Caja De Muertos Puerto Rico: Punta Borinquen Temma Pakala Point, HI Klawock, AK Ok’khovaya River Punaluu Bay, Hawaii, HI Playa Azul Zihuatanejo Cahuito-Puerto Viejo Masachapa Simpson Harbour, New Britain Camana
Maule River Talcahuano Bequia Island: Admiralty Bay
Location
7.60 1.50 0.90 0.90 3.00 9.00 1.50 4.50 5.00 10.00 4.60 15.00 7.60 2.50 2.50 2.00 6.00 1.20 8.00
3.50 9.00 1.80
Maximum water height
76.00 9.00 40.00 21.00 150.00 500.00 15.00 100.00 200.00 227.00 15.20 500.00 137.00 150.00 200.00 300.00 100.00 200.00 1000.00
140.00 4000.00 146.00
Maximum inundation distance
0.1005 0.1690 0.0225 0.0429 0.0200 0.0180 0.1005 0.0450 0.0250 0.0441 0.3175 0.0300 0.0556 0.0167 0.0125 0.007 0.06 0.006 0.008
0.0250 0.00225 0.0123
Beach slope (tan θ)
7.76 4.47 37.96 19.15 43.01 48.10 7.76 18.18 33.94 18.60 2.29 27.97 14.56 52.19 70.82 137.93 13.39 154.24 113.68
34.01 436.59 71.86
t (s) (assumed θ = θ1 )
3.11 0.63 0.34 0.35 1.11 3.32 0.61 1.75 1.88 3.89 2.02 5.70 3.00 0.92 0.90 0.69 2.38 0.41 2.81
1.32 2.92 0.65
Predicted tsunami height at 0 m depth (m)
Predicted tsunami travel time (t) from 1 m depth to 0 m depth near coastlines and tsunami heights (Hs ) of historical tsunamis of Pacific and Atlantic oceans.
Year
Table 12.11.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Validation of tsunami beach run-up height 141
142
G. Muraleedharan et al.
12.4
CONCLUSION
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
The predictive model for tsunami beach run-up heights based on work–energy theorem is validated for various tsunami events including the 26 December 2004 Indian Ocean Tsunami along the coasts of the rim of the Indian Ocean. The predicted tsunami heights for known beach run-up heights and distance to which seawater inundated inland (beach run-up) are reasonably good. The predicted heights are agreeing the logic that they can have a run-up height of more than double its height due to bottom topography and greater momentum of the terminal speed. This beach run-up height model is a guide to the public to what height to move from MSL in the event of a tsunami warning for safety. REFERENCES Department of Ocean Development (2005). Preliminary Assessment of Impact of Tsunami in selected coastal areas of India, Integrated Coastal and Marine Area Management Project Directorate, Chennai, India. Marchuk, A.G. and Anisimov, A.A. (2001). A method for numerical modeling of tsunami run-up on the coast of an arbitrary profile. ITS Proceedings, Session 7, No.7–27, 933–940. Muraleedharan, G., Sinha, M., Rao, A.D., and Murty, T.S. (2006). Statistical simulation of the Boxing Day tsunami of the Indian Ocean and a predictive equation for beach run-up heights based on work–energy theorem. Mar. Geod., Special issues on Tsunamis, Part I, 29(3), 223–231. Chandrasekar, N., Saravanan, S., Immanuel, J.L., Rajamanickam, M., and Rajamanickam, G.V. (2006). Classification of tsunami hazard along the southern coast of India: an initiative to safeguard the coastal environment from similar debacle. Sci. Tsunami Hazards, 24(1), 3–23.
CHAPTER 13
Normal Modes and Tsunami Coastal Effects
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
N. Nirupama Emergency Management, Atkinson School of Administrative Studies, York University, Toronto, Canada T.S. Murty Department of Civil Engineering, University of Ottawa, Ottawa, Canada A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology, New Delhi, India I. Nistor Department of Civil Engineering, University of Ottawa, Ottawa, Canada
13.1
INTRODUCTION
In a tsunami event, the most important aspects happen at the coastline, mainly because it is here that loss of life and damage occurs due to land inundation. The so-called normal modes are free oscillations of the coastal gulfs, bays, estuaries, inlets, lagoons and backwaters, and all these play an important role in determining the coastal behavior of the tsunami. Here we will briefly review some of the classical concepts of the normal mode dynamics, following Murty (1977), and Kowalik and Murty (1993). The role of the normal modes in tsunami coastal effects is evident from Kowalik (2005a,b), Murty et al. (2005a–c), Murty et al. (2006a–c), Nirupama et al. (2005) and Nirupama et al. (2006). Any water body (either completely closed or partially open) undergoes natural or free oscillations, which are referred to as normal modes. Several different physical phenomena can set a water body into oscillation (i.e. excite its normal modes). The frequencies of the fundamental normal mode and its higher harmonics can be determined solely from knowledge of the geometry of the water body and the water depths. The question of normal modes was first discussed in connection with tidal theories (LaPlace, 1775, 1776; Hough, 1898). 13.2
OSCILLATIONS OF THE FIRST CLASS AND OSCILLATIONS OF THE SECOND CLASS
Consider an artificial situation in which a thin layer of water covers the Earth’s surface entirely. We ask the question how this water can move freely, subject to gravity and the earth’s rotation. Hough (1898) showed that free motion can occur in either of two ways. Oscillations of the first class (OFC) are essentially gravity waves whose periods are modified by Earth’s rotation. However, OFC can exist even if the Earth does not rotate. Oscillations of the second class (OSC) owe their very existence to Earth’s rotation and have periods greater than 24 h. If the earth’s 143
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
144
N. Nirupama et al.
rotation tends to zero, OSC will lose their periodicity and will degenerate into steady currents. One manifestation of OSC is the so-called Rossby wave. If σ is frequency of oscillation and ω is the frequency of rotation, then OFC are those for which σ → σ0 ( = 0) as ω → 0 and OSC are those for which σ → O(ω) as ω → 0. Bjerknes et al. (1934) distinguished between these two types of oscillation by means of the ratio σ/2ω. Gravity modes (OFC) are those for which σ/2ω ≥ 1. Elastoid-inertia modes (OSC) are those for which σ/2ω ≤ 1. For the gravity modes, gravity appears in the frequency equation. In the case of the rotational (elastoid-inertia) modes, the frequency for a given mode is a function mainly of the ratio of the depth of the liquid to the radius of the container and gravity does not play an important role in the frequency equation. Here this discussion is restricted to gravity modes. In the mathematical analysis this restriction is imposed by introducing the approximations of the shallow water theory called the quasi-static approximation (Bjerknes et al., 1934). This means that in the vertical direction, equilibrium exists not only before the motion but also during the motion, with vertical accelerations of the liquid being considered negligible compared to that of gravity. In studying tidal motions on a rotating earth, Kelvin (1879) considered a shallow layer of water in a circular flat-bottomed cylinder and assumed the quasi-static approximation to the pressure field. Kelvin considered small rotations and neglected the curvature of the free surface due to rotation. If σ is the frequency of the rotating mode, σ0 is the frequency of the non-rotating mode and ω is the rotation frequency, then the result obtained is: σ 2 = σ0 + 4ω2 This result shows that the rotation increases the frequency and thus increases the restoring tendency of the system when disturbed. However, if the curvature of the free surface is taken into account, this is not always true, especially for the higher modes. Since one of the manifestations of normal modes in water bodies is in the form of seiches, we will start with a discussion of this phenomenon, following mainly Wilson (1972).
13.3 THE PHENOMENA OF COASTAL SEICHES Probably the first scientific study of seiches was that of Forel (1892), Chrystal (1905), Proudman (1953), Defant (1961), Wilson (1972) and Miles (1974). Merian (1828) gave a theory for free oscillations of water in a rectangular basin of length L and uniform depth h, the period T being given by: 2L T =√ gH
(13.1)
where g is gravity. Forel (1892) applied this formula to seiches in lakes. For real lakes with variable depth, he chose an average value of H to replace the variable depths. Lagrange (1781) showed that the velocity of c of a long wave is given by: c ∼ gH (13.2) From equations (13.1) and (13.2): T =
2L λ = c c
(13.3)
where λ is the wavelength of the oscillation (assuming it is in the form of a wave). Thus, the length of the wave is twice that of the water body (or basin). Forel explained this apparent paradox
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Normal modes and tsunami coastal effects
145
as being due to the superposition of two long waves whose length is twice that of the basin and traveling in opposite directions. In the following, an attempt will be made to visualize a seiche as a special type of standing wave. For this, consider two progressive waves traveling in opposite directions in water of uniform depth. At every quarter period, the crests and troughs are either in phase or out of phase. At halfwavelength intervals (x = λ/4, 3λ/4, 5λ/4, …) surface elevation is continuously zero with time. Such points are called nodes and the points that intermediate to these are the antinodes. This type of standing wave can also result if a progressive wave is reflected (without dissipation) at a vertical wall. Then, there will be an antinode of amplitude 2A (A being the amplitude of the progressive wave) at the wall and a first node at x = λ/4 from the wall. A seiche is a special case of a standing wave that would result from interposing a second vertical barrier at any of the points x = λ/2, 3λ/2, 2λ, … The standing wave or seiche exists due to repeated reflections (assuming no dissipation) from the two vertical walls, where it would have its antinodes. On the other hand, if the second vertical barrier were inserted at any point other than a multiple of λ/2, the standing wave would become an irregular motion of the water surface. Thus, one can think of a seiche as a standing wave that is commensurate with the basin length L. The seiche is uninodal for L = λ/2, binodal for L = λ, trinodal for L = 3λ/2, …. n-nodal for L = nλ/2. Hence, from equation (13.3), the period Tn of the nth mode of oscillation in a rectangular basin of length L and uniform depth H is: 2L Tn = √ n gH
(13.4)
Here n = 1, 2, 3, … For an open bay of rectangular geometry (length L) and uniform depth H , the period is given by: 4L Tn = √ n gH
(13.5)
Here n = 1, 3, 5, … This is a generalization of the Merian formula and is valid for a onedimensional oscillation (no transverse motions). Note that at the nodes the motion is purely horizontal and at the antinodes it is purely vertical. The higher nodal (binodal, trinodal, etc.) seiches that may occur simultaneously with the fundamental mode (i.e., uninodal oscillation) are higher harmonics of the fundamental. From equation (13.4) Tn 1 1 1 = 1, , , . . . , , n = 1, 2, . . . , n T1 2 3 n However, for irregular water bodies with variable depth (unlike in the case of a narrow rectangular basin of uniform depth), such a simple relation as above need not exist. Another point worth remembering is that neither the use of an average depth H nor a better version of this, as done by du Boys (see Defant, 1961): 2 Tn ∼ n
0
L
dx [gh(x)]1/2
improves the Merian formula significantly.
(13.6)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
146
N. Nirupama et al.
13.4
SEICHE AS A COMBINATION OF FREE AND FORCED OSCILLATIONS
Next, the concept of regarding seiches as a combination of free and forced oscillations will be developed. Any natural system, when displaced from its equilibrium position, will try to regain its equilibrium position (due to a restoring force) and will exhibit free oscillations once the disturbing force is removed. The nature of these oscillations depends on the system alone, the influence of the disturbing force being restricted to setting the initial amplitude of the oscillation. After some time, the free oscillations will gradually dissipate. In a water body or basin, the seiche is a type of free oscillation of the water, the restoring force being gravity. However, in nature the seiches could be of a forced nature because the disturbing force, instead of being instantaneous, can act over some period of time. The equation of motion for a linear vibrating mass spring system subject to a displacement X due to a disturbing force F(t), in the canonical form is: F(t) X¨ + 2βX˙ + ω2 X = m
(13.7)
where β is a non-dimensional damping coefficient, m is the mass of the vibrating body, ω is the angular frequency and mω2 is a spring constant for the restoring force. The solution of equation (13.7) can be visualized as the combination of a free and forced part of a transient and steady-state part. To obtain the solution for the free oscillation, put F(t) = 0. Then: X0 = e−βωt [a sin(γt) + b cos(γt)]
(13.8)
where a and b are amplitudes of the motion determined by the initial conditions. The natural frequency γ of the system is: γ = ω(1 − β2 )1/2
(13.9)
and the natural period T is given by: T =
2π γ
(13.10)
The frictional damping, which is given by β, makes the free oscillations decay at a rate such that the amplitude decreases in one cycle by e−δ where δ is the logarithmic decrement and is given by: δ = βωT
13.5
(13.11)
CONTRIBUTION FROM THE FORCED SOLUTION
For the forced solution, one must use equation (13.7) in complete form and take a periodic disturbing force as follows: F(t) = F cos(σt + ε) m
(13.12)
Normal modes and tsunami coastal effects
147
where ε is an arbitrary phase angle. Then the forced solution is: Xf =
Fµ cos(σt + ε − α) ω2
(13.13)
where:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
σ 2 2
µ=
1−
tan σ =
2β(σ/ω) 1 − (σ/ω)2
ω
σ 2 + 2β ω
−1/2 (13.14)
(13.15)
Here, µ is the dynamic amplification of the oscillation and α is a phase angle by which the forced oscillation lags the disturbing force. Thus, the total solution is: X = X0 + Xf
(13.16)
Here, X0 decays with time whereas Xf persists as long as the disturbing force is applied. One can deduce that: µ=
Xmax Xmax = (F/ω)2 Xi
(13.17)
where Xi is the amplitude of the input displacement. Ordinarily, µ and α are shown as ordinates versus σ/ω as abscissa. If the damping coefficient β < 12 , then from equation (13.9), the natural frequency γ of the system is approximately given by ω. Hence, the ratio σ/ω in equations (13.14) and (13.15) is effectively the ratio of the forced to the natural frequency. The dynamic amplification µ approaches its peak value when σ/ω ∼ 1. When this happens, resonance occurs and the amplitude of motion will be several times greater than the amplitude of the disturbing force. For small frequency ratios σ/ω 1, the magnification is small, µ ∼ 1 and the motion follows the excitation (i.e., α → 0). For σ/ω 1, the resulting motion is much smaller than that of the exciting force. Then, µ → 0 and the motion tends to become out of phase (i.e., α → 180◦ ). Hence, the degree of resonance is determined by the damping factor 2β. Miles and Munk (1961) defined the degree of resonance through the factor Q, which is the maximum value of the dynamic amplification µ. From equation (13.14), if σ/ω ∼ 1 (as occurs at resonance): µmax ≡ Q ≡
1 2β
(13.18)
In the frequency range (1 − β) < σ/ω < (1 + β), if damping factor 2β is small, the power amplification µ2 has a value greater than Q2 /2. Hence, the frequency band width (over which the power amplification exceeds half its maximum value Q2 ) is 1/Q. Thus, the sharper is the resonance, the narrower will be the spectral energy peak. This can be quantitatively expressed by stating that near resonance: Q2 σ 2 2 ∼ 1 + 4Q 1 − µ2 ω
(13.19)
148
N. Nirupama et al.
The following results can be easily deduced from the above relations. For low Q conditions (i.e., heavy dissipation), a large rate of absorption of energy from the disturbing force to the oscillating system is necessary whereas for high Q (small damping) only a small energy absorption rate is sufficient for resonance. Miles and Munk (1961) showed that for a water body with a rather regular topography, low damping prevails. Hence, the response is of the high Q type. Hence, a relatively small amount of energy (e.g., from atmospheric pressure gradients) at the correct damping is heavy and a low Q situation prevails.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
13.6 THEORETICAL ASPECTS OF FREE AND FORCED SEICHES Next, some theoretical aspects of free and forced seiches will be considered. With reference to a Cartesian coordinate system (x, y), let Mx and My be the components of the transport, ζ is the water level deviation from its equilibrium position, H be the water depth, Pa be the atmospheric pressure, τxs and τys be the wind stress components and r be the bottom friction coefficient. Omitting the Coriolis term and assuming that ∂y∂ = 0, and v = 0: ∂M ∂ζ + rM + g(H + ζ) = Fs(x, t) ∂t ∂x
(13.20)
where subscript x on M is omitted. The external force is given by: Fs(x, t) =
rxs H + ζ ∂Pa + ρ ρ ∂x
(13.21)
The continuity equation is: ∂ζ ∂M + =0 ∂t ∂x
(13.22)
Equation (13.20) and continuity equation can be transformed into two hyperbolic equations in the dependent variables M and ζ:
∂2 ζ ∂ζ ∂ ∂ζ ∂Fs + r − g (H + ζ) =− 2 ∂t ∂t ∂x ∂x ∂x
(13.23a)
∂2 M ∂Fs ∂M ∂2 M +r − g(H + ζ) 2 = ∂t ∂t ∂x ∂t
(13.23b)
The solutions of equation (13.23) with the right-hand sides set to zero give the solutions for the free oscillation, whereas the solutions for the complete equations are the forced oscillations. In a rectangular basin of length L and uniform depth H , in which a free oscillation is generated by equating the disturbing force Fs to zero, equations (13.22) and (13.23) give: 2 ∂2 ζ ∂ζ 2∂ ζ + r =0 − c ∂t 2 ∂t ∂x2
(13.24)
∂2 M ∂M ∂2 M +r − c2 2 = 0 2 ∂t ∂t ∂x
(13.25)
Normal modes and tsunami coastal effects
149
Since these equations have the same form in ζ and M , one can use the method of separation of variables to solve them: ζ (or M ) = X (x)T (t)
(13.26)
The solution can be shown to be: ζ (or M ) = e−rt/2 [A cos(kx) + B sin(kx)][C cos(γt) + D sin(γt)]
(13.27)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
where the angular frequency γ of the free oscillation is given by:
k γ =ω 1− 2ω
1/2
(13.28)
The wave number k and the angular frequency ω are related through: ω = kc
(13.29)
in which either k or ω must be determined. To determine the constants of integration A, B, C, D, the following boundary conditions must be used. At the ends of the basin, x = 0, L transport M must be zero for all time. Thus, M (x = 0) = 0
and
M (x = L) = 0
(13.30)
From the continuity equation and taking a as the amplitude of free oscillation, ζ ∼ ae−rt/2 cos (kx) cos(γt + ε) M∼
(13.31)
aγ −rt/2 sin(kx) sin(γt + ε) e k
The wave number k can be determined from the second equation of (13.30) to give kL = nπ
n = 1, 2, 3, . . .
(13.32)
Then γ and ω can be determined from equations (13.28) and (13.29). Since equation (13.31) represents a standing wave whose amplitude is a at t = 0 and decays exponentially with time, this oscillation is similar to the mechanical system discussed earlier. Thus, r = 2βω
(13.33)
In real applications, the disturbing force should be explicitly introduced. In the case of tsunamis, this disturbing external force could be the tsunami amplitude in shallow water that could excite edge waves. 13.7
SUMMARY
The normal modes are free oscillations of coastal water bodies, which play an important role in the coastal behavior of tsunamis. Not only these oscillations contribute to the transformation of the tsunami waves in the coastal region, but also they play a very dominant role in the secondary undulation that persists in the coastal water bodies up to several days after the main tsunami
150
N. Nirupama et al.
activity has died down. Here we briefly reviewed some classical concepts of these normal modes and coastal seiches in terms of the so-called OFC and OSC.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
REFERENCES Bjerknes, V., Bjerknes, J., Solberg, H., and Bergeron, T. (1934). Hydrodynamic Physique, Vol. II, Cap XI. Les Presses Universitatires de France, Paris, pp. 457–491. Chrystal, G. (1905). On the Hydrodynamical theory of Seiches (with bibliography on seiches). Trans. Roy. Soc. Edin., 41(3), 599–649. Defant, A. (1961). Physical Oceanography. Pergamon Press, NY, 598 pp. Forel, F.A. (1892). Le Leman (collected papers). Two Volumes, Rouge, Lausanne, Switzerland. Hough, S.S. (1898). On the application of harmonic analysis to the dynamical theory of tides, Part II, On the general integration of Laplace’s dynamical equations. Philos. Trans. Roy. Soc. A, 191, 138–185. Kelvin, Lord W. (1879). On gravitational oscillations of rotating water. Proc. Roy. Soc. Edin., 10, 92–109, Papers, 4, 141–148. Kowalik, Z. and Murty, T.S. (1993). Numerical Modeling of Ocean Dynamics World Scientific Publishers, Singapore, 481 pp. Kowalik, Z., Knight, W., Logan, T., and Whitmore, P. (2005a). Numerical modeling of the global tsunami: Indonesian tsunami of 26 December 2004. Sci. Tsunami Hazards, 23(1), 40–56. Kowalik, Z., Knight, W., Logan, T., and Whitmore, P. (2005b). TheTsunami of 26 December 2004: numerical modeling and energy considerations. In: G.A. Papadopoulos and K. Satake (eds.), Proceedings of the International Tsunami Symposium, Chania, Greece, June 27–29, pp. 140–150. Lagrange, J.L. (1781). Memoire sur la Théorie du Mouvement des Fluides. Nouv. Mem. Acad. R. Berli, Qeuvres 4. LaPlace, P.S. (1775, 1776). Reserches sur plusiers points du systeme du monde. Mem. Acad. Roy. Sci., 88, 75–182; 89, 177–267. Merian, J.R. (1828). Uber die Bewegung. Tropfbarer Flussigkeiten in Gebässen (Basle). Miles, J.W. (1974). Harbor seiching. Ann. Rev. Fluid Mech., 6, 17–35. Miles, J.W. and Munk, W.H. (1961). Harbor paradox. J. Waterway, Harbors Coast. Eng. Div., Proc. ASCE, 87, 111–130. Murty, T.S., Rao, A.D., and Nirupama, N. (2005a). Inconsistencies in travel times and amplitudes of the 26 December 2004 Tsunami. J. Mar. Med., 7(1), 4–11. Murty, T.S., Nirupama, N., Nistor, I., and Rao, A.D. (2005b). Conceptual differences between the Pacific, Atlantic and Arctic tsunami warning systems for Canada. Sci. Tsunami Hazards, 23(3), 39–51. Murty, T.S., Nirupama, N., and Rao, A.D. (2005c). Why the earthquakes of 26th December 2004 and the 27th March 2005 differed so drastically in their tsunami-genic Potential. Newslett. Voice Pac., 21(2), 2–4. Murty, T.S., Nirupama, N., Nistor, I., and Hamdi, S. (2006a). Far field characteristics of the tsunami of 26 December 2004. ISET J. Earthq. Techno., 42(4), 213–217. Murty, T.S., Nirupama, N., Nistor I., and Hamdi, S. (2006b). Why the Atlantic generally cannot generate trans-oceanic tsunamis. ISET J. Earthq. Technol., 42(4), 227–236. Murty, T.S., Rao, A.D., Nirupama, N., and Nistor, I. (2006c). Numerical modelling concepts for the tsunami warning systems. Curr. Sci. 90(8), 1073–1081. Nirupama, N., Murty, T.S., Rao, A.D., and Nistor, I. (2005). Numerical Tsunami Models for the Indian Ocean Countries and States, Indian Ocean Survey, 2(1), 1–14. Nirupama, N., Murty, T.S., Nistor, I., and Rao, A.D. (2006). The energetics of the tsunami of 26 December 2004 in the Indian Ocean: a brief review. Mar. Geod., 29(1), 39–48. Proudman, J. (1953). Dynamical Oceanography. London, Methuen, J. Willey, 409 pp. Wilson, B.W. (1972). Seiches. Adv. Hydrosci., 8, 1–94.
CHAPTER 14
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Helmholtz Mode and K–S–P Waves: Application to Tsunamis N. Nirupama Emergency Management, Atkinson School of Administrative Studies, York University, Toronto, Canada T.S. Murty and I. Nistor Department of Civil Engineering, University of Ottawa, Ottawa, Canada A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology, New Delhi, India
14.1
INTRODUCTION
During the Indian Ocean Tsunami of 26 December 2004, high water levels persisted in certain harbors (for example in the state of Kerala in India) in the Indian Ocean, even after the main tsunami was dissipated. For a detailed analysis of various aspects of this tsunami, see Murty et al. (2005a–c), Murty et al. (2006a–c), Nirupama et al. (2005) and Nirupama et al. (2006). The reason for this persistence is the so-called Helmholtz mode of resonance, in which the long gravity wave energy of the tsunami enters a wide harbor through a narrow entrance channel. The energy, once entered the harbor cannot easily get out of the harbor because each successive reflection from the harbor walls only leak out a small amount of energy. Another type of wave motion, referred to as K–S–P (Kelvin–Sverdrup–Poincaré) waves are generally invoked to account for the coastal behavior of long gravity waves, such as, tides, storm surges, and tsunamis. Specifically, the role of these waves is to explain the propagation near the coastlines of tides and tsunamis and the variation of their amplitudes from the coast in an offshore direction. Here we will explore some of the classical concepts about these waves as well as the Helmholtz mode, following Kowalik and Murty (1993).
14.2
HELMHOLTZ MODE – ACOUSTIC ANALOGY
Following mechanical and acoustical analogy, the so-called Helmholtz mode will be defined and then a hydrodynamic explanation invoked. With reference to Figure 14.1, the equation of motion for the mechanical system shown can be written as (Raichlen, 1966): M x¨ + c˙x + kx = F0 cos(ωf t)
(14.1)
where M is the mass of the oscillating body, c is a linear damping coefficient, k is a spring constant, ωf is a (circular) forcing frequency, and dots denote differentiation with respect to t. 151
N. Nirupama et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
152
Figure 14.1.
Resonance characteristics of a system with a single degree of freedom (Raichlen, 1966).
The following steady-state solution can be assumed: x = X cos(ωf t − φ)
(14.2)
where X is the maximum displacement and φ is a phase angle between the input and output functions. The parameters X and φ can be made non-dimensional as follows: X = Xst
1
2 2 1/2 ωf ωf 2 1 − ωn + 2ζ ωn
tan φ =
2ζωf /ωn 1 − (ωf /ωn )2
(14.3)
(14.4)
where F0 k 1/2 k ωn ≡ M c ζ≡ 2M ωn
Xxt ≡
(14.5)
Figure 14.1 represents graphically equations (14.3) and (14.4). First, consider the behavior of X /Xxt . For the case of small frictional dissipation, when the frequency is approximately equal to the undamped natural frequency of the system, the forcing function Xst is greatly amplified. As the damping ζ gets bigger, the difference between the resonant frequency and ωn increases.
Helmholtz mode and K–S–P waves: application to tsunamis
153
For low values of the frequency ratio, the amplitudes of the input and the output are approximately equal. However, for frequencies considerably above the resonant frequency, the response decreases substantially and the maximum displacement of the mass approaches zero. If the damping is zero, equation (14.3) gives infinite amplitude at resonance. However, this result, which is obtained from the linear theory, must be modified at great amplitudes to include the influence of the non-linear effects.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
14.3
DEGREES OF FREEDOM AND DEPENDENCY OF PHASE ANGLE ON FORCING FREQUENCY
Next, consider the behavior of the phase angle with respect to the forcing frequency. For low values of the frequency ratio, the forcing function is mainly in phase with the output displacement, and at higher values they are 180◦ out of phase. At resonance, the phase angle becomes 90◦ and hence, the Force F0 cos(ωf t) is in phase with the velocity x¨ . Thus, when the mass is going through its zero displacement position, a maximum force is impressed upon the system. The number of degrees of freedom of a system is the number of independent coordinates that are required to describe the motion of the system. Raichlen (1966) cited the vibration of a clamped circular membrane as an example of a system possessing infinite degrees of freedom, whereas the spring–mass–dashpot system considered here is an example of a system with a single degree of freedom. In acoustics, an example of a single degree of freedom system is the so-called Helmholtz resonator, which consists of a cavity of volume V connected to a tube of length l and area of cross-section A. The equation of motion for this system is: M x¨ + ra x˙ +
x = P cos(ωf t) B
(14.6)
where x is the volume displacement, c is the wave velocity, ra is the radiation loss coefficient, and: M≡
ρl A
B≡
V ρc2
(14.7)
The natural frequency of the Helmholtz resonator is given by: ωn ≡ c
A lV
(14.8)
Since equations (14.6) and (14.1) are similar, it can be seen that when the frequency is equal to ωn , the ratio of the volume displacement to the applied pressure will be ∞, when ra is zero. Also, the ratio of the volume displacement to the applied pressure varies, as shown in Figure 14.1. 14.4
HELMHOLTZ MODE IN THE CONTEXT OF HYDRODYNAMICS
Next, consider the Helmholtz mode in the context of hydrodynamics. Miles (1971) used the term “Helmholtz mode”, Platzman (1972) used “co-oscillating mode”, and Lee and Raichlen (1972)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
154
N. Nirupama et al.
referred to it as the “pumping mode”. Basically, Helmholtz resonance represents the balance between the kinetic energy of water flowing in through a narrow connecting channel and the potential energy from the rise in the mean water level within the harbor (Freeman et al., 1974). It is an additional gravitational mode of a substantially longer period than the fundamental free oscillation, as can be seen below. To conceptualize the Helmholtz mode, Platzman (1972) presented the following argument. Suppose that at the mouth of a rectangular bay an adjustable barrier exists and that this barrier is gradually moved from the two sides of the bay to the center, completely closing off the bay. The open modes with periods initially of the form 2T /(2n − 1), n = 1, 2, 3, . . . , will be transformed continuously into the closed mode periods of T /n, n = 0, 1, 2, . . . . It is obvious that the fundamental mode for the open bay transforms into the zeroth mode the closed basin, and as the barrier closes, this period approaches ∞. For small openings, the period of the Helmholtz mode is less than ∞ but greater than the period for a completely open bay. Platzman (1972) showed that rotation changes the period of the Helmholtz mode by, at most, 3%. The classic theory for the Helmholtz mode can be applied only to a single channel harbor. Freeman et al. (1974) extended this to a harbor (or basin) with multiple channels for openings. The dissipative forces (due to the eddy viscosity of the fluid and to the energy radiated from the mouth) are ignored. These forces affect the amplification factor at resonance and will shift the resonant frequency slightly. The solution developed by Freeman et al. (1974) for the frequency ω0 is: n g Si 1/2 ω0 = A i=1 Li
rad/s
(14.9)
where g is gravity, A is the surface area of the harbor, Si is the cross-sectional area of the ith channel, and Li is the length of the ith channel. Miles and Munk (1961) introduced the so-called harbor paradox in which they showed that narrowing of a harbor mouth (relative to the other dimensions) diminishes the protection from seiching. For a quantitative estimation of this in terms of the sharpness or Q at resonance, the reader is referred to their paper. Miles and Lee (1975) used equivalent electric circuit analysis to study this problem. Garrett (1970) showed that the harbor paradox, as originally postulated by Miles and Munk (1961), only holds for the Helmholtz mode.
14.5
KELVIN WAVES, SVERDRUP WAVES, AND POINCARÉ WAVES
There are classes of normal mode solutions with special properties that have been referred to as Kelvin waves, Sverdrup waves, and Poincaré waves. These wave types have been frequently invoked to explain the tidal phenomena in water bodies. For an excellent review on this topic, see Platzman (1971). Other relevant works are Defant (1961), Proudman (1953), Voyt (1974), and LeBlond and Mysak (1978). Simons (1980) has given a rather concise summary, and this discussion will essentially follow his line of argument. Earlier, the gravitational and rotational modes were introduced, also referred to as oscillations of the first class (OFC) and oscillations of the second class (OSC). It was also pointed out that OFC are motions with large divergence whereas OSC are essentially non-divergent. For introducing the concepts of different types of wave motion mentioned here, discussion begins with the linearized version of the vertically integrated equations, and these will be applied to a
Helmholtz mode and K–S–P waves: application to tsunamis
155
rectangular basin of uniform depth:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
∂U ∂ζ − fV = −c2 ∂t ∂x ∂V ∂ζ + fU = −c2 ∂t ∂y ∂ζ ∂U ∂V + + =0 ∂t ∂x ∂y
(14.10)
where U and V are the x and y components of the volume transport, ζ is the deviation of the water level from its equilibrium position, f is the Coriolis parameter, and c2 = gH where H is the uniform water depth. Assuming a time factor eiσt where σ is the frequency, and eliminating U and V from equation (14.10) gives the wave equation: (σ 2 − f 2 )ζ + c2 ∇ 2 ζ = 0
(14.11)
The boundary condition of zero normal transport to a boundary can be stated as: f
∂ζ ∂ζ + iσ =0 ∂S ∂n
(14.12)
where S and n are the coordinates along and perpendicular, respectively, to the wall. Equation (14.11) can be satisfied by: exp[i(kx + ly)] where the frequency σ is given by: σ 2 = f 2 + c2 (k 2 + l 2 )
(14.13)
For the non-rotating case, equation (14.12) is easily satisfied by standing waves with wave numbers k = mπ/L and l = nπ/B, where m and n are the integers and L and B are the length and breadth of the basin, respectively. However, in the rotating case, because of the compacted nature of the boundary condition (14.12), it is difficult to determine the normal modes. However, even in the rotating case, for an infinitely long channel, there are some elementary wave solutions that do not satisfy the boundary conditions. In equation (14.10), if V = 0, then the solutions to the resulting equations are: ζ = ζ0 e−fy/c eik(x−ct) U =−
c2 ∂ζ f ∂y
(14.14)
V =0 The wave speed is the same for the non-rotating and the rotating cases, but in the latter, the wave amplitude decreases exponentially from right to left for an observer looking in the direction of wave propagation. The rate of decrease of the amplitude from right to left is proportional to c/f , the Rossby radius of deformation. These waves are known as Kelvin waves. Another elementary
156
N. Nirupama et al.
solution can be obtained by setting ∂/∂x and ∂/∂y to zero. This will result in inertial oscillations with the frequency f (note that ζ is zero for these and there is only horizontal motion). If only the gradients in one horizontal direction are ignored, e.g., ∂/∂x = 0, then the solutions are: ζ = ζ0 ei(ly−σt)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
σ = f 2 + c2 l 2
(14.15)
These waves are referred to as Sverdrup waves and have horizontal crests. For a straight coast parallel to the x axis, two Sverdrup waves traveling in opposite directions may be combined to form a standing wave that satisfies the boundary condition (14.12). Thus standing Sverdrup waves with wave numbers l = nπ/β are the normal modes of an infinitely long rotating channel. The more general solutions of equation (14.11) are known as Poincaré waves. Pairs of progressive Poincaré waves can be combined into standing waves that display cellular patterns. For an infinite channel, in analogy with Sverdrup waves, Poincaré waves can be made to satisfy the boundary conditions by properly choosing the transverse wave numbers. At a transverse barrier in the channel, none of these waves could be made to satisfy the boundary condition (14.12). Hence, the analytical determination of the normal modes of a rotating rectangular bay is more difficult, and it is convenient to resort to numerical techniques. 14.6
SUMMARY
In certain tsunami events, it is known that the water levels in certain harbors persists for a long time even after the main tsunami waves have dissipated. The reason for this is the so-called Helmholtz resonance in which the long gravity wave energy of the tsunami enters the wide harbor through a narrow entrance channel, but cannot leave the harbor easily. Successive reflections of this energy from the harbor walls provide only a slow leak to outside the harbor region. The propagation of the tsunami (and tides) along coastlines and the variations of their amplitudes perpendicular to the coastlines can be explained through the so-called K–S–P waves. Here some of the classical concepts on the Helmholtz mode as well as the K–S–P waves have been briefly reviewed. REFERENCES Defant, A. (1961). Physical Oceanography. Pergamon Press, New York, 598 pp. Freeman, N.G., Hamblin, P.F., and Murty, T.S. (1974). Helmholtz resonance in harbours of the great lakes, Proceedings of 17th Conference on Great Lakes Research International Association, Great Lakes Res. Proc. 15, 399–411. Garrett, C.J.R. (1970). Bottomless harbors. J. Fluid Mech., 43, 433–449. Kowalik, Z. and Murty, T.S. (1993). Numerical Modeling of Ocean Dynamics. World Scientific Publishers, Singapore, 481 pp. LeBlond, P.H. and Mysak, L.A. (1978). Waves in the Ocean. Elsevier Oceanographic Series 20, Amsterdam, 602 pp. Lee, J.J. and F. Raichlen (1972). Oscillations in harbours with connected basins, J. Waterway, harbours and coastal engineering division, Proceedings of the ASCE, 98, 311–332. Miles, J.W. (1971). Resonant response of harbors: an equivalent circuit analysis. J. Fluid Mech., 46, 241–265. Miles, J.W. and Lee, Y.K. (1975). Helmholtz resonance of harbors. J. Fluid Mech., 67, 445–464. Miles, J.W. and Munk, W.H. (1961). Harbor Paradox. J. Waterway, Harbors Coast. Eng. Div., Proc. ASCE, 87, 111–130.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Helmholtz mode and K–S–P waves: application to tsunamis
157
Murty, T.S., Nirupama, N., and Rao, A.D. (2005a). Why the earthquakes of 26th December 2004 and the 27th March 2005 differed so drastically in their tsunami-genic potential, Newsletter. Voice of the Pacific, 21(2), 2–4. Murty, T.S., Rao, A.D., and Nirupama, N. (2005b). Inconsistencies in travel times and amplitudes of the 26 December 2004 tsunami. J. Mar. Medi., 7(1), 4–11. Murty, T.S., Nirupama, N., Nistor, I., and Rao, A.D. (2005c). Conceptual differences between the Pacific, Atlantic and Arctic tsunami warning systems for Canada. Sci. Tsunami Hazards, 23(3), 39–51. Murty, T.S., Nirupama, N., Nistor, I., and Hamdi, S. (2006a). Far field characteristics of the tsunami of 26 December 2004. ISET J. Earthq. Technol., 42(4), 213–217. Murty, T.S., Nirupama, N., Nistor, I., and Hamdi, S. (2006b). Why the Atlantic generally cannot generate trans-oceanic tsunamis. ISET J. Earthq. Technol., 42(4), 227–236. Murty, T.S., Rao, A.D., Nirupama, N., and Nistor, I. (2006c). Numerical modelling concepts for the tsunami warning systems. Curr. Sci. 90(8), 1073–1081. Nirupama, N., Murty, T.S., Rao, A.D., and Nistor, I. (2005). Numerical tsunami models for the Indian Ocean countries and states, Indian Ocean Survey, 2(1), 1–14. Nirupama, N., Murty, T.S., Nistor, I., and Rao, A.D. (2006). The energetics of the tsunami of 26 December 2004 in the Indian Ocean: a brief review. Mar. Geod., 29(1), 39–48. Platzman, G.W. (1971). Ocean tides and related waves, 239–291. In: W.H. Reid (ed.), Mathematical Problems in the Geophysical Sciences. AMS, Providence, Rhode Island. Platzman, G.W. (1972). Two-dimensional free oscillations in natural basins. J. Phys. Oceanogr., 2(2), 117–138. Proudman, J. (1953). Dynamical Oceanography. Methuen, J. Willey, London, 409 pp. Raichlen, F. (1966). Long period oscillations in basins of arbitrary shapes. In: Coastal Engineering Speciality Conf., Santa Barbara, CA, USA, 1965, ASCE, Part I. pp. 115–145. Simons, T.J. (1980). Circulation models of lakes and inland areas. Can. Bull. Fish. Aquat. Sci., Bull., 203, Ottawa, 146 pp. Voyt, S.S. (1974). Long waves and tides. In: P.S. Lineykin (ed.), ITOGI. Summaries of Scientific Progress. Oceanology, 2. G.K. Hall and Co., Boston, pp. 33–51.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 15
Numerical Models for the Indian Ocean Tsunami of 26 December 2004: A Brief Review
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
P. Chittibabu Baird & Associates, Ottawa, Canada T.S. Murty Department of Civil Engineering, University of Ottawa, Ottawa, Canada
15.1
INTRODUCTION
The devastating tsunami of 26 December 2004 in the Indian Ocean prompted many numerical modeling groups around the world to either develop or adapt their existing models to simulate this event. Following the Indian Ocean Tsunami of 26 December 2004, several numerical models for this tsunami, either were posted on the world wide web and or appeared in the literature. This is an attempt to summarize briefly the parameters and important results of these models. The order in which we arranged the review, has no particular significance, except to suggest that, that was the order in which we found them. Our cut-off date for this review is 1 November 2005, in the sense that any models that appeared after this date were not included this review. It is also quite probable that, there were some other models that appeared in the literature, but some how escaped our attention. Finally it should be pointed out that all these models mainly dealt with the tsunami generation and propagation aspects, and whatever results were included on the inundation aspects, are mostly from simple runup algorithms, rather than from exhaustive and detailed coastal inundation models. This is not a critical review of the models; it is a brief presentation of relevant information about the models for this tsunami event. The following models will be reviewed: 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Tsunami N2 Method of Splitting Tsunami Model TOAST Model Delft Model Italy Model E.C.J.R.C Model University of Frankfurt, Geophysical Institute Model NIO – India Model DCRC – TOHOKU Model AIST (SATAKE) Model Wakayama National College Model Model of Shunichi Koshimura Baird Model Kowalik et al. Model.
159
160
P. Chittibabu and T.S. Murty
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
15.2 TSUNAMI N2 MODEL The software for the TSUNAMI N2 model originally was developed by Fumihiko Imamura in Tohoku University, Japan and later modifications were done further development carried at Middle East Technical University by Ahmet Cevdet Yalçiner (2005) and his group and in the University of Southern California by Costas Synolakis. It is an outcome of IOC (Inter Governmental Oceanographic Commission of UNESCO) Tsunami Inundation Modeling Experiment (TIME) project. The generation and propagation of December 2004 Indian Ocean Tsunami was carried out using this model. Table 15.1. lists the fault parameters, based upon which the initial tsunami wave generation was simulated. The model also computed the maximum tsunami amplitudes and travel times to various locations, which are presented in Figure 15.1. Table 15.1. The fault data used to compute the tsunami source for simulation. Epicenter eastern coordinate Epicenter northern coordinate Fault length Fault width Strike angle Dip angle Slip angle Displacement Focal depth Maximum +ve amplitude at tsunami source Maximum −ve amplitude at tsunami source
93.13◦ N 03.70◦ E 443 km 170 km 329◦ 8◦ 110◦ 30 25 km +10.7 m −6.6 m
Figure 15.1. Tsunami travel times in hours (annotation “s” does not stand for second) Yalciner (2005).
Numerical models for the Indian Ocean Tsunami 2004
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
15.3
161
MOST MODEL
This model was developed at the Pacific Marine Environmental Laboratory (PMEL) of National Oceanic and Atmospheric Administration (NOAA) in Seattle, USA. Details of the model are described by Titov and Synolakis (1995, 1996), Titov and Gonzalez (1997) and Titov (2005). The MOST model simulates all three stages of the tsunami namely generation, propagation and runup. The generation process is based upon the elastic deformation theory (Gusiakov, 1978; Okada, 1985), which assumes an incompressible liquid layer on an underlying elastic half space, to characterize the ocean and the earth’s crust. The elastic fault plane model contains a formula for static sea floor displacement to compute the initial conditions required for further simulations of tsunami propagation and coastal inundation. Since tsunamis propagate over long distances, earth’s curvature must be taken into account. Hence the momentum and continuity equations are written in spherical polar coordinates (Murty, 1984). The other parameters that should be included are Coriolis force and dispersion. Because tsunami waves with different frequencies propagate with slightly different speeds, the shape of the wave changes due to dispersion. However, explicit inclusion of the dispersion terms makes the equations too complex. As a simple alternative, Shuto (1991) suggested that physical dispersion can be simulated through numerical dispersion, which is present in finite difference algorithms. The most model makes use of non-linear shallow water equations in spherical polar coordinates (Murty, 1984) and uses numerical dispersion and equations are solved using splitting method. Model simulation results are presented in Figures 15.2–15.4. Figure 15.2 shows amplitude and tsunami travel times for global ocean and Figure 15.3 shows same for Indian Ocean region. Figure 15.4 compares the observed and computed travel times. As can be seen, the agreement is good. Figure 15.5 shows the arrival time of first wave in the Indian Ocean and Figure 15.6 shows the observed versus model wave arrival time.
Figure 15.2.
Maximum wave amplitude for the Global Ocean.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
162
P. Chittibabu and T.S. Murty
Figure 15.3. Arrival time of first wave of the tsunami for the Global Ocean.
Figure 15.4.
Maximum tsunami wave height (cm) in the Indian Ocean.
15.4 TOAST MODEL This model was combination of different models developed under Advance Ocean State Forecast activity at MOG/SAC (Agarwal et al., 2005). The model is originally designed for ocean general circulation and also for the prediction of tides and storm surge coupled with a cyclone prediction model. Flexible grids in the model also allows to simulate coastal inundation due to storm surges. TOAST’s ocean general circulation model has separate barotropic and baroclinic modes of energy propagation. On 26 December 2004, Indian Ocean Tsunami was simulated by adapting this model
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Numerical models for the Indian Ocean Tsunami 2004
Figure 15.5. Arrival time of first wave in the Indian Ocean.
Figure 15.6.
Observed versus model wave arrival times.
163
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
164
P. Chittibabu and T.S. Murty
Figure 15.7.
Computed and observed inundation in Banda Aceh Indonesia from http://www.wldelft. nl/cons/area/ehy/flood/tsunami.html.
for tsunami generation and propagation. For tsunami generation elastic plate movement model was used. Agarwal et al. assumed the intensity of the quake as 9.0 Ritcher scale and also compared the simulation results with Jason altimeter pass in the central Bay of Bengal at 0256 UTC. According to them the model is working well and results are encouraging. 15.5
DELFT HYDRAULICS MODEL
There are two versions – a three-dimensional (3D) and a two-dimensional (2D). These models are mainly applied for simulation of tides and storm surges. In the 3D version, the vertical grid follows sigma coordinates and in the 2D version, a curvilinear boundary fitted grid is used. For the tsunami version, the computation started an initial disturbance of 650 km in length along the coast of Banda Aceh, Sumatra, Indonesia. The maximum tsunami amplitude in the open ocean is about 1 m. The computed tsunami travel times are some what less than the observed travel times. According to Delft, a high-resolution nested model of Aceh is applied to simulate the flooding. For the flooding simulations the “flood” scheme, which was recently incorporated in the Delft 3D system, has been applied. Figure 15.7 shows the computed flooded area in the coastal region of Banda Aceh. The satellite picture taken by IKONOS satellite show that the model has reproduced the flooded areas extremely well.
Numerical models for the Indian Ocean Tsunami 2004 15.6
165
ITALY MODEL
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
This model belongs to the Istituto Nazionale Geofisica e Vulcanologia. It is a 2D shallow water code in a finite difference frame work and used the Okada (1985) equations for the fault (slip amplitude of 20 m and dip angle of 15) plane model for tsunami generation with rupture dimensions of 700 km length and 100 km in width. The simulation produced some interesting results with the 2 min by 2 min grid. Seven hours of propagation took about 6 h of CPU time (on a laptop with a 1.7 GHz processor). The initial disturbance towards Thailand is depression or the trough of a wave. A tsunami wave reflected from the east coast of Sri Lanka returns towards the epi central area. The computed travel time of the tsunami to Maldives is slightly greater than the observed travel time. 15.7
EUROPEAN COMMISSION JOINT RESEARCH CENTRE MODEL
This is a simple model (Annunziato and Best, 2005) and was used to compute the travel times of the tsunami under the shallow water approximation. The grid used was based upon the ETOPO5 ocean bathymetry with a 5-min grid resolution. The main objective of this model is to predict travel times quickly so that early warnings for tsunami risk can be provided in the Earthquake alert system of GDAS. Diffraction and hydro dynamics are not included in the model. 15.8
UNIVERSITY OF FRANKFURT/MAIN, GEOPHYSICAL INSTITUTE MODEL
It is a non-linear shallow water model (Babeyko and Sobolev, 2005) in spherical coordinates with coriolis and bottom friction terms included. It is an explicit finite difference model with a grid size of 2 min for the coarse model and 30-arc seconds for a fine grid model. The time steps for these two models are 2 and 1 s respectively. Six hour simulation time is used. The ocean bathymetry used is ETOPO2. The authors, Babeyko and Sobolev remark that the model results degrade in the near shore region. According to authors, initial sea bottom displacement is calculated following Okada’s (1985) analytical solution for the surface deformation caused by deep planar fault of arbitrary size and orientation. In this scenario the fault zone follows the plate boundary between Indian and Sunda plates and is composed of 6 planar Okada’s segments with the following parameters. The rupture starts at time = 0 at the epicenter and moves northwards along the plate boundary with velocity Vr. As rupture moves along a segment, the slip at this segment linearly increases from zero up to its maximal value (here – 15 m). The total slip at a segment is being decomposed into vertical and horizontal parts depending on the segment orientation relative to the convergence direction of the Indian and Sunda plates (thick black arrow). Accordingly, the vertical bottom displacement, which is responsible for the tsunami generation, varies among the segments and is minimal at the northern segment. Current model of the seismic source corresponds to the total seismic moment of 1.2 × 1030 dyncm. 15.9
NIO MODEL
The tidal model of Unnikrishnan et al. (Proc. Ind. Acad. Sci. (Earth Planet. Sci.), volume 108, pp. 155–177, 1999) was adapted to simulate this tsunami by an initial water level elevation of 25 m along an arc encompassing the epicenter (which was at 3.4◦ N, 95.7◦ E) and the Andaman Nicobar Islands. The ETOPO5 bathymetry that was used provided a resolution of 5 min of arc,
166
P. Chittibabu and T.S. Murty
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Table 15.2.
Fault parameters as used in the DCRC model.
Fault-1
Fault-2
Fault-3
• 95.75◦ E, 2.5◦ N • (Strike, dip, slip) = (329, 15, 110) • (L,W) = (330 km, 150 km) • Dislocation = 11 m • Depth = 7.0 km
• 94.0◦ E, 5.0◦ N • (Strike, dip, slip) = (340, 15, 110) • (L,W) = (570 km, 150 km) • Dislocation = 11 m • Depth = 7.0 km
• 92.0◦ E, 10.0◦ N • (Strike, dip, slip) = (5, 15, 110) • (L,W) = (300 km, 150 km) • Dislocation = 11 m • Depth = 7.0 km
which is approximately 9 km (5 min) and the simulation carried out for 10 h, by which the tsunami spread more or less across the Indian Ocean. Refer to http://www.nio.org/jsp/tsunami.jsp 15.10
DCRC (DISASTER CONTROL RESEARCH CENTER, TOHOKU UNIVERSITY, JAPAN) TOHOKU MODEL
The model used ETOPO2 ocean bathymetry in computational domain of 30◦ N to 30◦ S and 60◦ E to 120◦ E. The fault zone was treated as made up of three different parts with the fallowing fault plane parameters as shown in Table 15.2. The simulated tsunami at Banda Aceh (Indonesia), Galle (Sri Lanka) and Chennai (India) are shown in Figure 15.8. 15.11 ADVANCED INDUSTRIAL SCIENCE AND TECHNOLOGY (AIST) MODEL SATAKE (2005) It was assumed that the aftershock area represents the tsunami source and a fault of 1200 km in length was used for the simulation. The tsunami reached Phuket (Thailand) and Sri Lanka in about 2 h and the coast of Africa in 8 to 11 h. East of the epicenter, the tsunami started as a receding wave, whereas as to the west, the initial tsunami wave form was a crest. 15.12 WAKAYAMA NATIONAL COLLEGE OF TECHNOLOGY MODEL Simulation of the tsunami model was done using a linear shallow water model in spherical polar coordinates with a grid size of 2 min arc by Nobuaki Koike (2005) Department of Civil and Environmental Engineering. Tsunami generation was modeled by moving a basic fault model (Yamanaka, 2004). The directivity of the tsunami energy can be seen clearly in Figure 15.9. 15.13
KOSHIMURA MODEL (2005)
Numerical modeling of tsunami was performed using TUNAMI, code of the Disaster Control Research Center, Tohoku University by Koshimura of Disaster Reduction and Human Renovation Inst., Japan. The model is based on the linear shallow water theory of spherical co-ordinate system. Seismic deformation modeling is based on the theory of Okada (1985). The computational grid size is approximately 2 and 5 min. Following model cases are mainly based on the revised CMT solution by Harvard University. Initial model parameters were given in Table 15.3.
Numerical models for the Indian Ocean Tsunami 2004 9.00
[m]
167
Banda Aceh
6.00 3.00 0.00 3.00 6.00 9.00
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
0 3.00
100
200
300
400
[min]
[m] Galle
2.00 1.00 0.00 1.00 2.00 3.00 0
100
200
300
[m]
3.00
400
[min]
Madras
2.00 1.00 0.00 1.00 2.00 3.00 0
100
200
300
400
[min]
Figure 15.8. Tsunami amplitudes at Banda Aceh (Indonesia), Galle (Sri Lanka), Madras (now Chennai, India).
Model-6 1st. segment (Southern part) • • • •
(Strike, dip, slip) = (329, 15, 90) (L,W) = (500 km, 150 km) Dislocation = 11 m Depth = 10 km
2nd. segment (Northern part) • • • •
(Strike, dip, slip) = (345, 15, 90) (L,W) = (400 km, 150 km) Dislocation = 11 m Depth = 10 km
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
168
P. Chittibabu and T.S. Murty
Figure 15.9.
Maximum water elevation. Table 15.3.
Model parameters.
Governing equation Numerical scheme Spatial grid size Temporal grid size Computational domain Bathymetry data
15.14
Linear shallow water equations in spherical co-ordinate system Leap-frog FDM 2 min/5 min 5 s/10 s 2-min grid: (15N, 70E)–(15S, 120E) 5-min grid: (25N, 20E)–(75S, 120E) 2-min grid: Sandwell-Smith sea floor topography 5-min grid: NOAA ETOPO 5
BAIRD MODEL
Numerical modeling of the tsunami in the Indian Ocean was done by Baird and Associates (Ottawa, Canada) using Danish Hydraulic Institute’s M21 Hydrodynamic flow module. It is a 2D shallow water model for free-surface flows. This model was extensively used to simulate hydraulic and environmental phenomena in lakes, estuaries, bays, coastal areas and seas. The hydrodynamic module is basic model in the flow model and is used to simulate water level variations and flows driven by wind, tide and other forcings. The model includes bottom shear stress, wind shear stress, barometric pressure gradients, Coriolis force, dispersion sources and sinks, evaporation, flooding and drying. A preliminary simulation of Indian Ocean Tsunami is carried out using a rectangular grid (with 15 km resolution) covering entire Indian Ocean. Three
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Numerical models for the Indian Ocean Tsunami 2004
Figure 15.10.
169
Grids used in the model.
fine grids as shown in Figure 15.10 (with resolution of 5, 1 and 500 m) nested within the main grid to resolve the tsunami propagation around Southern India, Sri Lanka and Indonesia coasts. ETOPO2 ocean bathymetry as well as hydrographic charts were used for the grid generation. Initial sea surface elevation is based on the data provided by USGS. Figure 15.11 show computed maximum surface elevations. 15.15
KOWALIK ET AL. MODEL
Kowalik et al. (2005) developed a comprehensive model for this tsunami and we reproduced below from their paper, their abstract (which summarizes the work) their source function (Table 15.4) and their table comparing computed and observed travel times (Table 15.5). A new model for the global tsunami computation is constructed. It includes a high order of approximation for the spatial derivatives. The boundary condition at the shore line is controlled by the total depth and can be set either to runup or to the zero normal velocity. This model, with spatial resolution of 1 min, is applied to the tsunami of 26 December 2004 in the World Ocean from 80_S to 69_N. Because the computational domain includes close to 200 million grid points, a parallel version of the code was developed and run on a supercomputer. The high spatial resolution of 1 min produces very small numerical dispersion even when tsunamis wave travel over large distances. Model results for the Indonesian tsunami show that the tsunami traveled to every location of the World Ocean. In the Indian Ocean the tsunami properties are related to the source function (i.e., to the magnitude of the bottom displacement and directional properties of the source). In the Southern Ocean surrounding Antarctica, in the Pacific, and especially in the
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
170
P. Chittibabu and T.S. Murty
Figure 15.11.
Computed maximum water elevation in the Indian Ocean.
Table 15.4.
Fault parameters used to generate vertical sea floor movement.
Earthquake parameter Strike Dip Slip Length Depth (SW corner) Moment Rigidity
Southern fault segment
Northern fault segment
335 8 110 300 km 8 km 3.2 × 1029 dyn/cm 4.2 × 1011 dyn/cm2
350 8 90 700 km 8 km 7.6 × 1029 dyn/cm 4.2 × 1011 dyn/cm2
Atlantic, tsunami waves propagate over large distances by energy ducting over oceanic ridges. Tsunami energy is concentrated by long wave trapping over the oceanic ridges. Our computations show the Coriolis force plays a noticeable but secondary role in the trapping. Travel times obtained from computations as arrival of the first significant wave show a clear and consistent pattern only in the region of the high amplitude and in the simply connected domains. The tsunami traveled from Indonesia, around New Zealand, and into the Pacific Ocean. The path through the deep ocean to North America carried miniscule energy, while the stronger signal traveled a much longer distance via South Pacific ridges. The time difference between first signal and later signal strong enough to be recorded at North Pacific locations was several hours. The generation mechanism for the Indian Ocean Tsunami is mainly the static sea floor uplift caused by abrupt slip at the India/Burma plate interface. Permanent, vertical sea floor displacement is computed using the static dislocation formulae from Okada (1985). Inputs to these formulae are fault plane location, depth, strike, dip, slip, length and width as well as seismic moment and rigidity. The earthquake’s total rupture extent can be estimated by several approaches.
Numerical models for the Indian Ocean Tsunami 2004
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Table 15.5.
171
Observed and calculated travel time.
Station location
Travel time observed
Travel time for 0.1 cm amplitude
Travel time for 5 cm amplitude
Chennai (80.17E, 13.04N) Male (73.52E, 4.18N) Hanimadhoo (73.17E, 6.77N) Diego Garcia (72.40E, 7.28S) Hillarys (115.73E, 31.82S) Salalah (54.00E, 16.93N) Pt. La Rue (55.53E, 4.57S) Lamu (40.90E, 2.27S) Zanzibar (39.18E, 6.15S) Portland (141.60E, 38.33S) Richard’s Bay (32.08E, 28.80S) Port Elizabeth (25.63E, 33.97S) Jackson (168.62E, 43.98S) Arraial de Cabo (42.02W, 22.97S) Arica (70.21W, 18.22S) Char. Amalie (64.55W, 18.20N) San Diego (117.12W, 32.45N) Halifax (63.59W, 44.66N) Atl.City (74.25W, 74.25W,39.21N) Toffino (125.55W, 49.09N) Adak (176.65W, 51.87N)
2 h 36 min 3 h 25 min 3 h 41 min 3 h 55 min 6 h 41 min 7 h 17 min 7 h 25 min 9 h 9 min 9 h 49 min 10 h 39 min 11 h 13 min 12 h 28 min 18 h 18 min 21 h 56 min 26 h 36 min 28 h 42 min 31 h 25 min 31 h 30 min 31 h 48 min 32 h 1 min 35 h
2 h 18 min 3 h 12 min 3 h 24 min 3 h 40 min 6 h 24 min 7 h 6 min 7 h 24 min 8 h 30 min 10 h 24 min 9 h 48 min 11 h 00 min 12 h 00 min 12 h 30 min 20 h 54 min 26 h 6 min 27 h 45 min 29 h 0 min 30 h 6 min 30 h 45 min 29 h 0 min 27 h
2 h 20 min 3 h 18 min 3 h 30 min 3 h 40 min 6 h 36 min 7 h 6 min 7 h 24 min 8 h 30 min 10 h 36 min 10 h 18 min 11 h 12 min 12 h 6 min 19 h 30 min 21 h 30 min 29 h 20 min 33 h 30 min 35 h 30 min 32 h 6 min 33 h 30 min 38 h 30 min 40 h
Finite fault seismic data inversion is one method which yield fault lengths on the order of 350– 650 km (e.g. Ji, 2004; Yagi, 2005). Another traditional method to delineate earthquake fault zones is plotting the aftershocks which occur in the first 24 h following the main shock. The aftershocks are expected to clusterwithin the slip zone. This approach leads to an estimate of 1200 km for the fault length (NEIC, 2004). In this study, the fault extent is constrained by observed tsunami travel times to the northwest, east and south of the slip zone. Figure 15.3 displays the tsunami arrival time constraints on the fault zone. Tsunami arrival times at Paradip–India (SOI, 2005), Ko Tarutao–Thailand (Iwasaki, 2005) and Cocos Island (Merrifield et al., 2005) tide gauges are plotted in reverse. That is, the observed travel time contour is plotted with the tide gage location as the origin point. This method indicates a fault zone approximately 1000 km by 200 km. The epicenter location lies on the southern end of the fault zone. To accommodate trench curvature, the fault plane is broken into two segments. Fault parameters for the two segments are listed in Table 15.1. Strike, dip and slip are based on the definitions from Aki and Richards (1980). Strike is determined by the trench orientation. Dip is taken from the Harvard CMT solution (HRV, 2005). The slip for the southern segment is based on the Harvard CMT solution while slip for the northern segment is set at 90_based on observed tsunami first motions on Indian tide gages (NIO, 2005). Depth is based on the finite fault inversion of Ji (2004). The total moment release (derived by assuming an average slip of 13 m and rigidity of 4.2 × 1011 dyn/cm2 ) in the two segments equals 1.08 × 1030 dyn/cm (Mw = 9.3) which is in good agreement to 1.3 × 1030 dyn/cm proposed by Stein and Okal (2005) based on normal mode analysis. 15.16
CONCLUSIONS
The numerical models simulating the generation and propagation of the tsunami of 26 December 2004, in the Indian Ocean, that were posted in the Internet were briefly reviewed here. The cut-off
172
P. Chittibabu and T.S. Murty
date for this review was 1 November 2005. The order in which models were reviewed here has no particular significance, other than it is the order, which we found the models. It should be noted that this review is only meant to provide information about the models posted on the web sites and no attempt was made to compare and contrast these models. Finally, it quite possible that there were several other models, that some how escaped our attention.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
REFERENCES Agarwal, K., Vijay, N., Agarwal and Kumar, R. (2005). Simulations of the 26 December 2004 Indian Ocean Tsunami using a multi-purpose ocean disaster simulation and prediction model. Curr. Sci., 88(3). Aki, K. and Richards, P.G. (1980). Quantitative Seismology Theory and Methods Volume 2, W.H. Freeman and Co., New York, 557pp. Annunziato, A. and Best, C. (2005). E.C. JRC Tsunami Propagation Model. Posted at http://tsunami.jrc.it/model/ Babeyko, A. and Sobolev, S. (2005). A Numerical Simulation of the Indian Ocean Tsunami 26 December 2004. Posted http://www.gfzpotsdam. de/news/recent/archive/20041226/TsunamiModelling/content.html DCRC Model http://www.tsunami.civil.tohoku.ac.jp/hokusai2/topics/04sumatra/index. html Delft3D simulation package of WL. Delft Hydraulics (2005). Posted at http://www.wldelft.nl/gen/news/ tsunami/ NIO, Goa, Tsunami Simulation (2005). Posted at http://www.nio.org/jsp/tsunami.jsp Gusiakov, V.K. (1978). Static Displacement on the Surface of an Elastic Space. Ill-Posed Problems of Mathematical Physics and Interpretation of Geophysical Data, Novosibirsk, VC SOAN SSSR, 23–51 (in Russian). Italy Model (2005). Posted at http://www.ingv. it/%7eroma/reti/rms/terremoti/estero/indonesia/indonesia. htm Iwasaki, S.I. (2005). Posting of Thailand tide Gage Data to Tsunami Bulletin Board, also posted at http://www.navy.mi.th/hydro/tsunami.htm Ji, C. (2004). Preliminary Result of the 04/12/26 (Mw 9.0), Off West Coast of Northern Sumatra Earthquake, posted at http://www.gps.caltech.edu/%7Ejichen/Earthquake/2004/aceh/aceh.html Koike Nobuaki (2005). Preliminary Report of Numerical Computation of Tsunamis Generated by the December 26, 2004 Off Sumatra Island Earthquake, Indonesia. Posted at http://www.wakayama.nct.ac.jp/gakkasyoukai/kan/staff/koike/sumatra.html Kowalik, Z., Knight, W. Logan, T. and Whitmore, P. (2005). Numerical modeling of the global tsunami: Indonesian Tsunami of 26 December 2004. Sci. Tsunami Hazards, 23(1), 40–56. Koshimura, S. (2005). DRI Preliminary Tsunami Modeling Report, Modeling a tsunami Generated by the December 26, 2004 Earthquake off the West Coast of Northern Sumatra, Indonesia. Posted at http://www.dri.ne.jp/koshimuras/sumatra/ Merrifield, M.A., Firing, Y.L., Brundrit, G., Farre, R., Kilonsky, B., Knight, W. and Kong L. (2005). Preliminary Report of Tsunami Observations, Survey of India, posted at http://www.surveyofindia.gov.in/tsunami4.htm Murty, T.S. (1984). Storm Surges – Meteorological Ocean Tides, Bulletin No. 212, Fisheries Research Board, Canada, Ottawa, 897pp. Okada, Y. (1985). Surface deformation due to shear and tensile faults in a half-space. Bull. Seismol Soc. Am., 75, 1135–1154. Satake Kenji. (2005). Posted at http://staff.aist.go.jp/kenji.satake/animation.gif Shuto, N. (1991). Numerical Simulation of Tsunamis, In: E. Bernard, (ed.) Tsunami Hazard, Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 171–191. Titov, V.V. (2005). Tsunami Event – 26 December 2004. Posted at http://www.pmel.noaa.gov/tsunami/indo_ 1204.html Titov, V.V. and Gonzalez, F.I. (1997). Implementation and Testing of the Method of Splitting Tsunami (Most) Model Noaa Technical Memorandum ERL PMEL-112 . Titov, V.V. and Synolakis, C.E. (1996). Numerical modeling of 3-D long wave runup using VTCS-3. In: P. Liu, H. Yeh, and C. Synolakis (eds.), Long Wave Runup Models, World Scientific Publishing Co. Pte. Ltd., Singapore, pp. 242–248.
Numerical models for the Indian Ocean Tsunami 2004
173
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Titov, V.V. and Synolakis, C.E. (1995). Modeling of breaking and nonbreaking long wave evolution and runup using VTCS-2. J. Waterways Ports Coast. Ocean Eng., 121(6), 308–316. Yagi, Y. (2005). Preliminary Results of Rupture Process for 2004 off coast of Northern Sumatra Giant Earthquake (ver. 1), posted at http://iisee.kenken.go.jp/staff/yagi/eq/Sumatra2004/Sumatra2004.html Yalciner A.C., Taymaz, T., Kuran, U., Pelinovsky, E. and Zaitsev. A. (2005). The Model Studies on December 26, 2004 Indian Ocean Tsunami, Posted at http://yalciner.ce.metu.edu.tr/sumatra/ Yamanaka (2004). Posted at http://www.eri.u.tokyo.ac.jp/sanchu/Seismo_Note/2004/EIC161.html
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 16
The Cauchy–Poisson Problem: Application to Tsunami Generation and Propagation
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
N. Nirupama Atkinson School of Administrative Studies, York University, Toronto, Canada T.S. Murty and I. Nistor Department of Civil Engineering, University of Ottawa, Ottawa, Canada A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology, New Delhi, India
16.1
INTRODUCTION
The classical Cauchy–Poisson (C–P) problem is the initial value problem for surface waves. The classical C–P problem as described by Lamb (1945) deals with one-dimensional standing waves in an ocean of infinite depth. Although the classical problem is hardly suitable for tsunami studies, it is simple and can be used to introduce certain concepts. Two different initial states are considered: initial elevation of the free surface with no motion, and a horizontal surface with an initial distribution of surface impulse. The classical C–P problem deals with only a symmetric source of tsunami generation. We will also consider asymmetric sources which are more common for real tsunami generation events. Other factors we will consider include sloping nature of the ocean bottom in the area of tsunami generation as well as the effect of viscosity of the fluid on tsunami generation. The Indian Ocean Tsunami of 26 December 2004 afforded an opportunity to test the validity of the classical C–P problem, with an initial surface elevation as well as an initial impulse. This tsunami has been studied in detail by (Kowalik et al., 2005a,b; Murty et al., 2005a–c; Murty et al., 2006a–c; Nirupama et al., 2005; Nirupama et al., 2006).
16.2
MOTION STARTING WITH INITIAL EVALUATION
First consider initial evaluation. Taking the origin at the undisturbed level of the surface, the water level, η, and the velocity potential, φ, can be written for simple harmonic standing waves: η = cos(ωt) cos(kx) φ=
(16.1)
g sin(ωt)φ kz e cos(kx) ω
(16.2)
where ω2 = gk
(16.3) 175
176
N. Nirupama et al.
The initial state is given by: η = f (x), φ0 = 0
(16.4)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
The Fourier double integral representation is: ∞ 1 ∞ f (x) = dk f (α) cos k(x − α) dα π 0 −∞ From equations (16.1)–(16.5) ∞ 1 ∞ η= cos(ωt) dk f (α) cos k(x − α) dα π 0 −∞ g φ= π
∞
0
sin(ωt) kz e dk ω
∞
−∞
f (α) cos k(x − α) dα
(16.5)
(16.6) (16.7)
Lamb assumed that the initially elevated region is small in extent and is confined to the immediate vicinity of the origin, so that f (α) is nonzero for infinitesimal values of α. Lamb expressed φ and η in a series as well as in another form involving Fresnel’s integrals. 16.3
MOTION STARTING WITH INITIAL IMPULSE
For initial impulse, the initial conditions are: ρφ0 = f (x)
(16.8)
η=0 A similar procedure was followed as above. For large gt 2 /4x, the following expressions hold approximately. For initial elevation: 2 2 g 1/2 t 2 gt gt η = 3/2 1/2 3/2 cos + sin 2 π x 4x 4x
(16.9)
and for initial impulse: η=
2 2 g 1/2 t 2 gt gt cos − sin 5/2 1/2 5/2 2 π ρx 4x 4x
(16.10)
One drawback of these solutions is that as the origin is approached the wavelength decreases monotonically, whereas the wave height increases asymptotically. 16.4
KELVIN’S METHOD OF STATIONARY PHASE AND AIRY INTEGRAL
Kelvin (1877) suggested that the C–P problem can be studied by more simple methods than were adopted by Cauchy and Poisson.
The Cauchy–Poisson problem: application to tsunami generation and propagation
177
Consider the integral:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
a
b
φ(x) eif (x) dx
and assume that f (x) varies much more rapidly than in a periodic manner. Kelvin’s method uses the fact that the various elements of the integral will for the most part cancel by annulling interference except in the neighborhood of x, if any, for which f (x) is stationary. For details on Kelvin’s method of stationary phase see Jeffreys and Jeffreys (1946) and Stoker (1957). Consider one-dimensional propagation of a disturbance due to an initial elevation over the water surface. Following Jeffreys and Jeffreys, represent the disturbance as: f (x) =
∞
0
f (K) cos(Kx − ωt) dK
(16.11)
Then at point x0 , at time t0 , the disturbance can be asymptotically represented by Kelvin’s method of stationary phase as: f (x0 , t0 ) ≈
fK0 π 1/2 cos K0 x − ωt ∓ dC 4 1 πt dKg 2
(16.12)
Here the subscript 0 signifies that the function has to be evaluated for the wave number, K0 . Also Cg is the group velocity and the ∓ sign should be taken when dCg /dK is positive or negative. However, equation (16.12) is not valid in the following cases: (a) when the√ group velocity is stationary or (b) at the head of a wave train where the long-wave formula (C = gD) holds. In these situations the method of stationary phase should be carried out to a higher approximation. Then the disturbance could be expressed, instead in equation (16.11), in terms of the Airy integral given by: 1 Ai (α) = π
0
∞
3 t cos + αt dt 3
(16.13)
where α is some variable. If one approximates 1 tanh(KD) ∼ KD − (KD)3 6
(16.14)
The group velocity of tsunami waves arriving at any point is: Cg =
1 gD 1 − K 2 D2 2
(16.15)
Let the initial form of the surface be such that: η = 1 in√the region L < x < L and η = 0 elsewhere. If α is identified with the phase velocity of a long wave, gD, the asymptotic form of η near x = αt is: η∼L
2 αtD2
1/3
Ai
2(x − αt)3 αtD2
1/3 (16.16)
178
N. Nirupama et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
The form of the Airy function is such that it increases monotonically to a maximum and then oscillates with decreasing amplitude. For positive α, Ai (α) can be expressed in terms of the Bessel functions of imaginary argument, whereas for negative α, it can be expressed in terms of the Bessel function, J . Based on the behavior of Ai (α), one can deduce that at any given point after the passage of the head of the wave train succeeding waves must be of the form of a dispersive wave train. The following asymptotic form is valid for η after the initial few oscillations.
1/12 2 1/3 αtD2 1 2 1/2 π 2 3/2 η=L ×sin + (αt − x) (16.17) αtD2 π1/2 (αt − x)1/4 2 3 αtD2 4
16.5
C–P PROBLEM WITH ASYMMETRIC SOURCE
Braddock and Van Den Driessche (1971) developed a theory for the C–P problem when the source is asymmetric with reference to the axis (of a cylindrical polar coordinate system) with the origin placed at the bottom. The ocean bottom is assumed to move with the velocity F(x, y, t) = X (x)Y (y)T (t)
(16.18)
Although this separation of variables is somewhat simplistic, some generality is achieved by expanding X (x), Y (y), T (t), in series of orthogonal functions. This type of bottom motion is more general than in the theory of Kajiura. 2 x X (x) = αn Hm (s) exp − for −∞ < x < ∞ 2 m=0 2 ∞ y Y (y) = βn Hn (y) exp − for −∞ < y < ∞ 2 n=0 2 ∞ t T (t) = γp Lp (t) exp − for 0 < t < ∞ 2 p=0 ∞
(16.19)
Here Hm (x) is the Hermite polynomial of degree, m, and Lp (t) is the Laguerre polynomial of degree, p. The coefficients αm , βn , and γp can be written as follows: 2 ∞ 1 x αm = m √ X (x) exp − Hm (x) dx 2 2 m! π −∞ 2 ∞ 1 y βn = n √ Y (y) exp − Hn (y) dy 2 2 n! π −∞ 1 ∞ t γp = T (t)Lp (t) exp − dt p! −∞ 2
(16.20)
Because of the presence of factors 2m , m!, n!, p! in the denominators of equation (16.20), the coefficients of αm , βn , γp decrease rapidly as m, n, p increase. Hence, in practice only a few terms will be necessary.
The Cauchy–Poisson problem: application to tsunami generation and propagation
179
It is possible that the source area for a tsunami could be very close to the shore such as a shallow submarine earthquake or an underwater nuclear explosion. In this case, the effect of the sloping nature of the bottom on the tsunami generated has to be taken into account. Slatkin (1971) used a three-dimensional model to study this problem. The origin of a Cartesian coordinate system, x, y, z, is taken at the equilibrium position of the free surface with the Z-axis pointing upward. Then the bottom is given by Z = −D(x, y, t) and the free surface perturbation is η(x, y, t). In the linear shallow-water theory for the nonrotating case, the wave equation is (see Lamb, 1945):
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
g {∇(D∇η)} −
∂2 η ∂2 D = − ∂t 2 ∂t 2
(16.21)
where ∇ is the gradient operator. Let the bottom profile be specified as: D(x, y, t) = D0 (y) + D1 (x, y, t)
(16.22)
where D1 << D0 and D0 (y) is the bottom profile in the equilibrium state, assumed to be prescribed in 0 ≤ y < ∞. The shore and the infinite extent of the ocean are represented by y = 0 and ∞, respectively. From equations (16.21) and (16.22) the wave equation to the lowest order is: ∂ ∂2 η ∂2 η ∂η ∂ 2 D1 g (16.23) D0 (y) + gD0 2 − 2 = − 2 ∂y ∂y ∂x ∂t ∂t The following boundary conditions must be satisfied at y = 0 ∂η = 0 if D0 (0) = 0 or (η)y=0 is finite if D0 (0) = 0 ∂y y=0
(16.24)
Take the Laplace Transform in t and the Fourier Transform in x of equation (16.23) and define: ∞ ∞ η¯ (K, y, S) = eikx e−St η(x, y, t) dt dx (16.25) 0
−∞
Then: g
d d η¯ D0 (y) − (gD0 k 2 + S 2 )¯η = F(k, y, S) dy dy
(16.26)
where: F(k, y, S) = −
∞
−∞
eikx
∞ 0
e−St
∂η¯ ∂ 2 D1 dt · dx + S η¯ (k, y, t = 0) + (k, y, t = 0) ∂t 2 ∂t (16.27)
The equivalence of the effects of ground motion of finite duration and initial surface displacement (Hwang and Tuck, 1970) makes the method of solution somewhat independent of the exact nature of the source. The following form was assumed for D0 (y): D0 (y) = D0 (1 − e−αy )
(16.28)
N. Nirupama et al.
180
where D0 and α are prescribed. This is finite at ∞ and is analytic in the range 0 < y < ∞. Define: V ≡ e−αy
(16.29)
Then equation (16.26) becomes
∂2 η¯ ∂η¯ K2 1 1 1 S2 V (1 − V ) 2 + V (1 − 2V ) + 2 (1 − V ) η¯ = 2 F K, n , S − ∂V ∂V α2 gD0 α α gD0 α V 2
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
≡ G(K, V , S)
(16.30)
Thus, the problem reduces to solving equation (16.30) under the following conditions: (a) η is finite at V = 0, (b) η is finite at V = 1, and (c) the radiation condition. Slatkin (1971) determined the eigenfunctions (different possible wave modes) of the homogeneous part of equation (16.30) to solve this problem, and then expanded G in terms of these functions. One result √ is that long waves can be trapped along the coast and can travel with deepwater wave speed, gD, and that the energy in these waves decays as x−1/2 rather than as x−1 , so that more energy would be observed on this coast than is expected on the basis of deep-water wave amplitudes. 16.6
C–P PROBLEM INCLUDING VISCOSITY
Sretenskiy (1941) considered the transient development of one-dimensional surface waves on a viscous fluid. Basset (1888) and Lamb (1945) studied the basic motions and classified them in three categories: (i) damped gravity waves which represent mainly a balance between gravitational and inertial forces, but with modification by viscous forces, (ii) a diffusive motion which represents a primary balance between viscous and inertial forces, and (iii) a creep wave, which represents a primary balance between gravitational and viscous forces. The relative importance of the three types of motion for any initial state depends on the viscous length: ≡ g −1/3 ν2/3
(16.31)
where ν is the kinematic viscosity. Miles (1968) developed a theory for the C–P problem, including viscosity, with the aim of applying it to examine a suggestion made by Van Dorn (1968) regarding the origin of the concentric circular ridges that surround the crater orientale on the moon. Miles considered an ocean of infinite depth and took the vertical axis, z, of the coordinate system pointing downward. The ocean is initially at rest and is subject to an impulse, ρφ(r), and an initial free surface displacement, η0 (r), at t = 0. The pressure, p, and the potential, φ, are related through: p(r, z, t) = ρ
∂φ (r, z, t) ∂t
(16.32)
where r is the radial coordinate. Let rψ(r, z, t) be the Stokes stream function (see Lamb, 1945). The radial and downward components of the particle velocity, W , are: ∂φ ∂ψ + ∂r ∂z ∂φ 1 ∂ w=− − (rψ) ∂z r ∂r u=−
(16.33)
The Cauchy–Poisson problem: application to tsunami generation and propagation
181
Under the assumption of infinitesimal disturbances, the linearized form of the Navier–Stokes (N–S) equations in vector form and the continuity equation are, respectively, ∂W p (16.34) = −∇ + ν∇ 2 W ∂t ρ and
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
∇W = 0
(16.35)
From the above equations: ∂2 φ 1 ∂φ ∂2 φ + + 2 =0 ∂r 2 r ∂r ∂z
(16.36)
∂2 ψ 1 ∂ψ ∂2 ψ 1 ∂ψ 1 + ψ + = − 2 2 2 ∂r r ∂r r ∂z ν ∂t
(16.37)
and
The initial conditions are: φ(r, 0, 0) = φ0 (r) η(r, 0) = η0 (r)
(16.38)
The total impulse acting on the surface at t = 0 is: ∞ I = 2πρ φ0 (r)r dr
(16.39)
The potential energy associated with the initial displacement is: ∞ ρE0 = πρg η20 (r)r dr
(16.40)
where η0 (r) satisfies the condition ∞ η0 (r)r dr = 0
(16.41)
0
0
0
provided the displaced volume is zero. The kinematic and dynamic boundary conditions, given in equations (16.42) and (16.43) respectively, at the surface assume the following forms for this situation: ∂h ∂ ∂h ∂ + = for z = h(x, t) ∂t ∂x ∂x ∂z
2 ∂ 1 ∂ ∂ 2 gh + = gD + + ∂t 2 ∂x ∂z
(16.42) for z = h
(16.43)
182
N. Nirupama et al. ∂η =w ∂t
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
∂φ ∂w + gη − 2ν =0 ∂t ∂z ∂u ∂w ν + =0 ∂z ∂r
(16.44) (16.45) (16.46)
Although, strictly speaking, these conditions should be satisfied at z = η, rather than at z = 0, the second alternative is taken because of its simplicity, with the understanding that the error caused is no more than the linearization of the equations of motion. Miles (1968) used Laplace and Hankel transforms to obtain a formal solution to this problem. He considered two cases: (i) point-impulse problem and (ii) initial cavity problem. In the first problem, it is assumed that the radius of the area over which the impulse is applied is negligible compared with the viscous length, , defined in equation (16.31), and that the initial displacement vanishes identically. In the case of the initial cavity, a nonzero initial displacement is prescribed. Results for both cases are similar, in that the three different regimes discussed above exist. Nikitin and Potetyunko (1967) studied the C–P problem with the inclusion of viscosity, for a water body of finite depth (unlike Miles’ study for infinite depth). 16.7
SUMMARY
The classical C–P problem can be applied for tsunami generation either from an initial surface elevation or from an initial impulse. The original C–P problem only dealt with symmetric sources for wave generation. Here we have reviewed approaches on how to overcome this limit of a symmetric source and extend to the more common cases of asymmetrical sources. Finally, the literature on the influence of viscosity on the C–P problem methodology is also briefly reviewed. REFERENCES Basset, A.B. (1888). A Treatise on Hydrodynamics, Vol. II. Dover Publishing Inc., New York, NY, 328 pp. Braddock, R.D. and Van Den Driessche, P. (1971). Tsunami Generation, Department of Mathematics and Applied Mathematics University of Queensland, Reprint 45: 13 pp. Hwang, L.-S. and Tuck, E.O. (1970). Wave Generating by Beach Displacement. Tetra Tech. Inc., Pasadena, CA, Report. TC-174-2: 86 pp. Jeffreys, H. and Jeffreys, B.S. (1946). Methods of Mathematical Physics, Revised edition 1965, 1960. Cambridge University Press, New York, NY, 679 pp. Kelvin, W. (1887). On the waves produced by a single impulse in water of any depth of in a dispersive medium. Proc. Roy. Soc. London, 42, 80–83. Kowalik, Z., Knight, W., Logan T., and Whitmore, P. (2005a). Numerical modeling of the global tsunami: Indonesian Tsunami of 26 December 2004. Sci. Tsunami Hazards, 23(1), 40–56. Kowalik, Z., Knight, W., Logan, T., and Whitmore, P. (2005b). The tsunami of 26 December 2004: numerical modeling and energy considerations. In: G.A. Papadopoulos and K. Satake (eds.), Proceedings of International Tsunami Symposium, Chania, Greece, June, pp. 27–29, 140–150. Lamb, H. (1945). Hydrodynamics, 6th edn. Dover Publishing Inc., New York, NY, 738 pp. Miles, J.W. (1968). The Cauchy–Poisson problem for a viscous liquid. J. Fluid Mech., 34, 359–370. Murty, T.S., Rao, A.D., and Nirupama, N. (2005a). Inconsistencies in travel times and amplitudes of the 26 December 2004 tsunami. J. Mar. Med., 7(1), 4–11. Murty, T.S., Nirupama, N., Nistor, I., and Rao, A.D. (2005b). Conceptual differences between the Pacific, Atlantic and Arctic tsunami warning systems for Canada. Sci. Tsunami Hazards, 23(3), 39–51.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
The Cauchy–Poisson problem: application to tsunami generation and propagation
183
Murty, T.S., Nirupama, N., and Rao, A.D. (2005c). Why the earthquakes of 26th December 2004 and the 27th March 2005 differed so drastically in their tsunami-genic potential, newsletter. Voice Pacific, 21(2), 2–4. Murty, T.S., Nirupama, N., Nistor, I., and Hamdi, S. (2006a). Far field characteristics of the tsunami of 26 December 2004. ISET J. Earthq. Technol., 42(4), 213–217. Murty, T.S., Nirupama, N., Nistor, I., and Hamdi, S. (2006b). Why the Atlantic generally cannot generate trans-oceanic tsunamis. ISET J. Earthq. Technol., 42(4), 227–236. Murty, T.S., Rao, A.D., Nirupama, N., and Nistor, I. (2006c). Numerical modelling concepts for the tsunami warning systems. Curr. Sci. 90(8), 1073–1081. Nikitin, A.K. and Potetyunko, E.N. (1967). Cauchy–Poissons Spatial Problem of Waves on the Surface of a Viscous liquid of Finite Depth. Academy Science, USSR, Doklady, 1974, 50–52. Nirupama, N., Murty, T.S., Rao, A.D., and Nistor, I. (2005). Numerical tsunami models for the Indian Ocean countries and states, Indian Ocean Survey, 2(1), 1–14. Nirupama, N., Murty, T.S., Nistor, I., and Rao, A.D. (2006). The energetics of the tsunami of 26 December 2004 in the Indian Ocean: a brief review. Mar. Geod., 29(1), 39–48. Slatkin, M.W. (1971). Long waves generated by ground motion. J. Fluid Mech., 48, 81–90. Stoker, J.J. (1957). Water waves: the mathematical theory with applications. In: Pure Applied Mathematics. Interscience Publishers, New York, 567 pp. Sretenskiy, L. (1941). Concerning waves on the surface of a viscous fluid, Tr. Tsentr. Aerol. Gidrodinam, 541, 1–34 (In Russian). Van Dorn, W.G. (1968). Tsunamis on the moon. Nature, 220, 1102–1107.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 17
A Review and Listing of Tsunami Heights and Travel Times for the 26 December 2004 Event
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
I. Nistor & K. Xie Department of Civil Engineering, University of Ottawa, Ottawa, Canada N. Nirupama Atkinson School of Administrative Studies, York University, Toronto, Canada T.S. Murty Department of Civil Engineering, University of Ottawa, Ottawa, Canada
17.1
INTRODUCTION
The 26 December 2004 tsunami can be considered a global tsunami, since it propagated not only throughout the Indian Ocean, in which it was generated, but also into the Pacific and Atlantic oceans (Kowalik et al., 2005a, b). Nirupama et al. (2005) which showed that about 90% of the tsunami energy stayed inside the Indian Ocean, while some 6% of the energy leaked into the Pacific Ocean, via mainly south of Australia and New Zealand. Another 4% of the energy leaked into the Atlantic Ocean via south of the Cape of Good Hope. It is known that roughly the ratio of tsunami energy propagation in a direction perpendicular to the fault versus in a direction parallel to the fault is 13:1 (Murty, 1977). This hypothesis, in addition to shadow zones and reflected tsunami waves from coastlines can generally account for the relatively low impact of this tsunami on Australia, Malaysia (to some degree), Myanmar, Bangladesh, Tanzania and countries further south on the African coast. The same conditions can, at the same time, explain why the tsunami was so violent and destructive in Indonesia, Thailand, India, Sri Lanka, Maldives, Somalia, and Kenya, etc. Figure 17.1 shows the locations of these countries in the Indian Ocean. 17.2
PHYSICAL OCEANOGRAPHIC PROCESSES
The following physical oceanographic processes (Murty et al., 2006) either individually or in combination can account for the observed characteristics of this tsunami in the global oceans. The detailed behaviour of tsunamis on the Indian coastlines can be explained through a combination of one or more of the following physical oceanographic processes (Kowalik et al., 2005a, b; Murty et al., 2005a–c, 2006a–c; Nirupama et al., 2005a,b, 2006).
185
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
186
I. Nistor et al.
Figure 17.1.
Geographical map of the Indian Ocean (Source: http://mapsherpa.com/tsunami/). Antinode Crest Node
Node
Node Trough Antinode
Figure 17.2.
Illustration of the antinode with highest positive amplitude at quarter wavelength of a sine wave.
(a) Quarter wave resonance amplification in bays and gulfs (Figure 17.2) It can be seen from Figure 17.2 that for a sine wave, the highest amplitude occurs at quarter wavelength. What this means is that, as the tsunami approaches from the deep ocean towards the coast, and as its wavelength is getting compressed, if the linear dimensions of a coastal gulf or bay matches one-fourth of a wavelength of the incoming tsunami, then the tsunami amplitude will be amplified, and this process is referred to as quarter wave resonance amplification. (b) Helmholtz resonance in harbours (Figure 17.3) During the tsunami of 26 December 2004 in the Indian Ocean, there were high water levels in certain harbours for several days after the tsunami which can be accounted for through the phenomena of Helmholtz resonance. The long gravity energy wave of the tsunami enters
A review and listing of 26 December 2004 tsunami
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
(c) (d)
(e)
(f)
187
the wide harbour through a narrow channel but cannot easily get out of the harbour, because successive reflections at various locations on the harbour boundaries would leak out a small amount of energy. Constructive interference Boundary reflections During the tsunami of 26 December 2004 in the Indian Ocean, tsunami waves with maximum amplitudes arrived at the central and northern part of the Kerala coasts several hours later than the direct tsunami waves, which arrived on the morning of 26 December. This delay of arrival can be explained as due to waves reflected from the Lakshadweep Islands (Figure 17.4(a)) and coming back to the Kerala coast. Some waves even were reflected from the coast of Somalia in Africa (Figure 17.4(b)). Interaction with astronomical tides (Figure 17.5) Figure 17.5 shows the computation of the water level due to tsunami tidal interaction. In the left panel, a maximum water level of 8.5 m is obtained when the tsunami and tide are linearly superposed. On the other hand, as shown in the right panel, when non-linear interactions are included, the maximum water level becomes 9.4 m, thus attributing an increase of 0.9 m in the water level due to non-linear interaction between tide and tsunami. Coupling with internal waves due to ocean density gradients (Figure 17.6)
HARBOR ENTRANCE CHANNEL
Figure 17.3.
HELMHOLTZ RESONANCE
Illustration of the phenomena of Helmholtz resonance in harbours.
(a)
Figure 17.4.
(a) Tsunami waves reaching Kerala after first diffracting around Sri Lanka and travelling northward and getting reflected from Lakshadweep Islands (for illustrative purposes only, not to scale; basemap adapted from www.mapsofindia.com). (b) Tsunami waves reaching Kerala after being reflected from the coast of Somalia (for illustrative purposes only, not to scale; basemap adapted from www.wikipedia.com).
188
I. Nistor et al. AFGHANISTAN
IRAQ
IRAN rs Pe n ia
PAKISTAN
f ul G
SAUDI ARABIA
UAE
Re
INDIA
ea dS
AN
M
O EN
Arabian Sea
ETHIOPIA
SO
M
AL
IA
SRI LANKA
KENYA
I n di an O c e an 0
500
(b)
Figure 17.4.
North 1000 Kilometers
2000
(Continued)
Maximum of level (cm)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
YEM
1000
1000
800
800
600
Linear superposition
600
400
400
200
200
0
0
800
900
1000
Distance (Km)
1100
Nonlinear superposition
800
900
1000
1100
Distance (Km)
Figure 17.5. Tsunami tidal interaction (personal communication, Prof. Z. Kowalik, 2006).
(g) Trapping of long gravity energy on continental shelves through Oscillations of the First Class (OFC) and Oscillations of the Second Class (OSC) via the mechanism of trapped and partially leaky modes. These are also referred to as Gravoid and Elastoid inertia modes (Table 17.1). During the tsunami of 26 December 2004 in the Indian Ocean, persistent high water levels were observed around the Andaman and Nicobar Islands. This can be explained through the phenomena of trapped and partially leaky modes which are due to the so-called OFC and OSC which are separated in periods in so-called pendulum day. As is shown in Table 17.1,
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
A review and listing of 26 December 2004 tsunami
189
Figure 17.6. Amplification of the tsunami at the ocean surface through coupling with internal waves. Table 17.1.
Length of pendulum day at different latitudes.
Latitude (degrees) 0 10 20 30 40 50 60 70 80 90
(h) (i) (j) (k) (l)
Length of pendulum day (hours) ∞ 139 71 48 37 31.5 28 26 24.5 24
the length of the pendulum day decreases as one moves from the equator towards the pole. The northern end of Andaman Islands is at 14◦ N latitude and the southern end of Nicobar Islands (Indira Point) is at 6◦ N latitude. As can be seen from Table 17.1, the pendulum day here at these latitudes is several days, and in reality it takes several pendulum days for all the long gravity wave energy of the tsunami that was trapped on the shelves around these islands, to completely leak out back into the ocean. Interaction with the strong tidal current gradients near regular and degenerate semi-diurnal and diurnal tidal amphidromic points (Figure 17.7). Extraction of energy from opposing ocean currents, through Reynolds eddy stresses. Interaction with the wind wave setup, considering the fact that wind waves (swell from the southern ocean) have the highest amplitudes in the Indian Ocean, as compared to the other oceans. Focusing and defocusing of tsunami energy due to ocean bathymetric features such as ridges and trenches. Phase or frequency dispersion and amplitude dispersion (non-linear effects). A most fundamental parameter in tsunami modelling studies is the so-called Ursell parameter defined below, where η is the tsunami amplitude, D is the water depth, and λ is the wavelength of the tsunami. The initial tsunami waves get dispersed due to ocean bathymetry, with a slight frequency dependence, which is referred to as frequency dispersion. The strong non-linear affects in the shallow water near the coastlines are referred to as amplitude dispersion. U=
ηλ2 D3
(17.1)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
190
I. Nistor et al.
Figure 17.7. Tidal amphidromic point in the Arabian Sea.
U=
1 O(1) 1
Amplitude dispersion can be ignored Both amplitude and phase dispersion are important Amplitude dispersion dominates
(17.2)
(m) Zones of convergence and divergence (n) Breaks in the continental shelves through which tsunami waves travelling through the deeper water interact with tsunamis in the shallow water. The main reason for the very high tsunami amplitudes and large scale loss of life and destruction on Indonesian and Thailand were their proximity to the epicentre of this earthquake as well as quarter wave resonance amplification. All the other physical oceanographic processes listed above also played a role to differing degrees in different areas. 17.3
LISTING OF TSUNAMI AMPLITUDES AND TRAVEL TIMES
The following data listed in Table 17.2 are all taken from various web sites.
A review and listing of 26 December 2004 tsunami Table 17.2.
Listing of observed tsunami arrival times and run-up. Physical location
Location (tide gauge)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
191
Survey point
Latitude
Longitude
Arrival time in UCT
Indonesia {local time = Indonesia spreads across three time zones (UTC +7, +8, +9)} Panjiang −5.270 10.517 4:00 Pelabuhan Ratu −9.590 103.090 ND Sibolga 1.450 98.460 2:20 Sigli 5.384 95.968 1:30 Sigli 5.384 95.968 1:30 Sigli 5.385 95.966 1:30 Sigli 5.389 95.960 1:30 Sigli 5.389 95.960 1:30 Krueng Raya port 5.596 95.526 1:30 Krueng Raya port 5.595 95.527 1:30 Krueng Raya port 5.588 95.525 1:30 Sabang 5.826 95.347 1:30 Sabang 5.882 95.324 1:30 Sabang 5.839 95.299 1:30 Center of Banda Aceh 5.556 95.284 1:30 Center of Banda Aceh 5.556 95.284 1:30 Center of Banda Aceh 5.554 95.291 1:30 Center of Banda Aceh 5.549 95.316 1:30 Center of Banda Aceh 5.549 95.317 1:30 Center of Banda Aceh 5.550 95.317 1:30 Center of Banda Aceh 5.552 95.317 1:30 Center of Banda Aceh 5.553 95.317 1:30 Center of Banda Aceh 5.553 95.317 1:30 Center of Banda Aceh 5.553 95.317 1:30 Center of Banda Aceh 5.547 95.317 1:30 Center of Banda Aceh 5.548 95.316 1:30 Center of Banda Aceh 5.548 95.315 1:30 Center of Banda Aceh 5.548 95.314 1:30 Center of Banda Aceh 5.548 95.313 1:30 Center of Banda Aceh 5.549 95.313 1:30 Center of Banda Aceh 5.549 95.312 1:30 Center of Banda Aceh 5.549 95.312 1:30 Center of Banda Aceh 5.553 95.317 1:30 Center of Banda Aceh 5.553 95.316 1:30 Center of Banda Aceh 5.553 95.316 1:30 Center of Banda Aceh 5.553 95.316 1:30 Center of Banda Aceh 5.553 95.316 1:30 Center of Banda Aceh 5.553 95.316 1:30 Center of Banda Aceh 5.553 95.315 1:30 Center of Banda Aceh 5.553 95.315 1:30 Center of Banda Aceh 5.553 95.315 1:30 Center of Banda Aceh 5.549 95.311 1:30 Center of Banda Aceh 5.549 95.311 1:30 Center of Banda Aceh 5.549 95.311 1:30 Center of Banda Aceh 5.550 95.310 1:30 Center of Banda Aceh 5.549 95.311 1:30 Center of Banda Aceh 5.554 95.319 1:30 Center of Banda Aceh 5.551 95.304 1:30
Wave run-up (m) 0.11 0.43 4.07 4.82 4.24 4.4 3.68 5.1 5.92 6.71 3.02 3.15 6.2 12 8.4 9 4.64 4.64 4.5 4.58 4.52 4.56 4.67 4.43 4.73 5.16 5.56 5.55 5.53 5.61 4.65 4.7 4.68 4.79 5.49 4.64 3.96 5.48 5.52 5.35 5.98 4.61 5.96 6.47 6.2 5.01 7.04 (Continued)
192
I. Nistor et al.
Table 17.2.
(Continued) Physical location
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Location (tide gauge)
Survey point
Latitude
Longitude
Arrival time in UCT
Wave run-up (m)
Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh Center of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh
5.551 5.551 5.551 5.552 5.552 5.553 5.550 5.551 5.549 5.551 5.550 5.550 5.559 5.549 5.542 5.541 5.541 5.547 5.536 5.604 5.583 5.477 5.443 5.443 5.458 5.452 5.451 5.460 5.405 5.414 5.415 5.415 5.418 5.420 5.466 5.460 5.453 5.462 5.496 5.492 5.485 5.482 5.477 5.472 5.422 5.429 5.429 5.440
95.304 95.304 95.304 95.303 95.303 95.304 95.307 95.307 95.306 95.307 95.308 95.308 95.284 95.311 95.285 95.287 95.282 95.278 95.283 95.346 95.350 95.245 95.240 95.242 95.247 95.244 95.243 95.246 95.254 95.249 95.250 95.250 95.248 95.248 95.242 95.246 95.245 95.243 95.229 95.231 95.228 95.236 95.235 95.242 95.244 95.241 95.236 95.241
1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30
4.34 3.52 6.26 6.11 4.4 6.76 6.58 6.59 6.43 5.87 6.78 6.59 12.19 6.21 5.04 6.87 4.7 7.68 7.78 7.12 6.58 12.42 19.96 27.86 34.85 18.79 21.56 31.91 20.13 17.04 17.25 20.08 24.86 21.57 27.66 21.97 18.46 23.83 29.68 17.48 32.72 15.42 18.06 34.25 28.61 20.49 20.8 29.98 (Continued)
A review and listing of 26 December 2004 tsunami Table 17.2.
(Continued) Physical location
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Location (tide gauge)
Survey point
Latitude
Longitude
West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh West Coast of Banda Aceh Rhiting
5.440 5.445 5.449 5.456 5.469 5.475 5.478 5.478 5.465 5.461 5.459 5.452 5.452 5.450 5.429 5.396 5.395 5.395 2.549 2.550 2.558 2.575 2.565 2.557 2.344 2.386 2.407 2.408 2.404 2.403 2.402 2.402 2.432 2.432 2.442 2.393 4.211 4.209 4.208 4.208 4.966 4.129 4.209 4.308 4.297 4.297 4.300 ND
95.241 95.242 95.242 95.245 95.240 95.237 95.246 95.246 95.244 95.245 95.246 95.242 95.243 95.243 95.234 95.261 95.256 95.256 96.335 96.330 96.308 96.269 96.288 96.301 96.468 96.489 96.483 96.482 96.482 96.481 96.481 96.481 96.260 96.260 96.243 96.337 96.065 96.040 96.040 96.039 96.310 96.129 96.037 95.972 95.948 95.947 95.987 ND
Lepung (Simeulue)
(Meulaboh)
193
Ganting Ganting Ganting Tsunami end point Tanjun Raya Senebu village Near Senebu Near Senebu village Labuhan Bakti Labuhan Bakti Labuhan Bakti Labuhan Bakti Near Labuhan Bakti Near Labuhan Bakti Laubang Lantik or Tembah Barat Salur Busung Suaktimah village Skoneda Skoneda Kuala Buban Bay Kuala Tadu Meulaboh port St 176, WPT17 Near Aronghan Aronghan Aronghan Near Aronghan Pentai Cermin
Arrival time in UCT
Wave run-up (m)
1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:30 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40
23.78 21.65 24.35 25.05 28.59 16.63 12.39 12.01 29.34 23.51 27.28 30.4 15.77 20.07 48.86 18.7 21.39 17.59 1.75 3 >4 <0.30 1.9 2 2.5 ND 4 ND 2.5 ND 2.5 2.5 1.5 1.5 1.5 1.5 >15 >15 >15 >10 >15 >15 ND ND 15 15 9 1.7 (Continued)
194
I. Nistor et al.
Table 17.2.
(Continued) Physical location
Location (tide gauge)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Medan East
Survey point Kuala Ruteri Belawan port Ferry port
Sumatra (local time = UTC +7) Lhokseumawe Lhokseumawe Sibolga (fishery port) Sibolga (fishery port) Sibolga (passenger port) Sibolga (tide gauge) South of Sibolga Pasarsorkam Barus Sigli Sigli Sigli Sigli Thailand (local time = UTC +7) Ranong Kuraburi Ta Phao Noi Krabi Kantrang Ta Ru Tao Tummarang Tummarang Yacht Mercator Karon Beach (South part) Karon Beach (Central part) Karon Beach (North part) Moodong Canal (Chalong Bay) Phalai village (Chalong Bay) Makham Bay (North, Pier) Sire village (Siray Is.) Khao Lak Khao Lak Chalong Bay Pier Leam Him Bang Rong Pier Phi Phi Don (North coast) Phi Phi Don (South coast) Phi Phi Don South–North Patong Beach Patong Beach Patong Beach Patong Beach Patong Beach Patong Beach
Latitude Longitude
Arrival time Wave in UCT run-up (m)
ND 3.784 3.787
ND 98.715 98.705
ND ND ND
5.249 5.235 1.718 1.719 1.729 1.729 1.664 1.871 2.008 5.387 5.388 5.388 5.388
96.913 97.060 98.797 98.795 98.785 98.785 98.826 98.565 98.403 95.964 95.963 95.962 95.962
1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40
2.9 1.7 1.6 1.63 1.5 2.6 1.5 1.0 1.7 4.4 4.0 3.5 3.1
9.967 9.133 7.667 8.167 8.667 6.083 6.083 6.083 7.750 7.829 7.842 7.821 7.842 7.839 7.870 7.873 8.637 8.637 7.819 7.943 8.047 7.739 7.748 7.738 7.884 7.888 7.888 7.887 7.887 7.894
98.583 98.083 98.033 99.200 99.867 100.000 100.000 100.000 98.280 98.298 98.297 98.298 98.375 98.373 98.417 98.425 98.251 98.253 98.403 98.401 98.419 98.777 98.772 98.773 98.292 98.296 98.296 98.296 98.296 98.299
4:38 4:10 3:20 4:20 5:50 4:00 6:10 6:10 2:38 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00
1.00 3.2 1.45 2.1 3.0 1.8 1.6 ND ND 4.04 4.49 ND 3.15 2.75 1.39 2.67 8.8 9.6 3.62 0.72 1.29 5.84 4.58 ND 5.09 4.88 5.44 5.33 5.28 5.48 (Continued)
A review and listing of 26 December 2004 tsunami Table 17.2.
(Continued) Physical location
Location (tide gauge)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
195
Survey point
Latitude
Longitude
Patong Beach Patong Beach Patong Beach Kamala Beach Kamala Beach
7.892 7.904 7.904 7.947 7.947
98.298 98.301 98.301 98.283 98.282
Bang Thao Beach
8.002
98.296
Nai Yang Beach
8.087
98.300
8.636
98.249
8.635
98.250
8.637 8.637 8.638 8.640 8.640
98.251 98.251 98.253 98.250 98.250
8.640 8.640 8.661 8.661
98.250 98.250 98.249 98.250
8.683
98.244
8.683
98.244
8.742
98.255
8.726 8.734 8.729
98.232 98.225 98.226
8.184
98.291
8.197 8.272
98.300 98.280
Khao Lak
Arrival time in UCT
Wave run-up (m)
3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00
5.02 4.79 4.90 4.85 5.29 4.47 3.41 4.04 3.76 4.41 3.95 4.05 4.07 8.76 9.34 9.5 9.45 9.28 8.67 8.71 ND ND 9.91 9.56 8.35 9.46 9.35 9.71 7.38 ND 8.27 8.30 7.99 8.59 10.62 8.50 8.17 6.46 6.06 6.11 6.24 4.48 6.30 8.86 4.83 3.32 3.11 4.05 (Continued)
196
I. Nistor et al.
Table 17.2.
(Continued) Physical location
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Location (tide gauge)
Latitude
Longitude
Arrival time in UCT
Kata Noi Beach Rawai Beach Frendship Beach Hotel (Rawai Rest Area) Kamala Beach Ban Na Tai Tap Lam Navy Base Ban Kao Lak
8.304 7.803 7.772 7.796
98.275 98.303 98.328 98.340
3:00 3:00 3:00 3:00 3:00
4.30 4.28 3.88 2.43 2.35
7.950 8.293 8.570 8.611
98.280 98.273 98.224 98.238
3:00 3:00 3:00 3:00
Ban Niang Beach
8.675
98.242
3:00
Ban Niang Beach Ban Niang Beach Ban Niang Beach Laem Pakarang
8.671 8.700 8.700 8.736
98.243 98.240 98.240 98.222
3:00 3:00 3:00 3:00
Ban Nam Kem Ta pou Noi Ta pou Noi Ta pou Noi Chalong Ban na Tai Rai Dan Nai Rai Ban Thung Wa Thai Muang Thai Muang, visitor center Thai Muang, Natural Conservation Park Ban Laem Po Ban Bang Phng Ban Num Kim Ban Ma Kap Ban Nok Na Ban Pak Ko Ko Koh Kao port Ban Nam Kim Ban Nam Kim Ban Nam Kim Ban Nam Kim Ko Yao Ban Pak Chok Ban Thung Dap Ban Ao Luk Tum Ko Yao, fishing village
8.864 7.834 7.834 7.834 7.821 8.274 8.297 8.310 8.378 8.399 8.436 8.484
98.274 98.422 98.421 98.421 98.345 98.278 98.272 98.273 98.255 98.265 98.238 98.228
3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00
5.40 4.3, 5.1 3.80 7.80 8.5, 8.6, 9.5, 8.8 7.60 5.8, 6.2 7.9 7.1, 8.7, 9.3,(14.6, 15.5) ND 3 6.40 2.53 4.70 1.70 3.90 4.801 6.77 5.29 6.78 6.07 6.25 5.19
8.573 8.812 8.857 8.923 8.999 8.882 8.872 8.860 8.860 8.858 8.857 9.222 9.160 9.028 9.203 9.222
98.225 98.266 98.269 98.258 98.257 98.270 98.275 98.275 98.279 98.279 98.268 98.375 98.271 98.257 98.272 98.375
3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00
ND 13.085 ND 6.0345 12.61 6.391 3.72 4.08 ND ND 15.77 ND 6.632 19.572 8.617 0.982
Survey point
Wave run-up (m)
(Continued)
A review and listing of 26 December 2004 tsunami Table 17.2.
(Continued) Physical location
Location (tide gauge)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
197
Survey point
Krabri Tide Station, National Park Office Krabri Tide Station, National Park Office Krabri Tide Station, National Park Office Ban Ko Dam Ban Tam Nang Ban Pak Nam Ban Pak Nam Ban Pak Nam Ban Pak Nam Ban Pak Nam fishering port Ban Pak Nam port Had Sai Dam (Ban La Ong) Ka Yu Harbor (Ban La Ong) Ka Yu Harbor (Ban La Ong) Ban Ao Khoei Hat Praphat Ban Thale Nok Ban Thale Nok Ramson Ramson Ramson Ban Chang Hak Ban Nam Kim Ban Nam Kim Khao Lak Malaysia (local time = UTC +8) Sungai Batu, Penang Is. Sungai Batu, Penang Is. Pasir Panjang Penang Is. Muara Sungai Pulau Betong, Penang Is. Muara Sungai Pulau Betong, Penang Is. Penang Is. Penang Is. Tanjung Bungah, Penang Is. Miami Beach, Penang Is. Gurney Drive, Penang Is. Mainland Kampung Tepi Sungai Kampung Paya Kampung Paya Sungai Chenang (Langkawi) Pelangi Beach Hotel Resort (Langkawi)
Latitude
Longitude
Arrival time in UCT
Wave run-up (m)
9.225
98.377
3:00
ND
9.225
98.377
3:00
2.43
9.225
98.377
3:00
ND
9.277 8.993 9.951 9.951 9.951 9.951 9.946 9.979 9.744 9.781 9.781 9.299 9.376 9.460 9.460 9.602 9.602 9.602 9.669 8.853 8.859 8.631
98.386 98.412 98.596 98.596 98.596 98.596 98.598 98.601 98.552 98.554 98.554 98.384 98.401 98.437 98.437 98.470 98.470 98.469 98.559 98.272 98.272 98.258
3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00 3:00
1.136 0.70 2.02 2.86 ND 1.16 1.66 1.67 3.727 1.979 −0.321 9.213 4.974 6.77 6.206 4.8 4.91 1.075 1.06 5.779 6.379 ND
5.280 5.283 5.295 5.306
100.239 100.236 100.183 100.192
5:15 5:15 5:15 5:15
1.5 1.5 2.2 ∼3.0
5.280
100.239
5:15
∼3.0
5.326 5.338 5.467 5.476 5.439 5.575 5.580 5.611 5.599 6.304 6.301
100.196 100.195 100.277 100.266 100.308 100.338 100.338 100.341 100.341 99.721 99.720
5:15 5:15 5:15 5:15 5:15 5:15 5:15 5:15 5:15 4:15 4:15
∼2.0 ∼3.0 ∼2.5 ∼3.0 ∼2.5 <1 ∼3.0 2.5 1.5 2.2 (Continued)
198
I. Nistor et al.
Table 17.2.
(Continued) Physical location
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Location (tide gauge)
Survey point
Sungai Kuala, Melaka River South of Kuala Triang Kuala Triang At land behind the shore of Kuala Triang India (local time = UTC +5.30) Vishakhapatnam Chennai Tuticorin Kochi Mormugoa Pulicat Pattinapakkam (Chennai) Kovalam Kalpakkam Periakalapet Puttupattnam Devanaampattinam Perangipettinam Tarangambadi Nagapattinam Vedaranniyan Sri Lanka (local time = UTC +6) Colombo Marawila, Lancigama Beach Vennapuwa North Waikkala Waikkala Negombo Negombo South of Negombo, Talahena Uswettakeiyawa Colombo, Mattakuliya Hikkaduwa Unawatuna Hambantota, in front of sand dunes Mahaseelawa Beach, Yala Patanangala Beach, Yala Bentota Seenigama Talpe Waligama Koggala Airport Galle Port Dodanduwa Hikkaduwa Fishery Harbour
Latitude
Longitude
Arrival time in UCT
Wave run-up (m)
6.201 6.356 6.361 6.351
99.718 99.715 99.709 99.721
4:15 4:15 4:15 4:15
17.65 13.10 8.75
83.280 80.320 78.200
13.384 13.021 12.791 12.506 12.026 11.860 11.743 11.516 11.027 10.763 10.393
89.333 80.279 80.250 80.161 79.865 79.815 79.787 79.766 79.856 79.849 79.867
3:35 3:35 4:27 3:35 ND
6.930 7.376 7.356 7.283 7.281 7.238 7.235 7.155 7.086 6.973 6.132 6.008 6.135
79.830 79.823 79.827 79.840 79.841 79.841 79.841 79.828 79.828 79.870 80.102 80.242 81.135
3:49 6:30 6:30 6:30 6:30 7:30 7:30 6:30 6:30 6:00 4:10 3:40 3:30
2.17 2.3 1.8 2.7 1.6 2 1.6 2.3 2.6 2.7 3.4 3.3 8.8
6.292 6.344
81.436 81.497
3:30 3:30 3:45 3:45 3:45 3:45 3:45 3:25 3:25 3:25
8.4 11.3
3.7
3.2 2.7 4.3 4.1 3.9 2.6 2.5 2.8 4.4 5.2 3.6
4.9 9.3 6.0 4.0 4.7 (Continued)
A review and listing of 26 December 2004 tsunami Table 17.2.
(Continued) Physical location
Location (tide gauge)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
199
Survey point
Kahawa Ambalangoda beach Beruwala Fishery Harbour North Beach of Beruwala Paiyagala Station Panadura Moratuwa Beach Oman (local time = UTC +4) Salalah Kenya (UTC +2) Lamu
North of Mombasa
Tanzania Zanzibar Mauritius (local time = UTC +4) Port louis Nias Island (local time = UTC +5.30) Muawe village Muawe village Medrehe Medrehe Medrehe Sirombu Sirombu Sirombu Sirombu Sirombu Hilirihono Hilirihono Gunungsitoli
Latitude
Longitude
Arrival time Wave in UCT run-up (m)
6.700 6.783
79.917 79.883
3:45 3:45 3:45 3:45 3:45 3:45 3:30
17.000
54.000
8:13
0.28
−2.267 −4.067
40.900 39.667
9:57 11:00
0.28 3.0
−6.150
39.183
10:45
0.28
−20.090
57.300
7:43–8:47
ND
97.212 97.170 97.405 97.405 97.393 97.424 97.419 97.405 97.409 97.421 97.732 97.709 97.609
1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40 1:40
2.8 3.8 1.5 3.5 4.5 2.0 4.6 5.3 4.5 4.65 2.5 2.9 1.5
1.4012 1.3958 1.003 1.008 1.016 0.957 0.951 1.008 1.004 0.951 0.574 0.568 1.288
10.0 4.7 2.4 4.8 6.0 5.6 4.4
ND: No Data.
The following sources are being used to gather some of the information given in Table 17.2: • • • • • • • •
Dr. Parluhutan Manurung Yoshinobu Tsuji (Earthquake Research Institute, The University of Tokyo, Japan) Hideo Matsutomi (Akita University, Japan) Yuichiro Tanioka (Hokkaido University, Japan) Yuichi Nishimura (Hokkaido University, Japan) Tsutomu Sakakiyama (Central Research Institute of Electric Power Industry, Japan) Takanobu Kamataki (National Institute of Advanced Industrial Science and Technology, Japan) Yoshikane Murakami (Kansai Electric Power, co., inc., Japan)
200 • • • • •
Andy Moore (Kent State University, USA) Guy Gelfenbanm (USGS, USA) http://ioc.unesco.org/iosurveys/Indonesia/yalciner/yalciner.htm http://www.drs.dpri.kyoto-u.ac.jp/sumatra/index-e.html Dr Viacheslav K. Gusiakov, Head, Tsunami Laboratory, Institute of Computational Mathematics and Mathematical Geophysics, Siberian Division, Russian Academy of Sciences Thai Marine Department Hydrographic Department of Royal Thai Navy http://www.drs.dpri.kyoto-u.ac.jp/sumatra/thailand2/ http://ilikai.soest.hawaii.edu/uhslc/iotd/ http://www.pmel.noaa.gov/tsunami/indo20041226/sibolga_nias.htm
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
• • • • •
I. Nistor et al.
5.882
#
Sabang
#
#
Krueng Raya Port
#
## #
#
#
# # # # # ## # # # ## ## ## # # ## # ## # # ## ## # #
# ##
# ### ##### ## # # #
## #
Centre of Banda Aceh
West of Banda Aceh Sigli
#
5.384 95.228
## #
95.968
#
N
3.15
Sabang 5.839
6.2
# #
#
5.583
7.12 6.58
#
0
Krueng Raya port 5.92 6.71 5.1 ## #
95.299 ## ####
# # # #
Figure 17.8.
3.02
### #
95.527
####### # # #####
6
Indonesia 12
18 Kilometers
Maximum tsunami amplitudes (m) in Sabang and Banda Aceh area.
A review and listing of 26 December 2004 tsunami 17.4
201
PICTORIAL REPRESENTATION OF SOME OF THE RUN-UP HEIGHTS (Murty et al., 2005)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Figures 17.8–17.11 shows the maximum tsunami amplitudes (m) based upon the survey by the Japanese team lead by Dr Y. Tsuji. The highest value reported by this team is 34.85 m, which is considerably smaller than the 50 m reported by another Japanese team lead by Dr Shibayama.
Figure 17.9.
Maximum tsunami amplitudes (m) in centre of Banda Aceh area.
202
I. Nistor et al.
5.882
#
Sabang
#
#
Krueng Raya Port
#
# #
# # # #
# ### #
# # #
# # ## # ## ## # # # # ## # ## # ## # ## ## # ## #
## ##### ## ## ##
Centre of Banda Aceh
West of Banda Aceh Sigli
#
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
5.384 95.228
95.968
Indonesia
5.496
#
29.68 #
#
## #
17.48
32.72
West Coast of Banda Aceh #
12.39 12.01 # 12.42
15.42
##
18.06 # 16.63 # 34.25 31.91 # 28.59 21.97 # 27.66 # 29.34 # 23.83 23.51# # # 27.28 # 34.85 # 25.05 30.40 # 18.46 ## # # 21.56 # # 18.79 20.07 #
#
#
#
#
# #
#
24.35 19.96 29.98 23.78 20.80#
#
# # ##
15.77
#
21.65
#
#
27.86 #
20.49
28.61# # #
21.57 24.86
17.25 20.08
# #
17.04#
N
0
0.8
5.405
#
95.228
Figure 17.10.
1.6 Kilometers
20.13
95.254
Maximum tsunami amplitudes (m) in West Coast of Banda Aceh area.
A review and listing of 26 December 2004 tsunami
5.882
#
#
Sabang #
#
Krueng Raya Port
#
#
## # # ## ## # #
# ## # # # ## # # ## # ## # # # #
5.384
# # # ## #
#### ### # ## ### #
# #
Centre of Banda Aceh
West of Banda Aceh Sigli
#
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
95.228
#
##
95.968
Sigli 5.389 5.384
4.40 3.68 #
#
4.24
95.960
###
4.07 4.82
95.968
Indonesia N
06
Figure 17.11.
12
Kilometers
Maximum tsunami amplitude (m) in Sigli area of Banda Aceh.
203
204
I. Nistor et al.
8.900
Khao Lak
8.742
### # # ##
Phuket
# # ##
Patong, Karon, Chalong Beaches # # ## ## #
#
## ##
##
Phi Phi Don
#
#
#
98.777
98.225
98.98
8.742
##
4.58
6.58 6.18 6.23
# #
# #
6.90 9.46
6.42
9.69 9.72 9.41 10.01 12.04 9.92 9.59
Kho Lak (North)
#
9.45
#
10.14 10.72 10.88 8.635
0
9.81 7.48
## ##
#
10.36 10.01
Figure 17.12.
THAILAND
Nai Yang Beach Bang Rong Pier Bang Thao Beach
7.002
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
#
8.80
9.91
9.90 ##
#
10.83 10.66 98.225 1
## #
#
2
## # # ## ## #
NO DATA
#
N
10.7
98.255
3 Kilometers
Maximum tsunami amplitude (m) in Khao Lak (north part) in Thailand.
Figures 17.12–17.14 show that distribution of maximum tsunami amplitudes, as surveyed by the Japanese team lead by Dr Y. Tsuji. Similar plot for Sri Lanka is shown in Figure 17.15 (amplitudes) for Indian mainland coast in Figure 17.16 (amplitudes) and for the Maldives in Figure 17.17 (amplitudes).
A review and listing of 26 December 2004 tsunami
8.9 00
205
Kh ao L ak
8.7 42
# ### # # # #
# # # #
Ph uk et
Pa to ng , K ar on, C hal o ng B ea ch es # # ## ## # # ## # ## #
#
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
TH A ILA N D
Nai Ya n g Be ac h Ba ng R ong Pie r Ba ng T h ao B ea ch
7.0 02
#
#
98. 225
8.27 2
Ph i P h i D on
##
98. 777 98. 98
4.28 4.53
#
#
Thai la nd
Kh ao L ak #
5.58 4.07
#
5.12 5.72 4.86
3.56
#
4.91 4.8 1.23
5.85 6.41 6.30 6.25
5.83 5.24 #
#
6.39
## # # # # ##
Ma kh am Ba y
#
4.99
4.92
#
##
#
3.28
# # #
#
5.3
#
Figure 17.13.
17.5
3.94
3.27 #
N
3.83
3.45 Frie n dship b ea ch #
#
#
98. 4 25
98. 2 80 10
#
#
3.53
7.77 2
2.43 #
Ka ro n be ac h
3.58
0
Lea m H im
#
#
Pa tong be ach
20
30
40 Kilo me t er s
Maximum tsunami amplitude (m) in Phuket beach area in Thailand.
CONCLUDING REMARKS
Tsunami arrival times and run-up values are listed based upon data posted by various survey teams on the Internet. The physical oceanographic processes that account for the behaviour of the Indian Ocean Tsunami of 26 December 2004 have been identified and it is suggested that these processes played a role to varying degrees in different places. Finally some pictorial presentations are also made of the run-up on some selected coastlines.
206
I. Nistor et al.
8.9 00
Khao Lak
8.7 42
# ### # # ##
Phuket
# # ##
Patong, K aron, Chalong B eaches # # # ## ## # ## # ## # #
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Phi Phi D on
Nai Yang B each Bang R ong Pier Bang Thao B each #
7.0 02
# ##
#
98.225
THAILAND
#
98.777 98. 98
Thailand #
#
# ### # # # #
## #
7.748
#
## #
5.32 # ## 6.89
4.36 4.99 4.71 5.36 4.90
5.00 5.02 2.50 2.46
Nai Y ang beach # #
# #
7.002
Figure 17.14.
Bang Rong Pi er
# #
Bang Thao beac h
N
98.772
98.296 0
THAILAND
Phi Ph i Don
40
80
120 Kilometers
Maximum tsunami amplitude (m) in Phi Phi Don area in Thailand.
A review and listing of 26 December 2004 tsunami
India
Sri Lanka
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
7.376
2.30 ## 1.80 1.60### 2.70 2.30 ## 2.60 # 2.70
11.30 # 8.40
13.384
# #
6.008
#
3.40 # 3.30
8.80
India
79.823
Figure 17.15.
100
6.607
81.497
200 Kilometers
Indonesia
Maldives
$
72.894
81.497
Maximum tsunami amplitudes (m) in Sri Lanka.
3.59 2.14 2.34 4.36 3.53 4.09 4.71
##
#
13.384
3.88 3.56 # #
4.93 4.01 3.16 3.25 2.93 4.94 5.33
##
#
#
3.39
#
# #
#
3.56 4.20 5.19 6.28 4.00
#
#
4.63 3.23 2.33 6.34
4.11 3.61 3.51 3.53
# #
India 5.06 4.21 3.69
#
# #
10.393
5.79 4.97
#
#
#
13.384
79.766
India
80.333 Sri Lanka
$$ $ $ $$ $ $ $$
N
90
0
90
180 Kilometers
6.607
% % % % % % % % % % % %
# ## # Sri L anka # ## ###
$ $
Figure 17.16.
Sri Lanka # # # # ## ##
$$ $ $$ $ $$ $ $ $
0
% % % % % % % % % % %
N
Maldives
$
72.894
Maximum tsunami amplitudes (m) in Indian east coast area.
81.497
Indonesia
207
208
I. Nistor et al. 3.43 2.92 2.88 2.43 2.43
1.91 2.93 2.76 2.45 2.85 2.22 2.49 2.13 1.34
# # # # # # #
2.37 1.59 2.02
6.766
2.45 2.96 2.13 2.83 2.37
Sri# Lanka # # ##
### #
#
# #
3.11 4.65 2.16 3.24 2.13 1.34 1.05
####
M a ld iv e
#
# #
# #
#
0.607
#
# ### #
1.32 2.00 1.40 72.894
# ## #
#
##
2.44 2.51 2.22
#
##
#
s
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
0.77 1.40 1.34 0.74 1.40 1.57
# # # #
India
1.60 1.90 1.75 1.52 1.59 1.60 1.93 1.98
2.68 3.06 2.69 3.10
2.89 3.40 3.68 1.94 2.30 2.67
2.36 3.25 2.91
13.384
N
India
73.582
Sri Lanka # ## # # ## ###
$ $ $ $$ $ $$$ $ $$
Figure 17.17.
0
300
600
6.607
900
1200
Indonesia
Maldives
$
72.894
300
% % % % % % % % % % % %
81.497
1500 Kilometers
Maximum tsunami amplitudes (m) in the Maldives.
REFERENCES Kowalik, Z., Knight, W., Logan, T. and Whitmore, P. (2005a). Numerical modeling of the global tsunami: Indonesian tsunami of 26 December 2004. Sci. Tsunami Hazards, 23(1), 40–56. Kowalik, Z., Knight, W., Logan, T. and Whitmore, P. (2005b). The Tsunami of 26 December 2004: Numerical Modeling and Energy Considerations, Proc. International Tsunami Symposium, Eds. G.A. Papadopoulos and K. Satake, Chania, Greece, June 27–29, 140–150. Murty, T.S. (1977). Seismic seawaves – tsunamis. Bulletin No. 198, Fisheries Research Board of Canada, Ottawa, p. 337. Murty, T.S., Rao, A.D. and Nirupama, N. (2005a). Inconsistencies in travel times and amplitudes of the 26 December 2004 Tsunami. J. Mar. Med., 7(1), 4–11. Murty, T.S., Nirupama, N., Nistor, I. and Rao, A.D. (2005b). Conceptual differences between the Pacific, Atlantic and Arctic tsunami warning systems for Canada. Sci. Tsunami Hazards, 23(3), 39–51. Murty, T.S., Nirupama, N. and Rao, A.D. (2005c). Why the earthquakes of 26th December 2004 and the 27th March 2005 differed so drastically in their tsunami-genic potential. Newslett. Voice of the Pacific, 21(2), 2–4. Murty, T.S., Nirupama, N., Nistor, I. and Hamdi, S. (2006a). Far field characteristics of the tsunami of 26 December 2004. ISET J. Earthq. Technol., 42(4), 213–217. Murty, T.S., Nirupama, N., Nistor, I. and Hamdi, S. (2006b). Why the Atlantic generally cannot generate trans-oceanic tsunamis. ISET J. Earthq. Technol., 42(4), 227–236. Murty, T.S., Rao, A.D., Nirupama, N. and Nistor, I. (2006c). Numerical modelling concepts for the tsunami warning systems. Curr. Sci. 90(8), 1073–1081. Nirupama, N., Murty, T.S., Nistor, I. and Rao, A.D. (2005a). Leakage of the Indian Ocean Tsunami Energy into the Atlantic and Pacific oceans, Canadian Association of Exploration Geophysicists Magazine “Recorder”, December. Nirupama, N., Murty, T.S., Rao, A.D. and Nistor, I. (2005b). Numerical tsunami models for the Indian Ocean Countries and States. Indian Ocean Survey, 2(1), 1–14. Nirupama, N., Murty, T.S., Nistor, I. and Rao, A.D. (2006). The energetics of the tsunami of 26 December 2004 in the Indian Ocean: a brief review. Mar. Geod., 29(1), 39–48.
CHAPTER 18
Overview and Integration of Part 2
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
N. Nirupama Atkinson School of Administrative Studies, York University, Toronto, Canada
18.1
MODELLING OF TSUNAMI GENERATION AND PROPAGATION
In this chapter, we will summarize the important points from the 12 chapters contained in Part 2 on modelling of tsunami generation and propagation. In Chapter 6, Nirupama et al. reviewed the classical concepts on the two types of dispersion that is relevant for tsunami dynamics. The frequency of phase dispersion occurs during the propagation of tsunamis, and it specifies the manner by which different frequencies propagate with slightly different velocities. On the other hand, amplitude dispersion also referred as non-linear effects dominates in the coastal aspects of tsunamis. The Ursell number sets out the criteria by which one can choose which types of models one should use in different phases of a tsunami event. One of the unsolved problems in tsunami research is the initial withdrawal or receding of the ocean before the arrival of the main tsunami waves. Part of the difficulty in explaining this stems from the fact that not all tsunamis give rise to initial withdrawal of the ocean. Even in a given tsunami, initial withdrawal occurs only at some locations, and not everywhere on all relevant coastlines. In Chapter 7, Nirupama et al. provide one possible explanation, invoking a mathematical model developed earlier by Spielvogel (1976). In Chapter 8, Nirupama et al. briefly reviewed the energetics of the Indian Ocean tsunami of 26 December 2004. This tsunamigenic earthquake is the second strongest in historical time, only surpassed by the Chilean earthquake of May 1960. This is the second global tsunami in historical time, the first one being the tsunami on 27 August 1983 from the eruption of the volcano Krakatoa in Indonesia. The 26 December 2004 event is the first truly global tsunami after modern instrumentation is put in place. Here, a global tsunami is defined as one that travelled at least into two other oceans, in addition to propagating fully throughout the ocean in which it is generated. In contrast to this event, the tsunami of 27 March 2005 with an epicenter somewhat south of the epicenter for the 2004 quake was insignificant. In Chapter 9, Nirupama et al. discuss the possibility of the amplification of the tsunami waves at the ocean surface through coupling with internal waves, generated through the density field due to thermal stratification of the ocean. It has been suggested that this amplification could be about 1 m. During the Indian Ocean tsunami of 26 December 2004, the observed tsunami amplitudes at certain locations on the coastlines around the Indian Ocean rim can be accounted for as having a contribution from possible amplification through coupling with internal waves. In Chapter 10, Kowalik et al. described a truly global ocean tsunami model, which they developed to simulate 26 December 2004 tsunami. They used a spherical polar coordinate system, to take the earth’s curvature into account and used a resolution of one-min arc in both the latitude and longitude. Their model covers the Indian, Atlantic and Pacific oceans, and stretches from 80◦ S to 69◦ N and has approximately 200 million grid cells to carry out this mammoth simulation, a parallel version of the model code was developed to run on a super computer. The 1 min of arc spatial resolution produced small numerical dispersion, even for global propagation of the tsunami. 209
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
210
N. Nirupama
In Chapter 11, Dimri and Srivastava described techniques for modelling the propagation of the Indian Ocean tsunami of 26 December 2004. They mentioned the importance of using better ocean bathymetry and coastal topography and the requirement of adequate resolution of the coastal geometry, through the use of fractals and finite element models with irregular triangular grids. They also discussed tsunami sources in the Andaman Sea as well as on the Makran coast of the Arabian Sea in addition to sources near Indonesia. In Chapter 12, Muraleedharan et al. discussed the application of the work energy theorem to compute tsunami runup heights. They validated their method by applying their technique to various coastal sections around the rim of the Indian Ocean, for the 26 December 2004 tsunami event. This work has a great practical utility in the sense that data such as this is very relevant for possible evacuation purposes during real tsunami events, as well as for long-term planning of coastal development against protection from tsunamis. In Chapter 13, Nirupama et al. describe how normal modes have a role in tsunami coastal effects. Normal modes of a water body are essentially free oscillations, which can be classified into two separate groups. The first group is the oscillations of the first class (OFC) and occur mainly due to gravity, and are somewhat influenced by earth’s rotation. These are also referred to as gravoid modes. Oscillations of the second class (OSC) owe their existence to earth’s rotation and have periods greater than one pendulum day. In Chapter 14, Nirupama et al. describe how the Helmholtz mode and the Kelvin–Sverdrup– Poincare (K–S–P) waves can be used to account for certain tsunami characteristics on the coastlines during the Indian Ocean tsunami of 26 December 2004. Certain harbours in the Indian Ocean experienced persistent high water levels up to several days after the tsunami event. This is due to Helmholtz resonance, in which long gravity wave energy of the tsunami enters a harbour through a narrow entrance channel, but successive reflections at the harbour walls leaks only a small amount of energy out of the harbour. K–S–P waves can be used to explain the propagation of tsunami energy near coastlines. In Chapter 15, Chittibabu and Murty provide information on all the models dealing with the Indian Ocean tsunami of 26 December 2004. All these models mostly deal with tsunami generation and propagation, and any coastal inundation that was simulated was only through simple algorithms and not through very detailed computations. The cut-off date for this review was 1 November 2005. The order in which the models are reviewed has no particular significance, except to point out that it is the order in which the authors saw these models, either on the Internet or in the published literature. The authors put in a caveat that there could easily be several other models that somehow escaped their attention. In Chapter 16, Nirupama et al. briefly reviewed some classical concepts of the Cauchy–Poisson (C–P) problem for tsunami generation and propagation. Tsunamis can be generated either through an initial elevation or an initial impulse (momentum), the classical C–P problem dealt with symmetric sources. The authors discussed the inclusion of asymmetric sources, viscosity, etc. In Chapter 17, Nistor et al. reviewed the tsunami travel times and runup data posted on websites by various groups that surveyed the aftermath of the 26 December 2004 tsunami in the Indian Ocean. They also discussed the various physical oceanographic processes that collectively can explain the tsunami behaviour on various coastlines of the Indian Ocean. These processes are: 1 2 3 4 5 6 7
Quarter wave resonance amplification in bays and gulfs. Helmholtz resonance in harbours. Constructive interference Boundary reflections. Interaction with astronomical tides. Coupling with internal waves due to ocean density gradients. Trapping of long gravity energy on continental shelves through OFC and OSC via the mechanism of trapped and partially leaky modes.
Overview and integration of Part 2
211
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
8 Interaction with the strong tidal current gradients near regular and degenerate semi-diurnal and diurnal tidal amphidromic points. 9 Extraction of energy from opposing ocean currents through Reynolds eddy stresses. 10 Interaction with the wind wave set-up, considering the fact that wind waves have the highest amplitudes in the Indian Ocean. 11 Focusing and defocusing of tsunami energy due to ocean bathymetric features such as ridges and trenches. 12 Phase or frequency dispersion and amplitude dispersion. 13 Zones of convergence and divergence. 14 Breaks in the continental shelves through which tsunami waves travelling through the deeper water interact with tsunamis in the shallow water. In summary, this section provides the theoretical background for numerical models dealing with tsunami generation, propagation and coastal inundation.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Part 3
Tsunami Detection and Monitoring Systems
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 19
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Satellite Detection of Pre-Earthquake Thermal Anomaly and Sea Water Turbidity Associated with the Great Sumatra Earthquake A.K. Saraf, S. Choudhury and S. Dasgupta Department of Earth Sciences, Indian Institute of Technology Roorkee, Roorkee, India J. Das Department of Earthquake Engineering, Indian Institute of Technology Roorkee, Roorkee, India
19.1
INTRODUCTION
Active tectonics along the Northeast moving Indian plate, which is subducting under the Burmese micro-plate with the Sunda trench as undersea surface expression, resulted in this century’s fourth largest earthquake of magnitude (Mw) 9.0 in Banda-Aceh, Sumatra on 26 December 2004 at 00:58 h (UTC). The earthquake originated at a shallow depth of about 30 km with the epicenter located at 3.32◦ N latitude and 95.85◦ E longitude (Figure 19.1). The tremors of the earthquake was felt at Banda-Aceh with intensity IX, at Meulaboh with intensity of VIII and at Medan, intensity IV, Sumatra and in parts of Bangladesh, India, Malaysia, Maldives, Myanmar, Singapore, Sri Lanka and Thailand with III–V intensity (http://neic.usgs.gov/neis/eq_depot/2004/eq_041226/). This mega-thrust earthquake generated a totally unexpected tsunami in the Indian Ocean and Bay of Bengal whose fury left more than 283,100 people dead, 14,100 missing and more than 1,120,000 displaced in numerous countries; the effects of which were felt as far as East Africa. The Indian Ocean and the Bay of Bengal has rare records of tsunamis in the past. This Tsunami is the direct consequence of about 1300 km long thrust faulting, which ruptured undersea displacing huge volumes of water in the Indian Ocean and the Bay of Bengal. Numerous aftershocks ranging in magnitude (Mw) from 3 to less than 7 rocked the region from Simeulue Island (Sumatra, Indonesia) in the south to near Landfall Island (north Andaman, India) in the north. Twenty-five aftershocks were of magnitude (Mw) 6.0 or greater, the strongest one being of magnitude 7.1 that occurred on 26 December 2004 at 04:21 h (UTC) (Figure 19.1). A major earthquake after the 9.0 magnitude (Mw) earthquake on Boxing Day, 2004 was unexpected, though this mega-event was followed by another powerful earthquake of magnitude (Mw) 8.7 on 28 March 2005 at 16:09:36 (UTC). This earthquake killed more than 1000 people and brought about huge damage in Simeulue. Several strong aftershocks followed the main earthquake event, including a 6.0 magnitude earthquake at 16:38 UTC and a 6.1 shock at 18:30 UTC (http://asc-india.org/events/050328_bob.htm). The 28 March 2005 earthquake was located at 2.074◦ N latitude and 97.013◦ E longitude, on a segment of the fault, 160 km to the southeast of the rupture zone of the magnitude 9.0 (Mw) Sumatra earthquake 215
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
216 A.K. Saraf et al.
Figure 19.1.
Location of the epicenter of the main shock of the 26 December 2004 mega-thrust earthquake in Banda-Aceh, Sumatra and the aftershocks. Also shows past seismicity of the region.
(http://neic.usgs.gov/neis/eq_depot/2005/eq_050328/neic_weax_ts.html). This earthquake is believed to be triggered by stress changes caused by the December 2004 earthquake. The movement of the Indian/Australian plate towards Eurasia is oblique and this movement is partitioned into thrust faulting and strike-slip faulting. The movement along the Sunda megathrust during the Sumatra earthquake has caused overriding of Burmese plate over the Indian plate (Figure 19.2). The rate of movement of the Indian plate in this zone is almost at a rate of 6 cm/year (http://neic.usgs.gov/neis/eq_depot/2004/eq_041226/neic_slav_ts.html). The rupture length (oriented parallel to the Sunda trench) of the thrust of the 26 December 2004 earthquake is estimated to be 1300 km, the width of the earthquake rupture measured perpendicular to the Sunda trench is about 150 km. The maximum displacement on the thrust plane is reported to be 20 m (http://neic.usgs.gov/neis/eq_depot/2004/eq_041226/neic_slav_ts.html).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Satellite detection of pre-earthquake thermal anomaly
217
Figure 19.2. Tectonics of the region around the epicenter of the devastating earthquake on 26 December 2004, the plate margin, where the India plate is being subducted beneath the Burma plate, along the Sunda trench. This active plate movement generates numerous earthquakes along the entire plate margin. The zone of plate movement stretches up to the Himalayan belt and results in the uplift of the Himalayan range.
19.2 THERMAL REMOTE SENSING IN EARTHQUAKE STUDIES The evolving techniques of remote sensing have the potential to contribute and assist human research studies in evaluating natural processes and events occurring daily on the earth’s surface on a global basis. Remote sensing now provides repetitive coverage, unbiased recordings of earth’s surface and events, cost effective technology, which consumes less time and multi-spectral information. The use of visible remote sensing in association with Geographic Information Systems (GIS) in earthquake studies has been a great stride in the last few years in disaster mitigation by delineating earthquake hazard zones and inducing management, analyzing spatial variables, monitoring of earthquake prone areas, in assessment and in identifying gap areas.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
218 A.K. Saraf et al. While GIS provides efficient analysis of natural hazards assessments, remote sensing apart from identification of damaged areas can monitor the earth on a real time basis for future signs for an impending danger. Now thermal remote sensing introduces more avenues for studying the earth. The earth’s emissivity can unfold many unknown natural processes associated with earthquakes. Any thermal anomaly in tectonically active regions occurring on the land surface can be monitored regularly and any abnormality, when other meteorological conditions, are normal may be an indication for an impending earthquake (Choudhury, 2005; Dasgupta, 2005). The history of the application of thermal remote sensing in natural resources perhaps started in Russia in the 1960s. Thermal data in seismic studies was first put to application in Russia in 1985 (Tronin, 2000) and the results were published in 1988 (Gorny et al., 1988). Case studies on earthquakes were carried out in Russia, Japan and China. This realization that earth’s surface temperature is significantly related to the earth’s physical processes has led to the development of an interesting trend of earthquake research.
19.3 THE CONCEPT OF THERMAL ANOMALIES The identification of a correlation between short-duration temporal thermal anomalies and earthquakes has been validated by the observation of the phenomenon for 12 past major earthquakes of different parts of the world (Saraf and Choudhury, 2003, 2004, 2005a–c). It has been observed in all these earthquakes there was a rise in the Land Surface Temperature (LST) around the epicentral region before the earthquakes. These short-duration temporal thermal anomalies went away along with the earthquake events. It has been discovered that there are sites in the earth’s crust near or around the epicenter of the impending earthquake, which most intensively change their characteristics with the processes preceding the earthquake, e.g., strain build up over an extended area. Such processes are known to be the cause of a number of observable effects, e.g., load uplift, changes in electrical properties of rocks, radon emission, etc. These phenomena are known as earthquake precursors. Recognition of thermal anomalies associated with earthquakes can lead to the identification of areas of increased seismic risk. Prior to an earthquake, crustal deformation is due to a stress field. It is a known fact that increases in pressure leads to an increase in temperature. Due to the acting stress field in tectonically active regions, sub-surface pressure increases with the consequent increase in temperature. Such deviation from normal in the thermal regime can pose to be an interesting observation in earthquake studies. It is also known that increase of stress might lead to release of green house gases like CO2 , CH4 , N2 , etc., trapped in the pore spaces of the rocks, which escape to the lower atmosphere and create a localized green house effect and thus augment the LST of the region (Figure 19.3). The crust disrupts under stress field prior to an earthquake and the existing gases in the crust such as CH4 are emitted above the ground or water surface (Zu-ji et al., 1999). Meanwhile, due to abrupt movement of underground electric water, electric particles transmit to the surface resulting in the mutation of the transient electric field. Gases like N2 , CO2 obtain energy from the electric field, change energy grade and release heat, thus creating impending earthquake precursors. Such study on temperature increasing mechanism is called Gas–Thermal Theory (Zu-Ji et al., 1999). Thermal changes due to stress fields have also been determined in laboratory studies of materials (Brady and Rowell, 1986; Zu-ji et al., 1997). A new theory of charge generation in rocks prior to earthquakes is given by Friedmann Freund (2002, 2003). This theory keeps parity with laboratory experiments (Zu-ji et al., 1997; Ouzounov and Freund, 2004) and also supports other explanations of pre-earthquake thermal anomalies and other unusual geophysical phenomena as well.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Satellite detection of pre-earthquake thermal anomaly
Figure 19.3.
219
Schematic model showing the generation of pre-earthquake thermal anomaly and detection by thermal remote sensing and meteorological stations in the ground.
Electronic charge carriers can be free electrons or sites of electron deficiency in the crystal structure, the latter known as p-holes. Crystallographically, rock-forming mineral structures practically are three-dimensional array of oxygen; where it is tacitly assumed that oxygen always occur in O2− oxidation state. Using different techniques (spectroscopically and using electrical conductivity, magnetic susceptibility, dielectric polarization measurements, etc.), presence of oxygen as O− state in silicate and oxide minerals has been confirmed (Freund, 2002, 2003). An O− is an anion or radical (also written as O• ) with an incomplete valence shell, 7 electrons 4+ instead of the usual 8. If the O− is part of a XO4+ 4 polygon (usually tetrahedron) (X = Si , 3+ 3+ O• − Al , etc.), it might be written either as XO4 or as O3 X/ . Being unstable radicals, O or O• O• and XO3+ form pairs, generating a positive hole pair (PHP), chemically equivalent to 4 or O3 X/ a peroxy anion, O− + O− = O2− 2 or a peroxy link as follows: •
O3 X/O +· O /XO3 = O3 X/OO \XO3 A PHP is devoid of any charge, as outer shells of both of the two O− atoms are filled. Such peroxy links can be thermodynamically stable, if they are associated with certain defect sites in the host crystal structure (Freund, 2002, 2003). Introduction of PHPs in minerals during rock-genesis and alteration has been explained by Freund (2002).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
220 A.K. Saraf et al.
Figure 19.4.
Concentration of positive charges (denoted by “+” sign) on the surface of the butterfly net, especially at the apex (maximum positive curvature). If the net is inverted by pulling up the thread, charges have the tendency to move towards the new outer surface.
Interestingly, O− –O− bond distance (1.5 Å) is almost half of O2− –O2− bond distance (2.8– 3.0 Å). It implies that the peroxy-bound O− has a small partial molar volume, thus having a tendency to be favored by high pressure. Being dissociated under high-accumulated stress a PHP introduces two holes (charge deficiencies) into the valence bond, causing the insulator to become a p-type semiconductor. Positive holes propagate though an oxide or silicate structure by electron hopping, whereby electrons from neighboring O2− can hop onto the O− site. The estimated maximum speed at which a positive hole could propagate by hopping is in the order of 100–300 m/s (Freund, 2002). Since the positive holes travel via the O 2p-dominated valence bond, they can easily cross grain boundaries without being scattered or annihilated. Laboratory experiments have proved the generation of heat and electric potentials on the surface of dry rock that is subjected to heavy stress (Ouzounov and Freund, 2004). Heat is especially generated along surfaces with maximum curvature, i.e., at the edges and corners. A thermal infrared camera can monitor the thermal emissions from the surface of the rock under stress. In rocks under high stress, PHPs dissociate into positively charged holes. In order to attain electrostatic stability, like any other movable charged particles, positive holes immediately rush towards the earth surface with high speed as mentioned above (Figure 19.4). Immediately after reaching a non-solid medium like atmosphere or water, a positive hole acquires an electron to become O2− from O− . The energy, therefore, emitted by this electron acquirement increases the surface temperature. 19.4
SATELLITE OBSERVATIONS OF THE GREAT SUMATRA EARTHQUAKE
Data from the sensor Advanced Very High Resolution Radiometer (AVHRR) on board the National Oceanic and Atmospheric Administration (NOAA) – series of satellites for the region around Sumatra and Myanmar were analyzed to detect any built up of temperature prior to the powerful earthquake in Banda-Aceh, Sumatra. Cloud free region showed a rise in temperature before the earthquake. This thermal anomaly appeared as a short-term temporal change in the land surface. 19.4.1
Data and methodology
Pre- and post-earthquake, NOAA–AVHRR Global Area Coverage (GAC) data, collected by National Environmental Satellite Data and Information Service (NESDIS) (Table 19.1) have been used to study the thermal regime of the land region surrounding the epicenter of the 26
Satellite detection of pre-earthquake thermal anomaly
221
Table 19.1. Time of acquisition of NOAA–AVHRR GAC data used to prepare LST time series maps to study pre-earthquake thermal anomaly.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
S. N. 1 2 3 4 5 6 7
Date
Time of GAC acquisition (UTC)
16 December 2004 17 December 2004 20 December 2004 21 December 2004 25 December 2004 26 December 2004 28 December 2004
03:11 to 05:04 02:46 to 04:41 03:19 to 05:12 02:55 to 04:49 03:05 to 04:58 02:41 to 04:36 03:36 to 05:30
Table 19.2. Time of acquisition of NOAA–AVHRR GAC data, used for analysis of seismically induced turbidity in the seawater by the mega-thrust earthquake of 26 December 2004. S. N. 1 2 3
Date
Time of GAC acquisition (UTC)
25 December 2004 26 December 2004 29 December 2004
03:05 to 04:58 02:41 to 04:36 03:13 to 05:07
Table 19.3. Time of acquisition of terra-MODIS data, used for analysis of tsunami induced turbidity in the seawater by the mega-thrust earthquake of 26 December 2004. S. N. 1 2 3
Date
Time of GAC acquisition (UTC)
22 December 2004 27 December 2004 29 December 2004
04:00 04:20 04:05
December 2004 earthquake in Sumatra. GAC data sets are available at 4 km spatial resolution. AVHRR has a temperature accuracy of 0.5◦ C. Channels 4 and 5 of NOAA–AVHRR are thermal channels, which can sense the earth’s emissivity during an overpass. Thermal channel 4 was used to calculate the LST of the study area. The LST calculation is based on the method provided in http://perigee.ncdc.noaa.gov/docs/klm/html/c7/sec7-1.htm#sec71-2. User specified temperature range of −50◦ C–35◦ C was used and temperature outside this range was masked. Cloud covers were delineated and avoided for any temperature calculation. For studying the turbidity conditions of the seawater after the fault rupture undersea, visible channel 1 (5.8–6.8 µm) of NOAA–AVHRR GAC datasets (Table 19.2) was used. This channel is best to measure the solar reflectance of the sea in the orange-red part of electromagnetic spectrum. Visible channel data from Moderate Resolution Imaging Spectroradiometer (MODIS) on board Terra satellite (Table 19.3) have been used to study the turbidity in seawater due to the devastating Tsunami generated by the earthquake.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
222 A.K. Saraf et al.
Figure 19.5.
19.4.2
NOAA–AVHRR data derived LST time series maps for the region around the epicentral region of the 26 December 2004 Sumatra earthquake. A thermal anomaly developed before the earthquake and went away along with the event. The LST was seen to be maximum on 25 December 2004, just one day before the earthquake.
Observations
LST time series maps derived from NOAA–AVHRR data sets were analyzed to study the thermal regime of the region around the Sunda trench (plate boundary on which overriding of the Burmese plate over the Indian plate has caused the mega-thrust earthquake). The entire region was quite cloudy around the period of the earthquake. AVHRR data cannot penetrate clouds. However, regions, which were free from clouds, were analyzed. The time series maps show that LST was normal in the beginning of December. Temperature started increasing around 16 December 2004 and was higher than normal beginning from 21 December 2004 (about 5 days prior to the main shock on 26 December 2004) in parts of Thailand (Figure 19.5). The temperature rose further on 25 December 2004 (1 day prior to the main shock). Clouds covered the entire region on 27 December 2004. But on 28 December 2004, the temperature was seen to be normal again. The
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Satellite detection of pre-earthquake thermal anomaly
Figure 19.6.
223
Seawater turbidity observed through NOAA–AVHRR data sets induced by the almost 1300 km fault rupture, which caused the great earthquake on 26 December 2004.
rise in temperature was about 4–6◦ C. The epicenter was offshore Banda-Aceh, Sumatra at a focal depth of 30 km and the surrounding region largely comprises of water bodies. The study of the behavior of Sea Surface Temperature (SST) before and after the earthquake was limited by the cloud covers all the time. The violent earthquake of magnitude 9.0 and associated aftershocks was caused by the vertical movement of the seafloor (estimated up to 20 m) over a large area (up to 1300 km along the Sunda trench, reaching up to Andaman Islands). Overriding of Burmese plate over the Indian plate has caused the deadly tsunami. The vertical and horizontal displacements from this earthquake caused slumping of the sea floor. Turbidity in the seawater was observed by AVHRR data as well as MODIS data. There were two kinds of turbidity in the water: (i) that which is introduced by the underwater slumping (with increase of total suspended solids (TSS) with landslides) in the rupture zone, which is visible along the line in which, the Sunda trench passes (Figure 19.6) and (ii) that which is brought about by the tsunami and is visible in the region along the coastlines and the shallow sea of Sumatra (Figure 19.7) and where ever the Tsunami hit.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
224 A.K. Saraf et al.
Figure 19.7. Turbidity observed through MODIS data sets near coasts in Sumatra induced by the tsunami after the great earthquake on 26 December 2004.
Channel 1 of NOAA–AVHRR was used to study the turbidity (because of the rupture and underwater slumping) conditions of the seawater. The GAC acquisition time was around 03:00 to 05:00 UTC; their common area of coverage around that time range from 2◦ N to 18◦ N latitudes and 92◦ E to 104◦ E longitudes. Cloud free scenes were analyzed for this study. The pre-earthquake scene (on 25 December 2004) shows that the seawater was clear before the earthquake (Figure 19.6). Turbidity in the seawater was seen to appear on 26 December 2004 in images acquired after the earthquake (Table 19.2). The clarity of the water was seen again 29 December 2004. Terra-MODIS data was used for analysis of the turbidity induced by tsunami after the earthquake. Satellite data on 22 December 2004 of the coastline of Sumatra shows a clear coast and a serene sea. On 27 December 2004 the seacoast was rough and turbid with suspended materials after the tsunami picked up sediments from the coast (Figure 19.7). Several coasts in the Indian Ocean were severely damaged, with human death and immense loss to properties. The suspended materials lessened, but stayed on till 29 December 2004. 19.5
DISCUSSION AND CONCLUSIONS
The thermal anomaly before the great Sumatra Earthquake appeared 5 days before the earthquake and intensified up to 1 day before the main event. On 25 December 2004, the thermal anomaly was seen to be maximum. The increase of temperature was seen to be around 4–6◦ C. The anomaly disappeared after the earthquake. The anomaly spread over a vast region, extending even up to
Satellite detection of pre-earthquake thermal anomaly
225
Thailand and Myanmar. Though such pre-earthquake thermal anomalies can provide important clues to future impending earthquakes, however, unlike thermal anomalies (increase in LST) observed before other earthquakes around the world, the anomaly observed before the Banda-Aceh (Sumatra) Earthquake has not been very convincing. The disadvantage in the study was due to presence of continuous cloud cover and less land parts surrounding the epicenter of the earthquake. Seawater turbidity induced by the fault rupture as well as induced by the tsunami due to the offshore mega-earthquake was observed through AVHRR and MODIS datasets.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
ACKNOWLEDGEMENT We acknowledge our thanks to Department of Science and Technology (Government of India), New Delhi, for their support and financial assistance for this study. REFERENCES Brady, B.T. and Rowell, G.A. (1986). Laboratory investigation of the electrodynamics of the rock fracture. Nature, 321, 488–492. Choudhury, S. (2005). Development of remote sensing based geothermic techniques in earthquake studies. PhD Thesis, Indian Institute of Technology Roorkee, India (unpublished). Dasgupta, S. (2005). Satellite geothermic techniques in earthquake studies. MTech Dissertation, Department of Earth Sciences, Indian Institute of Technology Roorkee, India (unpublished). Freund, F. (2002). Charge generation and propagation in rocks. J. Geodyn., 33, 545–572. Freund, F. (2003). Rocks that crackle and sparkle and glow: strange pre-earthquake phenomena. J. Sci. Explor., 17(1), 37–71. Gorny, V.I., Salman, A.G., Tronin, A.A., and Shilin, B.B. (1988). The earth outgoing IR radiation of the earth as an indicator of seismic activity. Proc. Acad. Sci. USSR, 301, 67–69. Ouzounov, D. and Freund, F. (2004). Mid-infrared emission prior to strong earthquakes analysed by remote sensing data. Adv. Space Res., 33, 268–273. Saraf, A.K. and Choudhury, S. (2003). Earthquakes and thermal anomalies. Geospatial Today, 2(2), 18–20. Saraf, A.K. and Choudhury, S. (2004). Thermal remote sensing technique in the study of pre-earthquake thermal anomalies. J. Indian Geophys. Union, 9(3), 197–207. Saraf, A.K. and Choudhury, S. (2005a). NOAA–AVHRR detects thermal anomaly associated with 26 January, 2001 Bhuj Earthquake, Gujarat, India. Int. J. Remote Sens., 26(6), 1065–1073. Saraf, A.K. and Choudhury, S. (2005b). Satellite detects surface thermal anomalies associated with the Algerian Earthquakes of May 2003. Int. J. Remote Sens., 26(13), 2705–2713. Saraf, A.K. and Choudhury, S. (2005c). Thermal remote sensing technique in the study of pre-earthquake thermal anomalies. J. Indian Geophys. Union, 9(2), 197–207. Tronin, A.A. (2000). Thermal IR satellite sensor data application for earthquake research in China. Int. J. Remote Sens., 21(16), 3169–3177. Zu-ji, Qiang, Kong, Ling-Chang, Zheng, Lan-Zhe, Guo, Muan-Hong, Wang, Ge-Ping, and Zhao, Yong (1997). An experimental study on temperature increasing mechanism of satellite thermo-infrared. Acta Seismologica Sinica, 10(2), 247–252. Zu-ji, Qiang, Chang-gong, Dian, Lingzhi, Li, Min, Xu, Fengsha, Ge, Tao, Liu, Yong, Zhao, and Manhong, Guo (1999). Satellite thermal infrared brightness temperature anomaly image – short-term and impending earthquake precursors. Sci. China, 42(3), 313–324.
URL LINKS http://neic.usgs.gov/neis/eq_depot/2004/eq_041226/ http://asc-india.org/events/050328_bob.htm http://neic.usgs.gov/neis/eq_depot/2005/eq_050328/neic_weax_ts.html http://neic.usgs.gov/neis/eq_depot/2004/eq_041226/neic_slav_ts.html http://perigee.ncdc.noaa.gov/docs/klm/html/c7/sec7-1.htm#sec71-2
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 20
Possible Detection in the Ionosphere of the Signals from Earthquake and Tsunamis
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
T.S. Murty Department of Civil Engineering, University of Ottawa, Ottawa, Canada N. Nirupama Atkinson School of Administrative Studies, York University, Toronto, Canada A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology, New Delhi, India I. Nistor Department of Civil Engineering, University of Ottawa, Ottawa, Canada
20.1
INTRODUCTION
The standard technique of identifying the existence of a tsunami event in real time is through its record on a tide gauge. However, since there could be an elapsed time interval of several minutes to an hour, or sometimes even longer, for the tsunami waves to arrive at a tide gauge, the available warning time for evacuation purposes, if needed, is somewhat reduced. One of such technologies is the possible detection of the earthquake and tsunami signals in the ionosphere. The basic mechanism is the possible amplification of the atmospheric pressure signal due to the tsunami in the ionosphere through atmospheric acoustic- and internal gravity waves. This is another area of tsunami research where most of the work was done earlier and the references are somewhat dated. After the 26 December 2004 tsunami in the Indian Ocean, there is renewed interest at present in this topic. In considering any wave motion in the atmosphere, both gravitational and compressional forces must be considered. Hines (1960) showed that even a simple model of the atmosphere (which is stationary and of uniform temperature and composition) can produce two distinct classes of wave motion with one feature in common; in the absence of dissipative forces both are unattenuated in the horizontal direction. However, in their vertical behavior they differ drastically. The class identified as surfaces waves, although having an exponential variation vertically, cannot support any phase propagation in the vertical direction. The class identified as internal waves can support significant phase propagation in the vertical direction. The internal atmospheric waves are of two types (Obayashi, 1963). In the high-frequency range they are acoustic (sound waves in the atmosphere), whereas in the low-frequency range they are internal gravity waves. Low frequency refers to a period of several minutes when gravitational forces introduce a profound anisotropy. 227
228 T.S. Murty et al. 20.2 TWO CHARACTERISTIC FREQUENCIES
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Two characteristic frequencies separate the two limiting cases. One is the Brunt–Vaisala frequency, ωg , also called the gravitational stability frequency, and the other is the modified acoustic stability frequency, ωa . For an atmosphere in hydrostatic equilibrium, the two frequencies are given by: g 1 ∂ρ g γ − 1 ∂H ωg2 = −g 2 + = + (20.1) c ρ ∂z H γ ∂z and ωa2
=
c ∂ρ 2ρ ∂z
2
γg = 4H
∂H 2 1+ ∂z
(20.2)
Here g is acceleration due to gravity, ρ(z) is the density of the atmosphere, c(z) is the speed of sound, γ is the ratio of specific heats, and H is the height (scale height, H , is the equivalent height of a homogeneous, isothermal atmosphere and is given by KθA /mg where K is Boltzmann’s constant, θA is the absolute temperature, g is gravity, and m is the mean molecular mass) of the atmosphere (a function of the vertical coordinate, z). The phase velocity, V , of the waves is given by: 2 V 1 − Y sin2 θ = c 1−X
(20.3)
where X ≡
ω 2 a
ω
and
Y ≡
ω 2 g
ω
(20.4)
and θ is the angle between the field of gravity and the wave normal. Here ω is the frequency and T = 2π/ω is the period. Waves with periods less than Ta are acoustic waves; those with periods greater than Tg are gravity waves. In the period range, Ta < T < Tg , no internal atmosphere waves exist. For acoustic waves period less than Ta , the phase velocity is approximately equal to the velocity of sound but tends to ∞ as the periods approach Ta . For gravity waves, the phase as well as the group velocities are zero at T = Tg , and the velocities increase with increasing T and approach (Y /X )sin θ for longer periods. The acoustic waves more or less propagate isotropically whereas the gravity waves propagate mainly in a horizontal direction and exhibit strong anisotropy.
20.3
FURTHER THEORETICAL DEVELOPMENTS
Although Hines’ original work involved only an isothermal atmosphere, subsequent models used many layers to model realistically the variation of temperature with height. For example, Harkrider (1964) used 40 layers, each a thin isothermal layer. Many numerical studies have been made for the wave-guide modes of acoustic and internal gravity waves for different atmospheric models, all more or less derived from the Air Research and Development Command (ARDC) standard atmosphere (Figure 20.1). For periods ranging from a few seconds to a few minutes, these studies were satisfactory in the sense that the calculated values agreed well with observed values.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Detection of signals from earthquake and tsunamis in the ionosphere
Figure 20.1.
229
Standard and extreme ARDC atmospheres (Harkrider, 1964).
Several authors introduced wind structure and considered its effects on the gravity waves. The work of Weston and vanHulsteyn (1962) is typical. They showed that wind increases the phase velocity of the gravity-wave mode. Some important results of the theoretical studies are: in the long-period wave range of 10–100 min, propagation is mainly in the form of gravity waves. There has been some confusion in the atmospheric propagation problems because the distinction between acoustic modes influenced by buoyancy effects, and internal gravity-wave modes influenced by compressibility effects, has not been clear in the commonly used term acoustic-gravity wave propagation. Actually, as far as the lower 100 km or so of the atmosphere is concerned, the behavior of gravity and acoustic modes is sufficiently different that no confusion arises. The difference in behavior could be easily visualized in terms of the two limiting frequencies already introduced; the acoustic low-frequency cutoff, ωa , and the Brunt–Vaisala frequency or the internal gravity wave, high-frequency cutoff, ωg . In the literature several notations have been used for these; Tolstoy and Pan (1970) used ω0 and N . In the regions where the ωa > ωg , there is no ambiguity in the two types of modes. However, centered at about 120 km in the atmosphere there is an anomalous zone in which ωg > ωa . This anomalous zone is probably important for ground-level detection of surface-gravity waves excited by atmospheric nuclear explosions. Tolstoy and Pan (1970) showed that simple models involving two to four layers in the atmosphere, can be used to determine the propagation and dispersion of the lowest two to three gravity modes of the atmospheric wave guide. Two important facts brought out by this paper are: these
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
230 T.S. Murty et al. simple models could explain the long-period portion of the 300 m/s group velocity plateaus in terms of coupling effects of the internal gravity wave guide. Tolstoy and Pan cautioned that, as this result is obtained from an atmospheric model with two incompressible layers, the 300 m/s value is somewhat fortuitously related to the velocity of sound in the air. Also, surface gravity-wave modes of the atmosphere are better understood by this approach. Another concept introduced pertains to the top boundary condition in the atmosphere. For example, Pfeffer and Zarichny (1962), Press and Harkrider (1962), and Harkrider and Wells (1968) used a free-surface condition, at several heights arbitrarily, to compute surface modes of the atmosphere, but did not explain the logic of this. Tolstoy (1967) showed that it is indeed impossible to justify the use of a free-surface boundary condition, and thereby make the selection of H , the effective atmospheric wave height, less arbitrary. For this he used the concept of vacuum, i.e., at height where the mean free path, , of the neutral gas molecules exceeds the wavelength, L, of the disturbance, the medium is considered to be vacuum. Hence, for heights where L, the usual continuum equations apply. Then the effective free surface of the atmosphere is at height, z = H , where = L. This means the effective height of the atmosphere depends on the wavelength of the disturbance. However, the region, Z = h, at which L is not infinitesimal, but is a transitional layer of thickness, d (say). In this transitional layer, strictly speaking, Boltzmann’s equations must be applied. For d L, the layer mainly behaves as a region with large viscosity and strongly attenuates waves with periods less than 10 min. Another consideration is that (at the heights relevant here) because of large ionization, hydromagnetic interactions could occur. However, Dungey (1954), Fejer (1960), and Hines (1955) showed that these interactions are effective only for periods greater than or equal to 3 h. Hence, it can be assumed that the atmosphere acts as a window for surface-gravity waves, with periods between 10 and 200 min. 20.4 TOLSTOY AND PAN’S MATHEMATICAL MODEL: THE CONCEPTS The simple mathematical model of Tolstoy and Pan (1970) for wavelengths L > 300 km will now be discussed, with particular attention to aspects of the propagation directly relevant to recent observations with microbarographs. They showed motions with 600 m/s and energy concentrated around the 15-min period. The logic behind using a simple model is given by Tolstoy and Pan (1970, p. 35): “Since in the present study we are interested chiefly in internal and surface gravity modes with periods >10 min and wavelengths >200 km, a small number of layers provides adequate approximations for discussing the propagation properties of atmospheric waves. One must remember that the frequency–wave number relationship and the group velocity may be written as quotients of two quadratic forms in the wave amplitudes (Boit, 1957). These are stationary with respect to variations of the amplitude, so that in most calculations relatively large errors in the amplitude distribution of the displacement field can be tolerated, without appreciably affecting the eigenvalue and propagations velocity calculations. Thus, even though amplitude of the vertical displacement predicted by an oversimplified model may be in error, the characteristic curve calculations can be quite accurate. We have limited ourselves to calculations on two and four-layer models, illustrating the effects of compressibility, of the upper boundary condition, of the earth’s rotation, and of layering upon the propagation of long periods. Within the limitations imposed by the small number of layers, we have used models that represent fits to the Vaisala frequency, the density, and sound velocity functions in the earth’s atmosphere.”
Tolstoy and Pan’s model allows interpretation of recent observations by microbarographs. Tolstoy and Herron (1970) also gave a new interpretation of the traveling disturbances in the ionosphere due to nuclear explosions. This is possible because in the moderately long-period range of 10–30 min there is considerable separation between the velocities of surface- and internal gravity waves.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Detection of signals from earthquake and tsunamis in the ionosphere
231
The main difficulty appears to be in the magnitude of the pressure signal. Harkrider and Wells (1968) showed that the observed pressure amplitudes from peak to trough of 10–100 µb would require unrealistically large vertical displacements of the atmosphere above 100 km. Other calculations with simple models essentially confirm their result. Tolstoy and Pan (1970) attributed this to ignoring the anomalous zone between 110 and 200 km, and ignoring the influence of the winds at high altitudes. When these are taken into account the difficulty disappears. That is, the vertical displacements of the atmosphere above 100 km need not be ridiculously large to be consistent with ground-level pressure amplitudes in the range of 10–200 µb. Next Tolstoy and Pan considered the question of the attenuation of the surface-gravity waves and showed that generally the attenuation is small for periods greater than 10 min for the modes m = 0 and 1. Liu andYeh (1971) studied the excitation of acoustic-gravity waves in an isothermal nonrotating atmosphere; because their model is highly idealized it is not included here. Yeh and Liu (1972) gave a good review of the propagation of waves in the ionosphere. 20.5
SIMILARITIES BETWEEN EARTHQUAKE (AND TSUNAMI) SIGNALS AND ACOUSTIC-GRAVITY WAVES FROM ATMOSPHERIC NUCLEAR EXPLOSIONS
Ionization is the phenomenon when ion pairs of separated electrons and positive ions are formed in the ionosphere. There are mainly three regions in the ionosphere – D, E, and F regions. In the daytime, and especially in summer, the F-layer splits in two, F1 and F2 . The electron density is a maximum in each region. However, the electron density increases generally with altitude, i.e. it is greater for the F region than for the E region, which has greater electron density than the D region. In the lower part of the atmosphere, i.e. below 48.3–56.3 km, the air is dense and the probability of collisions between free electrons and atmospheric atoms and molecules is great. Because of this, electrons are rapidly attached to the neutral particles and ionization cannot be produced. Even if it is produced, it is destroyed immediately because, on the average, the life of a free electron in this part of the atmosphere is less than a microsecond. In the ionosphere, electrons and ions are produced by interaction of solar radiation with atoms and molecules of the atmospheric constituents. They are destroyed by combining either with neutral particles or positive ions. Of the two destructive processes, the former dominates at higher levels where electron density is greater. The effect of a nuclear explosion on atmospheric ionization is mainly because of an increase in electron density in the surrounding region. The added electrons can effect all electromagnetic communication, either by attenuating the signal or by changing its direction of propagation through refraction. Donn and Shaw (1967) studied the atmospheric nuclear tests conducted by the USA and USSR during 1952–1962 and, based on the data recorded by a global network of stations maintained by the Lamont Geological Observatory of Columbia University, they arrived at interesting conclusions about pressure waves due to those explosions. In fact, effects of the high-yield atmospheric nuclear explosions are comparable to those of the enormous Krakatoa eruption of 1883 and the impact of the Siberian meteor in 1908, when pressure waves in the atmosphere traveled round the globe more than once, with appreciable amplitudes. Donn and Shaw used the data from sensitive microbarographs at 15 recording stations, and acquired 208 records of 45 nuclear explosions. The dispersion relations of the acoustic-gravity waves in the atmosphere were used to analyze the records and the authors concluded that: “The initial spherical wave at the source, which is modified into a cylindrical wave by the layered structure of the atmosphere, is composed of broad spectrum of pressure waves whose frequencies,
232 T.S. Murty et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
which propagate away at about the speed of sound in air, range from audible sound to about 0.002 cps. At distances of a thousand or more kilometers, the spectrum becomes considerably narrowed, with the highest frequency detectable being about 0.03 cps (30 s in period). It is more convenient to refer to period rather than frequency for these infrasonic waves. Such waves are also referred to as acousticgravity waves, since their propagation characteristics are controlled by both gravity and the acoustic properties of the atmosphere”. “Because the atmosphere is a dispersive medium for the long waves within this period range (30– 500 s), the initial impulsive because dispersed into a train of the temperature and wind stratification of the atmosphere along the propagation path.”
Daniels et al. (1960) studied vertically traveling shock waves in the ionosphere caused by nuclear explosions at the surface of the earth. They interpreted the distortions in ionosphere recordings as the result of retarded sound waves. Daniels and Harris (1961) reinterpreted the record (because the velocity of the shock wave was 115 m/s) as an ordinary hydrodynamic shock wave. They also mentioned that the pressure amplitude of this upward traveling shock wave was large enough to be detected by acoustic methods at ground level. Webb and Daniels (1964) reported ionospheric oscillations following the Soviet atmospheric nuclear test of 1 November 1962. They measured the rotations of the polarization plane of the radio waves during their travel through the ionosphere, and this parameter was proportional to the electron density. They used a 151-MC/S transmitter at Fort Monmouth, NJ, to reflect signals from the moon. The signals were received at the University of Illinois at Danville, between 11:00 and 18:00 CST on 1 November 1962. Thus the signals traveled twice through the ionosphere. From newspaper and other reports it was known that the Soviet Union had conducted a nuclear test in the atmosphere a few hours earlier. The record at Danville showed an oscillation with 30-min period, larger than the period recorded by the ground-level microbarographs. Webb and Daniels attributed the oscillations to acoustic waves in the atmosphere caused by the explosion. These acoustic waves are expected by the theory to have a group velocity of about 730 m/s, but the authors were unable to check this from the oscillations that may have started before the record began. Long-period ionosphere oscillations have been known to continue for several hours following a nuclear explosion (Daniels and Harris, 1958). Tolstoy and Herron (1970) studied the atmospheric gravity waves excited by nuclear explosions during 1967–1968. The data used were from the large aperture, i.e. a 250 km × 200 km array of 6–12 low-frequency microbarographs with period pass-band 1–60 min, in the New York–New Jersey area. The spectra of these gravity waves showed peaks at 15-min periods and the average group velocity was approximately 600 m/s. The authors deduced that these were indeed atmosphere surface-gravity waves, based on the agreement of the dispersion and attenuation of the recorded waves with expected theoretical relations. Tolstoy and Herron (1970, p. 59) concluded that the 600 m/s, 15-min period arrivals were probably surface-gravity waves traveling along what is effectively the top of the atmosphere. The height, H , of this “effective free surface” depends on the wavelength, L, as it corresponds approximately to the region where the mean free path, , of air molecules is of the order of L (indeed, the condition explicitly applied in making calculations is 2π /L = 1). They thought that a number of published ionosphere observations of fast traveling disturbances generated by US and USSR thermonuclear tests in the early 1960s could probably be explained in a similar way. It appears that reported ionospheric disturbances with horizontal group velocities ≥500 m/s are surface-gravity waves, whereas, lower velocities correspond to internal gravity waves (Hines, 1967). Acoustic modes of propagations (Wickersham, 1966) are possible for spectral components with periods shorter than 10 min. Earlier the problem of the pressure amplitude of atmospheric gravity waves was mentioned. To be consistent with the pressure amplitudes at the ground, it appeared at the outset that
Detection of signals from earthquake and tsunamis in the ionosphere
233
unrealistically large vertical displacements of the atmosphere above 100 km altitude would be needed. Later the problem was explained as neglect of an anomalous zone. Tolstoy and Herron (1970, p. 60) proposed that:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
“As much oversimplified one-layer model of the atmosphere (Tolstoy, 1967) can be displacement at these heights, a result which would be quite adequate where it is not for the fact that a more realistic two-layer model decreases this estimate to 0.2 µb or less. Proper inclusion of the effects of a layer between altitudes of 110 and 150 km, in which the acoustic cut-off frequency of the medium is lower than the Vaisala frequency, might eliminate the discrepancy; the problem is discussed in more detail elsewhere (Tolstoy and Pan, 1970).”
Hines (1967) clarified some confusion that arose in connection with the ionospheric disturbances caused by the Soviet upper atmospheric nuclear test over Novaya Zemlaya on 30 October 1962. Obayashi (1962, 1963) attributed these perturbations in the F-layer critical frequency to atmospheric surface-gravity waves with periods in excess of 10 min. Wickersham (1966) differed and proposed that the ionospheric disturbances were due to fully ducted acoustic-gravity waves in the atmosphere. Hines showed that Wickersham’s interpretations is not well founded and Obayashi’s original interpretation is correct.
20.6
SUMMARY
There was considerable work done in the 1950s and 1960s on the possible detection of atmospheric pressure signals in the ionosphere, generated from earthquakes and tsunamis. The linkage is, possible amplification of the signal through acoustic- and internal gravity wave spectrum. However, after the initial burst of earlier research, there was not much work done on this topic. After the Indian Ocean Tsunami of 26 December 2004, there is a renewed interest in this topic at present. REFERENCES Daniels, F.B. and Harris, A.K. (1958). Note on vertically traveling shock waves in the ionosphere. J. Geophys. Res., 66, 3964. Danniels, F.B., Bauer, F.J. and Harris, A.K. (1960). Vertically travelling short waves in the ionosphere, J. Geophy. Res., 65, 1848–1850. Danniels, F. and Harris, A.K. (1961). Note on vertically short waves in the ionosphere, J. Geophy. Res., 66, 3964. Donn, W.L. and Shaw, D.M. (1967). Exploring the atmosphere with nuclear explosions. Rev. Geophys., 5, 53–82. Dungey, J.W. (1954). The propagation of alfven waves through the ionosphere. Pennsylvania State University Ionosphere Research Laboratory Science Report, 57, 19p. Fejer, J.A. (1960). Hydromagnetic wave propagation in the ionosphere. J. Atmos. Terr. Phys., 18, 135–146. Harkrider, D.G. (1964). Theoretical and observed acoustic-gravity waves from explosive sources in the atmosphere. J. Geophys. Res., 69, 5295–5321. Harkrider, D.G. and Wells, F.J. (1968). The excitation and dispersion of the atmospheric surface wave. Proceedings of the Symposium Acoustic-Gravity Wave, Boulder, CO, pp. 299–313. Hines, C.O. (1955). Hydromagnetic resonance in ionospheric waves. J. Atmos. Terr. Phys., 7, 14–27. Hines, C.O. (1960). Internal atmospheric gravity waves at ionospheric heights. Can J. Phys., 38, 1441–1481. Hines, C.O. (1967). On the nature of traveling ionospheric disturbances launched by low altitude nuclear explosions. J. Geophys. Res., 72, 1877–1882. Liu, C.H. and Yeh, K.C. (1971). Excitation of acoustic-gravity waves in tan isothermal atmosphere. Tellus, 23, 150–163. Obayashi, T. (1962). Wide-spread ionospheric disturbances due to nuclear explosions during October 1961. Jpn. Ionosphere Space Res. Rep., 16, 334–340.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
234 T.S. Murty et al. Obayashi, T. (1963). Upper atmospheric disturbances due to high altitude nuclear explosions. Planet Space Sci., 10, 47–63. Pfeffer, R.L. and Zarichny, J. (1962). Acoustic-gravity wave propagation from nuclear explosions in the earth’s atmosphere. J. Atmos. Sci., 19, 256–263. Press, F. and Harkrider, D. (1962). Propagation of acoustic-gravity waves in the atmosphere. J. Geophys. Res., 67, 3889–3908. Tolstoy, I. (1967). Long period gravity waves in the atmosphere. J. Geophys. Res., 72, 4605–4622. Tolstoy, I. and Pan, P. (1970). Simplified atmospheric models and the properties of long-period internal and surface gravity waves. J. Atmos. Sci., 27, 31–50. Tolstoy, I. and Herron, T.J. (1970). Atmospheric surface gravity waves from nuclear explosions. J. Atmos. Sci., 27, 55–61. Webb, H.D. and Daniels, F.B. (1964). Ionospheric oscillations following a nuclear explosion. J. Geophys. Res., 69, 545–546. Weston, V.H. and VanHulsteyn, D.B. (1962). The effects of winds on the gravity waves. Can. J. Phys., 40, 797–804. Wickersham, A.F., Jr. (1966). Identification of acoustic-gravity wave modes from ionospheric range time observations. J. Geophys. Res., 71, 4551–4555. Yeh, K.C. and Liu, C.H. (1972). Propagation and application of waves in the ionosphere. Rev. Geophys. Space Phys., 10, 631–709.
CHAPTER 21
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Seismo-electromagnetic Precursors Registered by DEMETER Satellite A.K. Gwal Space Science Laboratory, Department of Physics, Barkatullah University, Bhopal, India S. Sarkar LPCE/CNRS, Orleans, France S. Bhattacharya Space Science Laboratory, Department of Physics, Barkatullah University, Bhopal, India M. Parrot LPCE/CNRS, Orleans, France
21.1
INTRODUCTION
Disturbances in the ionosphere linked with seismic activity have been studied from a long time. Many results have been published on satellite observation of electromagnetic and ionospheric perturbations apparently associated with seismic activity (Parrot et al., 1993; Gokhberg et al., 1995; Hayakawa, 1997; Liperovsky et al., 2000). Probably the first results on local plasma density and temperature variations measured onboard AE-C and ISIS-2 satellites were published by Gokhberg et al. (1983). Boskova et al. (1993, 1994) had observed the changes in ion composition before earthquakes over the earthquake preparation zone. Afonin et al. (1999) analyzed a large database of plasma densities recorded in the 3000 orbits of Intercosmos-24. They reported a reliable correlation between the global distribution of seismic activity and ion density variations in the ionosphere, as measured by the normalized standard deviation (NSD) and the relative-normal standard deviation (RNSD). The statistical studies of Afonin et al. (2000) based on Cosmos-900 data and the results of Pulinets and Legenka (2003) had shown the existence of large-scale irregularities in the ionosphere several days or hours before strong earthquakes. Furthermore, significant contributions in the past have reflected on detection of ULF/ELF/VLF emissions at the time of earthquakes by low altitude satellites. Investigation of data by Intercosmos-24 satellite has revealed strong ELF/VLF perturbations associated with earthquakes (Molchanov et al., 1993). Parrot (1994) has performed statistical study of ELF/VLF emissions connected to electric and magnetic field by AUREOL 3 satellite and found increase in the emissions during earthquakes. In spite of several results published, the database collected is not sufficient to reveal the nature of ionospheric earthquake precursors. It is essentially incomplete because only isolated ionospheric effects of individual seismic events have been detected until recently. Also, all these effects have been observed as an additional output of experiments aimed at study of other phenomena, 235
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
236 A.K. Gwal et al. mainly solar-terrestrial events. Today we cannot be certain if we have regular manifestations of lithosphere–ionosphere coupling or some accidental coincidences of seismic and ionospheric activities occurred (Parrot, 1999). The only way to proceed in the right direction is to carry out high sensitive regular satellite observations in the ionosphere over seismically active and quiet regions supported by ground operations. This will allow to create a sufficient database for statistical study of the seismo-ionospheric effects. It is with this aim that the French microsatellite DEMETER was launched on June 29, 2004 and one of its main scientific objective is to detect anomalous variations of the ionospheric parameters, which could be related to seismic activity. If it can be shown that such perturbations are real and systematic they could be considered as short-term precursors, occurring between a few hours and a few days before the earthquake. Up to now, DEMETER is the only satellite which has the capability to survey on a vast scale the Earth’s electromagnetic environment in the ionosphere. The database which will be built up during this mission will allow us to perform statistical analysis, which will be helpful in reliable detection of ionosphere-earthquake precursors. Some interesting cases of ionospheric irregularities over seismic regions prior to seismic activity identified by the DEMETER records are presented in this chapter. Additionally, some important results of ULF/ELF electric field perturbations found prior to seismic activities have also been presented to emphasize on the correlation of electromagnetic emissions with earthquakes. 21.2 THE DEMETER SATELLITE The French microsatellite DEMETER has been launched on June 29, 2004. The main scientific objective of this mission is to study the ionospheric perturbations, which are linked to seismic activity. The orbit of DEMETER is polar, circular with an altitude of 710 km. There are several sensors onboard DEMETER to survey the ionosphere. The Langmuir Probe Instrument ISL is designed to measure the electron density of plasma (in the range 102 −5.106 particles/cm3 ), electron temperature (in the range ±5 V). An energetic particle analyzer IDP gives the electron energy flux. An ion spectrometer IAP experiment measures the ion composition, density and temperature. A search coil magnetometer IMSC measures the three components of magnetic field in the frequency range from a few Hz up to 20 KHz. A set of four electric sensors ICE to perform a continuous survey of the DC and AC electric fields over a wide frequency range from ULF to HF. Details about these experiments can be found in Berthelier et al. (2006a, b), Lebreton et al. (2006), Parrot et al. (2006) and Sauvaud et al. (2006). DEMETER has two scientific modes of operation: (i) the survey mode collecting averaged data all round the earth (ii) the burst mode collecting data with a bit rate of 1.7 Mb/s above seismic regions. Data and plots are available through a web server dedicated to this mission (http://demeter.cnrs-orleans.fr). Different types of data and the associated products are available on this web server. For data selection we have used the quicklook (QL) images available on the web server. These images give a quick presentation of the data over one half-orbit. All the scientific experiments are presented in a portrait image (format postscript). The standard quicklook image is made from the quick view (QV) experiment frames plus a 14th frame containing earthquake events and with an addition information on the orbital parameters. To get an overview of the science DEMETER payload results we used the QL images. To identify the seismo ionospheric precursors we used the Level 2 data available on the DEMETER web server. Level 2 data processing corresponds to high-resolution plots of the calibrated data, which are obtained from Level 1. The Level 2 image is created by the user itself on the data server, which gives facilities to personalize the output image. To search for earthquakes close to the orbit of DEMETER we have used the web server. We have reduced the DEMETER orbits relative to the earthquake information. Earthquakes selected were greater than 5 in magnitude. Depth was taken up to 1000 km.The distance between the satellite and the epicenter was taken to be 900 km. The time interval between the time of quake
Seismo-electromagnetic precursors registered by DEMETER satellite
237
and data was taken to be 10 days. All these parameters can be changed by the user in the web server. All data files and plots are organized by half-orbits. After selecting the seismic events the QL images were checked for any seismo ionospheric variation. If some variation was found we plotted the Level 2 image. In these images we found good correlation of ionospheric perturbations with the seismic events.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
21.3
EVENTS
We discuss here a series of earthquakes that occurred in the Sumatra region of Indonesia. The main shock of the Sumatra earthquake (2.07◦ N, 97.01◦ E) occurred at 16:09:36 UT on March 28, 2005 with M = 8.7. It was followed by aftershocks. The time and location of these earthquakes is given in Table 21.1. Figure 21.1 presents the orbit of DEMETER above Indonesia on March 23, Table 21.1. Time, locations, depth and region of the earthquakes that occurred in the Sumatra region (from the web server http://www.iris.edu/seismon). Date 28-03-2005 28-03-2005 28-03-2005 28-03-2005 29-03-2005 29-03-2005 30-03-2005 30-03-2005 31-03-2005
Epicenter (Latitude, Longitude)
Time (UT)
Depth (km)
M
Region
2.07◦ N, 97.01◦ E 1.33◦ N, 97.39◦ E 0.95◦ N, 97.8◦ E 2.73◦ N, 95.96◦ E 2.61◦ N, 96.54◦ E 2.14◦ N, 96.62◦ E 1.92◦ N, 96.95◦ E 3.01◦ N, 95.37◦ E 1.80◦ N, 97.08◦ E
16:09:36 16:38:43 17:59:47 18:48:52 05:16:29 05:25:25 10:20:22 16:19:41 07:23:55
30 30 30 30 30 30 27.9 22 29
8.7 6.0 5.3 5.5 5.9 5.3 5.4 6.4 5.8
Sumatra Sumatra Sumatra Sumatra Sumatra Sumatra Sumatra Sumatra Sumatra
Demeter March 23, 2005 95
10
97
99
101
103
105
6
2
2
Figure 21.1. Track of DEMETER orbit on March 23, 2005 when the satellite was above the Sumatra region. Stars indicate the epicenters of the earthquakes.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
238 A.K. Gwal et al.
Figure 21.2.
From top to bottom the panels successively show the electron density, electron temperature, spectrogram of an electric component between 0 and 2 kHz and earthquakes seen by DEMETER along the orbit. The data are presented as a function of the universal time (UT), The local time (LT), geographic latitude and longitude values are also given.
2005. The corresponding data recorded by DEMETER along this orbit is shown in Figure 21.2. The first panel shows the variation of electron density recorded by the Langmuir Probe Instrument. Second panel gives the electron temperature. Ion density of the O+ ion is shown in the third panel and the spectrogram of electric component up to 2 kHz is shown in the fourth panel. The last panel indicates the satellite closest approach of past and future earthquake epicenters that are within 2000 km from the DEMETER orbit. The Y -axis represents the distances D between the epicenter and the satellite from 750 up to 2000 km. The symbols are filled green square for post seismic events, filled red triangle for pre-seismic events and filled blue circle for earthquakes occurring during the half-orbit. The color scale on the right represents the time interval
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Seismo-electromagnetic precursors registered by DEMETER satellite
Figure 21.3.
239
Spectrogram of ELF electric waveform above the latitude in which earthquake occurred obtained using Level 1 burst mode data for the orbit shown in Figure 21.1.
between the earthquakes and the DEMETER orbit with a color gradation from >30 days up to a [0–6 h] interval. The empty symbols have a similar description except that they are related to the conjugate points of the epicenters. The symbol sizes correspond to the earthquakes of magnitude [5–6], [6–7] and [7–]. Simultaneous disturbances are recorded by three instruments at this time. Rapid fluctuations in electron density (maximum up to 17%) are observed with similar fluctuations in O+ density (maximum up to 60%) near to the epicenters. Fluctuations in the electron temperature were also observed. Electrostatic turbulence was observed in the spectrogram of electric component up to 2 kHz at the same time. As this data was recorded during burst mode it was possible to perform a detailed analysis. We have analyzed the electric field for the same event in ULF/ELF range shown in Figure 21.3. Rapid variations are found in ULF/ELF starting around 15:27:50 h UT (1–60 Hz). Enhancement in the magnitude however is observed around 40–50 Hz. The magnitude of signal intensity on logarithmic scale rises to as high as 3.50 (µV2 m−2 Hz−1 ) around 15:29:00 h UT just above the epicenter. To confirm that the perturbations are associated with the corresponding earthquakes, we have taken the magnetic index Kp for that day (Figure 21.4) and found that the corresponding 3-h Kp values were one suggesting quiet geomagnetic activity. On March 26, 2005 we have another orbit of DEMETER above Sumatra (Figure 21.5). Data recorded along this orbit is shown in Figure 21.6. Electron density increase (∼14% from the unperturbed state) was accompanied with increase of main ion constituent O+ (∼25% from the unperturbed state) near to the epicenters. The three hourly Kp indices shown in Figure 21.7 indicate low and moderate geomagnetic activity.
240 A.K. Gwal et al. 23-03-2005
2.5 2
Kp
1.5 1
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
0.5 0
0 to 3
3 to 6
6 to 9
9 to 12 12 to 15 15 to 18 18 to 21 21 to 24 Time (hrs) UT
Figure 21.4. Three hourly Kp values for March 23, 2005. Demeter March 26, 2005 90
6
92
94
96
98
100
2
2
6
Figure 21.5. Track of DEMETER orbit on March 26, 2005 when the satellite was close to the Sumatra region.
For comparison, data related to another earthquake occurring in the same area but with a lower magnitude has been studied. These earthquakes occurred in the Indonesian region with maximum magnitude equal to 5.5 at 1.32◦ N, 97.20◦ E. Table 21.2 gives the position and time of these earthquakes. Figure 21.8 shows the ground track of DEMETER satellite on July 6, 2005. Figure 21.9 shows the spectrogram of electric field component up to 400 Hz. We have analyzed the electric field data (shown in Figure 21.10) for the lower frequency band comprising ULF/ELF range as this data was recorded during burst mode. Significant perturbations were observed in the ELF range
241
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Seismo-electromagnetic precursors registered by DEMETER satellite
Figure 21.6.
Data recorded by DEMETER along the orbit shown in Figure 21.5. The top panel shows the electron density, the middle panel shows the ion density and the bottom panel gives the earthquakes “seen” by the satellite. At the bottom, UT, LT, geographic latitude and longitude values are indicated.
(140–220 Hz). It is evident from the spectrogram that perturbations were exhibited at around 15:25:52 h UT and lasted for a few seconds. The magnitude on logarithmic scale was observed to be around 2.2 (µV2 m−2 Hz−1 ). We also find the corresponding day was a geomagnetically quiet day as the 3-h Kp values were less than equal to 1 (shown in Figure 21.11).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Kp
242 A.K. Gwal et al. 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0
23-03-2005
0 to 3
3 to 6
6 to 9
9 to 12 12 to 15 15 to18 18 to 21 21 to 24 Time (hrs) UT
Figure 21.7. Three hourly Kp values for March 26, 2005.
Table 21.2. Time, locations, depth and region of the earthquakes that occurred in the Indonesian region (from the web server http://www.iris.edu/seismon). Date July 8, 2005 July 11, 2005 July 11, 2005
Epicenter
Time (UT)
Magnitude
Depth (km)
Region
1.23◦ N, 97.20◦ E 1.32◦ N, 97.20◦ E 2.64◦ N, 94.33◦ E
21:28:23 14:36:10 01:07:55
5.0 5.5 5.0
30 23 30
Indonesia Indonesia Indonesia
Demeter July 6, 2005 96
100
104
4
0
4
Figure 21.8. Track of DEMETER orbit on July 6, 2005 when the satellite was above the Indonesian region. Stars indicate the epicenters of the earthquakes.
243
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Seismo-electromagnetic precursors registered by DEMETER satellite
Figure 21.9. The top panel gives the spectrogram of an electric component between 0 and 400 Hz and the bottom panel gives the earthquake information. At the bottom, UT, LT, geographic latitude and longitude values are indicated.
Figure 21.10.
Spectrogram of ULF/ELF electric waveform above the latitude in which earthquake occurred obtained using Level 1 burst mode data for the orbit shown in Figure 21.8.
1.2
Kp values for 06-07-2005
1 0.8 Kp Unit
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
244 A.K. Gwal et al.
0.6 0.4 0.2 0 0 to 3
3 to 6
6 to 9
9 to 12 12 to 15 15 to 18 18 to 21 21 to 24 Time ( UT)
Figure 21.11. Three hourly Kp values for July 6, 2005.
Seismo-electromagnetic precursors registered by DEMETER satellite
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
21.4
245
DISCUSSION
This chapter presents interesting cases in which ionospheric perturbations were observed before earthquake. These perturbations were recorded by the ISL Langmuir Probe experiment, IAP the Thermal Plasma Analyzer experiment and ICE the electric field experiment onboard the DEMETER satellite. From the insitu measurements of electron and ion density and ULF/ELF electric field changes, the variation of these parameters have been illustrated before earthquake. For the Sumatra earthquakes increase in electron concentration was observed close to the epicenter. Gokhberg et al. (1995) had demonstrated the penetration of seismic electric field into the ionosphere. This electric field leads to the formation of irregularities in the ionosphere and thereby affecting the electron density distribution in the ionosphere (Pulinets, 1998a). The electron concentration variations were found to be accompanied with variations in the main ion constituent O+ and fluctuations of the electron temperature. This is in agreement with previous events reported by Pulinets (1998b). Even small changes of ionospheric electric field can substantially modify the concentration of O+ at heights of the F2 layer maximum before earthquakes (Pulinets and Boyarchuk, 2004). They have reported several examples in which modification of the O+ ion was found above the region of the anticipated earthquake. ULF/ELF emissions are typically observed as bursts above the earthquake epicenters. These emissions may start several days before the earthquake (Pulinets and Boyarchuk, 2004). In both cases studied these emissions were observed several days before the event. 21.5
CONCLUSION
In this chapter we have shown interesting examples of variation of plasma parameters recorded when the satellite was flying over the region of anticipated earthquakes. The electric field perturbations associated with seismic events in the ULF/ELF range have also been presented. These examples have been automatically selected by a tool of the DEMETER mission center (Lagoutte et al., 2006) which sorts out satellite orbits at a selected distance to epicenters of earthquakes with magnitude larger than 6. One of the most intriguing and promising results of local plasma parameter measurements is the change in electron and ion composition before earthquakes over the earthquake preparation zone. In all cases presented in this chapter the variation of these parameters were observed close to the location of epicenters. More importantly these variations observed are of precursor type. The geophysical conditions during our period of observation have been also considered as these ionospheric parameters also show variation with solar and geomagnetic activity of the Earth. They were relatively quiet. ACKNOWLEDGEMENT The authors would like to thank J.J. Berthelier and J.P. Lebreton for the use of the DEMETER data, and to acknowledge the Indo French Centre for the Promotion of Advanced Research (IFCPAR), New Delhi for providing financial support for this research. The Kp values were provided by the International Service of Geomagnetic Indices from the web server http://www.cetp.ipsl.fr/∼isgi/. REFERENCES Afonin, V.V., Molchanov, O.A., Kodama, T., Hayakawa, M., and Akentieva, O.A. (1999). Statistical study of ionosphere plasma response seismic activity: search for reliable result from satellite observations. In: M. HayaKawa (ed.), Atmospheric and Ionospheric Electromagnetic Phenomena Associated with Earthquakes, Terra Scientific Publishing Company, Tokyo, pp. 597–618.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
246 A.K. Gwal et al. Afonin, V.V., Akentieva, O.A., Molchanov O.A., and Hayakawa, M. (2000). Statistical Study of equatorial anomaly from high apogee satellite APEX and low apogee satellite COSMOS-900, International Workshop on Seismo-Electromagnetics, Programme and Abstracts, NASDA, 19–22 Sep 2000. Berthelier, J.J., Godefroy, M., Leblanc, F., Malingre, M., Menvielle, M., Lagoutte, D., Brochot, J.Y., Colin, F., Elie, F., Legendre, C., Zamora, P., Benoist, D., Chapuis, Y., and Artru, J., (2006a). ICE, the electric field experiment on DEMETER, Planet Space Science, 54, 456–471. Berthelier, J.J., Godefroy, M., Leblanc, F., Seran, E., Peschard, D., Gilbert, P., and Artru, J., (2006b). IAP, the thermal plasma analyzer on DEMETER, Planetary Space Science, 54, 487–501. Boskova, J., Smilauer, J., Jiricek, F., and Triska, P. (1993). Is the ion composition of outer ionosphere related to seismic activity. J. Atmos. Terr. phys., 55(13), 1689–1695. Boskova, J., Smilauer, J., Triska, P., and Kudela, K. (1994). Anomalous behaviour of plasma parameters as observed by the intercosmos 24 satellite prior to the iranian earthquake of 20 June, 1990. Studia Geoph. Geod., 38, 213. Gokhberg, M.B., Morgounov, V.A., and Pokhotelov, O.A. (1995). Earthquake Prediction, SeismoElectromagnetic Phenomena. Gordon and Breach, Russia. Hayakawa, M. (1997). Electromagnetic precursors of earthquakes: review of recent activities. Rev. Radio Science, 1993–1995, Oxford University Press, 807–818. Lagoutte, D., Brochot, J.Y., de Carvalho, D., Elie, F., Harivelo, F., Hobara, Y., Madrias, L., Parrot, M., Pincon, J.L., Berthelier, J.J., Peschard, D., Seran, E., Gangloff, M., Sauvaud, J.A., Lebreton, J.P., Stverak, S., Travnicek, P., Grygorczuk, J., Slominski, J., Wronowski, R., Barbier, S., Bernard, P., Gaboriaud, A., and Wallut, J.M., (2006). The DEMETER science mission centre, Planetary Space Science, 54, 428–440. Lebreton, J.P., Stverak, S., Travnicek, P., Maksimovic, M., Klinge, D., Merikallio, S., Lagoutte, D., Poirer, B., Kozacek, Z., and Salaquarda, M. (2006). The ISL Langmuir Probe experiment and its data processing onboard DEMETER: scientific objectives, description and first results, Planetary Space Science, 54, 472–486. Liperovsky, V.A., Pokhotelov, O.A., Liperovskaya, E.V., Parrot, M., Meister, C.V., and Alimov, O.A. (2000). Modification of sporadic E-layers caused by seismic activity. Surv Geophys, 21, 449–486. Molchanov, O.A., Mazhaeva, O.A., Goliavin, A.N., and Hayakawa, M. (1993). Observation by the Intercosmos-24 satellite of ELF/VLF electromagnetic emissions associated with earthquakes. Ann. Geophys., 11, 431–440. Parrot, M. (1994). Statistical study of ELF/VLF emissions recorded by a low altitude satellite during seismic events. J. Geophy. Res., 99(23), 339–347. Parrot, M. (1999). Statistical studies with satellite observations of seismogenic effects. In: M. Hayakawa (ed.), Atmospheric and Ionospheric Phenomena Associated with Earthquakes, TERRAPUB, Tokyo, 685–695. Parrot, M., Achache, J., Berthelier, J.J., Blanc, E., Deschamps, A., Lefeuvre, F., Menvielle, M., Planet, J.L., Tarits, P., and Villain, J.P. (1993). High-frequency seismo-electromagnetic effects. Phys. Earth Planet. Interiors, 77, 65–83. Parrot, M., Beonoist, D., Berthelier, J.J., Blecki, J., Chapuis, Y., Colin, F, Elie, F., Fergeau, P., Lagoutte, D., Lefeuvre, F., Legendre, C., Leveque, M., Pincon, J.L., Poirier, B., Seran, H.C., and Zamora, P. (2006). The magnetic field experiment IMSC and its data processing onboard DEMETER: scientific objectives, description and first results, Planetary Space Science, 54, 441–455. Pulinets, S.A. (1998a). Strong earthquakes prediction possibility with the help of topside sounding from satellites. Adv. Space Res., 21(3), 455–458 Pulinets, S.A. (1998b). Seismic Activity as a Source of the Ionospheric Variability. Adv. Space Res., 22(6), 903–906. Pulinets, S.A. and Legen’ka, A.D. (2003). Spatial-temporal characteristics of large scale disturbances of electron density observed in the ionospheric F-region before strong earthquakes. Cosmic Res., 41(3), 221–229. Pulinets, S.A. and Boyarchuk, K.A. (2004). Ionospheric Precursors of Earthquakes, Springer Verlag Publication, Heidelberg. Pulinets, S.A., Legen’ka, A.D., Gaivoronskaya, T.V., and Dupuev, V.Kh. (2003). Main phenomenological features of ionospheric precursors of strong earthquakes. J. Atmos. Sol. Terr. Phys., 63, 1337–1347. Sauvaud, J.A., Moreau, T., Maggiolo, R., Treilhou, J.P., Jacquey, C., Cros, A., Coutelier, J., Rouzaud, J., Penou, E., Gangloff, M., (2006). High energy electron detection onboard DEMETER: the IDP spectrometer, description and first results on the inner belt, Planetary Space Science, 54, 502–511.
CHAPTER 22
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Web-Enabled and Real-Time Reporting: Cellular Based Instrumentation for Coastal Sea Level and Surge Monitoring A. Joseph and R.G. Prabhudesai National Institute of Oceanography, Dona Paula, Goa, India
22.1
INTRODUCTION
The 26 December 2004 monster tsunami that struck many South Asian countries, and the severe destruction and devastation wrought by that, have drawn attention to the urgent need for a versatile disaster warning system for the Indian Ocean rim countries. Presence of a network of sealevel gauges spread all along the coasts, with capability for providing real-time information on sea-level elevation and its trend, would provide the requisite data to the disaster management agencies for dissemination of disaster alert warnings to the coastal communities. Immediate alert and related information bulletins disseminated to appropriate local and central disaster management cells and the people by public communication channels such as commercial radio, television, and marine radio system available in almost all these countries, would save many lives. Unfortunately, such a network was not in place during the 26 December 2004 tsunami disaster. This lacuna, together with frequent storms that hit many coastal locations, suggests the urgent need for immediate deployment of a network of real-time integrated sea level and surface meteorological data communication systems (Joseph and Prabhudesai, 2005) for the benefit of the coastal communities, beach tourism agencies, and the local administrators. Given the popularity of Internet on a global scale, providing such state-of-the-art accessibility to sea-level data would mean that the current coastal sea-level scenario can be viewed in real-time from almost any part of the world. If Internet access to the sea-level web site is made available to television channels, then real-time visualization of the coastal sea level (e.g., during anomalous and disastrous state of the coastal seas) and its trend from the previous day to the present instant can be examined by everyone including the common people who are known to make good use of television when anything particularly important happens in any part of the world. Providing Internet accessibility to the sea-level gauge web site at other media centres such as radio stations and the press would also serve an equally important role in the quick dissemination of the current anomalous sealevel scenario to the navigators, and the travelling communities. Such a network would provide useful information also to the mariners sailing in the coastal waters. Moreover, the information obtained would be of great value to the scientific community. 22.2
EXISTING SYSTEMS
Various types of sea-level gauges have been developed over the years (Joseph, 1999). These include tide staff, float-driven gauge, pressure gauge, acoustic gauges, and most recently radar gauge. 247
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
248 A. Joseph and R.G. Prabhudesai It has been feasible to deploy seafloor-mounted tsunami detection sensors in the Pacific Ocean rim regions with the support of instrumented deepwater moored buoys in the vicinity of seafloor spreading centres. The primary sensors are seismic probes imbedded in the seafloor at varying depths (Hoffman, 1997). High precision pressure sensors that are located on the seafloor are also providing important data on the propagation of tsunamis in deep water (Filloux, 1982; Baba et al., 2004). In moored buoy systems, acoustic modems attached to the moorings transmit eventdetection data (i.e., seismic and bottom pressure data) to a module attached to a surface buoy. This surface module, in turn, relays the data to several land-based disaster warning centres via satellite transmission network. The tsunami buoy systems are aimed at quickly confirming the existence of potentially destructive tsunamis and reducing the incidence of false alarms. It is expected that a series of instrumented deep ocean buoys might provide the much needed and reliable early warning. Where the source motion is known, computer simulations have been found to agree quite accurately with observations (Van Dorn, 1982).
22.3
ENHANCEMENT OF EXISTING COASTAL SYSTEMS FOR TSUNAMI AND STORM-SURGE DETECTION
Tsunami is a special type of long wave (solitary wave) that is generated in the sea following a large scale impulsive disturbance (e.g., earthquake). In contrast, storm surge is a localized disturbance of sea level resulting from the action of a cyclone. Tsunamis and storm surges are the two natural calamities that have hit many Indian Ocean rim countries; and taken heavy toll of lives; and inflicted colossal damage to properties. Underwater earthquakes (seaquakes), which are the most potent cause for tsunamis, have been monitored successfully in limited areas of certain countries (e.g., California, south west Iceland) based on noticeable increase in the background seismic activities (Hoffman, 1997). However, with the present scientific knowledge and technology those are available even with the most advanced nations, reliability of seaquake prediction for warning purposes is considered to be rather poor. Although offshore instrumented buoys are important devices and can be used for detection of both tsunami and storm-surge events far away from land, they are expensive and might pose logistical challenges. Thus, a network of coastal-based systems that provides real-time or nearreal-time sea level and surface meteorological information can be implemented as an alternate system that complements the deep ocean moored system. A suitable network for real-time monitoring of storm surge and running of operational stormsurge models for predictive purposes must consist of real-time transmitting sea level and surface meteorological monitoring systems. A great advantage of storm-surge model is its usefulness in predicting the anticipated flooding at selected locations. The storm-surge model that is established on a geographical information system (GIS) platform would allow a realistic assessment of the impact of elevated/depressed sea levels on the regions covered by the storm-surge model. The prediction will enable the local administrators/planners to issue periodic warning of maximum likely flooding at a given location in a given coastal/estuarine region for a specific meteorological event. Subsequent to the 26 December 2004 tsunami episode, there appears to be a consensus in the Indian Ocean rim countries on the urgent need for establishment of a disaster warning system. A network capable of real-time reporting of integrated sea level and surface meteorological events from coastal sea-level gauges and weather stations is an important ingredient to complement the deep ocean systems that are currently contemplated for providing timely alerts. Such a network would provide a sufficiently large real-time database for running predictive models for stormsurge forecasting if ancillary database on the bathymetry and topography of the region of interest are available. The real-time database so obtained can also be used for real-time validation of the storm-surge model. Based on the above discussions, it would be desirable to go for development
Web-enabled and real-time reporting: instrumentation and monitoring
249
and deployment of a network of web-enabled wireless-transmitting real-time reporting integrated coastal sea level and surface meteorological monitoring systems incorporating the state-of-the-art wireless technology.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
22.4
DATA COMMUNICATION OPTIONS
Storm surge or tsunami warning systems would require the data to be reported to the authorities in a very short time. The Intergovernmental Oceanographic Commission (IOC) of UNESCO has recommended that for the Indian Ocean Tsunami warning system, data have to be reported within 5 min of being recorded at the gauge and would need to be made available on the Global Telecommunications System (GTS). There are various communication technologies that could be utilized for real-time reporting of sea-level data. A variety of real-time communication options are being used at present, and newer options are being examined. These include wired telephone connection, VHF/UHF transceivers, satellite transmit terminals, and cellular connectivity. In the past, the method of data communication depended largely on the distance over which the data had to be transmitted. Thus, for short links (e.g., harbour operations), a VHF/UHF radio link was often convenient. However, communication via VHF/UHF transceivers is limited by line-of-sight distance between transceivers and normally offer only point-to-point data transfer. Ideally, nationwide and even global scale links are necessary for storm surge/tsunami warning communication network systems. For countrywide links, Subscriber Trunk Dialling (STD) or dedicated telephone lines of the Public Switched Telephone Network (PSTN) have been successfully used (e.g., tide gauge network of Proudman Oceanographic Laboratory (POL), UK). However, wired telephone connections can be severely susceptible to loss of connectivity during natural disasters such as storm surges, primarily because of telephone line breakage under the force of storm-winds (e.g., uprooted trees and broken branches). Satellite communication via platform transmit terminals (PTT) has wider coverage and, therefore, allows data reception from offshore platforms. For the last decade or more, sea-level installations in a few countries have used satellite systems (ARGOS, GOES, ORBCOMM, IRIDIUM, METEOSAT, GMS, and INMARSAT) for data reporting. Nevertheless, data transfer speeds of many satellite-based communication systems are limited to 9600 baud or less. Many satellites (e.g., GOES, INSAT) permit data transfer only at predefined time-slots, thereby inhibiting continuous data access. To ensure that data transmission by the sea-level gauges takes place at these precise time-slots, the gauge electronics should have its clock derived from a common “time source”. This requirement has necessitated incorporation of GPS receivers with each data collection platform (DCP). A network of geostationary satellites comprising GOES (USA), METEOSAT (Europe), and GMS (Japan) offer overlapping longitudinal coverage and appreciable latitude coverage (75◦ ). However, the size of data that can be transmitted to a satellite during the specified time-slot is limited to 649 bytes. Another option is a network of satellite mobile telecommunications known as IRIDIUM, which is designed to enable data reporting regardless of the user’s location on land or at sea. Subscribers can use their Iridium modem to communicate with any other telephone anywhere in the world. The recently introduced Regional Broadband Global Area Network (RBGAN) by INMARSAT provides reliable and high-speed data communication and data file transfers at broadband speed. This network is expected to provide full world coverage in the near future. RBGAN is based on Internet Protocol (IP) and GPRS technology, which offers reliable and cost-effective access to the Internet. For the INMARSAT, there is a permanent connection with a static IP address, just like broadband but with a maximum speed of 144 kbit/s. Using the new INMARSAT RBGAN data terminal, it is thus possible to share a 144 kbit/s broadband connection, achieving more than double the transfer speed of current terrestrial GPRS mobile
250 A. Joseph and R.G. Prabhudesai networks. Because the service is based on IP packet technology, users only pay for the amount of data they send and receive, and not for the amount of time spent online. This enables them to stay “always connected” to the Internet. Perhaps, the biggest advantage of RBGAN over fixed-line broadband is its independence from local telephone infrastructure. This would mean that the communication network can be expected to continue operating during extreme weather conditions and the resulting storm-surge events.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
22.4.1
Cellular modem
While satellite communication is relatively expensive, proliferation of wirelesses networking infrastructure and ubiquity of cellular phones have together made cellular communication affordable. Low initial and recurring costs are an important advantage of cellular data communication. Various services exist for data communication: for example, Short Message Service (SMS), Data Call, and General Packet Radio Service (GPRS). SMS is a common method of sending text messages of up to 140 bytes between cellular devices. This mode of transmission is probably the easiest to implement when the data size is small. However, the recurring cost can be substantial if data needs to be reported frequently. Further, keeping track of lost SMS can easily override its simplicity. Data Call (similar in operation to the conventional STD call) is another alternative, which can also turn out to be expensive if data has to be frequently reported. Both SMS and Data Call services would require modems on the remote reporting device as well as on the receiving end. This adds to the hardware cost as well as software overheads on the receiving end to check the data integrity for transmission errors. The main benefit of cellular connectivity with GPRS technology is that it utilizes the radio resources only when there is data to send. In addition, GPRS offers improved quality of data services as measured in terms of reliability, response time, and features that are supported. Another advantage of GPRS over other data communication services is that it provides an “always on” communication channel without incurring time-based costs. That is, there is no difference in costs whether data is collected once a minute or once a day. Further, GPRS data transmission speeds are of the order of 3 to 4 times that of the traditional cellular data connection. The GPRS allows data rates of 115 k bit/s. Also, GPRS enables a constant TCP/IP connection with the Internet so that data can be easily uploaded. Since data can be posted on the Internet server, there is neither a need for a modem on the receiving side nor the requirement to have special software at the server-side to collect data from the remote site. 22.5
INITIATIVES AFTER THE INDIAN OCEAN TSUNAMI OF 26 DECEMBER 2004
At present, the Pacific Tsunami Warning Centre (PTWC) provides sea level and related information from the University of Hawaii Sea Level Centre for basin-wide tsunami warning to the Pacific Ocean rim countries. Prior to the 26 December 2004 Indian Ocean Tsunami, networks of daily or weekly reporting sea-level gauges have been in existence only in a very few countries. For example, UK had a network which reported sea-level data via land phone on interrogation from the Tide Gauge Directorate at POL. Subsequently, the UK network has been expanded to provide capability for real-time reporting and Internet-accessibility to sea-level data. POL has developed instrumentation that can take the output from a range of sensors, including radar and pressure types. The data are collected by a small Linux-embedded processor and sent back to base by e-mail or by Secure Copy Protocol (SCP). Broadband enabled test sites using a radar sensor connected to an embedded Linux system have been installed at Liverpool and Holyhead in the UK. One-min data values are available every 5 min in the form of an e-mail message. The resulting data are displayed on the NTSLF web pages. Chile established a network, which
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Web-enabled and real-time reporting: instrumentation and monitoring
251
transmits sea-level data via satellite (Fierro, 2005). Before December 2004, real-time reporting and Internet-accessible sea-level gauges were operational in USA, UK, and Hong Kong. The devastations brought by the 26 December 2004 Indian Ocean Tsunami proved to be an eye-opener internationally. Consequently, IOC of UNESCO established a Technical Committee, with 11 members covering India (Antony Joseph), Kenya (Charles Magori), South Africa (Ruth Farre), Australia (Bill Mitchell), Spain (Begona Perez and Lopez Maldonaldo), Norway (Daniel Hareide), France (Guy Woppelman), UK (Peter Foden), Chile (Juan Fierro), and USA (Bernie Kilonsky) to provide advice and recommendations concerning technical aspects of sea-level observations such as station configuration, data transmission, quality control and data processing, calibration and maintenance of sensors and stations, testing and evaluation of equipment, etc. Many of the Indian Ocean rim countries have plans to put in place networks of instrumented platforms for warning purposes in future. For example, the Government of India has taken initiative for an early warning system (Gupta, 2005) and decided to establish a network by September 2007, which would report sea-level data using India’s own satellite INSAT. In the meantime, the National Institute of Oceanography (NIO) of India conceived a substantially costeffective system for monitoring of sea level and other surface meteorological events (Joseph and Prabhudesai, 2005), and quickly developed and deployed a cellular-based real-time reporting and Internet-accessible coastal sea-level gauge. This system is functional at Mandovi estuary, Goa (India) since 24 September 2005 (Prabhudesai et al., 2006). This system would complement the national efforts towards the development of early warning systems for disaster management. The real-time reporting system developed at NIO is simple and cost-effective, and it can also be deployed in dams, reservoirs, and at any other water sources for water resource management. Deployment of this system would help in equitable sharing and distribution of the scarce water resources with much transparency, and would permit avoidance of many inter-state and international water related disputes which plague many states and countries at present. 22.6
SYSTEM DEVELOPED AT NIO, INDIA
Based on extensive experience in the use of pressure sensors for sea-level measurements in India (Joseph et al., 2006) andAfrica (Joseph et al., 2006), NIO in India quickly designed and developed a sea-level gauge incorporating temperature-compensated piezo-resistive semiconductor pressure transducer from Honeywell, having adequate features and performance (Vijay Kumar et al., 2005). The data is normally sampled at 2 Hz and can be averaged over duration of 15 s to 15-min interval in conformity with the revised requirements of Global Sea Level Observing System (GLOSS) to detect tsunami and storm-surge events. In the past, data recording at intervals less than 15 min was difficult in practice because of memory capacity limitations. However, this limitation has been overcome in the present design by incorporating multimedia cards, with storage capacity ranging from 128 to 1000 megabytes. Additionally, real-time/near-real-time data access capability is also implemented for effective operational applications during anomalous sea-level conditions such as those occurring during storm surges and tsunami episodes. The sea-unit of the gauge is mounted within a cylindrical protective housing, which in turn is rigidly held within the vertical legs and the interconnecting triangular collar of the mechanical structure used for mounting the entire gauge. Figure 22.1(a) shows the top portion of the gauge’s mounting structure, where battery, electronics, solar panel, and cellular modem are mounted. The pressure sensor and the logger are continuously powered on, and their electrical current consumption is 30 and 15 mA respectively. The cellular modem consumes 15 and 250 mA during standby and data transmission modes, respectively. Figure 22.1(b) illustrates the installation of the gauge at Verem Jetty in Mandovi estuary, Goa, India.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
252 A. Joseph and R.G. Prabhudesai
(a)
(b)
Figure 22.1.
22.6.1
(a) Top portion of the gauge’s mounting structure, where battery, electronics, solar panel, and cellular modem are fixed (after Prabhudesai et al., 2006). and (b) Illustration of NIO sealevel gauge installed at Verem Jetty, Mandovi estuary, Goa, India (after Prabhudesai et al., 2006).
Sea-level measurement
The methodology employed in the NIO sea-level gauge is to detect the hydrostatic pressure, and to estimate the water column height over the pressure transducer from knowledge of the effective depth-mean water density.
Web-enabled and real-time reporting: instrumentation and monitoring
253
An absolute pressure transducer senses the total pressure (i.e., atmospheric pressure + pressure exerted by the water column above the transducer). The total pressure, P(total) , over the pressure inlet, at depth d below the water surface is given by the relationship: P(total) = Pa + Pw
(22.1)
where Pa is the atmospheric pressure and Pw is the pressure exerted by the overlying water column on the transducer. The following well-known relation enables estimation of sea-level elevation:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Pw = (ρ × g × d)
(22.2)
In this relation, ρ is the effective depth-mean density of the overlying column of water, g is the acceleration due to earth’s gravity, and d is the depth of the pressure transducer below the sea surface. Thus, the value of d is estimated. In general, because of the complexities associated with effective density of suspended-sediment-laden shallow water bodies (Joseph et al., 1999, 2004), it is advisable to use alternate sensing methods in heavily suspended-sediment-laden water bodies (e.g., Hooghly estuary in India).
22.6.2
Cellular-based system
In India, practically all the populated areas are networked to cellular-transmitting stations and, therefore, the required dataset can be made available online for real-time or near real-time applications such as warning and predictive model-running applications. The monitored information and forecast can then be disseminated to the coastal communities and the general public through a variety of electronic media. Figure 22.2 illustrates implementation of cellular-based real-time coastal sea-level data reporting utilizing GPRS technology. The GPRS modem with built-in TCP/IP stack incorporated in the present design frees the host controller, which is responsible for acquisition of data from sealevel gauge, from communication overload. These modems need only a few simple commands to upload data on a remote File Transfer Protocol (FTP) server. Data communication is always initiated by the sea-level gauge. After communication link is established, a bidirectional data transfer is possible between Internet server and the sea-level gauge. This is because GPRS Gateway Support Node (GGSN) uses Dynamic Host Configuration Protocol (DHCP) to assign private IP addresses to cellular devices. This IP address is invisible to the Internet network. Although GPRS is termed as an “always on” network, if there is no data exchanged beyond predefined timeout periods, the connection is dropped and a new IP address is assigned to the cellular modem. Because of this scheme, the IP address of the cellular modem can change frequently. One obvious benefit of such a connection mechanism from the security point of view is that risk of attack from hackers and third party is reduced. The software consists of two main components, namely; (i) embedded system software for data acquisition and communication with Internet server, and (ii) web server-based user-interface to visualize the received sea-level data. In the present application, 5-min averaged sensor output data is logged by the embedded system at 5-min interval. Subsequently, all the data logged from the previous day to the current time is uploaded to the Internet server that is located at NIO, Goa, India. The data received at the Internet server is stored in its back-end database, and simultaneously presented in graphical format, together with the predicted fair-weather sea-level and the residual. The residual, which is the measured sea level minus predicted fair-weather sea level, provides a clear indication and a quantitative estimate of the anomalous behaviour of sealevel oscillation. The driving force for such anomalous behaviour could be atmospheric forcing (storm) or geophysical (tsunami).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
254 A. Joseph and R.G. Prabhudesai
Figure 22.2.
Schematic diagram illustrating implementation of realtime coastal sea-level data reception utilizing GPRS technology (after Prabhudesai et al., 2006).
Figure 22.3 shows display of predicted fair-weather sea-level, observed real-time sea levels, and residuals from Verem jetty (Mandovi estuary), Goa, India. The scale of real-time display is dynamically adjusted based on the maximum elevation observed. 22.6.3
Communication performance evaluation
In order to examine the performance of data upload from the sea-level gauge to the Internet server, 100 kilobytes of test data were uploaded continuously for 24 h using standard FTP. Figure 22.4 shows the distribution of time taken (throughput) for successive FTP uploads during the above experiment. The minimum and maximum transfer times were in the range 1.29–7.1 min for uplink (average time was 2.06 min). Data was transferred between sea-level gauge and the FTP server using Airtel GPRS service. On an average, 97.47% of data transfer attempts was found to be successful. 22.7
CONCLUSIONS
Because of the international attention bestowed on the December 26 episode, IOC of UNESCO established a Technical Committee to discuss and provide advice on various technical aspects of monitoring systems. In this effort, the already available expertise within the GLOSS community has been considered to be an asset. An obvious advantage that accompanies real-time or nearreal-time data reporting is the possibility to identify instrumental malfunctions, if any, and to
Figure 22.3.
255
Display of predicted fair-weather sea-level, observed real-time sea levels, and residuals from Verem jetty (Mandovi estuary), Goa, India (after Prabhudesai et al., 2006).
10 8 Upload time (mins)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Web-enabled and real-time reporting: instrumentation and monitoring
6 4 2 0
0
4
8
12
16
20
24
Time (hours)
Figure 22.4.
FTP up-load time during a period of 24 h for a 100 kb test data (after Prabhudesai et al., 2006).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
256 A. Joseph and R.G. Prabhudesai initiate remedial measures more rapidly. This permits remote diagnostics and might also provide the ability to re-program the system remotely. The general notion among the experts is that in addition to the conventional manner of averaging and sampling sequence, high-frequency motions such as waves and swells also need to be measured together with pertinent meteorological parameters such as barometric pressure. Such a scheme would cater to the optimal utilization of the future systems for operational applications in the events of both tsunamis and storm surges. These views are consistent with those espoused by Joseph and Prabhudesai (2005) in the Indian Ocean context. With the current advancement in communication technologies, various options are now available for real-time or near-real-time reporting of data for various operational and predictive applications. Technologies of data reporting via satellites have undergone a sea change recently in terms of frequency of reportage, data size, recurring costs, and so forth. Broadband technology has been identified to be one that can be optimally used for real-time reporting of data because of its many inherent advantages such as continuous two-way connection allowing high-speed data file transfer and near-real-time data reporting. It has been shown that an alternate and complementary cost-effective methodology for realtime reporting of data is cellular-based GPRS technology, which has been recently implemented by the NIO in India for real-time reporting of coastal sea-level data (Prabhudesai et al., 2006). In this system, by using the existing cellular phone network, continuous real-time updates of coastal sea-level elevations are realized on a web server. While satellite communication is expensive, wirelesses networking infrastructure and ubiquity of cellular phones have together made cellular communication affordable. Low initial and recurring costs are an important advantage of GPRS cellular communication. This methodology will soon be extended for real-time reporting of integrated coastal sea level and surface meteorological data as well to cater to effective monitoring and evaluation of trends of storm surge, which plague many nations in the Indian Ocean rim region. This development provides an excellent platform for real-time monitoring of coastal sea level and surface meteorological data; thus providing the requisite input for efficient implementation of any alert and warning mechanism in the event of oceanogenic hazards such as storm surge and tsunami. Considering the popularity of Internet on a global scale, providing such state-of-the-art accessibility to sea-level gauges would mean that the present coastal sea-level scenario can be viewed in real-time by anyone from any part of the world. Providing Internet accessibility to the sealevel gauge web site at other media centres such as television channels, radio stations, and the print media would serve an important role in the quick dissemination of the current anomalous sea-level scenario to the coastal communities, navigators, and the general public. Moreover, the information obtained would be of great value to the national and international scientific community. At the backdrop of the disastrous 26 December 2004 tsunami that caught us unaware, deployment and maintenance of a network of cost-effective monitoring system is an important step forward for the entire Indian Ocean rim regions. REFERENCES Baba, T., Hirata, K., and Kaneda, Y. (2004). Tsunami magnitudes determined from ocean-bottom pressure gauge data around Japan. Geophys. Res. Lett., 31, L08303. Fierro, J. (2005). Chilean Sea Level Network. IOC Manual on Sea Level Measurement and Interpretation, 4: An Update to 2005, pp. 111–114. Filloux, J.H. (1982). Tsunami recorded on the open ocean floor. Geophys. Res. Lett., 9, 25–28. Gupta, H. (2005). Mega-tsunami of 26 December 2004: Indian initiative for early warning system and mitigation of oceanogenic hazards. Episodes, 28, 1–4. Hoffman, C. (1997). Checking on seaquake. Sea Technol., 38(8), 74.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Web-enabled and real-time reporting: instrumentation and monitoring
257
Joseph, A. (1999). Modern techniques of sea level measurement. Encyclopedia of Microcomputers, Vol. 23. Marcel Dekker, Inc., New York, pp. 319–344. Joseph, A., Desa, E., Smith, D., Peshwe, V.B., Vijaykumar, and Desa, J.A.E. (1999). Evaluation of pressure transducers under turbid natural waters. J. Atmos. and Oceanic Technol., 16(8), 1150–1155. Joseph, A., Desa, E., Vijaykumar, Desa, E.S., Prabhudesai, R.G., and Prabhudesai, S. (2004). Pressure gauge experiments in India. In: S. Holgate and T. Aarup (eds.), Workshop Report No. 193, Intergovernmental Oceanographic Commission of UNESCO, pp. 22–37. Joseph, A. and Prabhudesai, R.G. (2005). Need of a disaster alert system for India through a network of real time monitoring of sea level and other meteorological events. Curr. Sci., 89, 864–869. Joseph, A., Mehra, P., Joseph, O., and Nkebi, E.K. (2006). Pressure Gauge Based GLOSS Sea Level Station at Takoradi Harbour (Ghana, West Africa) – Experiences over a Year. IOC Manual on Sea Level Measurement and Interpretation, 4: An Update to 2005, pp. 108–110. Joseph, A., Odametey, J. T., Nkebi, E. K., Pereira, A., Prabhudesai, R. G., Mehra, P., Rabinovich, A. B., Vijaykumar, Prabhudesai, S., and Woodworth, P. L. (2006). The 26 December 2004 Sumatra Tsunami Recorded on the Coast of West Africa, African J. Mar. Sci (under revision). Prabhudesai, R.G., Joseph, A., Agarvadekar, Y., Dabholkar, N., Mehra, P., Gouveia, A., Tengali, S., Vijaykumar, and Parab, A. (2006). Development and implementation of cellular-based real-time reporting and internet accessible coastal sea level gauge – A vital tool for monitoring storm Surge and tsunami. Curr. Sci., 90(10), 1413–1418. Van Dorn, W.G. (1982). Tsunami. McGraw-Hill Encyclopedia of Science and Technology, Vol. 14, pp. 140–142. Vijay Kumar, Joseph, A., Prabhudesai, R.G., Prabhudesai, S., Nagvekar, S., and Damodaran V. (2005). Performance evaluation of honeywell silicon piezoresistive pressure transducers for oceanographic and limnological measurements. J. Atmos. Oceanic Technol., 22(12), 1933–1939.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 23
Methodologies for Tsunami Detection
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
T.S. Murty Department of Civil Engineering, University of Ottawa, Ottawa, Canada N. Nirupama Atkinson School of Administrative Studies, York University, Toronto, Canada A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology, New Delhi, India I. Nistor Department of Civil Engineering, University of Ottawa, Canada
23.1
INTRODUCTION
Traditionally, tsunami detection has completely depended on a coastal tide gauge network. In recent years pressure sensors at the ocean bottom in the Pacific have been providing some additional valuable data in real time. Some other possible detection methods of the tsunamigenic earthquake and tsunami signals in the troposphere, as well as in the ionosphere are also being considered. The observation that certain animals are able to detect and react to earthquake and tsunami signals, is gaining some credibility in recent times. If this ability of animals could be exploited, it will provide some additional guidance in real-time tsunami prediction. The Pacific tsunami warning system came into existence in the late 1940s. Following the disastrous Aleutian Earthquake Tsunami of 1 April 1946, the Pacific Tsunami Warning Center (PTWC) was established in Ewa beach on Oahu Island of Hawaii, USA. At present there are 27 member countries in this system which is administered by the Intergovernmental Oceanographic Commission (IOC) of UNESCO in Paris, since 1965. During the Alaska earthquake tsunami of 28 March 1964, tsunami warnings from PTWC did not reach Alaska in an efficient manner. For this reason, the Alaska Tsunami Warning Center (ATWC) was established in Palmer, Alaska in 1967. Because tsunami occurrence is rare in the Atlantic and Indian oceans, as compared to the Pacific Ocean, until now, there have been no tsunami early warning systems for these two oceans. However, following the disastrous tsunami in the Indian Ocean on 26 December 2004, tsunami warning systems are being established, not only for the Atlantic and Indian oceans, but also for several marginal seas, such as the Caribbean Sea, the Mediterranean Sea, the East China Sea, etc. At present, because of very low population density, there is no particular priority for a tsunami early warning systems for the Arctic Ocean. 259
260 T.S. Murty et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
23.2 TRADITIONAL METHOD OF TSUNAMI DETECTION THROUGH TIDE GAUGES Since its inception in the late 1940s, the Pacific tsunami warning system has relied almost completely on a tide gauge network (Figure 23.1) for detection of trans-oceanic or ocean wide tsunamis. However, there are certain exceptions. In northern Japan, sometimes, the time interval between the occurrence of the earthquake and the arrival of the tsunami on the nearest coastline is at most a few minutes. In such situations, there is no time to wait for a confirmation of the existence of the tsunami through recording on a tide gauge. Hence, there is no real alternative, but to issue a tsunami warning, just based on the occurrence of the earthquake itself. Needless to say, this could lead to false alarms in some instances, which cannot be easily avoided. For ocean-wide tsunamis, usually there is at least some time available before a warning has to be issued, assuming that no warning will be issued unless and until the existence of a significant tsunami is confirmed at least at one tide gauge. Of course, the drawback is that, the people living on the coast near the tide gauge (which usually is the gauge closest to the epicenter of the earthquake) would have no warning at all, unless they hear about the earthquake (and possibility of a tsunami) through the news media. The strength of the traditional method is its robustness, i.e. the tide gauges will always record tsunami as long as the gauge can withstand the onslaught of the advancing tsunami. The weakness of the traditional technique, lies in the fact that in some instances, it may be too late to provide any warning, since the tsunami has already approached the coast and is recorded on coastal tide gauges.
Figure 23.1.
Sea level gauge network for the PTWS (McCreery, 2005).
Methodologies for tsunami detection
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
23.3
261
REMOTELY REPORTING TSUNAMI RUN-UP DETECTORS
Recently the PTWC has installed eight Remotely Reporting Tsunami Run-up Detectors (RRTRD) on the likely tsunami affected areas in Hawaii. Figure 23.2 (McCreery, 2005) shows a typical gauge and its operational links. These tsunami run-up detectors will trigger and send a message to PTWC within seconds of their being flooded by an advancing tsunami. This is definitive way of confirming that there is water on the land where the detector is located. These sensors are 2.1 to 4.4 m above the mean sea level and are located some 18 to 119 m from the shoreline, where, in the past, significant runups have been observed. These detectors are outside the normal surface runup and are not sensitive to rain or moisture, other than to a flood. According to McCreery (2005), these detectors which are based on home security alarm technology and cell phones communication, cost about one thousand US dollars each, are easy to install and maintain. The disadvantage of these detectors is that, they simply establish whether there is a tsunami or not, but cannot give a precise value of the tsunami amplitude, such as one can record on a tide gauge.
23.4
DART SYSTEMS AND THE DEEP WATER SIGNATURE OF A TSUNAMI
One of the serious problems facing numerical modelers of tsunami generation and propagation is the fact that, the rupture parameters of tsunami-genic earthquakes are difficult to obtain in real 160°
22°
159°
155°
156°
157°
158°
Hanalei Kauai Nawiliwili
22° Haleiwa Waianae
PTWC
21°
Oahu Mokuoloe Makapuu KPA Molokai Honolulu Kahului Lanai Maui
21°
Lahaina Hawaii
20°
SEA LEVEL INSTRUMENTATION
160°
– PTWC GAUGE (REAL TIME) – LARC GAUGE (DIAL UP) – NOS GAUGE (HOURLY / DIAL UP / TRIGGERED) – RUNUP DETECTOR (TRIGGERED)
159°
158°
157°
20°
HIL Honokohau KPHO Kapoho
HKH
19°
LPH
Kawaihae
MIL HPO
156°
Figure 23.2. Tsunami run-up detectors used by the PTWC (McCreery, 2005).
19°
155°
GMT
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
262 T.S. Murty et al.
Figure 23.3.
DART systems that were in operation for the Pacific Ocean (Bernard, 2005).
time, which is needed as input of the computer models. Barring this information, the next best piece of information is the deep water signature of the tsunami. Several attempts have been made (Murty, 1977) in the past, to deduce the deep water signature from coastal tide gauge records. These attempts were largely unsuccessful, mainly because of nonlinearity and the contaminations of coastal tide gauge records by local resonances in inlets, bays and gulfs. The inverse problem of obtaining deep water signature from coastal records is a mathematically ill-posed problem, and hence has no unique solution. To get around this difficulty, the Pacific Marine Environmental Laboratory (PMEL) of National Oceanic and Atmospheric Administration (NOAA), USA has installed six DART (Deep-Ocean Assessment and Reporting of Tsunamis) systems in the Pacific Ocean (Figure 23.3; Bernard, 2005). These are basically pressure gauge sensors on the deep ocean floor that can measure and detect the tsunami as it is racing across the deep ocean. Three tsunami meters have been installed offshore of the Alaska–Aleutian trench, two offshore of Washington and Oregon states, and one in the equilateral Pacific far offshore of Ecuador. Chile installed one tsunami meter off its coastline at latitude 20◦ S. Each instrument package consists of a sea floor pressure sensor, in acoustic contact with an anchored Buoy (Figure 23.4) that transmits the ocean bottom data to a Geostationary Operational Environmental Satellite (GOES), from which the data is sent to the tsunami warning centers (Bernard, 2005). Data from these tsunami meters, free of the coastal effects, provide accurate forecasts of tsunamis by assimilating real-time tsunami meter data into nested numerical models (Titov et al., 2005). The value of these tsunami meters to date has been summarized by Gonzalez et al. (2005) and shown in Table 23.1.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Methodologies for tsunami detection
263
Figure 23.4. The DART system as deployed in the ocean (Bernard, 2005).
Gower (2006) provided most recent information on these DART Systems. According to him, there are now a total of 11 systems and he mentions that they detected the tsunami of 26 December 2004 in the Indian Ocean.
23.5
DEEP WATER SIGNATURE OF A TSUNAMI
Reid and Knowles (1970) and Knowles and Reid (1970) are probably the first to use the terminology “inverse tsunami problem” to mean the determination of the deepwater signature of a
264 T.S. Murty et al. Table 23.1.
Use of the data from DART systems in tsunami warning (Gonzalez et al., 2005).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Date – magnitude, time (UTC), location
Tsunameter records
11 July 2000 – 6.5 M, 01:33, ∼70 km southwest of Kodiak, AK
No tsunameters were triggered.
10 January 2001 – 6.9 M, 16:03, ∼110 km southwest of Kodiak, AK
Seismic wave induced 3.2 cm signal that triggered tsunameter D157 at 16:11. Subsequent record was tsunami-free. Seismic waves induced signals that triggered three tsunameter stations. Subsequent records were tsunami-free. Seismic waves induced signals that triggered all six tsunameter stations. Subsequent records were tsunami-free. No tsunameter were triggered.
5 May 2002 – 6.5 M, 05:37, ∼160 km southwest of Sand Point, AK 3 November 2002 – 7.9 M, 22:13, ∼145 km south of Fair Banks, AK 23 June 2003 – 7.1 M, 12:13, Near rat Island, Aleutian Islands
17 November 2003 – 7.5 M, 06:43, ∼90 km southwest of Amchitka, AK
Seismic waves induced signals that triggered three tsunameter stations. Subsequent records registered maximum deep ocean tsunami amplitudes of 2 cm, 0.5 cm, and <0.2 cm.
Contribution to operational decisions Corroborative information for decisions not to issue warning. Hawaii Department of Emergency Management also requested and received information on tsunameter (Yanagi, 2000). Tsunameter data allowed PTWC personnel to confirm that potentially destructive tsunami waves were not propagating towards Hawaii or the rest of the Pacific (Goldman, 2001). Corroborative information for decisions not to issue warning.
Corroborative information for decisions not to issue warning.
The combinations of no trigger at tsunameter D165 with tsunami-free signal at the Adak coastal gauge, and exercise of the WC/ATWC forecast tool led to early cancellations of the WC/ATWC warning/watch and PTWC Hawaii advisory (McCreery, 2003). 07:07 – Alaska warning issued. 07:33–08:03 – Tide gauge at Shemya, AK, registers 25 cm maximum. 08:12 – Warning cancelled.
tsunami based on records made at or near an island station. The significance of this terminology will be broadened to include determinations in the deeper water of such parameters as source region, tsunami signature, and tsunami energy based on recordings made near land, whether a mainland coast or an island. Reid and Knowles (1970) determined the deepwater signature of a tsunami based on its record near an island and some of their assumptions are: the tsunami in deepwater can be represented by a plane wave and the distance between the epicenter and the island is much larger than the dimension of the island. Processed such as scattering, diffraction, refraction, and resonance were ignored and the linear long-wave theory was used. Let P be the location near an island where the tsunami is recorded (Figure 23.5). Reid and Knowles defined a transfer function as the ratio of the Fourier Transform of the response at P to
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Methodologies for tsunami detection
Figure 23.5.
265
Schematic diagram of plane-wave incident on a cylindrical island. Input is trace X , response is trace Y . (Reid and Knowles, 1970).
the Fourier Transform of the input, the deepwater signature of the tsunami. If this transfer function (generally a complex function of frequency and direction of incoming waves) is determined, then from the mareograms, the deepwater signature of the tsunami can be determined. This transfer, P, can be determined from, observational studies, laboratory experiments, or numerical models; the topography and other conditions duplicated as closely as possible in the latter two cases. Reid and Knowles used a simple geometry for the island so that analytical solutions of the problem could be obtained. Then they used a numerical model to incorporate several improvements such as broad-band input spectrum, real topography of the island, and different wave directions. With reference to Figure 23.5, let Rm be the radius of the circle surrounding an island such that, for radii greater than Rm , the region is deep and has uniform depth. Let X (t) and Y (t) be the water-level records at some point slightly outside Rm and at the point, P, near the island. Assume that at the source, Q, the wave input is of plane progressive form and that the response at P is linear, then X and Y can be expressed as convolutions of each other: Y (t) =
∞
−∞
K(λ) X (t − λ) dλ
(23.1)
G(λ) Y (t − λ) dλ
(23.2)
and X (t) =
∞
−∞
where λ is the wavelength and K(λ) and G(λ) are the kernel functions that depend on the bathymetry of the island, the wave propagation direction, and the location of P. Let Fx ( f ) and Fy ( f ), where f is the frequency, be the Fourier transforms of X (t) and Y (t) and let R( f ) be the Fourier Transform of K(t). Then the Fourier Transform of G(t) is R( f )−1 . By
266 T.S. Murty et al. definition, then: Fx ( f ) =
∞
i2πft
∞
i2πft
−∞ X (t) e
dt
Fy ( f ) = −∞ Y (t) e dt ∞ R( f ) = −∞ K(t) ei2πft dt
(23.3)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
In general these will be complex. The Fourier Transform of equation (23.1) gives: Fy ( f ) = R( f ) Fx ( f )
(23.4)
If X (t) and Y (t) are known then from equation (23.4), the transfer function R( f ) can be determined.
23.6
DETECTION BY SATELLITES
The 26 December 2004 tsunami in Indian Ocean has been detected by a NOAA satellite and is shown in Figure 23.6. This is quite encouraging, even though at present the satellite coverage of the global oceans for real-time detection of tsunamis is not adequate. It just so happened that a satellite was at the right time and correct location to detect this tsunami. However, we cannot count on a lucky coincidence at all the time. However, the mere fact that satellite detection of tsunamis is a definite possibility gives hope, that sometime in the future, when the satellite coverage of the global oceans is adequate, very useful and real-time detection of tsunamis from satellite is possible.
23.7
DETECTION OF EARTHQUAKE AND TSUNAMI SIGNAL IN THE TROPOSPHERE
It is quite conceivable that as the tsunami is racing across the deep ocean, it could generate some sort of a pressure signal in the planetary boundary layer of the atmosphere, in addition to the pressure signal produced in the troposphere by the earthquake itself. Benioff et al. (1951) presented data on atmospheric sound waves generated by an earthquake. A microbarograph at Pasadena recorded a train of waves with periods ranging from ¾ to 1 s, attributed by the authors to an earthquake that occurred on 24 January 1951. Donn and Posmentier (1964) studied the ground-coupled air waves from the Alaskan earthquake of March 1964, by the micro pressure fluctuations recorded on microbarovariographs at Honolulu, Berkeley, and Palisades. The microbarovariographs have a sensitivity in the range of a few to several hundred microbars and a flat response up to 300 s. The authors indicated that atmospheric pressure oscillations due to earthquakes may be produced by three mechanisms in the region directly surrounding the recording station, the epicentral region, and a region where there is resonant ground-air coupling. Rayleigh waves emanating from the epicentral area can generate large pressure fluctuations in the atmosphere by impulsive effect, provided their vertical displacement is large. The following relation holds between the air-pressure
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Methodologies for tsunami detection
267
Figure 23.6. A simulated satellite picture of the Indian Ocean tsunami of 26 December 2004 (Wilson (June 2005) http://www.physicstoday.org).
perturbation, p; density of air, ρ; velocity of sound in the air, c; and velocity of the vertical ground motion, ν˙ (Donn and Posmentier, 1964): p = ρ c˙ν
(23.5)
The formula assumes that the interference from sound waves in adjacent regions is negligible. This condition is met, provided resonant coupling between ground and air is not appreciable and wavelengths of the Rayleigh waves are much greater than the height of the receiver (microbarograph) above the ground. If ρ = 1.19 × 10−3 g/cm and C = 330 m/s, the above equation becomes: p = 247a/t
(23.6)
where p is peak-to-peak pressure in microbars, a is the double amplitude of ground motion in centimeters, and t is the period of motion in seconds. In the second region, the vertical displacement of the ground near the epicenter will generate atmospheric acoustic waves. Donn and Posmentier (1964) visualized this mechanism as somewhat analogous to generation of acoustic-gravity waves from nuclear explosions.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
268 T.S. Murty et al. In the third region, vertical displacement of the earth’s surface far from the epicenter might generate acoustic waves that could be detected at distances of a few hundred kilometers, provided the appropriate conditions for ground-air coupling exists. It is this type of mechanism that accounted for the ground-coupled air waves reported by Benioff et al. (1951) for the Imperial Valley earthquake of 24 January 1951. Donn and Posmentier estimated from the seismograms that the maximum vertical ground motion at Palisades was 4.2 cm, attained by the initial Rayleigh wave. The observed maximum pressure fluctuation at Palisades was 40 µb and from equation (23.6) (assuming a period for local vertical ground motion as 23 s, the period of strongest pressure waves), one gets 3.72 cm for the required vertical motion. Because this value approximates the observed value of 4.2 cm, it was concluded that the main source for the pressure fluctuations was indeed the local vertical ground displacement due to the earthquake. Donn and Posmentier concluded that dispersion curves for Palisades and Honolulu showed typical continental and oceanic Rayleigh waves, respectively, and the record for Berkeley showed both; mechanism 2 seemed to have worked only at Berkeley. The Berkeley record showed large amplitude, long-period waves in the period range of 3–5 min. The superimposed smaller waves were continuations for earlier ground-coupled air waves. The structure of the initial long-period waves is different from that of the background gravity waves. Donn and Posmentier attribute the dissimilarity between these waves and those caused by nuclear explosions (those due to earthquake show clearer dispersion) to the fact that in earthquakes the generation area is large, whereas an explosion is a point source.
23.8
DETECTION OF THE EARTHQUAKE AND TSUNAMI SIGNALS IN THE IONOSPHERE
It is quite possible that the atmospheric pressure signal in the troposphere could be amplified through the acoustic-internal gravity waves and could be detectable in the ionosphere. Davies and Baker (1965) detected ionosphere disturbances associated with the Alaskan earthquake of March 1964. They made observations at Boulder, CO, on frequencies of 4 and 5 Mc/s with vertical propagations, and on 10 Mc/s at WWVH in Hawaii, 5000 km from Boulder. During 4 years of observation, they recorded only one other such incident; that was due to the 9 July 1962, nuclear explosion at Johnston Island. Important contributions were made by Romanova (1970) and Petukhov and Romanova (1971) to understand generation and vertical propagation of acoustic waves in the atmosphere due to an earthquake, and subsequent heating in the atmosphere at different levels, caused by these waves. The earlier paper deals mainly with development of the theory, the later paper with its application. Results of the second paper are summarized here. Petukhov and Romanova applied Romanova’s theory to calculate the dissipation of infrasonics from earthquake in the atmosphere in the form of heat. In particular, they estimated the rate and amount of atmospheric heating due to acoustic waves traveling upward from two great earthquake, the Alaskan Earthquake of 1964 and the Katinoko, Japan, Earthquake of 1968. This atmospheric heating had a significant role in the observed ionospheric disturbances. The authors initially started with the same initial conditions as Donn and Posmentier (1964), i.e. with equation (23.5) or (23.6). This means the Rayleigh waves from the epicenter propagate over the earth’s surface with a velocity of roughly 3 km/s, which is supersonic for air. The vertical motion of the Rayleigh waves produces a vertical impulse effect in the atmosphere, and produces pressure perturbations. The relation between the pressure perturbation, p , and the displacement, a, of the ground surface is given (as in Donn and Posmentier’s work) by equation (23.5).
Methodologies for tsunami detection
269
Under the linear theory for small amplitude oscillations, the Strokes–Kirchoff equation for the attenuation coefficient, α , of acoustic waves is given by:
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
1 2πω2 ρµa 2 4 γ − 1 α = 260 + γP γn 3 Pn
(23.7)
where ω is the frequency expressed in Hertz, ρ is the air density , γ is ratio of specific heats, µa is the molecular weight of air, n is the density of air molecules, and Pn is the Prandtl Number (0.7 for air). The equation used for atmospheric heating is: ∂θ¯ a ∂ = − (cE) ∂t ∂z
(23.8)
where θa is the absolute temperature, E is the wave energy, c is the velocity of sound, and bar denotes averaging over a period. After some algebra, equation (23.8) becomes: z
∂θ¯ a = 2cα(z)(cp ρ)−1 exp −2 α(z) dz ∂t 0
(23.9)
where cp is the specific heat at constant pressure. This formula was used to calculate the temperature change for the atmosphere over Berkeley, Boulder, and Palisades for the 1964 Alaskan Earthquake and for the atmosphere over Maui for the 1968 Katinoko earthquake. The calculations showed that maximum heating occurred in altitudes of 170–190 km. At this altitude, the atmosphere over Berkeley was heated by 900◦ . Although this value appears large, the peak-to-peak amplitude of the ground motion with a period of 23 s was 12.6 cm. Ionospheric disturbances were observed associated with this earthquake (i.e. 1964 Alaskan Earthquake). An important method of identifying these disturbances was an increase of about 20 km in the altitude of the critical reflecting layers. The role of atmospheric heating in this could be significant, because, when the 170–190 km altitude layer (with maximum heating) expands with heat, it disturbs the layers above it. Under the assumption of an isobaric process, the rise in the critical layer z = z0 caused by heating and expansion of the underlying layers can be computed from the following relation: z(z0 ) =
0
z
θa (z) dz T0 (z)
(23.10)
Here T0 (z) is temperature at height, z, in the atmosphere before heating starts, and θ¯ a = (∂θa /∂z)t; ∂θ¯ a /∂z is calculated from equation (23.9). This calculation gives a rise of 20 km for the layer at 240 km altitude over boulder and the observed rise is of the same order. Over Mauii, the 200-km altitude layer theoretically should rise 15 km, although it was observed to rise only 5–6 km. In contrast to the linear theory, the nonlinear theory shows that lower atmospheric layers are heated more and higher layers are heated less, thus the nonlinear theory gives a smaller rise of the critical layers. Another important point Petukhov and Romanova (1971) emphasized is that because of the decrease of atmospheric density with altitude, temperature fluctuations will
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
270 T.S. Murty et al. mainly travel upward, and a new isothermal condition is rapidly established in the atmosphere above the heat source at the ground. To obtain this isothermal condition at the 170–200 km level would only take about 3 h. Next details on analysis of ionosphere disturbances due to seismic waves are presented. Yuen et al. (1969) studied the 16 May 1968 Hachinohe, Japan, Earthquake and analyzed and compared the seismic, atmospheric, and ionospheric data. Some important results are: acoustic waves generated in the atmosphere by seismic waves due to the earthquake traveled to 300 km altitude and created oscillatory disturbances in the ionosphere. Acoustic waves in the period range of 20–25 s were attenuated more strongly than those with longer periods of the order of 2 min. Value of the results lies in the fact that the observed pressure changes, determined from Doppler records, agreed well with theoretically expected values. Donn and Posmentier (1964), for the Alaskan Earthquake of March 1964, could not directly compare the seismograms with microbarograms because the vertical component seismograms were illegible, due to large ground displacements. For the Hachinohe Earthquake, the average double amplitude of the Rayleigh wave was 0.28 cm, as shown by the seismograph at the Hawaiian Institute of Geophysics (HIG). The average Rayleigh wave period was taken as 25 s, and this gives from equation (23.2), a value of 2.8 µb for the average peak-to-peak pressure variation, an order of magnitude less than the pressure variation associated with the 1964 Alaskan Earthquake. The pressure fluctuations for the Hachinohe Earthquake were too small for the microbarographs at Honolulu to record. However, Yuen et al. (1969, p. 2258) were able to deduce the pressure changes from the ionospheric Doppler record. They described the HF Doppler technique, reflected signal is measured. The Doppler frequency shifts indicate changes in the ionosphere caused by movements of the reflecting layer or by variations in the electron concentration below the point of reflection. A few minutes after the Rayleigh waves reached Hawaii, short-period oscillations were produced at heights near 200 km and long-period oscillations were produced near 300 km. The Faraday rotation of the polarization of continuous wave signals, transmitted from geostationary satellites, were measured to continuously obtain total electron content data. The period after the earthquake showed no unusual behavior. This indicated that no significant net production or loss of electrons were associated with disturbances on the Doppler and ionogram records, and implied that these disturbances were produced by electron concentration pressure waves in the ionosphere. According to Yuen et al. (1969): “An acoustic longitudinal pressure wave with a vertical velocity component would be expected to move the contours of constant electron concentration up and down with an oscillatory motion similar to the pressure wave itself. This would mean that the height of reflection for HF radio signals as those used in the Doppler system would also move up and down with an oscillatory motion. The pressure wave at ground level generated by a Rayleigh wave would be proportional to the seismic motions at the earth’s surface. Thus similar shapes would be expected on the seismic and Doppler records.”
To support their contentions that the ionospheric disturbances shown on the Doppler recordings were caused by acoustic waves generated by Rayleigh waves from the earthquake, Yuen et al. used ray tracing, and calculated the attenuation by the 1966 US standard atmosphere. In this calculation, the exospheric temperature was taken to be 1200◦ K over Hawaii at the time of the earthquake. Assuming adiabatic conditions, Yuen et al. (1969) calculated the pressure change from the change in the electron concentration. The effects of the Krakatoan eruption of August 1883, and the atmospheric disturbances that occurred globally pressure disturbances is presented here, briefly. Scott (1883) described the atmosphere pressure fluctuations observed following the Krakatoan eruption, and Strachey (1883) made deductions based on this work. Scott mentioned that the volcanic explosion probably occurred between 16:00 on 26 August and 05:00 on 27 August
Methodologies for tsunami detection
271
1883, local time (in Greenwich time, between 09:00 and 22:00 27 August 1883). On the average, the pressure wave took 36 h 37 min to travel around the earth from east to west, and 35 h 17 min from west to east. By working back with this data, Strachey deduced that the Krakatoan eruption occurred at 02:24 GCT (or 09:24 local time) on 27 August 1883. The atmosphere wave on the average traveled 674 mph from east to west and 706 mph from west to east.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
23.9 TSUNAMI DETECTION BY ANIMALS It has been observed for some time that certain land and marine animals might be sensitive to earthquake and tsunami signals. A notable point about the disastrous tsunami of 26 December 2004 in the Indian Ocean is the fact that, very few animals died in the tsunami. It is generally believed that, as evolution progressed, while the homo sapiens (humans) could out think all other animals, certain animals are more sensitive to vision, hearing, touch and smell. For example, frogs and fish can see part of the ultraviolet and probably some of the infrared, which human eye cannot see. Human eye can see only the visible part of the electromagnetic spectrum. While the human ear can only hear the audible part of the acoustic spectrum certain animals can hear parts of the ultra and infrasonics. Hence it is quite conceivable that animals may have some detection abilities for earthquakes and tsunamis. May be they can react to the p and s waves emanated by the earthquake. A lot more useful research could be done on this and hopefully in the future, additional real-time input can be obtained from animal behavior.
23.10
SUMMARY
The traditional method of tsunami detection is through a tide gauge network. Its strength is its robustness and dependability. Its weakness is, in certain instances, there is insufficient time to provide a warning. It is here to stay as one of the most important components of a tsunami warning system. However, other technologies are emerging, if not to replace the role of the tide gauge network, but to provide additional valuable data in real time. These technologies range from run-up detectors on the shore (which confirm land inundation by the tsunami) to DART systems to detect the deep ocean tsunami signature, to satellite detection in the troposphere and the ionosphere, to finally even detection of tsunamis through animal behavior. REFERENCES Benioff, H., Ewing, M. and Press, F. (1951). Sound waves in the atmosphere generated by a small earthquake. Proc. US Natl. Acad. Sci., 37, 600–603. Bernard, E.N. (2005). National tsunami hazard mitigation program: a successful state-federal partnership. Natural Hazard, 35, 5–24. Davies, K. and Baker, D.M. (1965). Ionosphere effects observed around the time of the Alaskan Earthquake of March 28, 1964. J. Geophys. Res., 70, 2251–2253. Donn, W.L. and Posmentier, E.S. (1964). Ground-coupled airwaves from the Great Alaskan Earthquake. J. Geophys. Res., 69, 5357–5361. Goldman, J. (2001). NOAA Tsunami Buoy “Feels” Alaska Earthquake. http://www.noaa.news.noaa.gov/ stories/s560.htm Gonzalez, F.I., Bernard, E.N., Meinig, C., Eble, M.C., Mofjeld, H.O., and Stalin, S. (2005). The NTHMP tsunameter network. Natural Hazard, 35, 25–39.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
272 T.S. Murty et al. Gower, J.F.R. (2006). U.S. warning system detected the Sumatra Tsunami. EOS, Transactions, Am. Geophy. Union, 87(10), 105, 108. Knowles, C.E. and Reid, R.O. (1970). The inverse tsunami problem for symmetric islands of simple shape. Texas A and M University Department of Oceanography, College Geo-Science, Technology Report, 69 pp. McCreery, C. (2003). Personal e-mail communication to E. Bernard on June 23, 2003. McCreery, C.S. (2005). Impact of the national tsunami hazard mitigation program on operations of the Richard H. Hagemeyer Pacific Tsunami Warning Center. Natural Hazard, 35, 73–88. Murty, T.S. (1977). Seismic sea waves – Tsunamis. Bulletin No. 198, Canadian Bulletin of Fisheries and Aquatic Sciences, Ottawa, Canada, 337 pp. Petukhov, V.K. and Romanova, N.N. (1971). Effects caused by acousto-gravitational waves in the upper atmosphere. Atmos. Oceanic Phys., 7, 219–223. Reid, R.O. and Knowles, C.E. (1970). An inverse tsunami problem. In: W.M. Adams (ed.) Tsunamis in the Pacific Ocean, East West Center Press, Honolulu, Hawaii, pp.399–406. Romanova, N.N. (1970). The vertical propagation of short acoustic waves in the real atmosphere. Atmos. Oceanic Phys., 6, 134–145. Scott, R.H. (1883). Note on a series of barometrical disturbances which passed over Europe between the 27th and 31st of August, 1883. Proc. R. Soc. London, Ser. A, 36, 139–143. Strachey, R. (1883). A Note on barometrical disturbances. Proc. R. Soc. London, 36, 139–151. Titov, V.V., Gonzalez, F.I., Bernard, E.N., Eble, M.C., Mofjeld, H.O., Newman, J.C., and Venturato, A.J. (2005). Real-time tsunami forecasting: challenges and solutions, Natural Hazard, 35, 41–58. Yanagi, B. (2000). Personal e-mail communication to WC/ATWC on July 12, 2000. Yuen, P.C., Weaver, P.F., Suzuki, R.K., and Furumoto, A.S. (1969). Continuous travelling coupling between seismic waves and the ionosphere evident in May, 1968, Japan Earthquake Data. J. Geophys. Res., 74, 2256–2264.
CHAPTER 24
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunami Travel Time Atlas for the Indian Ocean P.K. Bhaskaran Department of Ocean Engineering and Naval Architecture, Indian Institute of Technology Kharagpur, Kharagpur, India S.K. Dube Indian Institute of Technology Kharagpur, India T.S. Murty Department of Civil Engineering, University of Ottawa, Ottawa, Canada A. Gangopadhyay and A. Chaudhuri School of Marine Science and Technology, University of Massachusetts at Dartmouth, USA A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology Delhi, New Delhi, India
24.1
INTRODUCTION
Tsunami travel time charts have been prepared for 250 locations at the edge of the Indian Ocean (Bhaskaran et al., 2005), as well as some locations in the Pacific Ocean that are at the edge of the Indian Ocean. The reason some Pacific stations are included is due to the fact that some tsunami energy can leak from the Pacific Ocean into the Indian Ocean and vice versa. Table 24.1 lists the number of locations in each country and Figure 24.1 shows these 250 locations.
24.2 A LIST OF THE 250 LOCATIONS IN 35 COUNTRIES Table 24.2 lists the names of the locations, and their latitudes and longitudes in the 35 countries. 273
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
274
P.K. Bhaskaran et al.
Table 24.1.
Number of locations in each country for which tsunami travel time charts are prepared.
Country
Number of locations
Country
Number of locations
Country
Number of locations
19 1 3 1 2 3 47 31 3
Kenya Kuwait Madagascar Malaysia Maldives Mauritius Mozambique Myanmar Oman
3 1 10 12 1 1 9 9 5
Pakistan Philippines Qatar Reunion Saudi Arabia Seychelles Singapore Somalia South Africa
3 11 2 4 3 1 1 8 11
Australia Bahrain Bangladesh Brunei Comoros Egypt India Indonesia Iran
Figure 24.1.
Country
Number of locations
Sri Lanka Sudan Taiwan Tanzania Thailand UAE Vietnam Yemen
15 2 2 7 6 3 8 2
250 locations in 35 countries for which tsunami travel time charts are prepared.
Tsunami travel time atlas for the Indian Ocean
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Table 24.2.
275
Locations in each country for which tsunami travel time charts have been prepared.
Country
City
Latitude
Longitude
Australia
Adelaide Albany Bathurst Island Brome Bunbury Cape Leveque Cape Saint Lambert Carnarvon Crocker Dampier Downs Darwin Exmouth Geraldton Kalbarri Melbourne Nhulunbuy Onslow Perth Port Hedland
34.52◦ S 34.95◦ S 11.75◦ S 17.97◦ S 33.33◦ S 16.41◦ S 14.28◦ S 24.85◦ S 11.03◦ S 18.52◦ S 12.25◦ S 21.90◦ S 28.81◦ S 27.70◦ S 37.50◦ S 12.50◦ S 21.68◦ S 31.57◦ S 20.40◦ S
138.30◦ E 117.90◦ E 130.68◦ E 122.25◦ E 115.56◦ E 122.91◦ E 127.71◦ E 113.75◦ E 136.63◦ E 123.45◦ E 130.51◦ E 114.16◦ E 114.60◦ E 114.16◦ E 145.00◦ E 136.93◦ E 115.20◦ E 115.52◦ E 118.60◦ E
Bahrain
Manama
26.236◦ N
50.583◦ E
Bangladesh
Chittagong Cox Bazaar Dhulasar
24.05◦ N 21.26◦ N 21.87◦ N
91.00◦ E 91.59◦ E 90.23◦ E
Brunei
Bandar Seri Begawan
04.93◦ N
114.96◦ E
Comoros
Dzaoudzi Moroni
12.80◦ S 11.67◦ S
45.30◦ E 43.27◦ E
Egypt
Al-Ghardaqah Hurghada Suez
23.88◦ N 27.15◦ N 30.00◦ N
35.27◦ E 33.50◦ E 32.30◦ E
India
Andrath Island Bhatkal Bhavnagar Chennai Chetlat Island Chirala Cochin (Kochi) Coco Channel Dadar and Nagar Daman Diu Dwarka Gopalpur Haldia Haveli Henhoaha
11.00◦ N 13.96◦ N 21.45◦ N 13.08◦ N 11.76◦ N 15.98◦ N 09.58◦ N
73.16◦ E 74.58◦ E 72.10◦ E 80.19◦ E 76.83◦ E 80.08◦ E 76.20◦ E
14.08◦ N
93.30◦ E
20.25◦ N 20.40◦ N 22.25◦ N 19.27◦ N 22.03◦ N 20.05◦ N 06.80◦ N
72.57◦ E 71.02◦ E 69.05◦ E 84.95◦ E 88.03◦ E 73.00◦ E 93.81◦ E (Continued)
276
P.K. Bhaskaran et al.
Table 24.2.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Country
Indonesia
(Continued) City
Latitude
Longitude
Kakana Kandla Kanyakumari Karaikal Karwar Kavaratti Island Kozhikode Kumta Machilipatnam Mahim Malvan Mangalore Minicoy Island Misha Mumbai Murmagao Nachugo Nagapattinam Nellore Panaji Pondicherry Porbandar Port Blair Puri Quilon (Kollam) Rameswaram Rapur Ratnagiri Thoothukkudi Trivandrum Veraval Visakhapatnam
09.11◦ N 23.00◦ N 08.07◦ N 10.59◦ N 14.83◦ N 10.53◦ N 11.15◦ N 14.48◦ N 16.15◦ N 19.66◦ N 16.05◦ N 12.55◦ N 08.48◦ N 08.00◦ N 18.55◦ N 15.25◦ N 10.71◦ N 10.77◦ N 14.27◦ N 15.25◦ N 11.59◦ N 21.44◦ N 11.68◦ N 19.50◦ N 08.90◦ N
92.81◦ E 70.10◦ E 77.58◦ E 79.50◦ E 74.15◦ E 72.71◦ E 75.43◦ E 74.41◦ E 81.20◦ E 72.76◦ E 73.50◦ E 74.47◦ E 73.02◦ E 93.36◦ E 72.50◦ E 73.56◦ E 92.35◦ E 79.88◦ E 79.59◦ E 73.50◦ E 79.50◦ E 69.43◦ E 92.77◦ E 85.58◦ E 76.63◦ E
09.28◦ N 23.05◦ N 17.13◦ N 08.50◦ N 08.41◦ N 20.53◦ N 17.45◦ N
79.37◦ E 68.83◦ E 73.32◦ E 78.12◦ E 77.00◦ E 70.27◦ E 83.20◦ E
Ambon Balikpapan Banda Aceh Bandar Lampung Bandjarmasin Bengkulu Bula Denpasar Genteng Jakarta Jayapura Ketapang Kupang Langsa Majene Manado Mataram
03.43◦ S 01.25◦ S 05.35◦ N 05.30◦ S 03.20◦ S 03.50◦ S 03.12◦ S 08.39◦ S 07.35◦ S 06.09◦ S 02.28◦ S 01.83◦ S 10.22◦ S 04.47◦ N 03.55◦ S 01.29◦ N 08.35◦ S
128.12◦ E 116.83◦ E 95.20◦ E 104.30◦ E 114.35◦ E 102.12◦ E 130.45◦ E 115.13◦ E 106.33◦ E 106.49◦ E 140.38◦ E 109.98◦ E 123.63◦ E 97.98◦ E 118.98◦ E 124.51◦ E 116.07◦ E (Continued)
Tsunami travel time atlas for the Indian Ocean Table 24.2.
(Continued)
Country
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
277
City Mentok Meulaboh Namlea Padang Pontianak Semarang Sibolga Sumbawabesar Surabaya Tanjungbalai Tarakon Tobelo Ujung Pandang Yogyakarta
Latitude
Longitude
02.07◦ S 04.17◦ N 03.25◦ S 01.00◦ S 00.03◦ S 07.00◦ S 01.70◦ N 08.50◦ S 07.17◦ S 01.00◦ N 03.33◦ N 01.75◦ N 05.10◦ S 07.49◦ S
105.20◦ E 96.15◦ E 127.12◦ E 100.20◦ E 109.15◦ E 110.26◦ E 98.80◦ E 117.42◦ E 112.45◦ E 103.32◦ E 117.63◦ E 127.98◦ E 119.20◦ E 110.22◦ E
27.12◦ N 28.59◦ N 25.642◦ N
56.15◦ E 50.46◦ E 57.772◦ E
Iran
Bandar Abbas Bandar-e-Bushehr Jask
Kenya
Lamu Malindi Mombasa
02.28◦ S 03.12◦ S 04.02◦ S
40.90◦ E 40.05◦ E 39.43◦ E
Kuwait
Kuwait
28.59◦ N
47.52◦ E
Madagascar
Antalaha Antsiranana Cape Saint Marie Hell Ville Manakara Morondava Muhajanga Tambohorano Toamasina Toliary
14.88◦ S 12.25◦ S 25.57◦ S 13.40◦ S 22.15◦ S 20.32◦ S 15.40◦ S 17.50◦ S 18.10◦ S 21.50◦ S
50.27◦ E 49.20◦ E 45.17◦ E 48.28◦ E 48.00◦ E 44.28◦ E 46.25◦ E 43.59◦ E 49.25◦ E 43.74◦ E
Malaysia
Bintulu Georgetown Johor Bahru Klang (Kelang) Kota Bahru Kota Kinabalu Kuala Lumpur Kuala Terengganu Kuantan Kuching Melaka Sandakan
03.20◦ N 05.25◦ N 01.28◦ N 03.02◦ N 06.07◦ N 05.98◦ N 03.09◦ N 05.20◦ N 03.49◦ N 01.53◦ N 02.15◦ N 05.86◦ N
113.02◦ E 100.20◦ E 103.46◦ E 101.26◦ E 102.14◦ E 116.06◦ E 101.41◦ E 103.08◦ E 103.20◦ E 110.33◦ E 102.15◦ E 118.06◦ E
Maldives
Male
04.00◦ N
73.00◦ E
Mauritius
Port Louis
20.10◦ S
57.30◦ E (Continued)
278
P.K. Bhaskaran et al.
Table 24.2.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Country
(Continued) City
Latitude
Longitude
Mozambique
Angoche Beira Inhambane Maputo Nacala Pebane Pemba Quelimane Vilanculos
16.17◦ S 19.50◦ S 23.02◦ S 25.58◦ S 14.31◦ S 17.23◦ S 12.58◦ S 17.53◦ S 22.02◦ S
39.97◦ E 34.52◦ E 35.92◦ E 32.32◦ E 40.34◦ E 38.17◦ E 40.30◦ E 36.58◦ E 35.32◦ E
Myanmar
Kadonkani Kawthaung Kyaukpyu Mawlamyine (Moulmein) Mergui Sandoway Sittwe Tavoy Yangon
15.83◦ N 10.02◦ N 19.45◦ N
95.18◦ E 98.53◦ E 93.55◦ E
16.30◦ N 12.43◦ N 18.47◦ N 20.15◦ N 14.12◦ N 16.47◦ N
97.37◦ E 98.56◦ E 94.45◦ E 92.09◦ E 98.30◦ E 96.10◦ E
Oman
Duqm Masirah Muscat Salalah Sur
19.65◦ N 20.417◦ N 23.37◦ N 16.56◦ N 22.34◦ N
57.7◦ E 58.833◦ E 58.36◦ E 53.59◦ E 59.32◦ E
Pakistan
Gwadar Jiwani Karachi
25.10◦ N 25.117◦ N 24.53◦ N
62.18◦ E 61.733◦ E 67.00◦ E
Philippines
Bulan Cebu City Davao Iloilo Laoag Manila Mindaro Palawan Island Quezon City Siaton Zamboanga City
12.66◦ N 10.18◦ N 7.08◦ N 10.68◦ N 18.23◦ N 14.40◦ N 13.00◦ N 9.50◦ N 14.38◦ N 09.08◦ N 06.54◦ N
123.88◦ E 123.54◦ E 125.63◦ E 122.55◦ E 120.60◦ E 121.03◦ E 121.00◦ E 118.50◦ E 121.00◦ E 123.08◦ E 122.04◦ E
Qatar
Doha Dukhan
25.15◦ N 25.3◦ N
51.36◦ E 50.8◦ E
Reunion
Saint Benoit Saint Denis Saint Paul Saint Pierre
21.03◦ S 20.87◦ S 21.00◦ S 21.27◦ S
55.71◦ E 55.46◦ E 55.27◦ E 55.53◦ E
Saudi Arabia
Alqunfidha Jeddah Rabigh
19.03◦ N 21.53◦ N 22.50◦ N
41.04◦ E 39.17◦ E 39.05◦ E (Continued)
Tsunami travel time atlas for the Indian Ocean Table 24.2.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Country
279
(Continued) City
Latitude
Longitude
Seychelles
Victoria
04.63◦ S
55.47◦ E
Singapore
Singapore City
01.22 N
103.55◦ E
Somalia
Berbera Boosaso Eyl Kisimayo Merca Mogadishu Obbia Ras Hafun
10.47◦ N 11.28◦ N 08.00◦ N 00.22◦ S 01.48◦ N 02.06◦ S 05.33◦ N 10.48◦ N
45.03◦ E 49.18◦ E 49.82◦ E 42.32◦ E 44.50◦ E 45.37◦ E 48.50◦ E 51.33◦ E
South Africa
Cape Agulhas Cape Town Durban East London Grahamstown Ladysmith Mosselbaai Mtunzini Pietermaritzburg Port Elizabeth Port Saint Johns
34.83◦ S 33.93◦ S 29.87◦ S 32.97◦ S 33.19◦ S 28.02◦ S 34.18◦ S 28.97◦ S 29.36◦ S 33.96◦ S 31.62◦ S
20.00◦ E 18.47◦ E 30.99◦ E 27.87◦ E 26.31◦ E 32.66◦ E 22.13◦ E 31.77◦ E 30.23◦ E 25.59◦ E 29.53◦ E
Sri Lanka
Batticaloa Colombo Dehiwala-mount Galle Hambantota Jaffna Kankesanturai Lavinia Mannar Matara Moratuwa Mullaittivu Negombo Okanda Talawilla Trincomalee
7.72◦ N 05.56◦ N
77.73◦ E 79.58◦ E
06.05◦ N 6.12◦ N 09.45◦ N 9.85◦ N 06.51◦ N 8.98◦ N 5.95◦ N 06.45◦ N 09.25◦ N 07.12◦ N 06.65◦ N 08.13◦ N 08.38◦ N
80.10◦ E 81.12◦ E 80.02◦ E 80.08◦ E 79.52◦ E 79.92◦ E 80.55◦ E 79.55◦ E 80.80◦ E 79.50◦ E 81.77◦ E 79.70◦ E 81.15◦ E
Sudan
Port Sudan Suakin
19.38◦ N 19.08◦ S
37.08◦ E 37.17◦ E
Taiwan
Hsin-chu Kao-hsiung
24.80◦ N 22.60◦ N
120.98◦ E 120.28◦ E
Tanzania
Dar es Salaam Kilwa Lindi Mtwara Tanga
06.50◦ S 08.55◦ S 09.56◦ S 10.20◦ S 05.05◦ S
39.12◦ E 39.30◦ E 39.61◦ E 40.20◦ E 39.02◦ E
◦
(Continued)
280
P.K. Bhaskaran et al.
Table 24.2.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Country
(Continued) Latitude
Longitude
Wete Zanzibar
05.04◦ S 06.10◦ S
39.43◦ E 39.20◦ E
Thailand
Bankok Phuket Ranong Satun Songkhla Surat Thani
13.45◦ N 07.53◦ N 9.962◦ N 6.617◦ N 07.13◦ N 09.06◦ N
100.35◦ E 98.24◦ E 98.638◦ E 100.067◦ E 100.37◦ E 99.20◦ E
UAE
Abu Dhabi Dubai Sharjah
24.28◦ N 25.271◦ N 25.20◦ N
54.25◦ E 55.329◦ E 55.24◦ E
Vietnam
Da Nang Dong Hoi Haiphong Hon Gai Hue Nha Trang Qui Nhon Vinh
16.04◦ N 17.53◦ N 20.47◦ N 20.57◦ N 16.30◦ N 12.16◦ N 13.40◦ N 18.45◦ N
108.13◦ E 106.58◦ E 106.41◦ E 107.05◦ E 107.35◦ E 109.10◦ E 109.13◦ E 105.38◦ E
Yemen
Aden Al Mukalla
12.45◦ N 14.33◦ N
45.00◦ E 49.02◦ E
24.3
City
BACKGROUND ON TRAVEL TIME COMPUTATION OF TSUNAMIS
Tsunamis are long gravity waves, just like tides and storm surges. Since tsunamis can be categorized as long waves, tsunami travel times can be computed with water depth as the sole variable (Murty, 1977; Murty et al., 1987), atleast to zero order. Certainly, there are some first-order corrections, however for all practical purposes these corrections will be ignored here. As of today, no technology exists to predict a tsunami event well in advance until the earthquake happens. Contrary to popular belief, the tsunami travel times do not depend upon the magnitude of the under-ocean earthquake that generated the tsunami. For the Pacific Ocean, it has been clearly demonstrated that the computed tsunami travel times using the zero order approximation are correct to plus or minus one minute for each hour of travel. The advantage of this zero order of approximation is that tsunami travel times to selected locations around the rim of the Indian Ocean as well as to selected island sites can all be pre-computed in advance once and for all. This set of information can be stored in electronic as well as a tsunami travel time atlas format and can be quickly accessed in real tsunami events with a minimum effort. The software to produce this atlas was provided to us by Geoware Ltd. of Honolulu, USA (www.geoware.com). The ETOPO-2 ocean bathymetry, (www.gfdl.noaa.gov) with a resolution of 2 min of arc is the input data and the output is tsunami travel time contours in hours. The technique used to compute travel times over the entire grid is an application of Huygens principle which states that all points on a wave-front are point sources for secondary spherical waves. From the starting point, times are computed to all surrounding points. The grid point with minimum time is then taken as the next starting point and times are computed from
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunami travel time atlas for the Indian Ocean
281
there to all surrounding points. The starting point is continually moved to the point with minimum total travel time until all grid points have been evaluated. In the countries surrounding the Indian Ocean rim, 250 locations were selected for this study. The travel time of tsunami waves from the epicenter to various coastal regions will be evaluated for all the sample points identified. Following the disastrous tsunami in Hawaii on April 1, 1946 (Okal et al., 2002) from an earthquake in the Aleutian Islands of USA, a Pacific Ocean Tsunami warning system was established based in Ewa Beach, Oahu Island, Hawaii, USA. In the immediate vicinity of islands the catastrophe resulting from a tsunami is considered as enormous (Yeh et al., 1994). Although majority of the tsunami results from underwater earthquakes, there has been controversies pertaining to generation of tsunamis by submarine mass failure which undoubtedly is a rare phenomenon and virtually no attention was given to its hazard mitigation. The 1998 Papua New Guinea tsunami resulted from a large underwater slump. The most basic information a tsunami warning center requires is ETA’s (expected times of arrival) of the first tsunami wave at selected coastal locations, from the area of tsunami generation in the ocean. Almost always, the first wave in a tsunami event is not the wave with the greatest amplitude, nevertheless, tsunami travel time charts are generally constructed for the first wave, rather than the wave with the highest amplitude. Advance knowledge of travel times for the first wave, provides some additional valuable time for evacuation of people, if and when evacuation is needed. Also, tsunami travel times can be pre-computed, independent of the seismic moment magnitude of the earthquake, only for the first wave. The heights of the tsunami waves are not known till the event actually happens, and hence no pre-determination of the travel time of the highest wave can be made. It is useful to construct tsunami travel time charts for pre-selected coastal locations. Here the assumption is that the epicenter is located at this site. However, tsunami travel time charts are reversible, in the sense that the travel times are exactly the same, no matter in which direction the tsunami travels on a given chart, i.e. from an epicenter in the ocean to a coastal site of vice versa (i.e. from a coastal site to an epicenter in the ocean). Tsunami travel times in hourly contours are shown on a tsunami travel time chart. Even though contours can be constructed for 30 min intervals (or for that matter, at any desired time interval), one does not want to crowd the charts with too many contours. Once a reasonable number of tsunami travel time charts are prepared (say a total of about 250) for the Indian Ocean, for selected coastal and island locations, as well as for all historical tsunami events, it is quite probable that for any future tsunami events, the travel time information that is required could be quickly and effortlessly be obtained from these charts. 24.4
EPICENTERS OF TSUNAMI-GENIC EARTHQUAKES
Epicenters of historical earthquakes that generated tsunamis have been shown in Figure 24.2 and the latitude and longitude information as well as the area of the epicenters are listed in Table 24.3.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
282
P.K. Bhaskaran et al.
Figure 24.2.
Epicenter locations of the past earthquakes that generated tsunamis, which had some impact in the Indian Ocean.
Table 24.3.
Geographical coordinates and the source region of the epicenters used for the study. All longitudes are east and negative latitudes denote southern hemisphere.
Source region India, Rann of Kutch Off Sri Lanka Northern end Bay of Bengal West Car Nicobar, Bay of Bengal Andaman Sea Off Burma Coast Sumatra, Malay Peninsula Southwest Sumatra, Indonesia India, East of Java Philippines, Breueh Island, Bengal Bay Banda Atjeh South Indian Ocean Southwest Sumatra Java Trench
Latitude 23.583 8.570 21.000 9.000 12.500 18.500 5.500 2.000 5.200 5.700 5.500 −13.800 −1.000 4.370
Longitude 68.367 81.230 89.000 92.000 92.500 93.400 94.000 94.500 94.750 95.080 96.000 97.450 97.500 97.700 (Continued)
Tsunami travel time atlas for the Indian Ocean
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Table 24.3.
283
(Continued)
Source region
Latitude
Longitude
Southwest Sumatra (Banda Sea) Southwest Sumatra Southwest Sumatra Southwest Sumatra Sumatra Penang Island Sumatra Southwest Sumatra South Java Sea Djawa Java South Java Sea Hainan Island, Gulf of Tonkin Djawa (Java) Chilaiap, South Java Off Java Macau, South China Sea Indonesia, Bali Island Guam, Saip: Mariana Islands Djawa (Java), Flores Sea Indonesia, Flores Sea Indonesia, Makassar Strait Indonesia, Sunda Islands Indonesia, Makassar Strait Indonesia, Sumbawa Philippines, Luzon Islands Philippines, West Luzon Island Indonesia, Makassar Strait Philippines, West Luzon Island Celebes, Sea, Indonesia: Sulawes Off Badoc., South China Sea North Molucca Islands, Indonesia Indonesia, Celebes, Banda Sea Indonesia, Sulawesi Indonesia, Java South China Sea Philippines Philippines, Luzon Islands Philippines: Luzon: Central Philippines, East of Luzon Islands Manila, West Luzon Indonesia, Flores Sea Tolitoli, Sulawesi, Indonesia Philippines, Luzon Islands Philippines, Manila Philippines, Manila Sulu Sea Taiwan Philippines: Luzon: South Batan Indonesia Northeast Mindoro Islands, Sulu Sea
0.000 1.500 −0.150 −5.000 −2.000 5.383 −2.000 −5.410 −5.600 −6.500 −8.200 18.000 −6.500 −8.000 −13.080 22.000 8.000 15.120 −7.000 −8.200 0.000 −11.090 −3.180 −7.000 16.000 15.910 −3.470 17.470 −0.500 18.000 −0.380 0.060 0.730 −10.000 23.000 16.000 13.900 14.500 15.500 14.600 −6.800 1.060 18.230 14.600 14.500 13.740 23.770 14.400 15.000 12.900
98.000 98.000 98.190 100.000 100.000 100.250 101.000 102.450 105.300 105.380 107.300 108.000 108.500 109.000 110.420 113.300 115.400 147.600 117.000 118.000 118.100 118.460 118.800 119.000 119.000 119.060 119.070 119.150 119.500 119.500 119.510 119.700 119.930 120.000 120.000 120.000 120.400 120.500 120.600 120.700 120.800 120.800 120.860 120.900 120.900 120.940 120.980 121.000 121.000 121.000 (Continued)
284
P.K. Bhaskaran et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Table 24.3.
(Continued)
Source region
Latitude
Longitude
North part of Luzon Islands, Philippines Philippines, Luson Islands, Taal West Luzon Taiwan Islands, South China Sea Philippines: Mindanao, Mindor Philippines, Sulu Sea Manila Philippines, Mindoro Sulu Sea Philippines, Taal Philippines, South China-Sulu-Celebes Seas Luzon East China Sea Taiwan, Chilung Yellow Sea North Luzon Islands, Philippines: CAGA Indonesia, Flores Sea Philippines Taiwan Northeast of Taiwan Taiwan, east of Hwa-Lien East China Sea Taiwan Luson Islands, North Visayan, South Philippines Philippine Fault Indonesia, Flores Islands Philippines, West Luzon Islands Philippines: West Mindanao, Sulu Philippines: Mindanao: Zamb Indonesia, Celebes Sea Philippines Philippines: Batan Islands Taiwan Philippines, East Luzon Islands Sulu Sea East Taiwan-Ryukyu Islands Northwest of Taiwan East Luzon Islands, Philippines: Tayaba East Taiwan: Taitung, Hsinkong, TA. Philippines, Mindanao Islands Indonesia, Timor Sulu Sea Naga, East Samar Indonesia, Celebes Sea Philippines: South Negros: Duma Philippines, Sulu Sea Indonesia, Kupang, Timor Islands Philippines: Sorsogon, Masbat Philippines, Luzon Islands
18.600 14.000 14.200 24.000 13.700 13.500 18.000 13.520 13.400 13.600 12.780 15.680 25.300 25.100 31.500 18.000 −8.300 17.000 22.280 23.380 23.900 23.200 23.720 15.770 14.440 −8.480 12.520 6.000 6.500 3.000 10.500 20.720 24.380 16.380 7.100 24.370 24.280 14.000 22.550 8.250 −8.400 12.300 13.630 9.000 9.200 13.300 −10.100 12.400 11.000
121.000 121.000 121.000 121.000 121.000 121.000 121.000 121.070 121.100 121.100 121.130 121.170 121.400 121.400 121.500 121.500 121.500 121.500 121.510 121.520 121.570 121.600 121.630 121.660 121.680 121.900 121.990 122.000 122.000 122.000 12.000 122.010 122.080 122.080 122.100 122.100 122.180 122.300 122.320 122.400 123.080 123.100 123.170 123.200 123.200 123.400 123.500 123.500 123.500 (Continued)
Tsunami travel time atlas for the Indian Ocean Table 24.3.
(Continued)
Source region
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
285
Indonesia, Sulawesi Celebes Celebes Sea, Indonesia, Mindana Timor Sea Philippines, East Samar Islands Philippines, Moro Gulf, Mindanao Japan, SW of Ryukyu Islands Philippines, West Luzon Islands Indonesia, Timor Philippines, Mindanao Ishigaki Islands Ryukyu Archipel Philippines Indonesia, North Molucca Islands Indonesia, Manado, Sulawesi Philippines Philippines, Sulu-Celebes Seas Cotabato Cotabato West northwest of Dili, East Timor Indonesia, South Molucca Sea Indonesia, Belang, North Molucca Islands Indonesia, Minahasa Pen, Sulawesi Philippines, Takloban Philippines, Sulu Sea Philippines, Balere, Visayan Islands JAPAN, Miyakojima, Ryukyu Trench Indonesia, Timor Philippines, Luzon East of Taiwan Sulawesi – North Molucca Islands Japan, Miyakojima Sulawesis – North Molucca Islands Indonesia, North Molucca Islands South China-Sulu-Celebes Sea Indonesia, Banua Wuhu Philippines, Samar Indonesia, North Molucca Islands Sulawesi – North Molucca Islands Indonesia, Ceram Island, Maluku Off Samar, Philippine trench Indonesia, North Molucca Islands Indonesia, Sulawesi, North Molucca Islands Philippine Trench Indonesia Philippines Philippines Indonesia, North Molucca Islands Indonesia, North Molucca Islands Indonesia, North Molucca Islands
Latitude
Longitude
−1.110 6.300 5.770 −9.000 12.800 6.260 23.350 10.200 −8.250 6.030 24.000 13.750 −1.300 1.500 12.540 5.600 6.870 7.630 −8.170 −2.070 1.000 1.500 11.200 9.500 10.820 25.250 −8.380 12.630 22.310 2.300 25.000 3.500 0.000 9.000 3.100 12.060 1.000 3.700 −2.460 12.540 1.000 0.000 9.000 −2.400 6.790 9.040 1.420 2.000 1.000
123.570 123.600 123.640 124.000 124.000 124.020 124.090 124.100 124.150 124.250 124.300 124.360 124.500 124.700 124.740 124.800 124.840 124.850 124.860 124.890 125.000 125.000 125.000 125.000 125.010 125.060 125.130 125.300 125.310 125.400 125.500 125.500 125.500 125.500 125.500 125.580 125.600 125.600 125.960 125.990 126.000 126.000 126.000 126.000 126.080 126.180 126.260 126.500 126.500 (Continued)
286
P.K. Bhaskaran et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Table 24.3.
(Continued)
Source region
Latitude
Longitude
Molucca Passage Indonesia, Sulawesu – North Molucca Philippines, Mindanao Islands Indonesia Philippines, Davao Indonesia, Buru Islands, Banda Sea Montenegro Philippines, Mindanao, Hinat Philippines, Mindanao Philippines Indonesia, Makian Islands Indonesia, Ambon, Banda Sea Indonesia, Sulawesi – North Moluccas Philippines, East Mindanao Taiwan Ceram Philippines, Mindanao Philippines Indonesia, Ternate Islands, North Moluccas Indonesia, Ternate Islands Indonesia, Ternate Islands Indonesia, Ternate Islands, Molucca Sea Indonesia, Halmahera Japan, Okinawa Ceram Indonesia, Banda Sea Indonesia, Lomblen Islands Indonesia, Oma Indonesia, Halmahera Ceram Indonesia, Banda Sea Indonesia, Halmahera, Obi Islands Indonesia, Banda Sea Indonesia, Banda Sea Japan Japan, Okinava Islands, Ryukyu Indonesia, Ambon Japan, Torishima, Ryukyu Java Trench Banda Sea Indonesia, Banda Sea Ceram Indonesia, Banda Sea Indonesia, Banda Sea Indonesia, Ambon, Saparua, Banda Sea Saparua Islands, Banda Indonesia, Amahai, Banda Sea Banda Sea Indonesia
4.500 4.180 6.470 2.410 6.790 −3.500 2.850 7.950 7.190 7.280 0.000 −4.000 4.000 8.000 25.100 −3.000 9.500 8.260 0.800 0.780 0.780 −0.780 −1.320 26.200 −2.500 −3.000 −7.530 −3.700 1.010 −3.000 −4.060 −1.260 −3.400 −3.000 32.000 26.500 −3.700 27.800 −3.600 −3.800 −3.500 −1.400 −3.000 −3.500 −3.700 −3.600 −3.340 −4.000 −10.000
126.500 126.550 126.650 126.670 126.680 126.700 126.700 126.760 126.760 126.880 127.000 127.000 127.000 127.000 127.100 127.200 127.200 127.280 127.300 127.380 127.400 127.400 127.440 127.500 127.500 127.500 127.500 127.700 127.730 127.800 127.920 127.980 128.000 128.000 128.000 128.000 128.150 128.200 128.300 128.300 128.500 128.500 128.500 128.600 128.600 128.600 128.920 129.000 129.000 (Continued)
Tsunami travel time atlas for the Indian Ocean
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Table 24.3.
287
(Continued)
Source region
Latitude
Longitude
Indonesia, Banda Sea Japan, Amami, Oshima, Ryukyu Trench Indonesia, Maluku: Bandanai Java Trench Indonesia, Banda Sea Indonesia, Bandanaire, Banda Sea Indonesia, Banda Sea Indonesia, Moluccas Indonesia, Lantor Islands, Banda Sea Japan, Ryukyu Islands Japan, Ryukyu Islands Banda Sea Indonesia, Banda Sea Japan, Kagoshima Bay, Seikaido Japan, North Kagoshima Bay Irian Jaya Japan, Amami-Oshima, Ryukyu Trench Japan, Kyushyu Islands Japan, Kyushyu Islands Japan, Hiuganada Japan, Kyushyu Islands Banda Sea Japan Japan, Hiuganada, Seikaido Japan, Chichi Jima, Ryukyu Japan, Hiuganada, Seikaido; SHIK Indonesia, Kairana, Irian Jaya Indonesia, Irian Jaya Indonesia, Irian Jaya Japan, Nankaido West of Mariana Islands Indonesia, Northwest Irian Indonesia, Northwest Irian Yap Islands, Caroline Islands Japan, Baiones Nampo Islands New Guinea, Bismarck Sea Guam, Mariana Islands Mariana Islands
−6.900 28.300 −4.300 −6.930 −4.300 −4.500 −6.700 −5.000 −4.500 28.090 27.930 −5.000 −5.500 31.600 31.600 −5.580 28.000 30.570 31.790 31.900 31.890 −5.100 31.720 31.790 27.000 32.000 −4.000 −1.760 −4.430 32.000 20.900 −1.000 −3.500 9.500 31.550 28.250 −4.060 12.980 18.700
129.200 129.300 129.500 129.550 129.600 129.900 130.000 130.000 130.000 130.150 130.180 130.500 130.500 130.600 130.700 130.790 131.000 131.090 131.310 131.400 131.470 131.530 131.540 131.720 132.000 132.000 133.430 134.300 134.480 134.500 134.800 135.000 136.000 138.000 139.550 141.000 141.430 144.800 145.400
24.5 TSUNAMI TRAVEL TIME CHARTS Some examples of tsunami travel time charts for a few selected locations are shown in Figures 24.3–24.8.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
288
P.K. Bhaskaran et al.
Figure 24.3. Tsunami travel time chart for a location in the Rann of Kutch in India.
Figure 24.4. Tsunami travel time chart for a location off the coast of Sri Lanka.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunami travel time atlas for the Indian Ocean
Figure 24.5. Tsunami travel time chart for a location in the Bay of Bengal.
Figure 24.6. Tsunami travel time chart for a location in the Andaman Sea.
289
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
290
P.K. Bhaskaran et al.
Figure 24.7. Tsunami travel time chart for the city of Cota Raja, Banda Aceh in Indonesia.
Figure 24.8. Tsunami travel time chart for a location in the Indian Ocean.
Tsunami travel time atlas for the Indian Ocean 24.6
291
EPILOGUE
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
It should be noted that this atlas provides travel times for the first wave of the tsunami and nothing more. It does not provide information on the arrival times of subsequent waves, nor it provides information on how many waves will be in the tsunami event, which wave will be the highest, at what time each wave will arrive at a given location on the coastline, how far each wave will penetrate on the land and cause inundation, how strong the currents will be in each wave, exactly at what locations should people and cattle be evacuated, how long should they be evacuated, at what time it will be safe for them to return, etc. To obtain detailed information on all the above parameters, separate numerical models of tsunami generation and propagation, and of coastal inundation should be developed. REFERENCES Bhaskaran, P.K., Dube, S.K., Murty, T.S., Gangapadhyay, A., Chaudhuri, A., and Rao, A.D. (2005). Tsunami Travel Time Atlas for the Indian Ocean. Indian Institute of Technology Kharagpur, India, p. 279. Murty, T.S. (1977). Seismic sea waves tsunamis. Bulletin No. 198, Department of Fisheries and the Environment. Fisheries and Marine Service, Ottawa, Canada, 337pp. Murty, T.S., Saxena, N.K., Sloss, P.W., and Lockridge, P.A. (1987). Accuracy of tsunami travel time charts. Mar. Geod., 11, 89–102. Okal, E.A., Plafker, G., Synolakis, C.E., and Borrero, J.C. (2002). Near-field survey of the 1946 Aleutian Tsunami on Unimak and Sanak Islands. Bull. Seismol. Soc. Am., 93, 1226–1234. Yeh, H., Liu, P.L., Briggs, M., and Synolakis, C.E. (1994). Tsunami catastrophe in Babi Island. Nature, 372, 6503–6508.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 25
Overview and Integration of Part 3
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
N. Nirupama Atkinson School of Administrative Studies, York University, Toronto, Canada
25.1 TSUNAMI DETECTION AND WARNING Part 3 summarizes the techniques for tsunami detection and monitoring systems. In Chapter 19, Saraf et al. described the satellite detection of pre-earthquake thermal anomaly and the turbidity of seawater. Because of the accumulation of stress in the fault plane of the tectonic plate margins, the thermal regime near the ground surface can be affected. Monitoring of land surface temperature (LST) near earthquake zones can yield valuable clues about possible future tsunamis. Several recent earthquakes have indeed shown these pre-cursors in the form of increased LST. The authors have analyzed the NOAA–AVHRR data and deduced that the thermal anomalies started appearing about 5 days prior to 26 December 2004 and the LST anomaly reached a maximum of 4–6◦ C on 25 December 2004. The authors also analyzed the TERRA–MODIS data set and noted the seawater turbidity generated by the tsunami. In Chapter 20, Murty et al. reviewed some classical work on a theoretical framework for the detection of signals from earthquakes and tsunamis, in the ionosphere. The linkage between the troposphere and the ionosphere mainly occurs through the possible amplification of the acoustic and internal gravity waves in the atmosphere after the initial burst of interest on this topic during the 1950s and 1060s, not much work has been done on this topic till recently. The Indian Ocean Tsunami of 26 December 2004 renewed great interest in this topic. In Chapter 21, Gwal et al. discussed seismic precursors in the ionosphere as registered by the DEMETER satellite. The authors suggest that ionospheric disturbances associated with earthquakes can occur from several minutes to several days prior to the occurrence of the earthquake. The acronym DEMETER stands for “detection of electromagnetic emissions transmitted from earthquake regions”. This chapter provides a preliminary analysis of the data collected by the DEMETER satellite when it is flying over seismic regions. Analysis and interpretation of this data is very useful in the possible prediction of earthquakes and tsunamis. In Chapter 22, Joseph and Prabhudesai provide an interesting concept based upon Webenabled and real-time reporting and cellular-based instrumentation for coastal sea level and storm surge monitoring. They discussed the possible role of the gloss community of the IOC (InterGovernmental Oceanographic Commission). They also suggest that in addition to the traditional monitoring of the sea level, high-frequency variations of sea level associated with wave motion, such as storm surges and tsunamis also should be routinely monitored. In Chapter 23, Murty et al. discussed the various methodologies for tsunami detection. Some of these techniques are now routine, while certain others are in the research mode. It should be noted that, at present the detection methods are mainly for tsunamis generated by underocean earthquakes, and may be some limited capability for tsunamis from volcanic eruptions in the ocean, and practically no capability at present to detect tsunamis generated by submarine landslides. The problem is not so much with the detection of the tsunami, as more with detecting the landslide itself, unless the slide is associated with an earthquake. However, this may not be a serous issue, since a majority of the tsunamis are from under-ocean earthquake. 293
294
N. Nirupama
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
In Chapter 24, Bhaskaran et al. discussed the tsunami travel time atlas prepared by them. This is the first tsunami travel time atlas for the Indian Ocean and covers some 37 countries, which have a coastline on the Indian Ocean. The most basic information a tsunami warning center would require is information on tsunami travel times. This atlas provides travel times to some 250 coastal locations in the Indian Ocean, for tsunamis generated anywhere in the Indian Ocean. In addition to this, travel time charts were also prepared for historical tsunami events in the Indian Ocean, as well as some events in the western part of the Pacific Ocean that could have an impact in the Indian Ocean. These charts are expected to be accurate to ±1 min for each hour of tsunami travel. It is possible to compute travel times, independent of the size of the earthquake, because, to a first order, tsunami travel times depend mainly upon ocean depths.
Part 4
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Biophysical and Socio-Economic Dimensions of Tsunami Damage
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 26
Performance of Structures Affected by the 2004 Sumatra Tsunami in Thailand and Indonesia
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
M. Saatcioglu Department of Civil Engineering, University of Ottawa, Ottawa, Canada A. Ghobarah Department of Civil Engineering, McMaster University, Hamilton, Canada I. Nistor Department of Civil Engineering, University of Ottawa, Ottawa, Canada
26.1
INTRODUCTION
A team of structural and coastal engineers from Canada visited Thailand and Indonesia shortly after the December 26, 2004 Earthquake and Tsunami to investigate the performance of infrastructure during the disaster (Ghobarah et al., 2006). For the case of Thailand, the visit focused on urban areas and coastline stretches where engineered and non-engineered structures were affected. The observed engineered structures were mostly in the form of reinforced concrete frame buildings with masonry infill walls. A large number of them were hotel buildings since areas most affected by the tsunami were popular tourist resorts. Both concrete block and clay brick masonry were used as non-structural elements. Nonengineered structures were either in the form of low-rise reinforced concrete frames with masonry infill walls or timber frame buildings. These types of construction were used predominantly for shops, hotels and residential accommodation. In Indonesia, where both the effect of the seismic shocks and the tsunami waves were felt with high intensity, the entire infrastructure in coastal regions was devastated. Banda Aceh particularly was the major city that suffered seismic damage. In addition, a large segment of the city, including residential and business districts, where the common form of construction was non-engineered reinforced concrete, confined masonry and timber framed buildings, suffered widespread structural failures and total collapses caused by the tsunami wave forces. 26.2 26.2.1
EFFECTS OF THE TSUNAMI ON STRUCTURES IN THAILAND General observations
The Beaches of Rawai, Kata Noi, Kata, Patong, Nai Thon and Kamala were visited first on the Thai island of Phuket, followed by Phi Phi island, about 48 km south east of Phuket and the coastal town of Khao Lak about 100 km north of Phuket. Figure 26.1 illustrates the locations where site investigations were carried out. Damage along the southern coast of Phuket island was limited to coastal erosion and partial failures of non-engineered reinforced concrete and timber frame structures near the coast. Figure 26.2 297
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
298
M. Saatcioglu et al.
(a)
Figure 26.1.
(a)
Figure 26.2.
(b)
(a) Phuket, Thailand and (b) Khao Lak, Thailand, visited during reconnaissance investigations. (Figures reprinted from Mapsoftworld, 2005, Khao Lak Promotions.)
(b)
Coastal erosion in Rawai Beach on Phuket Island, Thailand.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Structures affected by the 2004 Sumatra tsunami
(a)
Figure 26.3.
(a)
Figure 26.4.
299
(b)
Damage to (a) roof tiles and (b) timber columns, Kata Beach on Phuket Island, Thailand.
(b)
Damage to first-story masonry walls, Patong Beach on Phuket Island, Thailand.
illustrates the coastal erosion observed in Rawai Beach. Similar soil erosion was observed along the coast of Kata Noi Beach, which caused minor foundation damage in the reinforced concrete frame building that was operated as Club Med. The degree of damage increased towards north. The water height in Kata Beach was measured from water marks to be approximately 6.0 m from the sea level, covering single-story buildings entirely and inflicting damage to low-rise buildings. Examples of damaged buildings are shown in Figure 26.3. The largest beach town on the west coast of Phuket Island was Patong Beach, which had a building inventory of non-engineered 1–2-story reinforced concrete and timber frame shops and hotels. There were also a number of multi-story engineered reinforced concrete hotels. Measured water marks on buildings varied between 4.0 m and 6.0 m from the sea level. Limited damage occurred in reinforced concrete structural elements, but significant damage occurred in timber structures. The entire shopping district of Patong Beach was heavily damaged within an area extending approximately 2 km inland from the shore as illustrated in Figure 26.4. Significant structural and non-structural damages were observed in Nai Thon Beach, further north on Phuket Island. The water height in the region was in excess of 10 m, especially in areas between the shore and the nearby hilly terrains, which led to water runups. A large number of buildings sustained heavy damage as illustrated in Figure 26.5. An area that was entirely devastated by the tsunami is Khao Lak Beach, about 100 km north of Phuket. The water height near the beach was measured to be in excess of 10 m, causing
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
300
M. Saatcioglu et al.
(a)
(b)
(c)
(d)
Figure 26.5.
(a)
Figure 26.6.
Damage to buildings in Nai Thon Beach on Phuket Island, Thailand.
(b)
Extensive damage to columns, masonry walls and roofs of low-rise buildings in Khao Lak Beach, Thailand.
extensive structural damage. The failure of first-story masonry infill walls and structural collapse of low-rise reinforced concrete frame buildings are shown in Figure 26.6. Many resort hotels in Khao Lak were completely destroyed. Some multi-story reinforced concrete hotels survived the tsunami pressure, with damage limited to the first-story infill walls. Further north of Khao Lak Beach is the harbor, which was also hard hit by the tsunami and wave
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Structures affected by the 2004 Sumatra tsunami
(a)
301
(b)
Figure 26.7.
(a) Damage to the harbor and (b) boat floated inland, just north of Khao Lak Beach, Thailand.
(b)
(a)
Figure 26.8.
Damage to buildings in Phi Phi Island, Thailand.
pressure destroyed the harbor and floated boats inland as shown in Figure 26.7. The town near the harbor was devastated. Phi Phi is a small island located about 48 km south east of Phuket. The topography of this island is such that east and west sides are entirely covered with steep hills with a low laying area between the two, where most of the island’s inhabitants live. This area is only a few meters above the sea level and was hit by the tsunami from both sides. Most structures on the island were destroyed, with the exception of a few well-built reinforced concrete frame buildings, a steel frame building and some non-engineered construction. Many of the 1- and 2-story nonengineered reinforced concrete and timber frame buildings of the island collapsed entirely due to tsunami wave pressures. Figure 26.8 illustrates the extent of damage on the island. 26.2.2
Hydrodynamic pressure of the tsunami wave
Tsunami waves imposed dynamic water pressures on coastal structures as well as buildings and bridges near the coastline, inducing serious damage to the entire surrounding infrastructure located approximately up to 4 km inland. The resulting impulsive pressures of breaking waves and hydrodynamic pressures associated with water velocity inflicted partial and full collapses of non-structural and structural components. The damage observed in Thailand was almost entirely due to water pressures that varied from impulsive pressures of breaking waves at the shore, to reduced dynamic pressures inland as water velocity decreased due to surface friction. Equation 26.1 gives an empirical expression for the estimation of impulsive water pressure of breaking
M. Saatcioglu et al. 35
35
30
30
25 20 15 10 5 0
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Wave height (m)
Wave height (m)
302
25 20 15 10 5
0
500 1000 2000 1500 Hydrodynamic pressure (kN/m2) (a)
0 0
100 300 400 500 200 Hydrodynamic pressure (kN/m2) (b)
Figure 26.9. Wave pressure as per: (a) equation (26.1) (Goda, 1995) and (b) equation (26.2) (Hiroi, 1919).
waves, developed by Goda (1995). Accordingly, wave impulse pressure is a function of specific gravity of water, water celerity, wave height, and impact duration. The variation of wave pressure obtained from equation (26.1) is plotted in Figure 26.9(a) as a function of water height. pmax =
πγcHw 4gτ
(26.1)
where γ is specific gravity of sea-water (10.3 kN/m3 ), c is celerity of wave in m/sec (estimated from recorded video to be 13 m/s), τ is impact duration (approximately equal to 0.2 s) and Hw is wave height in m. The above expression was developed to compute the effects of breaking waves on coastal structures at the shore. Therefore, it produces high values for building and bridge infrastructure located away from the shore. For such structures, the effect of the wave impact becomes very small. Furthermore, hydrodynamic pressure associated with wave celerity is reduced due to surface friction. Therefore, the use of equation (26.1) in investigating structural performance due to tsunami loading within some distance from the shore may be extremely conservative. A simpler expression that is more applicable to the mechanism of tsunami wave impact was suggested by Hiroi (1919) for the estimation of uniform dynamic pressure as a function of sea-water specific gravity and wave height. The expression is given in equation (26.2). The variation of uniform lateral pressure as computed from equation (26.2) is plotted in Figure 26.9(b). p = 1.5 γHw
(26.2)
The hydrodynamic tsunami wave pressure obtained from equation (26.2) was compared with wind design pressure specified in the National Building Code of Canada (NBCC-95) on a building located at the coastal city of Vancouver, B.C., Canada. It was found that the tsunami pressure at the first floor level could be approximately 26 times the design wind pressure in Vancouver, explaining the widespread damage to masonry walls observed within the first stories of most buildings. Design lateral forces due to wind and seismic effects were compared with the estimates of lateral forces generated by a tsunami of 5 m water height. A sample 6-story, 3-bay reinforced concrete frame structure in Vancouver Canada was used for this purpose. The comparisons of lateral forces are shown in Figure 26.10. It was found that the total base shear due to tsunami was about twice the base shear due to wind and 60% of that caused by elastic seismic forces and 2.5 times the inelastic seismic design base shear for a ductile moment resisting frame building. While this comparison is presented to provide a feel for the magnitude of tsunami forces on buildings in Thailand, it is important to recognize that the same comparison could give vastly different results
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Structures affected by the 2004 Sumatra tsunami E.Q.
Wind
120 kN
18 kN
100 kN
36 kN
81 kN
36 kN
64 kN
36 kN
45 kN
34 kN
26 kN
39kN
303
Tsunami
1090 kN
VEQ = 436 kN
Vw = 199 kN
VTS = 1090 kN
(VEQ)Elastic= 1744 kN
Figure 26.10.
Comparisons of lateral forces due to earthquake, wind, and tsunami for an interior frame of a 6-story reinforced concrete building in Vancouver, Canada for 5.0 m tsunami water height, 6.0 m transverse span length and seismic force reduction factor of R = 4.0.
for buildings with different heights, different exposure conditions and different structural mass, especially for different tsunami water heights. Another important aspect of tsunami pressures on structures is the nature of the exposed area. Because of the relatively high level of tsunami pressures generated, non-structural elements, like infill masonry walls, may collapse immediately after the application of tsunami wave pressure, and reduce the portion of the load transmitted to the structural elements. The non-structural elements, though suffer extensive damage under tsunami pressures, tend to act like a fuse for the load carrying system, reducing the amount of lateral loads transferred to the structure. Consequently, a more detailed investigation of the mechanism of tsunami load transfer among exposed elements should be conducted before the tsunami loads can be assessed accurately. 26.2.3
Performance of timber construction
Coastal towns in southern Thailand have a large number of low-rise timber frame buildings. These buildings have timber columns and beams, supporting timber joist floor systems. The roofs either have light corrugated metal coverage or clay roofing tiles. These buildings suffered damage to their columns, which led to the failure of joist floor systems. Also damaged were roofs, especially when the water height was high enough to reach the roof level, sweeping roofing tiles and exposing timber roof trusses. Figure 26.11 illustrates typical framing systems used and the damage observed in low-rise timber frame buildings. 26.2.4
Performance of unreinforced masonry walls
The majority of buildings in Thailand had either frames infilled with unreinforced masonry walls or confined masonry with lightly reinforced concrete beams and columns. The masonry units used were consistently of the same type, with 50 mm thickness. Both hallow clay bricks and concrete masonry blocks of the same thickness were used as illustrated in Figure 26.12.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
304
M. Saatcioglu et al.
(a)
Figure 26.11.
(b)
Damage to timber frame structures in Thailand.
(a)
(b)
(c)
(d)
Figure 26.12.
(a) Damaged wall exposing masonry units in Patong Beach; (b) typical 50 mm thick concrete block masonry units; (c) and (d) construction of confined masonry in Khao Lak harbor town.
This figure also illustrates a typical confined masonry wall during construction between small size non-engineered reinforced concrete columns to be cast. The same types of walls are quite common between concrete and timber framing elements as non-load-bearing infill walls. These walls suffered extensive out of plane failures when subjected to tsunami pressures perpendicular to the wall plane. The water pressure resulted in large holes in lower story walls, sometimes removing
305
(b)
(a)
Figure 26.13. Typical punching failure of masonry walls: (a) Kamala Beach and (b) Phi Phi Island. V = 52 kN M
(a)
1200 1000 L = 3.0 m 800 600 400 200 M V = 52 kN 0 Axial Force (kN)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Structures affected by the 2004 Sumatra tsunami
200 mm ρ = 0.5%
0
10
(b) M = 52 x 1.5 = 78 kN.m
Figure 26.14.
20
30
40
Moment (kN.m)
(a) Column failures in a 2-story reinforced concrete frame units in Khao Lak Beach and (b) Nominal moment-axial force interaction diagram for a 200 mm square column with 0.5% reinforcement and approximate moments imposed due to assumed tsunami pressure.
the masonry almost entirely. The remaining walls around the frames did not show any sign of diagonal cracking. Figure 26.13 shows typical punching failures observed in masonry infill walls. 26.2.5
Performance of non-engineered reinforced concrete buildings
The majority of 1–2-story low-rise buildings were constructed using site-cast concrete without much evidence of engineering design. The columns were of very small cross-section (about 200 mm square), containing 4–8 mm diameter smooth or deformed corner bars, resulting in approximately 0.5% reinforcement ratio. Their flexural capacity was computed to be significantly below the moments imposed by tsunami waves, accounting for the majority of failures. Figure 26.14 illustrates a column failure in a non-engineered reinforced concrete frame in Khao Lak Beach and the column interaction diagram plotted for the column, on the basis of measured geometry and estimated material strengths. The story shear was calculated based on a 4-column frame resisting the tsunami pressure calculated with equation (26.2), acting over 3 m square panels on both sides of the frame without considering the secondary moments due to P– effects. It was assumed that the entire tsunami load acting on the panels was fully transferred to the frame. This may be a conservative assumption, as panels may sustain immediate partial failure before the full load is transferred to the columns. However, the interaction diagram indicates that the columns could not even sustain half the moments imposed, explaining the extensive damage observed in Khao Lak Beach for similar
M. Saatcioglu et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
306
(a)
(b)
(c)
(d)
Figure 26.15.
(a) and (b) Column failures in Khao Lak Beach; (c) and (d) frame damage on Phi Phi Island.
types of buildings where the first-story columns were incapable of maintaining overall strength and stability. Column failures were blamed for the collapse of entire structural systems in non-engineered low-rise reinforced concrete buildings when the tsunami water height was 4.0–6.0 m and the columns received significant lateral pressure due to the attached panels. Figure 26.15 shows partial and full structural collapses in Phuket and Phi Phi Island. However, many similar columns survived the tsunami without much damage, especially those that were away from the shore and those that had additional lateral bracings provided by in-plane infill walls. Figure 26.16 illustrates non-engineered frame buildings that remained intact during the tsunami, some with damaged non-structural walls. 26.2.6
Performance of engineered reinforced concrete buildings
There were many low- to mid-rise reinforced concrete frame buildings which appeared to have been engineered in the visited areas of Thailand. These frame buildings survived the tsunami pressure without structural damage, though they suffered damage to non-structural elements,
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Structures affected by the 2004 Sumatra tsunami
(a)
Figure 26.16.
(b)
Non-engineered concrete frames that survived the tsunami in Khao Lak Beach.
(a)
(b)
(c)
(d)
Figure 26.17.
307
Engineered reinforced concrete frame buildings: (a) and (b) on Phi Phi Island; (c) in Nai Thon Beach; (d) in Khao Lak Beach, that survived the tsunami without structural damage.
especially the first-story masonry infill walls. Figure 26.17 shows examples of reinforced concrete hotel buildings in Thailand that survived the tsunami loading without any sign of structural damage, although nearby non-engineered buildings were either partially or fully collapsed. Although engineered frame buildings performed well, there were some exceptions to this observation in NaiThon Beach, where water run-up effects due to the topography of the area between the ocean and the hilly terrain behind, resulted in increased water height. A large number of structural
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
308
M. Saatcioglu et al.
(a)
(b)
(c)
(d)
(e)
(f)
Figure 26.18.
Damage to reinforced concrete frame buildings in Nai Thon Beach.
failures were observed in this region. Column failures due to insufficient flexural and shear capacities, as well as inadequate detailing of reinforcement to protect structures beyond the elastic range of deformations resulted in partial or full collapses of complete frames. Examples of structural damage in Nai Thon Beach are illustrated in Figure 26.18, which depicts poor column behavior. Figure 26.18(a) and (b) show rear views of Khao Lak shopping center that suffered serious structural damage. The building was three stories high at the back, and a single-story high facing the street at the front because of the grade. The tsunami waves hit the building from the ocean side at the back, covering the entire building with water up to the roof level as evidenced by the
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Structures affected by the 2004 Sumatra tsunami
(a)
(b)
(c)
(d)
Figure 26.19.
309
Damage to reinforced concrete frame building under construction in Nai Thon Beach.
missing roofing tiles. The structural system consisted of reinforced concrete frames in the first 2 stories with slender circular columns, supporting a timber framing system at the third-story level. The slender columns became unstable at the second-story level, probably due to high secondary moments and failed. In some frames the continuity between the timber and concrete framing elements was poor and led to failures as illustrated in Figure 26.18(b). A reinforced concrete hotel complex was under construction in Nai Thon Beach when it was hit by the tsunami. The frame elements, first-story stone masonry walls and second floor reinforced concrete slab were mostly completed at the time. Figure 26.19 shows various views of the structure, exposing different types of structural failures. Figure 26.19(a) shows the failure of the first-story masonry walls and the second floor shoring for the corner slab, exposing slab reinforcement in Figure 26.19(b). Figures 26.19(c) and (d) illustrate the failure of a corner column, which resulted in the failure of the two-way slab system immediately above. It is clear that the tsunami pressure in the region was high enough to induce severe structural damage, all being triggered by the failure of weak columns, supporting deep and strong beams and attached two-way slab systems. A common precast slab system that is used in Thailand consists of prefabricated reinforced concrete strips, supported by cast-in-place beams. These strips typically have 50 mm thickness, 300 mm width and 2.0 m length, reinforced with 4–6 mm diameter smooth wires, equally spaced in the center of the section (Figure 26.20). Because of lack of proper connection to the supporting beams, these strips lifted up due to water pressure, causing slab failures. One good example was a shopping center in Patong Beach on Phuket Island where the below grade garage was filled up with water, lifting and destroying the first-story slab panels as illustrated in Figures 26.20(a), (b) and (c). Other examples of damaged precast slab systems were observed in a number of
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
310
M. Saatcioglu et al.
(a)
(b)
(c)
(d)
Figure 26.20.
(a)
Figure 26.21.
(a) and (b) Damage to precast slab strips in Phaton Beach shopping center; (c) Close-up view of a precast slab strip; (d) lifting of precast concrete slab due to water pressure in Nai Thon Beach.
(b)
Damage to precast slab strips of the concrete dock in Khao Lak.
different buildings in Thailand. Figure 26.20(d) shows a building in Nai Thon Beach that suffered dislocation of the precast slab system. A similar type of slab failure was also observed in the concrete dock of the Khao Lak Harbor. The slab strips used in the harbor dock was thicker. The failure was in the form of lifting of individual strips due to water pressure, or lifting of the entire slab panel, as shown in Figure 26.21, when the strips had cast-in-place topping.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Structures affected by the 2004 Sumatra tsunami
(a)
(b)
Figure 26.22.
26.2.7
311
Soil erosion and related foundation problems: (a) buildings in Kamala Beach and (b) a building in Nai Thon Beach.
Soil erosion and shallow foundations
Many buildings were observed to have shallow foundations, in the form of isolated footings. Where the soil erosion was significant, some of these buildings suffered from soil settlement. Figure 26.22 illustrates examples of foundation problems associated with soil erosion near the coast. 26.3 26.3.1
EFFECTS OF THE TSUNAMI ON STRUCTURES IN INDONESIA General observations
The impact of tsunami is a function of the topography of coastal area. In flat areas of coastal Sumatra, including the city of Banda Aceh, the tsunami waves reached 4–5 km inland, which affected a large population. In these areas the maximum wave height was 4–6 m. However, the water run-up in hilly terrains resulted in significantly higher water levels. A coastal engineering team from Japan measured the maximum tsunami run-up height in Rhiting, south west of Banda Aceh, to be 48.8 m (Shibayama, 2005). The city of Banda Aceh, with a population of about 300,000 before the tsunami, has a flat topography with a number of rivers and creeks passing through the city. The coastal areas of the city were entirely swept away by tsunami waves, leaving behind piles of timber as the remains of the building infrastructure. Residential wood houses appeared to survive the earthquake but were completely disintegrated when hit by the tsunami. The residential neighborhoods in coastal areas were entirely devastated and the houses were broken down to small pieces of debris. Figure 26.23 illustrates these areas and the remains of residential buildings. A common form of low- and mid-rise construction in Banda Aceh is non-engineered lightly reinforced concrete buildings. These buildings suffered significant structural damage that led to partial or full collapses. The business district of the city, where this type of construction was very common, was entirely non-functional because of widespread structural failures of these substandard buildings that appeared to have been first damaged by seismic forces and then swept away by strong hydrodynamic forces of the tsunami. Figure 26.24 shows the extent of damage to non-engineered reinforced concrete buildings in Banda Aceh. 26.3.2 Tsunami loading Two types of tsunami related loads were applied on structures in Banda Aceh: (i) tsunami water pressure and (ii) the impact of floating debris. Tsunami waves imposed water pressures due to the impulse of breaking waves along the shore, and dynamic pressures that varied with water celerity
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
312
M. Saatcioglu et al.
(a)
(b)
(c)
(d)
Figure 26.23.
(a)
Destruction of residential timber framed construction in Banda Aceh.
(b)
Figure 26.24. Tsunami damage to non-engineered reinforced concrete buildings in central Banda Aceh.
and height. The impulse component diminished as water moved inland. The hydrodynamic pressure also decreased due to reduced water velocity caused by surface friction. It became evident during the field investigation that some of the damage was caused by floating debris. The flat topography of Banda Aceh and sufficient water height and pressure allowed not only the remains of collapsed building components to float but also large and heavy objects; such as cars, trucks and fishing vessels to float and impact on the physical infrastructure. It was clear that many small size building columns were damaged significantly by impact forces caused by
Structures affected by the 2004 Sumatra tsunami
(b)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
(a)
F
(c)
Figure 26.25.
(d)
Impact loading on columns due to floating debris, Banda Aceh.
(a)
(b)
(d)
(c)
Figure 26.26.
Impact of fishing boats on buildings, Banda Aceh.
313
M. Saatcioglu et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
314
Figure 26.27.
Power generating vessel that floated 3.5 km inland in Banda Aceh.
(a)
(b)
Figure 26.28. Timber framed structures in Banda Aceh.
floating debris, resulting in the collapse of slabs supported by such critical elements. Figures 26.25 and 26.26 illustrate the impact loading due to debris. A unique aspect of Banda Aceh, from impact loading perspective, was the presence of a large number of fishing boats which were spread around various harbors, as well as rivers that pass through the city. These boats floated during the tsunami, impacting on buildings as illustrated in Figure 26.26. A large floating power generator, the same size and appearance of a large size oil tanker docked at the harbor was freed up by tsunami forces and floated 3.5 km inland, destroying residential buildings on its way. This is shown in Figure 26.27 in its current location, which happens to be on top of several residential buildings that were completely crushed. 26.3.3
Performance of timber construction
Low-rise timber construction, shown in Figure 26.28, is typically used for low cost and affordable housing in Banda Aceh. The roofs either have light corrugated metal coverage or clay tiles. These houses appeared to have survived the earthquake with minor damage but many collapsed completely under tsunami wave pressures. The building components disintegrated into smaller pieces, contributing to debris that impacted on other structures. Figure 26.23 illustrates the
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Structures affected by the 2004 Sumatra tsunami
Figure 26.29.
315
Punching failure of masonry infill walls under tsunami pressure, Banda Aceh.
devastation due to massive failures of timber buildings in residential areas. Some of the very same timber buildings did survive the tsunami, especially if they were away from the coast, protected by surrounding buildings and laterally braced by masonry walls. 26.3.4
Performance of unreinforced masonry walls
The majority of buildings in Indonesia had frames infilled with unreinforced masonry walls of either 50 mm clay brick units or the same thickness concrete block units. These walls sustained punching failures under out of plane tsunami pressures. First-story walls in most buildings experienced failures in the form of large circular openings. In some buildings, where the tsunami water height was in excess of 4 m, second-story walls were also damaged. Figure 26.29 shows damaged masonry walls in reinforced concrete frames. 26.3.5
Performance of non-engineered reinforced concrete buildings
The majority of low-rise buildings in Banda Aceh were constructed using site-cast concrete without much evidence of engineering design. Low-rise buildings are often in the form of confined masonry with small-size concrete structural elements. The reinforced concrete elements are lightly reinforced, with small cross-sectional dimensions. The columns are often no larger in cross-section than 200 mm square, containing 4–8 mm diameter smooth or deformed corner bars, resulting in approximately 0.5% reinforcement ratio. The beams are often larger in size, resulting in frames with strong beams and weak columns. Figure 26.30 shows typical non-engineered concrete construction which is commonly used in Banda Aceh. Some of these framed buildings were damaged by the earthquake and approximately 15 min later completely or partially destroyed by the tsunami, as reported by survivors. Column failures were often responsible for the overall structural collapse. In spite of the widespread damage near the shore, similar buildings were able to survive the disaster, especially if they were away from the coast and had additional lateral bracings provided by in-plane infill walls. Figure 26.31 illustrates non-engineered framed buildings that remained intact during the tsunami, some with damaged non-structural infill walls. 26.3.6
Performance of engineered reinforced concrete buildings
The engineered and well-constructed reinforced concrete framed structures survived the tsunami with minor damage. There were many low- to mid-rise reinforced concrete framed buildings which appeared to have been engineered in the central part of BandaAceh. Some of these buildings were government buildings and were damaged by the earthquake. Many survived the tsunami with
M. Saatcioglu et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
316
(a)
(b)
Figure 26.30.
(c)
Damage to non-engineered reinforced concrete framed buildings, Banda Aceh.
limited non-structural damage at the first floor level. Figure 26.32 shows examples of reinforced concrete framed buildings, braced by masonry infill walls that survived tsunami pressures. 26.3.7
Structural failures resulting from lack of anchorage
The tsunami pressure and wave height were sufficient to displace structures and structural components from their foundations and float them hundreds of meters away. Figure 26.33 shows a single-story reinforced concrete framed house with masonry walls that was displaced and floated away from its foundation, resting on the nearby street. The lack of anchorage was attributed to poor and shallow foundation on soft soil. Fuel supplies in Banda Aceh were disrupted due to damage to oil storage tanks. The fuel storage tank depot suffered extensive damage from the tsunami. The retreating water displaced three large tanks a distance of approximately half a kilometer as shown in Figure 26.34. The tanks were torn from their connecting pipes, lost their fuel content and impacted and damaged many houses as seen in Figure 26.34. Lack of anchor bolts, which is the common practice for these tanks, was blamed for the failure. 26.3.8
Performance of bridges
Damage to bridges in Aceh province was extensive. This severely constrained rescue and relief efforts. Tsunami forces caused the collapse of a number of bridges that formed vital links to
Structures affected by the 2004 Sumatra tsunami
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
(a)
(b)
(d)
(c)
Figure 26.31.
(a)
Figure 26.32.
317
Non-engineered reinforced concrete framed buildings that survived the tsunami.
(b)
Engineered reinforced concrete framed buildings that survived tsunami in downtown Banda Aceh, away from the coastal region.
population centers in the area of Banda Aceh. Temporary Bailey bridges were constructed at several locations. Many of the bridges on the coastal road to Meulaboh were washed away and sections of the road disappeared, which isolated many small communities where survivors could only be reached by helicopters.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
318
M. Saatcioglu et al.
Figure 26.33. A single-story house, displaced by water pressure due to lack of proper anchorage, Banda Aceh.
(a)
Figure 26.34.
(b)
Displaced fuel tanks due to tsunami water pressure that were swept by half a kilometer, also destroying houses in their way, Banda Aceh.
A 2-span steel truss bridge on the west coast of Banda Aceh was displaced approximately 50 m by the tsunami as shown in Figure 26.35. Since the bridge was important for the road to a cement plant, industrial facilities and to communities to the west, a temporary Bailey bridge was constructed using the undamaged pier and abutments of the displaced bridge. Similarly, a single-span reinforced concrete bridge, with precast girders, was swept away by the tsunami pressure near downtown Banda Aceh, and was temporarily replaced by a Bailey bridge, as shown in Figures 26.36(a) and (b). In eastern Banda Aceh a 2-lane, multi-span reinforced concrete bridge was swept off its piers, as illustrated in Figure 26.37(a). Two of the bridge piers were also destroyed while the others remained in place. Another multi-span, reinforced concrete bridge over the same river, further away from the ocean, survived the tsunami wave pressure as shown in Figure 26.37(b). This bridge is likely to be of the same type as that collapsed (shown in Figure 26.37(a)), judging by the spans and the piers, though this point could not be confirmed. A simply supported 3-span concrete bridge that led to the Banda Aceh harbor in an area that was totally devastated by the tsunami survived the disaster with dislocations of the girders
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Structures affected by the 2004 Sumatra tsunami
(a)
319
(b)
Figure 26.35. The failure of a steel truss bridge in eastern Banda Aceh.
(a)
Figure 26.36.
(a)
Figure 26.37.
(b)
(a) and (b) The failure of a precast concrete bridge near downtown Banda Aceh and replacement by a temporary bridge.
(b)
Multi-span reinforced concrete bridges in eastern Banda Aceh, crossing the same river; (a) a bridge close to the ocean that was completely swept off by the tsunami and (b) a bridge approximately 3 km away from the ocean that survived the tsunami.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
320
M. Saatcioglu et al.
(a)
(b)
Figure 26.38.
(a) A 3-span reinforced concrete bridge in Banda Aceh that was displaced on its abutments due to tsunami and/or seismic ground shakings and (b) close-up view of the separation of girders at a pier.
from the abutments, as well as the separation of girders at the piers. Figure 26.38(a) shows the bridge that has precast “I” girders supporting a reinforced concrete deck. Figure 26.38(b) depicts approximately 150 mm of separation of the girders at a pier. 26.4
CONCLUSIONS
The following conclusions can be drawn from the reconnaissance conducted in Thailand and Indonesia to assess the engineering significance of the December 26, 2004 tsunami. • Lateral forces generated by tsunami wave pressure can be significantly higher than typical design wind pressures, generating out of plane forces high enough to damage unreinforced masonry walls within the wave height. The observations indicated widespread failure of masonry infill walls within the first-story level of most framed buildings. These failures were often in the form of large circular openings. • In addition to tsunami wave pressures that can induce base shears higher than those generated by wind pressures, approaching elastic seismic base shears of multi-story buildings, significant impact forces can be imposed on structures due to floating debris. • Low- and medium-rise timber construction is vulnerable to tsunami forces. These buildings experienced partial or total collapse under tsunami induced water pressures, especially if they were located near the shore, where they were subjected to higher water pressures. The failure of these frames was triggered by column damage and lack of sufficient connection between the framing elements. • Non-engineered low-rise reinforced concrete frame and confined masonry buildings, as well as timber houses are vulnerable to lateral tsunami pressures. Columns of such buildings are further vulnerable to impact forces generated by floating debris, resulting in flexural failures of columns within their mid-heights. Many residential districts with timber residential buildings in Banda Aceh were entirely damaged by tsunami waves. • Prefabricated reinforced concrete slab strips, commonly used in Thailand, suffered from uplift forces caused by water pressure. Lack of proper anchorage to the supporting beams and lack of continuity was blamed for the failure of these slab systems. • Engineered reinforced concrete frames appear to have sufficient strength against tsunami forces. There was very little damage observed in structural components of well designed
Structures affected by the 2004 Sumatra tsunami
321
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
concrete buildings. Often, non-structural elements failed before the effects of tsunami pressure reached a critical level for structural components of such buildings, relieving pressure on structural elements. However, in regions where the water height was high due to run-up effect and when the columns supporting the buildings did not have sufficient strength and connection details, engineered reinforced concrete frames suffered serious structural damage, resulting in partial and total collapses. • Bridge infrastructure was devastated by tsunami forces. Many bridges were swept away from their supports, disabling the transportation network. • Storage tanks and other light structures should be well anchored to their foundations to resist tsunami pressures. Many steel storage tanks, as well as other unanchored structures floated away long distances due to uplift pressures generated by the tsunami. REFERENCES Geocities (2005). web page: www.geocities.com Goda Y. (1995). Random Seas and Design of Maritime Structures. University of Tokyo Press, Tokyo, Japan, 323 pp. Hiroi, I. (1919). A formula for evaluating breaking wave pressure intensity in the case of breaking waves, J. Coll. Eng., Tokyo Imperial University, Tokyo, pp. 11–21. Mapsoftworld. http://www.mapsoftworld.com (Accessed in March, 2005) Ghobarah, A., Saatcioglu, M. and Nistor, I. (2006). How to design structures for tsunami. J. Eng. Struct., Elsevier, 28, 312–326. Shibayama, T. (2005). The December 26, 2004 Sumatra Earthquake Tsunami, Tsunami Field Survey in Banda Aceh of Indonesia. Scientific Report, http://www.drs.dpri.kyoto-u.ac.jp/sumatra/indonesiaynu/indonesia_survey_ynu_e.html
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 27
Field Observations on the Tsunami Impact Along the Kerala Coast, Southwest India
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
N.P. Kurian, T.N. Prakash and M. Baba Centre for Earth Science Studies, Trivandrum, India
27.1
INTRODUCTION
The December 2004 tsunami generated by the M9 Sumatra–Andaman earthquake devastated many parts of the Kerala coast too (Kurian et al., 2005; Narayana et al., 2005). The Kerala coast (Figure 27.1) is located in the shadow zone with respect to the direction of propagation of the tsunami, and in that sense its severity was rather unexpected. Nearly 200 people were killed and hundreds injured in addition to the loss of houses and properties worth several crores of rupees (INR one crore = ∼USD 220,000). Although there are reports of some previous tsunamis (1881, 1833, 1941, to mention a few) in the past, generated by earthquakes in the Andaman– Sumatra region, there is no documented evidence of any such events affecting the Kerala coast. An earthquake of 1945 M8.0 in the Mekran coast is believed to have generated significant tsunami run-up in some parts of Gujarat (Berninghausen, 1966; Wadia, 1981), and it is the only documented report of any tsunami affecting the west coast. To the best of our knowledge, the 2004 tsunami is the first of its kind to have affected the Kerala coast. In this paper, we document the results of the post-tsunami surveys (Kurian et al., 2005, 2006; Prakash et al., 2005) along the Kerala coast and evaluate geomorphological changes in the inner shelf, shores and backwaters of the Kayamkulam inlet region, where the impact of tsunami was most severe.
27.1.1 About the Kerala coast The 560 km long Kerala coast (Figure 27.1) is by and large of submergent nature. Lateritic cliffs, rocky promontories, offshore stalks, long beaches, estuaries, lagoons, spits and bars are characteristic of this coast (Soman, 2002). The sand ridges, extensive lagoons and barrier islands are indicative of a dynamic coast with transgression and regression in the geological past. The coastline of Kerala is more or less straight trending in north northwest–south southwest direction from north till the Thangassery headland near Quilon. The coastline orientation south of Thangassery is in the northwest–southeast direction. The offshore continental shelf bathymetry is steeper to the south. While the 100 m-contour is at a distance of around 40 km off Trivandrum from the shore, it is 58 km off Kasargod, south of Mangalore. The variation in the slope of the inner shelf is more pronounced towards south. While the inner shelf is very steep with the 50 mcontour at a distance of only 11 km off Trivandrum, it is very gentle off Calicut with this depth occurring at a distance of 42 km. This change in the bottom slope has significant implications in the hydrodynamic and sedimentological characteristics of the inner shelf of Kerala. The nearshore wave intensity decreases from south to north (Baba and Kurian, 1988). While the maximum wave height recorded is 6 m at Valaithura near Trivandrum it is only 2.6 m at Tellichery. Incidentally Trivandrum coast has 323
324
N.P. Kurian et al. 72˚0'0"E
74˚0'0"E
76˚0'0"E 10
100 0
20
1000
9
1000
0
LAKSHADWEEPISLANDS
0
0
10 00
200 20 00
10
1000
00
0
100
100 km 12˚0'0"N
CALICUT 10
00 10 00
10
274
00
10
00
20
00
100
0
1000
00
50
10
10˚0'0"N
COCHIN
10
10˚0'0"N
100
00
2000
20
0"
1000
QUILON TRIVANDRUM
20
0
100
00
0
1000
20
1000
8˚0'0"N
8˚0'0"N
2000
72˚0'0"E
Figure 27.1.
2000
2000
3000
MALDIVE ISLANDS
00 20
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
50
Isobaths in meters Shallow waters < 50 m depth
0 10
00
25
30
12˚0'0"N
10
.
MANGALORE
00
20
74˚0'0"E
76˚0'0"E
Location map.
the highest wave intensity along the Indian (excluding Lakshadweep and Andaman and Nicobar islands) coast (Shahul Hameed et al., 1994). The coastal currents are generally known to be southerly during monsoon and northerly during the fair weather period. Deviations from this are observed and are attributed to the formation of an anticyclonic eddy off the southwest coast during the northeast monsoon and cyclonic eddy during the (southwest) monsoon (Kurian et al., 2002). The coast is in general micro-tidal with the tidal range decreasing from north to south. While the parts of the coast north of Calicut have a tidal range of around 1.5 m, the tidal range is around 0.50 m in the south along the Trivandrum coast. The surficial sediments of the continental shelf and slope of Kerala can be divided into three types – terrigenous, biogenous and chemogenous. In the shelf and slope of Kerala, terrigenous sediments mostly occur as sands in the nearshore (up to 10–12 m water depth) followed by a zone of silty clays on the inner shelf (Rao et al., 1997). An admixture of abundant terrigenous and biogenic constituents carpets the outer shelf. The continental slope sediments are clayey silts with abundant carbonate tests. The surface sediment distribution in the inner continental shelf also shows considerable diversity with respect to the bottom slope conditions (Rao et al., 1997). The mud banks of Kerala are unique transient nearshore features appearing during monsoon. Purakkad, Alleppey, Vypin and Koilandy are locations well known for the occurrence of
Field observations on the tsunami impact: Kerala coast
325
mudbanks. The coast is also characterized by the presence of rich heavy mineral deposits at Chavara, which is being mined by the Indian Rare Earths Ltd. The high density of population along this coast brings in all the adverse consequences arising out of it. Karumkulam, south of Trivandrum, has a density of population of 15,000/km2 . Coastal erosion, though limited to certain pockets is a very sensitive issue. Settlements, tourism and mining are adversely affecting the barrier islands of this coast (SoE Report Kerala, 2005).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
27.2
IMPORTANCE OF POST-TSUNAMI FIELD SURVEYS
A significant observation associated with the 2004 tsunami effects along the Kerala coast is its localized amplification in some regions and total attenuation elsewhere. Understanding the spatial pattern of the tsunami and its effects on the coastal morphology has important implications for assessing future scenarios of inundation. Since most of the evidences left by tsunamis are temporary, it is important to carry out post-tsunami surveys to assess the run-up heights, inundation limits, arrival time of waves and also to assess the impact on the coastal life and property, flora and fauna, geomorphology, etc. Such information is important for future hazard projections, and to develop inundation models. Tsunami mitigation strategies have to be formulated based on such database (Preuss et al., 2001). Thus field visits were conducted at different locations of Kerala coast starting from 27 December 2004. Altogether 83 locations spread all over the Kerala coast, reported to have been affected by the tsunami, were visited. The Post-Tsunami Survey Field Guide (Web site, ITIC) was taken as a guide in the field trip. Considering the geomorphic set up of the coast, the run-up was estimated as the elevation at the local maximum of the horizontal inundation measured relative to mean water level at each location. For estimation of run-up level, field signatures such as trapped flotsam in plants/trees/buildings, flood mark or damaged windows and doors of buildings, etc. were relied upon. In addition, local people were also interviewed to collect eyewitness reports. Beach profile measurements were carried out from the established stations in the worst affected Kayamkulam inlet region. Bathymetric survey was carried out in the inner shelf upto 50 m water depth, for a coastal stretch of about 50 km length extending over both the sides of the Kayamkulam inlet. Surficial sediment samples from the same region were collected for study of changes in sediment characteristics. Siltation of backwaters of the same area was also studied by depth sounding using an echosounder. Impact of the tsunami on flora and fauna was not covered in this survey.
27.3
RUN-UP LEVEL
The run-up level shows wide variations along the coast (Figure 27.2). In the Pozhiyur to Vizhinjam (south of Trivandrum) sector, the run-up level was only up to 1.5 m, whereas in the Vizhinjam– Varkala sector it was 2–2.5 m. In the Thangasseri harbour area of the Quilon coast, the run-up was about 2.5 m, whereas in the segment to its north, it was up to 3 m. The run-up level increased further north, reaching about 3.5 m. In the Cheriyazhikkal area south of Kayamkulam, run-up up to 4.5 m is reported. In Azhikkal and up to Kayamkulam inlet, the severity of attack of tsunami was further intensified with a run-up of up to 5 m. In the sector immediately to the north of Kayamkulam inlet also the tsunami onslaught was severe with run-up level up to 5.0 m. Further north, in the Arattupuzha region and up to Thottappally, the run-up level got reduced to 3.5 m. From Thottapally onwards there was a further decrease till south of Anthakaranazhi inlet. In the zone around Anthakaranazhi inlet, there was an increase in the run-up level reaching up to 3.5 m. Further north, in the Chellanum–Puthuvype
N.P. Kurian et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
326
Figure 27.2.
Run-up level along the Kerala coast (after Kurian et al., 2006).
Field observations on the tsunami impact: Kerala coast
327
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
region around Cochin, run-up level decreased to 3 m. However, in the Edavanakkad region, the run-up level increased drastically and went up to 4.5 m. There was a reduction in the run-up level further north with a drastic reduction in the zone immediately north of the Munambam inlet. However, further north the level increased showing up to 3 m around Vadanapally. There is again a drastic decrease in the sector south of the Ponnani inlet. An increase in the level is found north of Ponnani inlet and run-up level upto 2.5 m is found in Beypore inlet, south of Calicut. In the northern parts of Kerala coast comprising of Calicut, Cannannore and Kasargod districts, the run-up levels were generally low, up to 2.5 m. However, a short sector around Choottad north of Cannanore was notable for a high run-up level of 3–3.5 m which is not reported anywhere in the northern Kerala. A short sector north of Nandhi and the Kanjhagad–Manjeswaram sector (Kasargod) showed run-up level of only up to 1.0 m.
27.4
DAMAGE TO LIFE AND PROPERTY
The damage to life and property was maximum in the sectors adjoining Kayamkulam inlet in the southern Kerala where the run-up was maximum. Towards north and south of this zone, the damage decreased. A brief description of the damages due to tsunami along the coast is given below. In the southern-most, Pozhiyur to Vizhinjam sector, there was no damage, but just 1 km south of Vizhinjam harbour, some country boats were damaged. However, in Poonthura, Beemapally, Valiathura, Sankumugham, Veli and Paravur areas, further north, the water level increased only up to the monsoonal berm and no damage was reported. In the coastal stretch between Mayyanad and Eravipuram near Quilon, seawalls collapsed due to the tsunami. Further north, near Thangassery, tsunami caused severe damage to the harbour area, which is protected by two breakwaters. Areas north of the harbour were also affected, despite the protection by the seawall. John Brito colony, south of Sakthikulangara harbour area was also badly hit killing one person and washing away 46 houses; country boats and other fishing vessels were damaged inside the harbour. The sector immediately to the north of Sakthikulangara till Kovilthottam is well protected by a seawall, and hence there was no hinterland inundation. However, at Kovilthottam, hinterland inundation took place through a gap in the seawall and caused some damage. In the sector north of Kovilthottam, till Cheriazhikkal, damage was minimal in spite of higher inundation, probably due to the good quality of construction of houses. Further north, the damage was extensive (see Figure 27.3). Many houses collapsed completely and almost all the houses were partially damaged. Many deep pits were formed on the eastern side of the coastal road. In the Alappad Panchayat, which covers the Kovilthottam–Kayamkulam inlet barrier beach sector, 132 people were killed, mostly women and children, who were trapped in the collapsed houses. More than 1250 people were injured and more than 2400 houses completely damaged. The floodwater that inundated the whole barrier beach flowed towards TS canal, which runs parallel to the shore. Water rushed through Kayamkulam inlet and together with the overflow from the barrier beach, water level rose to about 4.5 m, causing severe damage in the adjoining areas of Kayamkulam backwaters. Another notable feature of the tsunami onslaught in the area was the deposition of black sand in the sector north of Cheriazhikkal up to the inlet, with the maximum thickness of about 1 m. The devastation was quite extensive in the sector immediately to the north of Kayamkulam inlet also (Figure 27.4(a),(b)), though not as much as at Alappad, probably because this area was not as thickly populated. Just north of the Kayamkulam inlet, a deep trench was formed adjacent to the coastal road. As in the case of the southern part of the inlet, here also extensive deposits of black sands were found in the hinterland area. Maximum damage was observed in the Tharayilkadavu area, about 4 km north of Kayamkulam inlet. Many houses were completely damaged and 30 people died in this area. More than 800 houses were completely damaged and about 1500 were partially damaged. The incidents at Tharayilkadavu and Alappad
328
N.P. Kurian et al.
N5 Kayamkulam
Quilon
Nov. 2004 Jan. 2005
Figure 27.3.
Beach profile stations.
Elevation (m)
(a) Kovilthotam 4 3 2 1 0 -1 -2 -3 -4 -5
2001 2005 Mean Water Level
0
20
40
60
80
100 120 140 160 180 200 220 240 260 Distance (m)
(b) Vellanathuruthu 6 4 Elevation (m)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
N1 S1 S4
2 0
2001 2005
Mean Water Level
-2 -4 -6 0
20
40
60
80
100
120
140
160
180
200
220
240
Distance (m)
Figure 27.4.
SLED profiles across two stations off Chavara.
show that, the breakwater constructed for fishing harbour interfered with the tsunami waves. No serious damage was reported along Arattupuzha–Thottappally–Alleppey sector, further north. The inundation and loss of life and property was much less in the sector from Alleppey to south of Anthakaranazhi inlet. In Arattuvazhi, about 3 km south of Anthakaranazhi, inundation
Field observations on the tsunami impact: Kerala coast
1015000
329
Arattupuzha
1010000
Kayamkulam Inlet
1005000
Vallikkavu
Northing (m)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
1000000
Vellanathuruthu
995000
990000
Neendakara Inlet
985000
Thangasseri 980000
975000
2000 2005 Shoreline
970000 640000 645000 650000 655000 660000 665000
Easting (m)
Figure 27.5.
Bathymetric changes along Arattupuzha–Thangasseri coast.
caused damages to many houses. Further north, in the zone around Anthakaranazhi inlet, three deaths are reported in addition to other damage and loss of fishing gadgets. In the zone north of Anthakaranazhi up to Cochin inlet, there was no significant inundation or damage. However, at Puthuvype, just north of the Cochin, many shops arranged for a beach carnival were damaged. Further north, in the Edavanakkad region, the inundation was high causing wide spread damage in the area. Many houses were completely damaged and five people were killed, a number that would have gone up, had the area been more densely populated. Deep pits were formed along the eastern side of coastal road. Huge boulders of seawalls of this area were thrown far inland by the tsunami waves. The impact clearly shows that, even seawall cannot protect a coast from tsunami. There was a reduction in the inundation further north, till Munambam inlet, where the waves entered upstream up to about 500 m and flooded up to a level of 2.5 m in the eastern parts, causing heavy damage to houses. Further north, in the Ponnani, Calicut, Cannannore and Kasaragode coasts, the inundation was of lower intensity. However the flood waters indeed caused some damage at a few locations (Figure 27.5) along the northern Kerala coast.
330
N.P. Kurian et al.
Table 27.1.
Run-up level and maximum inundation time along Kerala coast (after Kurian et al., 2006).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Location Chootad (Cannanore) Dharmadam (Tellichery) Calicut Ponnani Edavanakad (Cochin) Andakaranazhi Alleppey Valiazhikkal (Kayamkulam) Azhikkal Thangasseri (Quilon) Paravur Vizhimjam (Trivandrum)
Run-up level (m)
Maximum inundation time (h)
3.0–3.5 2.0–2.5 1.5–2.0 0.5–1.0 4.0–4.5 3.0–3.5 2.5–3.0 4.5–5.0 4.5–5.0 2.5–3.0 2.0–2.5 2.0–2.5
2145 2130 2330 +0200 1430 1400 1300 1240 1130 1400 1300 1400
27.5 TIME OF ARRIVAL OF THE TSUNAMI The time of arrival of tsunami waves at different coastal locations was recorded during field visits by interviewing local people. The time of arrival of highest wave is reproduced in Table 27.1 from Kurian et al. (2006) for selected locations. Along the Trivandrum coast where no damage occurred, the time of maximum inundation was reported to be during 13:00–14:00 h, which coincided with the high tide. However, along the Quilon coast, the time of maximum inundation was earlier by about 1 h indicating the link with the flood tide (the time of occurrence of high tide advances towards south along this coast while the tsunami waves propagated to the north). Along the Alleppey and Cochin coasts, the time of maximum inundation was reported to be in the afternoon. However, towards north of Munambam and the whole of the northern Kerala coast, the time of maximum inundation barring one or two locations was late in the evening or around midnight coinciding with the second high tide of the day when two high waves were also recorded (Kurian et al., 2006). It is quite possible that the highest waves in the afternoon, which progressed towards the northern Kerala coast, went unnoticed due to the low tide. 27.5.1
Beach profile variations
In order to understand the impact of tsunami on the beach morphology, post-tsunami beach profiles were measured at five stations in the sector north on 14 January 2005 and four stations south of Kayamkulam inlet on 15 January 2005 (Figure 27.3). For these stations pre-tsunami beach profiles (16 November 2004) were available and the reference stones were intact without any damage due to the tsunami. Beach profile measurements were carried out using the standard level and staff method. The volume changes computed from the beach profiles (Table 27.2) are reproduced from Kurian et al. (2006). It is seen that at N1 (just north of Kayamkulam inlet) nearly 53 m3 of erosion has taken place whereas at N2, further north of N1, an erosion of 16 m3 has taken place. The highest quantum of erosion has taken place at N3 which is 2 km north of the inlet. At N4 and N5 (which are further north of the inlet) also erosion took place though the quantum is much less. At S1, which is just on the southern side of the inlet adjacent to the breakwater, high quantum of deposition equal to 91 m3 has been noticed. But at S2 which is about 200 m south of the inlet, erosion equal to 38 m3 has occurred. However, at S3, about 1 km south of the inlet, deposition of 65 m3 is seen. At S4, which is further south of the inlet a modest deposition of 13 m3 is noted.
Field observations on the tsunami impact: Kerala coast
331
Table 27.2. Volume changes at different stations adjoining the Kayamkulam inlet (after Kurian et al., 2006).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Stations N1 N2 N3 N4 N5 S1 S2 S3 S4
Status
Volume change (m3 /m of beach)
Erosion Erosion Erosion Erosion Erosion Deposition Erosion Deposition Deposition
53.4 16.1 66.5 4.0 7.3 91.4 38.1 64.8 12.6
The erosion/deposition pattern obtained has to be seen in the backdrop of the coastal sedimentation processes prevalent in the area, in addition to the erosional impact of the tsunami waves. The breakwaters at the inlet, jetting out into the sea is acting as a groin, ever since the construction started a couple of years ago. Thus huge accretion has been taking place in the southern side of the inlet due to the predominant northerly longshore currents during fair weather. Erosion has been taking place in the northern side due to the groin effect of the breakwater. In the present case, the pre-tsunami beach profiles were taken on 14 November 42 days before. Thus the beach in the southern side of the inlet must have got considerably accreted with respect to the pretsunami profile and thereafter till the post-tsunami beach profiling on 15 January 2005. The field signatures on both the sides of the inlet showed scouring and erosion. However, the erosional effect of the tsunami was not sufficient enough to offset the depositional trend in the southern side except at station S2. In a similar way, the erosion observed in the northern side may not be entirely due to the tsunami. 27.5.2
Nearshore profile
Nearshore profiles of three locations were measured using Sliding Level Estimation Device (SLED). The SLED has a 6 m high graduated staff mounted vertically on its base. Nearshore profile measurement using SLED involves its transportation offshore and its positioning at a depth of about 5 m off the station at which profile measurement is required. Then it is pulled shoreward, taking care to stop at regular intervals to facilitate level measurement using a theodolite, which is already positioned on the beach. Simultaneous to the SLED profiling of nearshore, (from 5 m to shoreline) beach profiling (from backshore to shoreline) using dumpy level and staff was also carried out. The SLED profiles for two locations are given in Figure 27.4. The profiles have a distinct character with a steep shore face at 1 : 20 above 3 m depth and then a much lower-gradient profile beyond 3 m depth. The nearshore profiles for the two locations have been compared with the profiles for 2001 from Kurian et al. (2002). Since the zone under the study is a dynamic one prone for seasonal and yearly changes, no impact of tsunami could be assessed with the data of 2001. 27.6
BATHYMETRIC CHANGES
In order to assess the impact of tsunami on the bathymetry of the inner shelf, a bathymetric survey was carried out during March–April 2005, in the Neendakara–Thottapally sector, which
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
332
N.P. Kurian et al.
encompasses the worst affected Kayamkulam inlet area. A Bathy-500DF echo sounder was used for sounding together with a Leica DGPS system for position fixing integrated through a hydrographic software. The survey was carried out along transects at 1 km interval limited to 50 m water depth with additional transects at 500 m intervals in the Kayamkulam inlet area. The Hydrographic data editing software was used to view all the depth profiles and to identify spurious depths or depth spikes. All spikes were removed from the raw data using automatic batch processing tool and the data were cross-checked from the analogue data recorder. The bathymetric map so prepared is presented in Figure 27.5. In order to assess the impact of tsunami on the bathymetry, the bathymetric data available for the year 2000 (Kurian et al., 2002) is superimposed. Since inner shelf profiles barring the nearshore zone are usually not amenable to any significant changes except at the mouth of large rivers, it is reasonable to assume that the changes noticed in the offshore zone beyond 10 m correspond to the tsunami effect. It can be seen that, in general, there is a shifting of depth contours towards shore indicating erosion of sediments and deepening of inner shelf due to the tsunami. The impact seems to be maximum in the northern region, particularly in the zone off Vallikavu, south of Kayamkulam inlet where the bathymetric changes are seen prominently even upto 45 m depth. In the southern sector off Neendakara–Vellanathuruthu region too, shifting of contours offshore are seen in the 40–50 m isobath region, though it is missing in the shallower locations. The shallow region of this sector is known for its rocky substratum (GSI, 1997) and that could be the reason for lack of any impact in the inshore of this region. The bathymetric survey confirms the erosional tendency of the tsunami waves even in the inner shelf.
27.7
SEDIMENT CHARACTERISTICS
The sediment distribution in the shelf depends on the wave-current regime of the area. Since there was considerable change in the bathymetry, particularly in the offshore region between 30–50 m depth, surficial sediment samples were collected in the same region during May–June, 2005 (CESS, 2005). Totally 110 surficial sediment samples were collected at different transects limited to 50 m water depth off Neendakara–Thottapally coast and were analysed for coarse and fine fraction sizes. The micro-geomorphological study of quartz grain from the shelf region of Kerala coast also indicates that they are the beach deposits of Holocene period (Prithiviraj and Prakash, 1991). The post-tsunami sediment distribution (Figure 27.6) indicates that these offshore sandy sediments become more linear with occasional pinching out towards the shore. The sandy patch off Chavara seen in the pre-tsunami map is seen to be extending towards Kayamkulam inlet. It is quite possible that the tsunami waves might have taken away the finer sediments leading to dominance of sandy sediments. This is corroborated with the erosion reported from the bathymetric survey. The analysis of the results indicates changes in the inner shelf sediment characteristics with shoreward migration of sandy sediments off Kayamkulam where the impact of tsunami was most severe with run-up level up to 5 m.
27.8
SILTATION OF BACKWATERS
The inundation due to the tsunami caused considerable flooding of backwaters bordering barrier beaches. The flooding occurred either due to the entry of water that crossed over the shores or due to inundation through the tidal inlets. In locations where the run-up heights were higher than the level of barrier beaches such as the Neendakara–Thottapally region in southern Kerala and Vypin region in central Kerala the inundation were due to both. It is quite possible that these
333
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Field observations on the tsunami impact: Kerala coast
Figure 27.6.
Distribution of sand in the inner shelf (a) during 1987 and (b) during 2005.
inundations could cause considerable siltation of the backwaters due to the sediments brought in by the inundation, particularly in the backdrop of the erosive tendency of the tsunami as is seen already. It is already reported that the inundation generated deposits as thick as about a metre on the coastal roads of Kayamkulam inlet area (Narayana et al., 2005; Kurian et al., 2005).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
334
N.P. Kurian et al.
Figure 27.7.
Bathymetric changes along the TS canal.
Thus a study of the siltation in the TS canal bordering the Neendakara–Thottapally region, which is part of the National Waterway from Kollam to Kottapuram, was undertaken (Figure 27.7). As is already seen the shore region of this sector witnessed run-ups in the range 2.5–5 m. The sector north of Chavara is a barrier beach with width varying from <50 m to 300 m. The land elevation of this barrier beach is hardly 3 m above mean sea level (MSL). In order to assess the siltation of this backwater, bathymetric survey was carried out in May 2005 before the onset of the southwest monsoon. The availability of sounding data for this canal for the month of July 2004 (Inland Waterways Authority of India), about 5 months prior to the tsunami enabled the comparison of the soundings (Figure 27.7) and estimation of siltation that has taken place in the canal. It has been observed that the sector from Kollam to Chavara has undergone siltation ranging up to about 1 m. A small sector further north up to Vattakayal has not undergone any significant change, apparently due to lack of any inundation in the hinterland region. The sector from Vattakayal to Kayamkulam inlet of length 14 km witnessed both siltation and erosion. It may be noted that this was the zone where the inundation was high with a maximum run-up of 5 m. Erosion took place in the Azhikal–Shrayikad stretch in this sector which also happens to be the worst affected area in terms of human causalities and loss of property. The entire sector north of Kayamkulam inlet, which has a depth between 2 and 4 m during pre-tsunami period, has shown siltation. Siltation is prominent at Valaizhikkal, just north of Kayamkulam inlet, Cochin
Field observations on the tsunami impact: Kerala coast
335
yard jetty and near National Thermal Power Corporation (NTPC) with the values ranging upto a maximum of 3.5 m. Other locations of the canal in this sector witnessed siltation varying from 0.1 to 0.6 m.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
27.9
DISCUSSION AND CONCLUSIONS
The December 2004 Tsunami had a devastating effect on some sectors of Kerala coast, notably along the Neendakara–Arattupuzha sector of Kerala coast bringing about significant changes in the coastal geomorphologic setting along this region. The run-up level distribution along the Kerala coast shows wide variation. It shows a maximum at Kayamkulam inlet, recording upto 5 m which incidentally is lower than the highest reported along the Tamil Nadu coast (Chadha et al., 2005). While the tsunami that hit the northern Tamil Nadu coast were directly from the source, the ones that reached Kerala coast are those diffracted at the southern tip of the Sri Lanka and India. These waves while travelling northward along the southwest coast of India must have undergone combined refraction–diffraction bringing in convergence zones (such as Colachel in Tamil Nadu, and Kayamkulam in Kerala) and divergence zones (as at Trivandrum). Reflection from the Lakshadweep–Maldive ridge could be another important process compounding the refraction–diffraction process. The occurrence of a few high waves several hours after the main waves points to the possibility for reflected waves from as far away as Somalia. Other physical oceanographic processes also could have played a role in the tsunami intensity variation along the coast. The interactions of the tsunami with tides, currents and waves could be crucial. Growth of tsunami waves by drawing energy from these hydrodynamic forces through Reynolds’ eddy stresses is possible (Murty and Rao, 2005). The arrival of tsunami waves at high tide is a factor that compounded the inundation, leading to higher intensity of damage around Kayamkulam inlet (Kurian et al., 2006). In the same way, the low tide minimized the effect, as observed in the northern tracts of the coast, where the tsunami waves arrived in the afternoon. The highest waves along the northern Kerala coast occurred in the midnight, coinciding with the next high tide and the occurrence of two major waves at that time. In addition to the above processes, local geomorphic setup also could play some role in the intensity of inundation. A factor that could have contributed to the higher inundation at Kayamkulam inlet area is the two breakwaters jetting out into the sea at the inlet blocking the northward propagation of the tsunami waves. The observed run-up level distribution along the coast could only be understood by setting up inundation models, which relies on the characteristics of the approaching waves, finer data on the inner shelf bathymetry, hydrodynamic characteristics and coastal topography. The erosive tendency of the tsunami is clearly reflected in the shelf and beach profiles. Estimation of the quantum of erosion in the beach has limitations with regard to the pre-tsunami data which was 42 days before the tsunami. Nevertheless, the data conclusively shows the erosion on the beach. In a similar way, the bathymetric data shows erosion that has taken place in the inner shelf region. The inner shelf is also an area where there are occurrences of sandy sediments of Holocene period deposited during the lower stand of sea level (Nair and Hashimi, 1980). It is quite possible that the tsunami waves might have stirred the bottom and taken ashore the finer sediments leading to dominance of sandy sediments. The run-up distances of seawater to the inland region is directly related to the beach slope, land elevation, settlement pattern, infrastructure facilities, etc. Tsunami inundation into the hinterland regions is relatively less in Kerala due to the occurrences of coast parallel backwaters/lagoons. It can be surmised that the sediments resuspended in the inner shelf by the tsunami is transported onto the barrier beach resulting in its deposition inland. The heavier ones among sediments are the first to be deposited with the lighter ones carried further inland leading to its deposition
336
N.P. Kurian et al.
finally in the backwaters. The huge deposits of heavies seen in the hinterland areas and siltation seen in the TS canal support this. An increase of depth in the canal to the south of Kayamkulam inlet possibly due to the direct entry and flushing out of the tsunami water and siltation to the north due to the outwash and inundation is deciphered. Spatial variations in the inundation also decided the amount of siltation of the backwater.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
ACKNOWLEDGEMENT This work was supported by Department of Ocean Development (DOD), Ministry of Environment and Forests (MoEF) and Department of Science and Technology (DST, Government of India). The Inland Waterways Authority of India has provided the sounding chart (pre-tsunami) of TS canal. Thanks are due to M/s K. Rajith, B.T. Muralikrishnan, Abhilash P. Pillai., P. Kalaiarasan and Tiju Varghese, M.S. Saran, K.O. Badarees and K.P. Indulekha for contributions during the field surveys, processing of data and preparation of manuscript. REFERENCES Berninghausen, W.H. (1966). Tsunamis and seismic seiches reported from regions adjacent to the Indian Ocean. Bull. Seisml. Soc. Am., 56(1), 69–74. Baba, M. and Kurian, N.P. (1988). Ocean Waves and Beach Processes of the SW Coast of India. Monograph, Centre for Earth Science Studies, Trivandrum, 249 pp. Centre for Earth Science Studies (CESS) (2005). Characterisation of Indian Placer deposits with special reference to the offshore region of Southwest Coast of India, Annual Report 2004–2005, submitted to Central Mining Research Institute, Dhanbad, India, 34 p. Chadha, R.K., Latha, G., Yeh, H., Peterson, C., and Katada, T. (2005). The tsunami of the Great Sumatra earthquake of M 9.0 on 26 December 2004 – Impact on the east coast of India. Curr. Sci., 88, 1297–1301. Geological survey of India (GSI) (1997). Report on Detailed Exploration for Heavy Mineral Placer Deposit off Chavara, Kerala (Northern and Southern Block). Marine Wing, West Coast Division, Mangalore, pp. 1–39. Kurian, N.P., Prakash, T.N., Thomas, K.V., Shahul Hameed, T.S., Chattopadyay, S., Baba, M., Black, K.P., and Joseph Mathew (2002). Heavy mineral budgeting and management at Chavara. Project Report No. CESS-PR-19-2002. Centre for Earth Science Studies, Trivandrum, 2002, 513 pp. Kurian, N.P., Rajith, K., Murali Krishnan, B.T., Kalairasan, P., and Pillai, A.P. (2005). December 2004 tsunami: runup level and impact along the Kerala Coast. In: Tsunami: The India Context. S.M. Ramasamy and C.J. Kumanan (eds.), Allied Publishers Pvt. Ltd, Chennai, India, pp. 111–127. Kurian, N.P., Rajith, K., Pillai, A.P., Murali Krishnan, B.T., Kalairasan, P. (2006). Inundation characteristics and geomorphological impacts of December 2004 tsunami on Kerala coast. Curr. Sci., 90, 240–249. Murty, T.S. and Rao, A.D. (2005). The tsunami of 26th December 2004 in the Indian Ocean. Paper presented in the Seminar on Tsunami and Coastal Protection, Trivandrum, 11 February 2005. Nair, R.R. and Hashimi, N.H. (1980). Holocene climatic inferences from the sediments of the Western Indian continental shelf. Proc. Indian Acad. Sci., (Earth Planet. Sci.), 89, 239–315. Narayana, A.C., Tatavarti, R., and Shakdwipe, M. (2005). Tsunami of 26 December 2004: observations on Kerala coast. J. Geol. Soc. India, 65, 239–246. Prakash, T.N., Kurian, N.P., Rajith, K., Pillai, A.P., Murali Krishnan, B.T., Kalairasan, P., and Vargese, T.I. (2005). December 2004 tsunami: some results of field surveys along the Kerala coast. Proceedings of National Workshop on Tsunami Effects and Mitigation Measures, pp. 307–320. Preuss, P., Raad, P., and Bidoae, R. (2001). Mitigation strategies based on local tsunami effects. In: G.T. Hebenstreit (Ed.), Tsunami Research at the End of a Critical Decade. Kluwer Academic Publishers, The Netherlands, pp. 47–64. Prithiviraj, M. and Prakash, T.N. (1991). Surface microtextural study of detrital quartz grains of innershelf sediments off central Kerala coast. Indian J. Mar. Sci., 20, 13–16.
Field observations on the tsunami impact: Kerala coast
337
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Rao, V.P. and Wagle, B.G. (1997). Geomorphology and surficial geology of the western continental shelf and slope of India: a review. Curr. Sci., 73(4), 330–350. Shahul Hameed, T.S., Kurian, N.P., and Baba, M. (1994). Wave climate and power off Kavaratti, Lakshadweep. Proceedings of the Indian National Conference on Harbour and Ocean Engineering, Pune, Vol. 1. pp. A63–A72. State of Environment (SoE) Report Kerala (2005). Kerala State Council for Science, Technology and Environment, 349 pp. Soman, K. (2002). Geology of Kerala, Text Book Series, Geological Society of India, Bangalore, India, 335 pp. Wadia, D.N. (1981). Geology of India. Tata-McGraw-Hill, Delhi, 409 pp.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 28
Ecological Impact of Indian Ocean Tsunami
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
C.S.P. Iyer Centre for Marine Analytical Reference and Standards (C–MARS), Regional Research Laboratory (CSIR), Thiruvananthapuram, Kerala
28.1
INTRODUCTION
The impact of tsunami, hit the Indian Subcontinent on December 26, 2004, was moderate on the Southwest coast compared to the Southwest coast of India (Jayakumar et al., 2005; Narayana et al., 2005; Yalciner et al., 2005). However, there was severe damage to properties and loss of life. It also affected considerably the fishing activities. The major scientific questions, which were addressed, included the following: 1 Are there any major changes in the marine environment as an aftermath of the Killer waves? 2 If there are any major changes, has the sea the potential to rejuvenate itself in course of time? 3 How is it that certain coastal areas have been spared, where as some adjacent areas have been affected drastically? The answers to these questions came from an intense monitoring of the coastal waters of southwest India. For the purpose, the Department of Ocean Development (DOD) Research Vessel, MV Sagar Purvi, managed by the National Institute of Ocean Technology (NIOT) was utilized. The first Cruise was undertaken just after the tsunami from 7th to 17th of January 2005, which covered the coast from Muttam in the south to Thottapally in the north. A second set of measurements was undertaken in May from 13 to 22, 2005. 28.2
IMPACTS OF TSUNAMI ON THE SOUTHWEST COAST
Only the satellite imageries of the Kerala Coast, taken before and after tsunami are available (Figures 28.1(a) and (b)). Around 12:30 h on December 24, 2004, a series of waves hit the S.W. coastal areas. Then the sea receded up to 1 km revealing the seabed. This was followed by a very large wave of approximately 5 m in height, which had a devastating effect. Most of the houses, property, fishing vessels were destroyed. In addition, boulders constructed to protect the coast were thrown up to a distance of 150 m. 28.2.1
Choice of transects/stations
Based on the intensity of tsunami, seven transects were selected (Figure 28.2). The transects thus chosen were Thottapally, Valiyazhikkal, Vizhinjam, Kolachel, and Muttam. At each transect, stations were chosen at 5 km intervals, up to a distance of 25 km from shoreline. 339
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012 Figure 28.1.
Indian Remote Sensing Satellite (IRS) imageries of the Kerala coast, captured before and after tsunami. (a) During January 2004 using IRS 1D and (b) December 27, 2004.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
(a)
(i)
(ii)
(b) (iii)
Figure 28.2.
(iv)
(a) Selected transects and stations. (b) Isolines of various physical parameters, January 2005: (i) temperature in ◦ C, (ii) pH, (iii) salinity, and (iv) dissolved oxygen.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
342
C.S.P. Iyer
28.3
SAMPLING AND ANALYSIS
At each station, online measurements were taken for the parameters – conductivity, temperature, and chlorophyll-a, at 1 metre interval, from surface to bottom. The depths of the stations were assessed using the echo sounder, available on board the ship and the profiles scanned for any abnormalities on the ocean floor. Samples of water were collected from surface, mid depth, and bottom, using the Hydro Bios water sampler and analyzed for dissolved oxygen and nutrients (Grasshoff, 1983a,b; Koreleff, 1983). Separation of the phytoplankton was carried out by filtration of water samples using a 55 µ net (UNESCO, 1978). For zooplanktons, oblique hauls for a fixed time period were made using a Heron Transfer Net of 200 µm size, with attached flow meter. The samples of planktons, thus collected, were analyzed for the composition and population counts (UNESCO, 1968). Primary productivity measurements were carried out using the carbon 14 method (UNESCO, 1994). Sediment samples were collected using a Van Veen grab sampler (Sverdrup et al., 2003). About 100 g of the sample was taken after careful removal of large shells and shell fragments by hand picking. The samples were treated with 100 ml of H2 O2 (6%) to remove organic matter and thoroughly washed with distilled water to remove salts. Decantation after addition of sufficient water was carried out repeatedly till all silt and clay were removed. After air-drying, sub-samples were taken by coning and quartering. Both sand and silt+clay were weighed separately; the former was subjected to sieving using meshes with ½ phi interval and the weight percentages of each fraction were computed. Graphic measures as mean size, standard deviation, skewness, and kurtosis were also calculated (Folk, 1954, 1966). 28.4
RESULTS
Though data has been collected at all the stations and at different depths, both in January and May, only those of the surface at 5 km distance have been presented. To highlight the impact of tsunami, the data wherever available with us for December 2003 (pre-tsunami) have also been included. The data on the physico-chemical parameters for the transects are given in Table 28.1 for January and May 2005 along with those of pre-tsunami. For effective presentation of the data on nutrients, isolines are drawn as shown in Figures 28.3(a)–(f). Similarly, isolines are prepared for biological measurements – primary productivity, chlorophyll-a, and zooplankton – for the pretsunami, immediate post-tsunami (January 2005), and May 2005 periods (Figures 28.4(a)–(i)). The data is also entered in Table 28.2. For sediment samples, textural parameters are measured (Table 28.3), and plotted in Figure 28.5. Regarding bathymetry, only the depth profile off Muttam is shown, as the rest of the bathymetric studies did not show any significant variation (Figure 28.6). 28.5
DISCUSSION
In order to assess the impact of tsunami; the pre- and post-tsunami data is discussed for each transect separately. 28.5.1 Thottapally For comparison, we do not have any earlier data on the Thottapally transect. However, data is available for Alleppey, 10 km north of Thottapally. Therefore, as an approximation, a comparison has been made at Thottapally at 5 km offshore, with that reported for Alleppey at the same distance from shore. As can be seen, among the physico-chemical parameters, there is considerable
Water temperature (◦ C) Salinity pH DO (mg/l) NO− 2 (µg/l) NO− 3 (µg/l) SiO3− 4 (µg/l) PO3− 4 (µg/l)
Parameters
33.55 8.25 5.85 0.13 0.35 2.07 0.6
33.63 8.22 5.72 0.61 1.92 2.51 1.78
Pre-tsunami
27.68
Post-tsunami January 2005
28.23
May 2005
33.22 8.23 5.45 0.27 1.48 – 1.04
28.16
Pre-tsunami 33.62 8.21 4.84 0.82 3.25 2.34 1.58
28.26
Post-tsunami January 2005 33.21 8.29 5.61 0.04 3.06 1.70 0.18
28.16
May 2005 33.27 8.24 5.48 0.23 3.18 – 1.00
27.94
Pre-tsunami 33.72 8.16 5.22 1.89 4.26 2.89 1.16
28.42 33.26 8.26 5.95 1.69 3.76 2.65 0.3
28.00
Post-tsunami January 2005
Karunagappally
33.56 8.06 5.46 1.92 3.53 – 0.36
28.62
May 2005
Valiyazhikkal Pre-tsunami 33.82 8.16 4.69 1.21 4.25 2.52 1.86
28.26
Vizhinjam
33.24 8.30 3.16 0.04 1.29 1.30 0.19
28.26
Post-tsunami January 2005
Thotappally May 2005 33.50 8.62 4.12 0.44 1.16 – 0.65
28.64
Kolachel
33.34 8.32 4.85 0.18 0.16 1.37 1.08
31.16
Post-tsunami January 2005
Stations
33.62 8.62 4.85 0.21 0.29 – 1.22
30.24
May 2005
Comparison of physico-chemical parameters. Muttam
33.12 8.24 4.55 0.01 0.21 1.51 0.46
30.46
Post-tsunami January 2005
Table 28.1.
33.42 8.41 4.74 0.23 0.31 – 0.55
31.00
May 2005
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Ecological impact of Indian Ocean Tsunami 343
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
344
C.S.P. Iyer
(a)
(b)
(c)
(d)
(e)
(f)
Figure 28.3.
Isolines of various nutrients. (µmol/l). (a) Nitrite – January 2005, (b) nitrite – May 2005, (c) nitrate – January 2005, (d) nitrate – May 2005, (e) phosphate – January 2005, and (f) phosphate – May 2005.
decrease in the concentrations of phosphate, nitrite, nitrate, and to a smaller extent silicate in the immediate post-tsunami scenario (Table 28.1). The more glaring difference is in biological parameters. At all the stations in the Thottapally transect, primary productivity, chlorophyll-a, phytoplankton cell counts, zooplankton biomass, and zooplankton population have decreased just after Tsunami. Encouragingly, the results of the samples collected in May 2005 show that the primary productivity has increased, as also chlorophyll concentration. The population of phytoplanktons and zooplankton has also increased. However, the benthos population has not fully recovered from the impact of tsunami, even by May 2005 (Table 28.2).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Ecological impact of Indian Ocean Tsunami
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
345
(i)
Figure 28.4.
Isolines of various biological parameters. Primary productivity in mgC/m3 /h for (a) pretsunami, (b) January 2005, and (c) May 2005. Chlorophyll-a in mg/m3 for (d) pre-tsunami, (e) January 2005, and (f) May 2005. Zooplankton biomass in ml/m3 for (g) pre-tsunami, (h) January 2005, and (i) May 2005 periods.
11,880
2125
0.12
196
–
4920
0.13
501
5854
Mean
1.8350
1.7588 1.4338
Vizhinjam 15
Vizhinjam 25 Muttam 25
1.0956 0.7995
0.8570
Standard deviation
Pre-tsunami −0.0125 −0.3424
−0.0400
–
143 1166
163
0.16
14,918
1.13
20.42
May 2005 0.8733 1.1525
0.8951
Kurtosis
0.12
1955
0.43
8.04
Post-tsunami January 2005
Skewness
3231
168
0.11
3862
1.19
15.52
Granulometric data on sediments.
1333
156
0.13
0.85
0.51
20.5
1.84
Post-tsunami January 2005
7.30
May 2005
11.36
Pre-tsunami
Sample No.
Table 28.3.
Primary productivity (mgC/m3 /h) Chlorophyll-a (mg/m3 ) Phytoplankton (Nos./l) Zooplankton biomass (ml/m3 ) Zooplankton population (No./m3 ) Benthos (No./m2 )
Parameters
–
166
0.12
2680
0.97
7.22
0.00 0.46
0.50
Granule (%)
2043
189
0.14
2952
1.19
9.86
Pre-tsunami
Karunagappally Post-tsunami January 2005
Valiyazhikkal
100.00 99.54
99.50
Sand (%)
1041
158
0.15
18,652
1.21
23.75
May 2005
Thotappally Pre-tsunami 0.00 0.00
0.00
Silt (%)
1260
248
0.14
2640
1.4
12.13
Vizhinjam
562
126
0.09
9346
0.63
10.16
May 2005
0.00 0.00
0.00
Clay (%)
–
52
0.03
65
0.07
0.85
Post-tsunami January 2005
Stations
Kolachel
583
189
0.14
16,428
0.96
18.75
Sand –
–
Name (Shepherd)
–
76
0.07
235
0.14
1.52
Post-tsunami January 2005
Comparison of biological parameters.
May 2005
Table 28.2. Muttam
708
212
0.21
11,546
0.77
13.33
Slightly gravelly sand Sand Slightly gravelly sand
Name (Folk)
–
76
0.05
265
0. 12
1.55
Post-tsunami January 2005
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
May 2005
346 C.S.P. Iyer
Ecological impact of Indian Ocean Tsunami Sedimentological parameters off 25 km – Vizhinjam
100 80 Pre tsunami Jan-05
60 40 20 0 Sand
Silt
Clay
Texture (%)
Figure 28.5. Textural variations in sediments off 25 km – Vizhinjam. Depth Profile-Muttam to Kolachel
0 10 7.8 80
6 84 7.0 80
80
6.4
28
8
1 93 5.8 80
0 55 5.1 80
8 57 4.2 80
2 47 3.8 80
1 77 3.3 80
80
3.4
82
3
Latitude
Depth in meters
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Percentage composition
120
347
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
Figure 28.6.
Latitudinal depth profile along Muttam to Kolachel transect, note the channelized flow at 8.03 latitude.
28.5.2 Valiyazhikkal Looking at the chemical parameters of the water column, it is seen that the phosphate concentration just after tsunami has decreased compared to the earlier data at the near by Kayamkulam. Regarding the other nutrients, there is decrease in the concentrations of nitrite and silicate, where as nitrate has not changed (Table 28.1). However, the analysis of the samples collected in May 2005 indicates that there is improvement in the concentrations of nitrite and phosphate. Nitrate is almost at the same level as in January 2005. The biological parameters of primary productivity, chlorophyll-a, phytoplankton, zooplankton counts and biomass have come down just after tsunami (Table 28.2). However, the subsequent samples collected in May 2005 indicate that the
C.S.P. Iyer
348
primary productivity has considerably improved as also phytoplankton population. There is some improvement in the case of zooplankton, whereas benthos population has still to improve to the levels of pre tsunami days.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
28.5.3
Karunagapally
In the case of this transect, a comparison has been made with the measurements taken at the nearby transect, Neendakara. Though the phosphate levels show a steep decrease, just after tsunami, there is only marginal decrease for other nutrients in the samples collected in January 2005. All the biological parameters, though perceptibly decreased after tsunami show signs of considerable improvement in May 2005. However, as in the case of the other transects, the benthos population has still to recover from the effects of tsunami. 28.5.4 Vizhinjam The concentrations of all the nutrients taken just after tsunami show considerable decrease, compared to the earlier pre-tsunami data (Table 28.1). This is also reflected in the biological parameters of primary productivity, chlorophyll-a, phytoplankton and zooplankton cell counts, and zooplankton biomass (Table 28.2). In May 2005, there is improvement in the nutrient concentration as also the biological parameters. The benthos population has not improved, even after 5 months of the occurrence of tsunami. 28.5.5
Kolachel and Muttam
These transects have not been included under the COMAPS program and as such there is no earlier data available. Therefore no comparisons could be made. For the post-tsunami period in January 2005, compared to Vizhinjam, the concentration of nitrate is lower whereas nitrite and phosphate are higher (Table 28.1). There is considerable improvement in the nutrient concentrations and biological productivity in May 2005 (Table 28.2). 28.5.6
Sediment distribution
The majority of samples collected from the area consists of fine sediments of silt and clay except a few samples, which are mixed with sand and gravelly sand (Krumbein and Pettijohn, 1938; Friedman and Johnson, 1982). The latter are mainly from the area, off Vizhinjam and Muttam (Table 28.3). Sand is generally medium to coarse at Muttam and fine to medium off Vizhinjam. In general, the seabed in the northern part off Karunagapally and Thottapally is predominantly covered by clay and in the southern part by sand. The sand comprises of shell fragments and coarse detritals like quartz, feldspar, pyroxene, etc. Heavy minerals are also found in minor amounts, even up to a distance of 25 km offshore. 28.5.7
Ocean bathymetry
Bathymetric studies just after tsunami showed a sudden drop in the seabed of the order 5 m off Muttam at latitude of 8.03◦ . This has been confirmed during the subsequent cruise in May 2005 (Figure 28.6). 28.6
CONCLUSIONS
The following conclusions are drawn from the present study: 1 The concentrations of nutrients had come down at all transects just after tsunami. However, these picked up in the period from January to May 2005.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Ecological impact of Indian Ocean Tsunami
349
2 Primary productivity had drastically reduced in the wake of tsunami. This also has improved as evident from the samples collected in May 2005. 3 Though there was a lowering of plankton species diversity just after tsunami, it shows improvement. 4 The maximum impact was on the benthos community, which had not recovered even after 5 months of tsunami. 5 Due to the fall in primary productivity, subsequent to tsunami, the fish catch was affected. This also shows improvement. 6 The drop observed off Muttam indicates flow of water along with sediments to develop certain channels in the ocean bed. 7 The sediment samples collected offshore, have more of coarse sands, indicating their recent transportation from the coast. 8 The presence of heavy minerals in the sediment samples collected as far as 25 km offshore indicates that along with coarse sands these have also been transported due to high-energy backwash. 9 The impact of Tsunami was maximum at Vizhinjam, Kolachel, and Valiyazhikkal. Impact was least from Veli to Quilon and north of Thottapally. Summarising, the immediate impact of tsunami on the southwest coast was the draining of nutrients with the consequent lowering in chlorophyll concentration and thereby of primary productivity. Naturally, this affected the food web and resulted in depletion in fish. With time, there was perceptible recovery in the marine system, as evident from the monitoring in May 2005. It is also seen that the benthos being bottom dwelling, suffered maximum from tsunami. Naturally, it may take more time to recover. The maximum impact of tsunami was on those coastal regions, which had inland basins as Muttam, Vizhinjam and Valiyazhikkal. On the other hand, the impact was least from Veli to Quilon, due to long stretches of plains. It is heartening to note that the marine environment is slowly recovering from the impact of tsunami. This is evident from the improvement in biological productivity of this coastal stretch. REFERENCES Folk, R.L. (1954). The distinction between grain size and mineral composition in sedimentary-rock nomenclature. J. Geol., 62, 344–359. Folk, R.L. (1966). A review of grain-size parameters. Sediment, 6, 73–93. Friedman, G.M. and Johnson, K.G. (1982). Exercises in Sedimentology. Wiley, New York. Grasshoff, K. (1983a). Sampling and sampling techniques. In: K. Grasshoff, M. Ehrhardt, and K. Kremling (eds.), Methods of Seawater Analysis. Verlag Chemie, Weinheim, pp. 1–19. Grasshoff, K. (1983b). Determination of Oxygen. In: K. Grasshoff, M. Ehrhardt, and K. Kremling (eds.), Methods of Seawater Analysis. Verlag Chemie, Weinheim, pp. 61–72. Jayakumar, S., Ilangovan, D., Naik, K.A., Gowthaman, R., Tirodkar, G., Naik, G.N., Ganeshan, P., Murali, R.M., Michael, G.S., Ramana, M.V., and Battacharya, G.C. (2005). Run-up and inundation limits along southeast coast of India during the 26 December Indian Ocean Tsunami. Curr. Sci., 88, 1741–1743. Koreleff, F. (1983). Determination of nutrients. In: K. Grasshoff, M. Ehrhardt, and K. Kremling (eds.), Methods of Seawater Analysis. Verlag Chemie, Weinheim, pp. 125–187. Krumbein, W.C. and Pettijohn, F.J. (1938). Manual of Sedimentary Petrography. Appleton-Century-Crofts, New York. Narayana, A.C., Tatavarti, R., and Shakdwipe, M. (2005). Tsunami of 26 December 2004: observation on Kerala Coast. J. Geol. Soc. India, 65, 239–246. Sverdrup, K.A., Duxbury, A.C., and Duxbury, A.B. (2003). An Introduction to the World’s oceans, 8th edn. McGraw-Hill Publishers.
350
C.S.P. Iyer
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
UNESCO (1968). Zooplankton Sampling. Monograph on Oceanographic Methodology, UNESCO Publ., No. 2, 174 pp. UNESCO (1978). Phytoplankton Manual. UNESCO, Paris. UNESCO (1994). Protocols for the Joint Global Ocean Flux Study (JGOFS) Core Measurements. Intergovernmental Oceanographic Commission, UNESCO, Paris, Manual and Guides 29. pp. 128–134. Yalciner, A.C., Perincex, D., Ersoy, S., Presateya, S., Hidayat, R., and McAdoo, B. (2005). Report on 26 December 2004 Indian Ocean tsunami, Field survey during 21–31 January North of Sumatra by ITST, UNESCO IOC.
CHAPTER 29
Tsunami Damage to the South Eastern Coast of India
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
N. Chandrasekar School of Technology, Manonmaniam Sundaranar University, Tirunelveli, India R. Ramesh Planetary Geosciences Division, Physical Research Lab, Navrangpura, Ahmedabad, India
29.1
INTRODUCTION
On 26 December 2004, a great earthquake of magnitude of 9.3 had struck the active subduction corridor along the eastern margin of Indian lithosphere. Its epicentre, 3.7◦ N 95◦ E, lies close to northwest boundary of Sumatra where the trench appear to bend a little and be intersected by nearly northsouth running oceanic feature. Such features entering subduction way act as barriers to cause large stress to build, which eventually released in form of major earthquake (Raval, 2005). As a consequence of great energy release and associated land movements in the marine region, tsunami was produced in the eastern part of the Indian Ocean and spread with speed of approximately 800–900 km per hour on its eastern path. It reached the Andaman–Nicobar Islands, east coast of Sri Lanka and Tamil Nadu and then further north along the east coast up to Orissa and in the west coast up to Quilon. On reaching shallow water along the coast line, the large energy of the seismic waves get transformed into very forceful tidal waves of great height causing vast devastation there.
29.2
GEOMORPHOLOGY OF THE COAST
The Kanyakumari coast (Figure 29.1) displays different geomorphological units (Figure 29.2): 1 Along the shoreline between Colachel and Kanyakumari, there are number of rocky cliffs projecting into the Arabian Sea and forming as headlands. In between the headlands wide beaches are noticed with high concentration of black sands. 2 Between Kanyakumari and Rajakkamangalam, linear calcareous terraces covered by aeolian sands with an enrichment of heavy minerals, are observed. The pronounced calcareous sandstones at Chothavilai beach and Kanyakumari beach corresponds to sea level stand associated with 4–5 m high wave cut notches (Stearns, 1968). The sandy beaches are built by the development of foredune belt as well as high sand dunes. The sand dunes are fine in nature and well rounded with broken molluscan shells. 3 Numerous intertidal mudflats and lacustrine ponds are present along the study area. They are not received by any river flow and probably a swale system. 4 Dunes are rich amount of loose sand formed by aeolian activity. The coastal dunes are found in dry sands onshore blown to the back of the beach. The coastal dunes are noticed in the area of Chothavilai, Periakadu, Pozhikarai, Rajakkamangalam and Colachel. A large parabolic 351
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
352
N. Chandrasekar and R. Ramesh
Figure 29.1.
Location map.
dune along with dune complex is observed in Chothavilai beach. It has a length of 5 km and the width ranges from 5 to 2 km. The height of the dune rose up to 2–3 m near Periakadu, Rajakkamangalam and Chothavilai. Migration of dune is also noticed between these regions. 5 Mudflat area is seen in Manakudi, Pallam, Rajakkamangalam. These mudflats, contains silt, clay and water. They are always associated with sheltered environments like estuaries and embayments. 6 The most remarkable landforms like mangrove, salt marshes are also observed in Manakudi estuary. This estuary varied in length and width scored by tidal currents. Flooding of water by high tides control this kind of marshy land in this region. 7 Beach ridges are also seen in the study area with intervening sandy plains occurring parallel or subparallel to the shore formed by periodic wave impounding actions (Short et al., 1989). They are followed in the backshore by sandy plains and are discontinuous in nature. 29.3
INUNDATION EXTENT
As the tsunami approaches the shore it begins to slow and grow in height just like other water waves. Tsunami begins to lose energy as they rush onshore – part of the wave energy reflected offshore while shoreward propagating waves’ energy dissipated through bottom friction and turbulence that is, it can travel thousands of kilometre per hour with tremendous amount of
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunami damage to the south eastern coast of India
Figure 29.2.
353
Coastal geomorphology of the Kanyakumari District.
energy and have devastating effects on land. The tsunami has great erosional potential, stripping beaches of sand that may have taken years to accumulate and in the inland and coastal vegetation area, capable of flooding thousands of metre inland. The inundation can crush homes and other coastal structures by fast moving tsunami water. The field observation clearly indicates the coast of Kanyakumari was severely damaged. The Manakudi area and Colachel were badly hit by the tsunami waves and developed the inundation up to 1–1.5 km from the high tide line with the tsunami wave height of 10 m (Figure 29.3). The run-up distance was noticed in the field visit from the imprints available on the wall and trees and they were recorded 200–1500 m depending on the coastal topography and configuration. The observed values of the inundation limit and the run-up are shown in Table 29.1. The areas of tidal flats, mud flats, estuaries and river inlets were more inundated and the casualty was very high. Tsunami intensity for all the areas visited is based on the tsunami intensity scale (Ambrasey, 1962). During the field enquiry with the eyewitness people, it was known that the wave approaches the Manakudi estuary in an oblique form from the Kanyakumari seashore as diffracted waves. The wave height was about few metre but as it enters the estuary, the wave increases its size almost double as it breaks on the bar and the whole estuary and the low lying coastal lands remained flooded about several metres high than the usual and the entire residents had been washed off. On the other hand, the mouth of the estuary was dumped with sediments by the Tsunami waves from the offshore. Although the entire Kanyakumari coast experience the effect of the Tsunami waves, a stretch of few kilometres along the coast of Manakudi, Chothavilai, Pallam, Pozhikarai, Azhikkal, Pillaithoppu, Kadiapattinam, Kottilpadu and Colachel were most affected in terms of inundation, runup and erosion. On this stretch, the maximum devastation has occurred. The geomorphic features like beach terraces, sea
354
N. Chandrasekar and R. Ramesh 77°25′
77°30′
77°35′ N W
Inundation extent
E
8°0′
8°0′
S
8°5′
PERIAKKADU POZHIKKARAI KESAVANPUTHENTHURAI PUTHENTHURAI PALLAM
CHINNAMUTTOM MANAKUDI COVALAM
8°5′
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
AROCKIAPURAM
KANYAKUMARI
9 Kilometers
0
77°35′
77°30′
77°25′
77°15′
77°20′ N
Inundation extent W
E
COLACHEL
8°0′
8°0′
S
KOTILPADU PUDUR PERIAVILAI CHINNAVILAI
KADIAPATANAM MUTTOM MELATHURAI PILLAITHOPPU AZHIKKAL
7 Kilometers
0 77°15′
Figure 29.3.
8°5′
8°5′
RAJAKKAMANGALAM
77°20′
(a) Map showing the inundation between Arokiapuram to Dharmapuram, (b) Map showing the inundation between Rajakkamangalam to Colachel.
Tsunami damage to the south eastern coast of India Table 29.1.
Inundation extent and run-up level along the study area.
Location
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
355
Inundation limit from high tide line (m)
Run-up level (m)
40 50 50 50 400 350 100 50 50 50 50 150 250 300 200 75 50 50 100 100 150 250 300 450
1–2 1–2 1–2 1–2 5–6 4–5 3–4 2–3 2–3 2–3 2–3 2–3 4–5 4–5 2–3 3–4 3–4 2–3 2–3 2–3 2–3 3–4 4–5 5–6
Arokiapuram Chinnamuttam Kanyakumari Covalam Melamanakudi Keezhamanakudi Pallam Puthenthurai Kesavanputhenthurai Pozhikarai Periakadu Rajakkamangalam Azhikkal Pillaithoppu Melathurai Keezhamuttam Melamuttam Keezhakadiapatinam Melakadiapatinam Chinnavillai Periavillai Pudur Kottilpadu Colachel
Table 29.2. Amount of accretion/erosion of beach volume (m3 ) due to the tsunami. Location Arokiapuram Chinnamuttam Kanyakumari Chothavilai Azhikkal Pillaithoppu Muttam Kadiapattinam Kottilpadu Colachel
Accretion
Erosion
Overall result
0.77 0.16 1.91 1.38 0.91 0.49 0.07 0.17 0.45 0.24
1.55 1.52 0.00 0.00 0.10 0.15 0.56 0.80 0.06 0.30
−0.78 −1.36 −1.91 +1.38 +0.81 −0.34 −0.48 −0.63 −0.38 −0.06
Note: + Sign indicates accretion; − Sign indicates erosion.
cliffs which are perpendicular to the wave propagation and the area covered by thick vegetation of shrubs and mangroves are protected from the inundation effects. Similarly, the structures of groins erected in Arokiapuram, Chinnamuttam, Leepuram has protected the coast from tsunami wave inundation (Table 29.2).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
356
N. Chandrasekar and R. Ramesh
29.4
CHANGES IN THE COASTAL SETTING
It is well known that a tsunami attack caused substantial erosion and scour on the coastal landforms. Topographic survey was conducted in the area 1.5 m wide from the high tide line of the beach. The beach profile variations are measured and compared with previous survey records available with the author. Beaches of the Kanyakumari coast is in a dynamic system that resists inundation and erosion by two processes – storage of material on the foreshore and dune complex; and storage of sand in the offshore through creation of offshore sand bars that also act to reduce the erosive energy reaching the beach by breaking the incident waves. During tsunami waves which are characterized by higher waves and water level, sediments moves from the beach face, from dunes to offshore bars. This is understood during our field trip after the tsunami event along the coast between Arokiapuram and Colachel. The purpose of the present observation is to picture the beach profile modification due to tsunami havoc. To examine the tsunami force, we have compared the beach profile carried out on December 2004 with the profile carried out on January 2005. The generated tsunami wave was hitting the coast severely in the southeast coast of Kanyakumari and the water was receded for a longer distance in Manakudi, Chothavilai and Colachel where the coast is forcefully hit and damaged the landforms and properties. Moreover, the Chothavilai coast has elevated dunes at a height of 2–4 m above sea level. This has led to the removal of sand from the fore dunes. This is due to the angle of the incident wave caused by tsunami surge in the area. Here inundation has occurred due to the wave runup overtopping the berm and breaching the calcareous sandstone mined region. As the tsunami approaches the shore, the wave energy is lost due to the portion of the wave energy reflected in the offshore whereas the shoreward progressing wave dissipated through bottom friction and turbulence. Therefore, it has induced great devastation on the landscape and created great erosion. Each beach profiling in the tsunami affected area are distinct in nature and they are not similar to one another. The pre-tsunami beach profile reveals that most of the beaches along the study area are depositional in nature. The profile shows that the beaches are stretched one with flat and broad berm except at Arokiapuram and Chinnamuttam (Figure 29.4). The trend in the beach profile clearly reflected the nature of post-tsunami beach profile and we could estimate the rate of erosion roughly to the maximum of 1.5 m. The profile indicates the steep foreshore that joins to a gently sloping profile at a small distance seaward of the shoreline representing the region of most active sand transport due to the tsunami wave breaking (Figure 29.5). The seasonal variation and posttsunami variation were superimposed on this trend. The difference in the sub-aerial sand volume change from upper berm and dune to the offshore was known in the profile. The variation is much pronounced in the low water level due to a single factoral dimension and fluctuation at various scales due to tsunami inundation. Thus, the shoreline position between Arokiapuram and Colachel has displayed distinctly different temporal pattern of change volume due to tsunami inundation. 29.5
SEDIMENTATION DUE TO TSUNAMI
Field survey on tsunami-borne sediment deposits and the nature of erosion were made immediately after the 26 December 2004 tsunami in Kanyakumari (Figures 29.4 and 29.5). Measurements were made for land elevation, tsunami deposit thickness and character along several locations in the affected area. Numerous small pits were made along these transects to observe the tsunami deposits thickness and grain size. Tsunami deposits were identified as pebbles, single boulders of beach rocks, shells above on backshore dunes and brown rooted soil. It is inferred that the tsunami penetrated over the berm of the sandy beach and swept away the broken structures on its foundation particularly in Chothavilai beach. The calcareous sandstone exposed in Kanyakumari,
CHOTHAVILAI
5
(d)
3 1 1
0
10
20
(g)
3 1 1
0
10
20
COLACHEL
1
0
10
20
(e)
2 0 2
0 10 20 30 40 50 Distance from the Reference Point (M) 4.5
(h)
KADIAPATTINAM
2.5 0.5 1.5 5
10
15
(c)
1.5 0.5
0 5 10 15 20 25 30 35 45 Distance from the Reference Point (M)
30
AZHIKAL
4
0
30
40 50 Distance from the Reference Point (M)
4.5
Elevation Point (M)
Elevation Point (M)
MUTTOM
1
KANYAKUMARI
357
3.5
40 Distance from the Reference Point (M)
30
5
Elevation Point (M)
0 5 10 15 20 25 30 35 40 Distance from the Reference Point (M)
(b)
CHINNAMUTTOM
3
Elevation Point (M)
1.5
5
PILLAITHOPPU
4.5
(f)
2.5 0.5 1.5 0 5 10 15 20 25 30 35 40 Distance from the Reference Point (M) Elevation Point (M)
0.5
Elevation Point (M)
2.5
Elevation Point (M)
(a)
40 50 Distance from the Reference Point (M)
Elevation Point (M)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
AROKIAPURAM
4.5
Elevation Point (M)
Elevation Point (M)
Tsunami damage to the south eastern coast of India
20
25 30 35 Distance from the Reference Point (M)
KOTTILPADU
4.5
(i)
2.5 0.5 1.5
0
5 10 15 20 25 30 35 40 Distance from the Reference Point (M)
(j)
2.5 0.5
PRE-EVENT
POST-EVENT
1.5
0 10 20 30 40 50 Distance from the Reference Point (M)
Figure 29.4.
Impact of tsunami on the beach profile along the study area. COLACHEL KOTTILPADU
N W
KANYAKUMARI DISTRICT
S
MUTTOM KADIAPATANAM
E
PILLAITHOPPU AROKIAPURAM
AZHIKKAL CHOTHAVILAI
CHINNAMUTTOM Erosion Accretion
ARABIAN SEA
1 0 1 2 Kilometers KANYAKUMARI
Figure 29.5.
Proportion of erosion and accretion of beach volume due to tsunami along the study area.
Covalam and Chothavilai were dislocated and broken by the giant wave and carved with tsunami flow to the inland area for about 50–100 m. Smaller boulder fragment occurred as boulder ridges in the area and it is the result of wave runup overtopping the berm at the height of approximately 5 m above the sea level. The tsunami waves transported the sand with shells from the shoreline and offshore and deposited the sand in the buildings, on top of rocky cliffs, coconut tree and on the irrigated lands. Tsunami deposits were noticed at all affected areas. Black sand mixed with shells deposited further inland is believed to have come from offshore and beach. The deposits thickness is about 8 cm for over 200–300 m in the cross-shore direction whereas huge accumulation of coarse grained white and black sands were piled up over a distance of 200 m
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
358
N. Chandrasekar and R. Ramesh
Figure 29.6.
Physical damages along the Kanyakumari District.
from estuarine mouth in Manakudi and Valliyar River (Kadiapattinam). The deposits become thinned from 300 m inland and shows evidence for inundation level limits. Similarly, the grain size decreased from the shore to inland. Similar features were recorded in Papua New Guinea Tsunami. The locations affected severely by the tsunami reflected the large tsunami deposits in the inland and beach elevation was lowered due to erosion. The width of erosion zone increased and black sand distribution was better sorted in the areas of Colachel and Manavalakurichi. Here we could notice two or more layers of tsunami deposits. These layers were formed by different tsunami waves, variation in flow depth and direction within the waves. Detailed sedimentalogical analysis of tsunami wave deposits is not yet completed and the work is in progress. Complete analysis will provide the impact of tsunami wave in the shore and will help us to characterize the tsunami hazard. 29.6
PHYSICAL DAMAGES
Figure 29.6 and Table 29.3 provides the details of all the physical damages caused by the tsunami surge including loss of human lives, human residents and properties like boats, fishing nets, etc. It discloses the fact that casualty and settlement damage had occurred horribly in the beaches of Manakudy, Kottilpadu, Colachel, Azhikkal, where inundation occurred towards the hinterland (Chandrasekar, 2005; Narayana et al., 2005). The coast of Kanyakumari, Covalam, Pallam, Muttam, Kadiapattinam, Periavillai, Pudur, Rajakkamangalam and Azhikkal have witnessed severe casualty, house damage and fishing gears due to direct exposure to the waves and settlement in the CRZ (Coastal Regulation Zone). Those coasts have witnessed severe damages in spite of low inundation. The force of tsunami could be judged by these damages. 29.7
CHANGES IN WATER QUALITY
Figure 29.7 displays the effect on the inundated seawater to convert the groundwater into saline. The level of pH and TDS (Total Dissolved Solids) in some of the coasts like Mankudy and
Tsunami damage to the south eastern coast of India Table 29.3.
Loss of life and properties due to the tsunami.
Location
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
359
Arokiapuram Chinnamuttam Kanyakumari Covalam Keezhamanakudi Manakudy Pallam Puthenthurai Kesavanputhenthurai Pozhikarai Periakadu Rajakkamangalam Azhikkal Pillaithoppu Melathurai Muttam Kadiapattinam Chinnavilai Periavillai Pudur Kotilpadu Colachel
Human death
Houses
Boats
Fishing nets
0 0 0 0 34 139 6 0 1 0 0 2 57 0 0 52 31 3 4 24 215 227
0 0 0 35 295 400 13 0 1 0 0 25 300 0 0 187 7 20 25 60 402 495
0 0 675 150 427 600 540 0 21 31 25 102 333 0 0 374 276 150 50 150 400 1200
0 0 900 250 200 400 1000 0 135 25 10 40 40 0 0 750 800 200 70 200 400 800
Colachel were modified. It has been estimated that the pH level varied between 8.1 to 8.7 and the TDS varied between 2000 and 7000 mg/l along the coast. Local enquiry revealed that the quality of groundwater was very much affected by the tsunami as salt as seawater. 29.8
DAMAGES IN COASTAL VEGETATION
As it is beyond the scope of any satellite imagery to delineate the minor changes in the coastal vegetation like salt marshes, mangroves, etc., due to the tsunami surge, only the observations during field work has been taken into consideration to decode the biological damage along the study area and it has been observed that excluding the negligible damage to the mangroves of Manakudy estuary (Besana et al., 2004; Keating et al., 2004) and to the salt marshes of some of the beaches like Colachel, Chothavilai, etc., no significant changes in the biological scenario has been observed along the study area due to the tsunami surge. 29.9 VULNERABILITY All the estimated parameters were incorporated including geological, physical, chemical, economical and biological aspects for preparing the vulnerability map (Figure 29.8) following standard procedures (Walsh et al., 2000; Chandrasekar and Immanuel, 2005). It demarcates the study area into three different zones of vulnerability. The criteria adopted for the vulnerability map is given in Table 29.4. The total casualty, damaged buildings, boats and fishing nets were taken into account. Damage percentage was then calculated and then the percentage was divided into three range of almost equal interval. Then weightage has been assigned based on the type of damage. The classification was then given rank and
360
N. Chandrasekar and R. Ramesh 77°20
77°15
77°25
77°30
77°35
W
8°15
8°15
N E S
8°10
8°10
COLACHEL
8°5
8°5
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
MANAKUDI
Negligible Variation Moderate Variation High Variation 2
0
2
4 kilometers
8°00
8°00
77°15
77°20
77°25
77°30
77°35
77°15
77°20
77°25
77°30
77°35
N 8°15
8°15
W
E S
COLACHEL
8°10
8°10 8°5
8°5
MANAKUDI
Negligible Variation Moderate Variation High Variation
2
0
2
4 kilometers
8°00
8°00
77°15
Figure 29.7.
77°20
77°25
77°30
77°35
(a) Impact of tsunami in the pH level along the Kanyakumari District, (b) Impact of tsunami in the TDS level along the Kanyakumari District.
then the sum of weightage and the rank was calculated which expose the coast which are highly vulnerable and the coast of less vulnerable. In the physical damages which include casualty, house damage, boats, fishing net, etc., the coast like Kanyakumari, Covalam, Manakudy, Pallam, Azhikkal, Muttam, Kadiapattinam, Kottilpadu
Tsunami damage to the south eastern coast of India 77°15
77°20
77°25
77°30
361
77°35 N
8°5
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Less Vulnerable
Kanyakumari District Melamanakudi
Arokiapuram Chinnamuttom
Pallam u
ak
an
di
Covalam
m
a zh
Arabian Sea
e Ke
Kanyakumari
Vulnerable
1
0
1
2 km
8°5
Pudur Petiavilar i ra Chinpavilai hu at Melakadiapattinam el M Azhikkal Keezhakadiapattinam pu m m op to ut utto laith Rajakkamangalam m l Pi u ela am M ezh ad ai i kk ar ai ra Ke ria ikk hur hu Pe ozh ent nt P e h t th u Pu np va sa Ke
8°10
8°10
Colachel Kotilpadu
Kanyakumari
Indian Ocean
Highly Vulnerable 77°15
77°20
77°25
77°30
77°35
Figure 29.8. Tsunami vulnerability map of the study area.
Table 29.4.
Criteria adopted for the preparation of vulnerability map. Damage percentage
Range in percentage
Classification
Weightage
Rank
W×R
Total death
Max: 28.55 Min: 0.13
Fully damaged house
Max: 21.85 Min: 0.04
3
Partly damaged house
Max: 28.95 Min: 0.32
4
Boat damaged
Max: 21 Min: 0.42
5
New damaged
Max: 15.08 Min: 0.15
High Moderate Less High Moderate Less High Moderate Less High Moderate Less High Moderate Less
30
2
>18 9–18 <9 >14 7–14 <7 >18 9–18 <9 >14 7–14 <7 >10 5–10 <5
3 2 1 3 2 1 3 2 1 3 2 1 3 2 1
90 60 30 75 50 25 60 40 20 45 30 15 30 20 10
S. No.
Type of damage
1
25 20 15 10
Note: High: Highly vulnerable; Moderate: vulnerable; Less: Less vulnerable.
and Colachel fall on the highly vulnerable area. The coast like Kesavanputhenthurai, Chinnavilai, Periavillai and Pudur falls on vulnerable areas. As far as casualty is concerned the coast like Manakudy, Kotilpadu and Colachel are identified as highly vulnerable and the coast like Pallam, Azhikkal, Muttam, Kadiapattinam and Pudur were the vulnerable areas and the remaining coast are less vulnerable. Regarding building damage, the coast like Manakudy, Azhikkal, Muttam, Kotilpadu and Colachel are the highly vulnerable areas and the coast like Covalam, Pallam, Rajakkamangalam,
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
362
N. Chandrasekar and R. Ramesh
Chinnavilai, Periavillai and Pudur are identified as vulnerable areas. The other entire coast is less vulnerable. As far as economic properties like fishing boats and fishing nets are concerned, the coast like Kanyakumari, Covalam, Manakudy, Pallam, Kesavanputhenthurai, Azhikkal, Muttam, Kadiapattinam, Chinnavilai, Pudur, Kotilpadu and Colachel are the highly vulnerable areas and the coast like Puthenthurai, Pozhikarai, Periakadu and Periavillai are identified as vulnerable areas. As far as changes in the beach sediment volume is concerned, excluding the coast of Chothavilai and Azhikkal, all the other coast had endured significant erosion. The tsunami surge had not shown any partiality in amending the coastal geomorphology and in the beach profile especially the intertidal zones were brutally affected. Except notable variation in the pH andTDS level of the groundwater along some of the coast like Manakudy and Colachel, the water quality has not been affected beyond the bearable limit in other areas of the Kanyakumari District. Fortunately, the coastal vegetations like coconut plantations, mangroves, salt marshes, etc., were least affected by the inundated seawater excluding minimal damage to the mangroves and salt marshes of Manakudy, Colachel and Chothavilai. Having incorporated the above parameters, the overall impact of the seismic waves along the study area has been prepared which shows that the coast like Manakudy, and Colachel were the highly vulnerable areas whereas the coast like Kanyakumari, Pallam, Azhikkal and Muttam find their place in the vulnerable area. Remaining other coast like Arokiapuram, Chinnamuttam, Covalam, Puthenthurai, Kesavanputhenthurai, Pozhikarai, Periakadu, Rajakkamangalam, Pillaithoppu, Melathurai, Chinnavilai, Periavillai and Pudur were identified as less vulnerable areas.
29.10
CONCLUSION
The damage assessment executed on the various impacts of the tsunami surge on the coastal environment divulges the varied nature of the effect of tsunami on the study area. The evaluation of the coastal landforms before and after the event reveals that the uprush of the tsunami surge had brought enormous amount of sediments but the backwash had altered the beach profile especially at the intertidal zone. The estimation of beach volume discloses the fact that in spite of considerable accretion in some of the beaches mostly erosion had dictated the sand budget. The appraisal of the physical damages reveals that the casualty occurred more in the areas of high inundation. The buildings and the money-spinning properties like boats and fishing nets have been damaged brutally even in the beaches of minimal inundation. It is well evident from the chemical assessment done along the study area that excluding the increase of pH level and TDS level of groundwater in some of the coastal zones where there was much inundation, most of the coastal zone were found to be less affected or almost so by the flooding of the seawater during inundation as the duration of flooding would have not be well enough for the seawater to disturb the groundwater. The biological damage assessment proves the fact that the tsunami surge could not succeed in affecting the coastal vegetation such as salt marshes, mangroves, coconut plantations except in some of the estuaries like Manakudy with negligible degree of damage. Though the inundated seawater flooded the coconut plantations, the net result is not any note worthy damages to those plantations. The damage assessment map provides a panoramic view of the destruction rendered by the tsunami surge along the study area and urges the need of proper hazard management system especially along the brutally affected coast. To summarise the net effect of the event along the study area, most of the coastal zones were highly affected and the devastation induced by the tsunami surge on the coastal settings and human community may take some period to recover to the pre-event setting.
Tsunami damage to the south eastern coast of India
363
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
REFERENCES Ambrasey, N. (1962). Data for investigation of the seismic waves in the Eastern Mediterranean. Bull. Seismol. Soc. Am., 52(4), 895–913. Besana, M.G., Ando, M., and Hannah Mirabueno, M. (2004). The May 17, 1992 event: tsunami and coastal effects in eastern Mindanao, Philippines. Sci. Tsunami Hazards, 22(2), 61–68. Chandrasekar, N. (2005). Tsunami of 26th December 2004: observation on inundation, sedimentation and geomorphology of Kanyakumari coast, South India. In: Papadopoulos and Satake (eds.), Proceedings of the 22nd International Tsunami Symposium, Chania, Greece. pp. 49–56. Chandrasekar, N. and Immanuel, J.L. (2005). GIS supported categorisation of tsunami experienced beaches along the southern east coast of India: usage in mitigation activities. In: Proceedings of the National Seminar on GIS Application in Rural Development, Hyderabad, India, pp. 349–362. Keating, B., Whelan, F., and Bailey-Brock, J. (2004). Tsunami deposits at Queen’s Beach, Oahu, Hawaii – initial results and wave modelling. sci. Tsunami Hazards, 22(1), 23–43. Narayana, A.C., Tatavarti, R., and Shakdwipe, M. (2005). Tsunami of 26th December 2004: observations on Kerala coast. J. Geol. Soc. India, 65(2), 239–246. Raval, U. (2005). Some factors responsible for the devastations in Nagapattinam region due to Tsunami of 26th December 2004. J. Geol. Soc. India, 65(5), 647–649. Short, A.D., Buckley, R.C., and Fortheringhum, D.G. (1989). Preliminary investigations of beach ridge progradation on Eyre Peninsula and Kangarao Island. Trans. Roy. Soc. South Australia, 113, 145–161. Stearns, H.T. (1968). Quaternary shorelines in the Hawaiian Islands. B.P. Bishop Mus. Bull, 237, 34–46. Walsh, J.T., Charles Caruthers, G., Anne Heinitz, C., Edward Myers, P., Antonio Baptista, M., Garnet Erdakos, B., and Robert Kamphaus, A. (2000). Tsunami Hazard Map of the Southern Washington Coast: Modelled Tsunami Inundation from a Cascadia Subduction Zone Earthquake. Geologic Map GM-49. Report. Washington Division of Geology and Earth Resources, 12 pp.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 30
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Hydrophysical Manifestations of the Indian Ocean Tsunami Y. Sadhuram, T.V. Ramana Murthy and B.P. Rao National Institute of Oceanography, Regional Centre, Visakhapatnam, Andhra Pradesh, India
30.1
INTRODUCTION
The horrendous destruction caused by the tsunami waves due to the earthquake in the west coast of northern Sumatra (3.307◦ N; 95.947◦ E) at 30 km depth with a magnitude of 9.3 on Richter scale at 00.59 UTC on 26 December 2004, has been described in detail by several authors (e.g. Krishamoorthy et al., 2005). This chapter summarises the results of our investigations on the hydrophysical manifestations (salinity and temperature, coastal currents, internal waves, etc.) of the tsunami on the coastal environments in India. The National Institute of Oceanography (NIO), Goa, its Regional centre at Visakhapatnam and the National Centre for Antarctica and Ocean Research (NCAOR), Goa, jointly organised two cruises on board ORV Sagar Kanya during 3–15 January and 16 January–21 February 2005, to study the impact of tsunami in the Bay of Bengal and Andaman region. From the post-tsunami cruise on board Sagar Kanya during 16 January–21 February 2005, from Chennai to Andamans, it was observed that the sea surface temperature (SST) was 27◦ C off Chennai. The mixed layer depth (MLD) varied from 50 to 100 m between 80◦ 52 E and 87◦ E and thereafter it decreased to 70 m towards Andamans. Salinity varied from 32.8 to 33.9 psu along the west–east section (13◦ N) and the low salinity (31.7 psu) was observed near the Andaman Islands. High values of salinity (35.2 psu) and temperature (29◦ C) were observed around 100 m depth in the region, 83–84◦ E (Murthy, 2005). The Visakhapatnam Centre organised a short cruise on board CRV Sagar Sukthi during 3–10 January 2005, to study the impact of tsunami on temperature, salinity and currents in the coastal waters off Visakhapatnam coast. The effect of the tsunami waves on internal waves, acoustic losses and propagation are also studied. 30.2
IMPACT OF TSUNAMI ON SALINITY AND TEMPERATURE
Figure 30.1 shows the time series data on temperature and salinity at 90 m water depth off Visakhapatnam, collected during 4–6 January 2005, on board CRV Sagar Sukthi, a few days after the tsunami. Temperature was less than 26.5◦ C in the top layer (0–20 m) while pockets of high temperature (27.5–28.0◦ C) are seen in the middle layer (40–50 m). At bottom, the temperature was 22◦ C. Salinity at surface and bottom was 29 and 33 psu, respectively. Strong oscillations could be seen in the middle layer. Average salinity and temperature profiles from the above data are plotted and compared with the nearest Levitus data for the month of January. Present data shows a strong inversion of 2◦ C while the Levitus data shows a very mild inversion (<0.5◦ C). The temperature at the bottom was 3.5◦ C 365
366 Y. Sadhuram et al. (a) Temperature (˚C) 0 26
Depth (m)
10 20 30 40
28
50
26
60
24
70
23 0
5
10
15
4-1-05 17:00hrs
20
25
30
35
40
45 6-1-05 17:00hrs
Time (hrs)
(b) Salinity (psu)
0
.5
Depth (m)
29
20
30
30 40
33 34.
34
50
5
60 70 0
5
10
15
4-1-05 17:00hrs
20
25
30
35
40
45 6-1-05 17:00hrs
Time (hrs)
Figure 30.1. Time series (hourly) data on (a) temperature (◦ C) and (b) salinity (psu).
1020
28 30 32 34 36
22 24 26 28
Depth (m)
(c) Density (kg/m2)
(b) Salinity (psu)
(a) Temperature (°C) 0
0
0
20
20
20
40
40
40
60
60
60
80
80
80
1024
(d) Sound velocity (m/sec) (e) Brunt Vaisala frequency (N2 ; cph) 1624 1528 1532 1536 1540 0
0
5
10 15 20 25
Levitus 20 Depth (m)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
10
Post tsunami
40
60
80
Figure 30.2. Average profiles of (a) temperature (◦ C), (b) salinity (psu), (c) density (kg/m3 ), (d) sound velocity (m/s) and (e) Brunt Vaisala frequency (N 2 ; cph) based on the above data (solid line). Profiles from Levitus for the month of January are plotted (dotted line).
less compared with the Levitus data. At surface, the salinity was 3.5 psu less than the Levitus data. Large variations could be seen in sound velocity and Brunt Vaisala Frequency (N 2 ) (Figure 30.2). An inversion of 1.5◦ C at 30 km away from the coast was observed along the transect off Visakhapatnam harbour on 8 January 2005. A dome like structure is seen in salinity in the middle
Hydrophysical manifestations of the Indian Ocean tsunami (a) Temperature (°C)
(b) Salinity (psu) 0
10
10 26.5
20
27.5
50
27.5
26.5
60
25.5
90
5 10 15 20 25 30 35
33
50 60
23.5 5 22. 21.5
80
0
40
Depth (m)
40
70
31
34
70 35
80 90 100
0
Distance from coast (km)
5 10 15 20 25 30 35 Distance from coast (km)
(c) Density (kg/m3) 0 10
1018.5
20 30 Depth (m)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
30
30 32
Depth (m)
30
100
29 29
0 20
367
1019.5
40 50
1021.5 1022.5
60 70
.5
20
10
3.5
102
80 90 100 0
5 10 15 20 25 30 35 Distance from coast (km)
Figure 30.3. Variations of (a) temperature (◦ C), (b) salinity (psu) and (c) density (kg/m3 ) along the transect off Visakhapatnam harbour on 8 January 2005.
layer (30–50 m) close to the coast, which may be due to the oscillations created by tsunami. Normally we expect the isohalines to run perpendicular to the coast. This is reflected in the density structure (Figure 30.3). Fortunately, we have a cumulative trauma disorder (CTD) profile on 19 December 2004 (before tsunami) at one location (50 m) off Visakhapatnam coast, which could be compared with a profile taken on 7 January 2005 (after Tsunami) close to the above location. It may be noted that both the profiles were collected using the same equipment (SBE 25; Seabird Electronics, USA). Cooling of temperature could be seen throughout the water column with a maximum of 1.5◦ C at 40 m depth. The inversion, which existed at this depth before tsunami almost, disappeared due to the mixing caused by tsunami waves on 26 December 2004. Under normal conditions, we expect the intensification of inversion by January in this region. Salinity after tsunami increased by 1 psu at surface while at bottom it decreased by more than 1.5 psu compared with that before tsunami. The large gradient in salinity very much reduced after tsunami. These variations are reflected in density in the top 35 m layer (Figure 30.4).
368 Y. Sadhuram et al. (a) Temperature (°C) 27
28
26 28 30 32 34
1018 1020 1022
0
0
0
10
10
10
20
20
20
30
30
30
40
40
40
50
50
50 19/12/05 (Before Tsunami) 07/01/06 (After Tsunami)
U (m/sec) Speed (m/sec)
Direction(°) Temperature(°C)
Figure 30.4. Variations of (a) temperature (◦ C), (b) salinity (psu) and (c) density (kg/m3 ) before (20 December 2004) (solid line) and after (7 January 2005) (dotted line) tsunami.
V (m/sec)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Depth (m)
26
(c) Density (kg/m3)
(b) Salinity (psu)
a (10 m)
b (60 m) 26
27 26
24
25
22
260 250 240 230 220
260 240 220
0.7 0.6 0.5 0.4 0.3
0.6 0.4 0.2
0.2
0.2
0.4
0.4
0.3 0.4 0.5 0.6
0.4 0.6 0
4/1/05
1000
2000
Time (in minutes)
3000
0
6/1/05
4/1/05
1000
2000
Time (in minutes)
3000 6/1/05
Figure 30.5. Time series data (2 min interval) on temperature (◦ C) direction and speed (m/s) of the current, U and V components at (a) 10 m and (b) 60 m below surface.
30.3
IMPACT OF TSUNAMI ON COASTAL CURRENTS
Time series data (2 min interval) on temperature and currents at 10 m (surface) and 60 m (bottom) depths at the same location are presented in Figure 30.5. The speed of current at 10 m was mostly
Spectral estimate (°C)2/cph
Hydrophysical manifestations of the Indian Ocean tsunami 1.00E001
(a) Surface (10 m)
8.00E002 6.00E002 4.00E002 2.00E002 0.00E000
Spectral estimate (°C)2/cph
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
0.00
1.00E001
4.00 8.00 Frequency (cph)
12.00
(b) Bottom (60 m)
8.00E002 6.00E002 4.00E002 2.00E002 0.00E000 0.00
Figure 30.6.
369
4.00 8.00 Frequency (cph)
12.00
Spectral analysis of temperature (2 min interval) at (a) 10 m and (b) 60 m depths at the above location.
between 0.4 and 0.6 m/s and the direction was 220–240◦ . (almost parallel to coast). Strong oscillations are seen in V (meridional component), which varied between −0.3 and −0.6 m/s. The flow was almost similar at 60 m depth also, which indicates a strong and consistent flow throughout the water column (Figure 30.5). Normally, the speed of current in this region is about 0.2 m/s during winter (December–February). It appears that the coastal current was strengthened by the tsunami waves which hit east coast of India. 30.4
IMPACT OF TSUNAMI ON INTERNAL WAVES
Spectral analysis of temperature data (2 min interval) collected at 10 m and 60 m depths indicate high-frequency internal waves during the observational period (4–6 January 2005). The dominant frequencies are varying between 0.2 and 2.0 (cph) and 7.2–8 (cph) at both the depths (Figure 30.6). Transmission losses and acoustic propagation characteristics are reported by Murthy et al., (2005). 30.5
IMPACT OF TSUNAMI ON SALINITY AND TEMPERATURE OBSERVED FROM ARGO DATA
We have tried to examine the impact of tsunami on temperature and salinity at the deep sea using the data from Argo floats. Fortunately, there was one Indian float (2900459) at 15.184◦ N; 82.17◦ E at 11.37 GMT on 26 December 2004 (i.e., approximately 8 h after tsunami waves hit east coast
370 Y. Sadhuram et al. Temperature (°C) 24 26 28
31
0
0
20
20
40
40
60
Depth (m)
Depth (m)
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
22
Salinity (psu) 32 33 34
60
80
80
100
100 20/12/2004 (Before tsunami) 26/12/2004 (After tsunami)
Figure 30.7.
Changes in temperature (◦ C) and salinity (psu) in the top layer as seen from the Argo data before (20 December 2004) (solid line) and after (26 December 2004) (dotted line) tsunami.
of India). We have plotted this data in the top 100 m layer (the data is available upto 1500 m) and compared with the data on 20 December 2004 (03.13 GMT) close to the above location, (14.98◦ N; 82.15◦ E) before tsunami. It is interesting to see that the temperature in the top layer (0–10 m) was about 1.8◦ C cooler on 26 December, and an increase could be seen from 10 to 20 m. Salinity was 2.5 psu less than that observed on 20 December 2004 and a step like structure is seen in the salinity. There were no major changes in temperature and salinity in the layer 50–100 m (Figure 30.7). This location is south of Machilipatnam and close to the mouth of Krishna River. We surmise that these changes in the salinity and temperature could be due to the tsunami waves which hit this coast. The cooler water near the coast as a result of tsunami waves might have advected to this location, which may be possible in 8 h. From this, the speed of current generated by the tsunami waves was estimated as >3.0 m/s which is quite abnormal. This might have caused cooling in temperature and changes in the salinity of top 20 m layer. It may be noted that the run-up heights and the intensity of tsunami waves were higher at Machilipatnam compared with that at Visakhapatnam. However, further studies are necessary to confirm this view. 30.6 VULNERABILITY OF THE INDIAN COAST FOR THE DAMAGES DUE TO TSUNAMIS Preliminary assessment of the Indian coast for the damages due to tsunamis was reported by Sadhuram (2005) prior to the post-tsunami surveys. This study showed the possibility of assessing the vulnerability of the Indian coast based on the earlier results on sea level rise due to Greenhouse effect (Shetye et al., 1990).
Hydrophysical manifestations of the Indian Ocean tsunami
371
3000
Metres
2000
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
1000
0 25
20 °N 15 10 WEST COAST
Gujarat
Goa Maha rastra
Kerala Karnataka
5 5
10 15 °N 20 EAST COAST T.N.
A.P.
25
W.B. Orissa
Figure 30.8. Vulnerability of the Indian coast for the damages due to tsunamis/storm surges, inferred from the shoreline displacement for 1 m rise in sea level (Shetye et al., 1990).
The studies (Chadha et al., 2005; Jayakumar et al., 2005 and Ramana Murthy et al., 2005) on run-up heights and inundation limits based on the post-tsunami surveys along the coastal stretch 10–14◦ N, indicate maximum run-up heights of 5–7 m at Nagapattinam and minimum (2 m) at Chennai, in Tamil Nadu coast. In Andhra Pradesh, the run-up heights at Krishnapatnam and Visakhapatnam were 2.5 m and 1.51 m respectively. These run-up heights are compared with the displacement of shoreline for 1 m rise in sea level shown in Figure 30.8. It is interesting to see that the highest peak coincides with the Nagapattinam coast where 6065 people were killed and about 40,000 houses were damaged (Ramana Murthy et al., 2005). Lower values of run-up heights at Chennai and near Krishnapatnam coincide with the minimum displacement of shoreline. The peak is around 16◦ N (near Machilipatnam), where many people who went for a holy bath on that fateful day lost their lives. Unfortunately, no information is available on run-up heights at this place. Since this coast is shallow, one can expect at least a run up height of about 3 m. The tsunami energy was dissipated by the time it arrived the southern part of the Andhra Pradesh and northward from that point, the amplitudes decreased quickly. Hence, the tsunami height was low at Visakhaptnam and Orissa and West Bengal were not affected (Murty, 2005). Since the slopes are generally less on the east coast of India compared with the west coast, the damages due to storm surges and tsunamis can be expected to be more on the east coast. Even a small storm surge of 1 m height could inflict considerable damage near Nagapattinam area (Gonnert et al., 2001). A super cyclone which hit Paradip on 31 October 1999, resulted a storm surge of 8 m height which travelled about 30 km inland. More than 10,000 people were killed and the total loss was estimated as Rs. 5000 crores (Sadhuram et al., 2004). This is close to the total loss (Rs. 5500 crores) (INR one crore = USD 220,000) estimated due to the tsunami of 26 December 2004. On the west coast, 12–18◦ N appear to be safe for the sea level rise due to Greenhouse effect (Shetye et al., 1990). Even though the tsunami waves hit the west coast, there was no loss of life in Karnataka, Goa and Maharashtra, which are located in the above belt. North of this belt, there
372 Y. Sadhuram et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
is a peak at 22◦ N close to Kandla (Gulf of Kutch), where about 300 people were killed due to an 11.8 m height tsunami, which hit this area on 27 November 1945. From the above discussion and based on the post-tsunami surveys, it may be inferred that 10–12◦ N; (south Tamil Nadu coast), 14–16◦ N (south Andhra coast) and >20◦ N (West Bengal) on the east coast; 9–10◦ N (Kerala) and 21–24◦ N (Gujarat) appear to be more vulnerable for the damages due to tsunamis/storm surges/sea level rise due to Greenhouse effect. This is a preliminary assessment which can be modified if the detailed information on nearshore bathymetry and topography at very close intervals along the Indian coast is available. These digital bathymetry and topography maps are essential for tsunami and storm surge modelling. ACKNOWLEDGEMENTS The authors are thankful to Prof. Tad Murty for giving the opportunity for writing this chapter and for his suggestions. They would like to thank Dr. S.R. Shetye, Director, NIO, Goa and Dr. K.S.R. Murthy, Scientist-In-Charge, NIO, Regional Centre, Visakhapatnam for their interest and encouragement in this work. REFERENCES Chadha, R.K., Latha, G., Yeh, H., Peterson, C., and Katada, T. (2005). The tsunami of great Sumatra earth quake of M9.0 on 26 December 2004 – Impact on the East coast of India. Curr. Sci., 88(8), 1297–1300. Gonnert, G., Dube, S.K., Murty, T., and Siefert, W. (2001). Global Storm Surges; Theory, Observations and Applications, German Coastal Engineering Research Council, Germany. Jayakumar, S., Ilangovan, D., Naik, K.A., Gowthaman, R., Tirodkar, G., Naik, G.N., Ganeshan, P., Mani Murali, R., Michael, G.S., Ramana, M.V., and Bhattacharya, G.C. (2005). Run up heights and inundation limits along south east coast of India during the 26th December 2004 Indian Ocean tsunami. Curr.Sci., 88(11), 1741–1743. Krishnamoorthy, K., Harichandra Kumar, K.T., Krishna Kumari, A., and Das, P.K. (2005). Years of life lost and productivity loss due to tsunami in India. Curr.Sci., 89(5), 739–740. Murthy, K.S.R. (2005). First oceanographic expedition to survey the impact of the Sumatra earth quake and the tsunami of 26 December 2004. Curr.Sci., 88(7), 1038–1039. Murthy, T.V.R ., Sadhuram, Y., Rao, M.M.M., Rao, B.P., Sarma, V.V., Rao, K.M., Surya Prakash, S., and Chandramouli, P. (2005). Identification and modeling of internal waves, NIO Regional Centre, Visakhapatnam. Murty, T.S. (2005). Personal communication. Ramana Murthy, M.V., Sundaramoorthy, S., Pai, Y., Ranga Rao, V., Mishra, P., Bhat, M., Usha, T., Venkatesan, R., and Subramanian, B.R. (2005). Inundation of sea water in Andaman and Nicobar Islands and parts of Tamil Nadu coast during 2004 Sumatra tsunami. Curr.Sci., 88(11), 1736–1740. Sadhuram, Y., Rao, B.P., Rao, D.P., Shastri, P.N.M., and Subrahmanyam, M.V., (2004). Seasonal variability of cyclone heat potential in the Bay of Bengal. Nat. Hazards, 32, 191–209. Sadhuram, Y. (2005). Tsunami of 26 December 2004. Curr.Sci., 88(10), 1530–1531. Shetye, S.R., Gouveia, A.D., and Pathak, M.C. (1990). Vulnerability of the Indian coastal region to damage from the sea level rise. Curr.Sci., 59(3), 152–156.
CHAPTER 31
Tsunamis and Marine Life
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
D.V. Subba Rao ERD, Bedford Institute of Oceanography, Dartmouth, Nova Scotia, Canada B. Ingole Biological Oceanographic Division, National Institute of Oceanography, Dona Paula, Goa, India D. Tang, B. Satyanarayana, and H. Zhao Key Laboratory for Tropical Marine Environmental Dynamics (LED), South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou, PR China
31.1
INTRODUCTION
Submarine volcanic eruptions generate tsunamis which remind us of the power of nature, and the fragility of nature. In a sense, the tsunami as a natural disaster is analogous to Hiroshima atomic bomb blast. In both the cases, the destructive impulse is of short duration, but its consequences are felt for decades. The impact of tsunamis on life and property per unit time could be greater and more long-lasting even than those of longer duration such as pollution due to industries and sewage, beach erosion, and sea level variations (Figure 31.1, Krishna, 2005). For example surveys in French Polynesia have revealed plausible associations of outbreaks of ciguatera poisoning with disturbances to live corals caused by hurricanes and tsunamis (Bagnis, 1994). Unfortunately their ephemeral and unpredictable nature precludes any lead time for any planned investigation on the impact of tsunamis in a specific geographic region. The 26 December 2004 tsunami in the Indian Ocean exerted far reaching temporal and spatial impacts on marine biota. Our synthesis was based on satellite data acquired by the Laboratory for Tropical Marine Environmental Dynamics (LED) of the South China Sea Institute of Oceanology, China, near-shore as well as deep-sea observations by the National Institute of Oceanography (NIO), India, National Institute of Ocean Technology, India (NIOT), Central Institute of Brackish Water Aquaculture (CIBA, 2005), India, augmented by observations made by agencies in Sri Lanka and Indonesia. The tsunami impacted both the oceanic waters and the near-shore waters. The massive dislocation of sub-surface deep waters was similar to an upwelling, and was characterized by a decrease in sea surface temperature (SST) by about 1◦ C, increase in suspended particles, and increased nutrients which probably caused an increase in phytoplankton biomass to the northeast of Sumatra, and off Chennai. The time-series data for chlorophyll a compiled for 2000–2005 showed an increase in phytoplankton biomass (>0.35 chl a µg l−1 ) between midJanuary and February 2005 soon after the tsunami. Two weeks after the tsunami, in January 2005 a phytoplankton bloom developed with chlorophyll a (>0.5 µg l−1 ) in a 300 × 300 km area to the southeast of Sri Lanka and north of Aceh Province of Indonesia in the Andaman Sea. Similarly, in the near-shore waters near Chennai, a bloom dominated by the diatom Lauderia annulata developed and could be attributed to nutrient enrichment. The submarine land slides, and geomorphic changes resulted in extensive losses in coastal population, structures, mariculture 373
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
374
D.V. Subba Rao et al.
Figure 31.1.
Schematic representation of the duration of a natural hazard and its impact on life and property per unit time (based on Krishna, 2005).
operations, and coral reefs. A schematic analysis of the impacts of the tsunami on various habitats and biotopes is presented. 31.2
MATERIAL AND METHODS
Data to analyse the various aspects for this study were drawn from the following. 31.2.1 The oceanic region Study site: The sea surface area bound between 5S◦ –25◦ N and 75◦ –105◦ E (box-S in Figure 31.2(a), adjoining southeast Asian countries was investigated. To understand the (preand post-) consequences of earthquakes comparable with other years, an area of 6◦ × 6◦ (2◦ – 8◦ N and 90◦ –96◦ E) (box-T in Figure 31.2(b)) was examined further for time-series (daily and 8-day intervals) of SST between 2002 and 2005 (October–May). The LED of the South China Sea Institute of Oceanology, China, utilizing past and present satellite data, considered impacts of the December 2004 Tsunami on chlorophyll a (chl a), a measure of the algal biomass that sustains the rest of the marine populations. These data and analyses kindly made available, form part of the data base we have used in our analysis. 31.2.1.1 Satellite data Chlorophyll a: Data acquired from MODIS (Moderate-resolution Imaging Spectroradiometer; spatial resolution, 4 km) aboard the Aqua of Earth Observing System (EOS) was obtained from NASA’s Ocean Colour Group (USGS, 2004, 2005; http://oceancolor.gsfc.nasa.gov). The data processing was carried out through log-10 transformation and images were plotted using Matlab v. 6.5 and GrADS (Grid Analysis and Display System) v.1.8. SST: SST data were obtained from geophysical data sets derived from the observations collected by AMSR-E (Advanced Microwave Scanning Radiometer for EOS) instrument onboard NASA’s Aqua satellite. We used daily SST available at ftp://ftp.discover-earth.org/ sst/daily/amsre/ with a 0.25 × 0.25 degree grid. 31.2.2 The coastal region Additional data are available from NIO, Goa, India (NIO) based on their earlier cruises of Oceanographic Research Vessel Sagar Kanya in the Bay of Bengal and adjacent areas (Figure 31.3).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunamis and marine life
375
Figure 31.2.
(a) Study area (box-S) showing southeastAsian countries affected by tsunami on 26 December 2004. (b) Study area in detail. The epicentres of earthquakes are marked with star/numbers – 1 for the disaster on 26 December 2004 and 2 for 28 March 2005. Box-T is the region for which time-series data (daily and 8-day) of and SST were examined. Provinces of Indonesia labelled: Aceh, N.S: North Sumatra, S.S.: South Sumatra, W.S.: West Sumatra, Riau and Jambi.
Figure 31.3.
Cruise track and sample location map of the study area (based on Murthy, 2005).
D.V. Subba Rao et al.
376
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Besides the oceanographic observations during the Sagar Kanya 217 cruise conducted by NIO (Murty, 2005), the near-shore surveys conducted by various agencies revealed extensive ecological damage in the Andaman–Nicobar area. For near-shore studies, particularly for the coral reefs, photographs of flow depths and flow directions were used to assess the trail of destruction. To augment, rapid assessment of the damage to coral reefs in Sri Lanka by the National Aquatic Resources Research and Development Agency (NARA), reliance was placed on information analyses of the Coral Reef Degradation in the Indian Ocean (CORDIO), Sri Lanka Sub-Aqua Club (SLSAC), The World Conservation Union (IUCN), Global Coral Reef Monitoring Network (GCRMN). 31.3 31.3.1
RESULTS AND DISCUSSION Chlorophyll a in the oceans
Variations in spatial and temporal variability of phytoplankton biomass measured as chlorophyll a may occur as a result of large scale disturbances in the ocean circulation and sudden changes in the thermal structure of the water (Chaturvedi and Narain, 2003; Tang et al., 2002, 2004). During December 2004, chlorophyll a levels (Figure 31.4 (a)) increased and showed perceptible differences during the tsunami period. Phytoplankton biomass increased from the second week of December (Figure 31.4(b)), and continued throughout the third week (Figure 31.4(c)) before the tsunami event. Subsequently higher levels of chlorophyll were present in a larger area and attained a maximum in the fourth week (Figure 31.4(d)). However, on the northeast of Sumatra Island (circles in Figure 31.4(a,b)), the coastal waters adjacent to southeast Asian countries such as India, Sri Lanka, Myanmar, and Thailand (X and Y regions in Figure 31.2(d), Figure 31.3(d)) after the tsunami chlorophyll levels decreased abruptly. In general, the Bay of Bengal shows a greater enhancement of phytoplankton biomass during the summer monsoon (May–September) than in winter (November–February), owing to input of nutrients from rivers and the entrainment of sub-halocline layers through turbulent mixing (Ittekkot et al., 1991). However, chlorophyll levels were usually high during winter in inshore waters due to reverse Ekman transport towards onshore (Naidu et al., 1999). The coastal chlorophyll values off the southeast Asian countries (bordering Bay of Bengal) are usually high, but seem to be diluted in the wake of the giant tsunami waves (X and Y regions, Figure 31.4(d)).
Figure 31.4.
Spatial distribution of chlorophyll a (mg/m3 ) (derived from MODIS) in the Indian Ocean during December 2004. Position of the earthquake on 26 December 2004 was indicated with red star in Figure 31.4(d). The circles in Figure 31.4(a,b) show high on northeast of Sumatra Island. Regions X andY (Fig. 31.4(d)) represent changes in the coastal chlorophyll associated with tsunami waves. The interruption of satellite coverage (white patches over sea surface area) is due to cloud cover.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunamis and marine life
377
Note the presence of high levels of phytoplankton biomass even after the tsunami at the borders of Bangladesh, Myanmar, and Thailand (Y region in Figure 31.4(d)); these could be due to the propagation of stronger tsunami waves in east–west directions than in north–south (DOD, 2005). The observations of January 2005 (Figure 31.5) revealed the occurrence of a large phytoplankton bloom (300 × 300 km) with high chlorophyll values (>0.5 mg m3 ) southeast off Sri Lanka (Bay of Bengal), about 2 weeks after the tsunami (circle 1 in Figure 31.5(c)). Later, it developed into a long narrow band (Figure 31.5(d)) extending almost between Sri Lanka and Indonesia. There was also another bloom to the north of Aceh Province of Indonesia in the Andaman Sea (circle 2 in Figure 31.5(c)) that seemed to be progressing in an east-west direction. These features were probably associated with the nutrient enrichment from the land provided by the receding waves. Nevertheless, the chlorophyll levels decreased from February through March 2005, perhaps as a result of the onset of the spring–winter monsoon (March–May) (Kumar et al., 2002).
31.3.2
Chlorophyll a: time-series analysis
The annual variations of chlorophyll a for 2002 to 2005 (Figure 31.6) revealed substantial changes during December 2004 and March 2005. An analysis of the daily data for the tsunami period (2004–2005/October–May) (Figure 31.6(a)), showed significantly higher values before the earthquake (up to 0.3 mg m3 ) on 22 December 2004 (1 in Figure 31.6(a)). During the tsunami period the values decreased for 1 week, but subsequently increased (2 in Figure 31.6(a)). During March 2005, a similar trend was observed with two peaks of chlorophyll (3 and 4 in Figure 31.6(a) and line-T in Figure 31.6(a)) at the time of earthquake (28 March). Previous studies on earthquake and ocean productivity (Singh et al., 2002) also indicated a similar increase. The chlorophyll a data based on 8-day averages (Figure 31.6(b)) shows variations over the 3 years. Interestingly, the values are lower for the current period (October 2004–May 2005) than for the other 2 years, except during the tsunami. Between mid-January and February, chlorophyll a reached a peak (P in Figure 31.6(b)) during 2002–2003 and 2003–2004, but not during 2004– 2005 when the tsunami occurred. Figure 31.6(c) shows the correlation between chlorophyll a and SST. While the SST decreased from 30.1 to 28.7◦ C (4–27 December 2004), chlorophyll increased from 0.13 to 0.23 mg m3 ). Immediately after the earthquake, SST increased to 29.9◦ C (28 December 2004–4 January 2005) while the phytoplankton decreased to 0.13 mg chl a m3. The reasons for low chlorophyll levels for October 2004–May 2005 and for higher concentrations between January and February for both 2002–2003 and 2003–2004 (P in Figure 31.6(b)) are not known.
Figure 31.5.
Spatial distribution of chlorophyll a (mg/m3 ) during January 2005. Circles 1 and 2 are the phytoplankton blooms.
D.V. Subba Rao et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
378
Figure 31.6. Time-series data: (a) Daily concentrations of chlorophyll a derived from MODIS. Line-T passing through the bars is the trend line. Discontinuity of graph could be noticed due to missing values. Numbers 1, 2, 3, and 4 are the peaks of concentration during the events of earthquakes. (b) Comparison of (8-day average) among 3 years. P is the maximum encountered between mid of January and February during 2002–2003 and 2003–2004. (c) and SST during October 2004–May 2005. The dates of earthquakes are indicated with down arrows in all panels.
Tsunamis and marine life
379
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
31.3.3 Temperature Temperature and availability of nutrients are some of the important factors affecting the growth of phytoplankton (Dey and Singh, 2003; Tang et al., 2003, 2004). Compared to 20 December (Figure 31.7(a)), the daily SST images (Figure 31.7(b)) show that overall temperatures decreased by about 1◦ C in the southern area on 26 December 2004. This trend continued for nearly another week (Figure 31.7(c)). The coincidence of low (∼1◦ C) SST over an extensive area could be related to incursion of cold subsurface waters; any upwelled water would also enrich the surface layers (Singh et al., 2002; Dey et al., 2004; Ouzounov and Freund, 2004). From 1 January 2005 (Figure 31.4(d)), SST increased and attained levels comparable to the pre-tsunami event (Figure 31.7(a)). It is of interest to note that the SST on the west coast of northern Sumatra (close to epicentre of earthquake) was appreciably high (∼30–31◦ C) (straight arrows in Figure 31.7(a–d)), with a variation of 0.5–1.0◦ C on 26 December (straight arrow in Figure 31.7(b)). This may be due to the release of thermal energy from the point of earthquake (Ouzounov and Freund, 2004). On the other hand, the extent of this area with high SST seems to have diminished (Figure 31.5(b)) probably because of subsequent incursion of sub-surface cold waters generated by the tsunami. Further lowering of temperature (27–28◦ C) was noticed from northeast to the southeast coast of India in Bay of Bengal (down arrow in Figure 31.5(c)). A decrease in SST by about 1◦ C on the day of tsunami was also reported from the Andaman and Nicobar Islands, India (DOD, 2005). 31.3.4
Near-shore studies
The impact of a tsunami on life and property per unit time can be far greater and long-lasting than threats of a longer duration such as pollution, sewage, beach erosion, and sea level variations (Figure 31.1). Unfortunately their ephemeral and unpredictable nature precludes any lead time for any planned investigation on their impact. 31.3.5 Tsunamis: upwelling effect In the coastal waters of the Andamans and Nicobars, there was a decrease up to 1◦ C (DOD, 2005) coinciding with occurrence of the tsunami. On the east coast of India, off the Chennai and Nellore coasts, and off the Andaman Islands following the tsunami event the suspended sediment concentration (SSC) increased from 9–21 mg/m3 on 25 December 2004 to 4–36 mg/m3 on 27 December 2004. Also the zone of high SSC extended from 50 m depth (located in the 15 km) to 1000 m depth (in the 45 km). Off Chennai, the total nitrogen before the tsunami was 90.6 µm
Figure 31.7.
SST in the period of tsunami. The position of earthquake was indicated with red star in Figure 31.7(b). Downward arrow (Fig. 31.7(c)) denotes the gradient of lowering temperature from northeast to the southeast coast of India in Bay of Bengal. Variation of temperature close to the epicentre of earthquake was pointed with straight arrows.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
380
D.V. Subba Rao et al.
that increased to 165 µm after the tsunami. In the coastal waters off Kanyakumari, and Aleppey phosphate levels before the tsunami were 0.30 and 0.60 µm respectively; corresponding values after the tsunami were 1.58 and 1.96 µm. The waters off Vizhinjam registered an increase in the NO2 from 0.04 to 1.69 µm, NO3 from 1.29 to 4.25 µm, SiO4 from 1.30 to 2.52 µm, and PO4 from 0.19 to 1.86 µm. The nutrients at best may act as a pulse and result in rapid growth of certain marine phytoplankton. For example, off the Chennai and Nellore coasts the chlorophyll level was 0.1 µg l 1 on December 25, 2004 and increased to 0.3 µg l 1 on 31 December 2004 associated with blooms of the diatom Lauderia. Pre-tsunami concentrations of this diatom were 1.05 × 103 l−1 and post-tsunami levels were 3.22 × 103 l−1 and could be attributed to nutrient enrichment of these waters. Periodic inundation or exposure to air for few hours is a common experience of the intertidal biota and therefore exposure to short-term cataclysmic changes in temperature and salinity may not seriously affect them. The benthos and interstitial fauna and flora may be temporarily dislodged but they are adapted to that sort of environmental perturbation. Of interest is the increase in density of benthic fauna off Chennai from the usually low 1–9 individuals 0.08 m2 and 15 species to ∼146/0.08 m2 and 35 species after the tsunami. First time occurrence of the polychaete Polydontes melanonotes (∼2500 individuals/0.08 m2 ) off Pondicherry could be probably due to temporary dislodgement and redistribution (DOD, 2005). 31.3.6 Subsidence and submergence Comparisons of the pre- and post-tsunami data showed considerable physical changes. These include land subsidence up to 3 m (Indira point near Great Nicobar), sea level rise (1.3 m near Campbell Bay), sea level retreat (0.8–1.0 m Aerial Bay-Diglipur), land uplifts (1.2 m near Ariel jetty–North Andaman Island). Comparison of pre-and post-tsunami data shows the overall situation of island submergence near Katchal Island (Andaman–Nicobar area) with effects of local erosion and deposition particularly along the western coastline. Major submergence of land is noted south of Lamoh–a creek basin that had existed earlier was totally submerged. Coastline recession also took place along the axis of the basin to an extent of 3.2 km and a width of 2.5 km. These underwater alterations would cause an up thrust of coral beds and rock strata in the Middle Andaman Island coupled with submergence of coral beds, beach and forest area in Southern Andaman Island (Figure 31.8). Out of the total area of the Nancowry group of islands, 16% has undergone major changes (Ramachandran et al., 2005). Nearly 7% of the area is submerged and 9% heavily damaged. Maximum change (more then 42%) has been observed in Trinkat Island. More than 23% of Katchal and 9% Camorta and Nancowry and Tarasa group have undergone major changes. About 95% of the mangroves in Katchal Island and 52% inTrinkat Island, 43% in Camorta and Nancowry Islands have been submerged in deep water. In addition to the mangrove forests, >62% of littoral forests in Katchal Island have been submerged (27%) or converted to a sandy area (36%). About 13% of evergreen forest in the Trinkat, 5% in Tarasa/Chowra groups, and 3% in the Tillangchong have suffered severe to moderate damage (NRSA, 2005). 31.3.7
Run-up
During a tsunami sea water penetrates the coast at a high speed and causes extensive inundation, which is called run-up. Run-up is usually expressed in metres above normal tide or mean sea level. A detailed tsunami run-up survey was carried out along a 350 km long stretch of the shoreline in Tamil Nadu from Pulicat (13.3◦ lat.N) in the north to Vedaraniyam (10.3◦ lat.N) in the south by the National Geophysical Research Institute, India during 7–11 January 2005. Normal shore profiles were studied to estimate the run-up heights of the tsunami (Ram Mohan, 2005). The runup heights ranged from 1 to 10 m and the inundation varied between 50 and 2000 m. Maximum
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunamis and marine life
Figure 31.8.
381
Submerged coral beds, beach and forest area along the eastern coast of Southern Andaman Island near Baratang Island.
surge elevations were also measured and were found to vary between 3.8 and 6.0 m (mean tidal level) (Chadha et al., 2005). Tsunami sand deposits ranged from coarse upper (700–1000 mm) to very fine upper (88–125 mm) in grain size, based on comparisons with grain-size cards. On open coast beaches, bays and harbours, as a result of shoaling effects, the wave increases caused rises in sea level as high as 30 m at the coast. Historical run-up level data for the Andaman and Nicobar Islands and the Tamil Nadu coast indicate 4 m in Port Blair during 1868, 0.76 m in Car Nicobar during 1881 and 1.22 m in Port Blair and Nagapattinam in1881 (http:\\www.ngdc.noaa.gov). The Andaman and Nicobar Islands located in the subduction zone of Burma Plate are classified as seismic zone 5, indicating a high level of risk from earthquakes. Run-up measurements at different sites along the Andaman and Nicobar and Tamil Nadu coasts were made using Real Time Kinematic Global Positioning System (RTKGPS) during January–February 2005 by the Scientists of ICMAM-PD, Department of Ocean Development (DOD) in association with Andaman and Nicobar Centre for Ocean Science Technology, NIOT, Port Blair and Institute for Ocean Management, Anna University, Chennai are summarized in Table 31.1 (Ramanamurthy et al., 2005).
382
D.V. Subba Rao et al. Table 31.1.
Satellite based data on inundation of seawater in Nicobar Islands during tsunami.
Island
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Trinkat Camorta Nancowry Katchal Little Nicobar Great Nicobar
Figure 31.9.
Population (2001census)
Geographical area (ha)
Villages affected
Area affected (ha)
432 3412 927 5312 353 7566
3630 18820 6690 17440 15910 104510
4 20 9 13 20 29
360 665 26 1432 235 993
Map showing Malacca in Car Nicobar and settlement areas close to coast.
In Andaman and Nicobar area, in the north and south Andaman group of islands, the run-up levels varied between 1.5 and 4.5 m and the distance penetrated from the coast ranged from 100 to 250 m (Table 31.1). Little Andaman recorded a run-up of 5 m, with a penetration of 1200 m. In the two Nicobar Islands, the run-up levels varied from 3 to 7 m, with penetration ranging from 50 to 1000 m and with higher run-up levels and longer penetrations noted in Car Nicobar. The National Remote Sensing Agency (NRSA) through IRS-1D LISS III image and other satellites has estimated the area affected due to tsunami (area inundated in a few Nicobar Islands (Figure 31.9 and Table 31.1). The satellite-based data indicate that Great Nicobar, Trinkat, Katchal, and Camorta were the worst affected (Table 31.1). The run-up levels along the Tamil Nadu coast, showed trends very similar to those in the Andaman and Nicobar Islands. The worst affected area (Nagapattinam) showed penetration of sea water (750 m) up to an elevation of 3.9 m because of the gentle slope of the coastal land combined with the effect of tsunami wave diffraction caused by the northern tip of Sri Lanka. Suganthi Devadason Marine Research Institute (SDMRI, 2005), while making an assessment of the shoreline changes along the Tamil Nadu coast has reported that at Chennai that an approximately 5 m width of the shoreline of Adayar disappeared. The mouth of Adayar creek, which had no connection with the sea before the tsunami, was connected with the sea after tsunami. In some stretches of Chennai coast, the regression has been in the range of 5–50 m (Figure 31.10). 31.3.8
Impact on coastal ecosystems
The coastal ecosystem generally experiences frequent perturbations, which affect the structure of the ecological communities by removing established species and allowing fugitive species to colonize the disturbed area. Geomorphological changes, inundation by oceanic water, submarine land slides resulting from tsunamis impact the shoreline, pelagic and sea bed habitats (Figure 31.11).
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunamis and marine life
Figure 31.10.
Returning wave of tsunami at Chennai, India.
Figure 31.11.
Schematic representation of impact of tsunami on the various marine biotopes.
383
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
384
D.V. Subba Rao et al.
Populations ranging from meiofauna to large aquatic mammals inhabiting the backwaters, salt marshes, tidal waters, mangroves, coral reefs, sandy shores, rocky shores, mud flats, and shoreline structures will be adversely affected. In the coastal waters off Pondicherry the sewage mix extended even up to 10 km evident from growth of Escherichia coli, faecal coli forms, and Salmonella. In the Marina beach, Chennai, vertical and horizontal displacement of meiofauna was reported (Altaff et al., 2005). Prior to the tsunami, during December meiofauna numbers were 2175 ± 336/10 cm2 but decreased to 744 ± 14/10 cm2 (5% significant level) a day after the tsunami. Two days after tsunami the numbers increased to 6028 ± 932/10 cm2 . Concomitantly there were changes in the biodiversity. Harpacticoids, nematodes, oligochaetes, gastrotrichs, and polychaetes were dominant in the meiofauna prior to tsunami but by the third and fourth day after the tsunami changed to foraminiferans, cnidarians, nemertines, gastrotrichs, rotifers, kinorhynchs, ostracods, isopods, halacarids, and insects (Altaff et al., 2005). The following are some of the possible impacts of tsunami on coastal and marine ecosystems. The physical structure damaged or removed by the force of the wave itself, can result in physical removal of flora and fauna; the increased sediment load kills sediment sensitive corals and sea grasses by smothering. The extent of this damage will likely vary considerably depending on the local topography and hydrology. Chemical changes have included saltwater intrusion, eutrophication (enrichment) of the water resulting from increased runoff, raw sewage, and decomposition of flora and fauna including unrecovered bodies. Additionally there will be the decomposition of timber from mangroves, fishing boats, and buildings. Non-biodegradable wastes such as plastics washed from the shore have contributed to a build-up in marine debris during the tsunami. This may in the long-run propagate settlement of bryozoans, barnacles, polychaete worms, hydroids, and molluscs (Barnes, 2002) and may cause introduction of non-indigenous species into the coastal Indian Seas (Subba Rao, 2005). Environmental damage to the inter-tidal and sub-tidal areas appears to be extensive. This includes drastic changes in the health of coastal marine ecosystems, with potentially irreversible destruction of some areas, as well as the immediate loss of living coastal resources such as fish, lobsters, and crabs. This will have serious implications for fisheries. Many coral reefs have lost both their structure and biota, and are now reduced to rubble due to mechanical damage. The force of the tsunami can move enormous boulders and sections of reef, as well as thousands of tons of smaller fragments, sand and silt, which dislodge, crush and kill marine biota. Further, there is significant contamination by run-off from land, with large quantities of wastes and pollutants, debris, soil, and organic matter. This will further increase the damage and hamper recovery. Similarly devastation of vast stretches of coral reef beds, mangroves, sea grass, and sea weed beds which act as feeding and nursery grounds for a myriad species of finfish and shellfish species would affect the capture and culture fishery for long. Increased turbidity in the wake of the tsunami will smother large areas; clogging of gills and respiratory tracks by silt may kill many organisms that may have survived the wave itself (e.g. on reefs and sea grass beds). The consequences are serious and may have lasting effects. In addition to these immediate effects, there are a number of long-term implications. The effect of loss of breeding fish populations, habitat and nursery grounds would have severe implications for nutrition and livelihoods among coastal populations over years to come. Nesting beaches for the five species of globally threatened marine turtles in the Andaman and Nicobar Islands region are reported to be damaged. The tsunami has struck ecosystems that in many cases are already stressed by unsustainable resource use, such as over fishing, and habitat destruction, including development or indiscriminate cutting down of mangroves for various developmental activities. In the sub-arctic on the mainland Norwegian coast at Kvennavatnet, because of a catastrophic natural disaster such as the deglaciation and land upliftment, it took between 80 and 120 years before a comparatively diverse invertebrate community developed (Solem et al., 1997). But in
Tsunamis and marine life
385
the tropics where the biological wheels turn faster unlike in the sub-arctic, recovery of the marine environment from the tsunami impacts could be faster and this needs be monitored.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
31.3.9
Impact on mangrove forests and swamps
The mangrove swamps of the Andaman and Nicobar constitute about 18% mangrove forests of the islands. Mangroves facing sea seem to have undergone severe destruction compared to those along the creeks. Many mangrove areas remained submerged for a long time resulting in their destruction. In the Andaman Islands, about 3850 ha mangroves were totally lost, while 7750 ha were damaged. In Nicobar Islands, viz. Camorta, Katchal, Nancowry, and Trinkat Islands, located between 7◦ 51 50 – 8◦ 14 43 , and 93◦ 17 41 and 93◦ 37 30 E, the impact was very severe (Ramachandran et al., 2005); about 3900 ha (out of 7000 ha) area mangroves were damaged. Totally lost mangrove areas were 377 ha in the Car Nicobar, 102 ha in the Tarasa and Bompoka and 376 ha in the Katchal Islands. Land-cover estimates of the four islands showed that the mangrove areas were affected; 336 ha (51%) in Camorta, 153 ha (100%) in Nancowry, and 240 ha (68%) in Trinkat (Ramachandran et al., 2005). Mangroves of the Tamil Nadu coast and the Kerala coast were marginally affected (SAC, 2005). Such a major damage in mangrove area will not only affect the coastal productivity and destabilize coastal areas, but will accelerate shoreline erosion. The impact of tsunamis on mangrove forests will be severe in these islands and as the natural regeneration of mangroves is slow (10–15 years) their total recovery may take a few decades. Recent studies on the mangroves of Andaman and Nicobar Islands revealed that majority of the mangrove species were impacted heavily either by exposure or by submergence (Table 31.2). Since the rise and fall in water level due to tides is a significant factor it influences salinity, sedimentation, geomorphology, and nutrient cycles in the mangrove environment. Consequently many of the mangrove species may not recover from changes in the salinity regime (especially those adjusted to low saline conditions). They may also find it difficult to adapt to the permanent change in the submergence level resulting from the subduction of land. It may be possible to find temporal succession in various mangrove zones eventually. Some of the species like Avicennia and Sonneratia would thrive well in the proximal zone; in the middle zone, where spring tide does not reach the mangrove plants, it has been postulated that Bruguiera sp. would replace Rhizophora sp., as they are better adapted to the dry conditions (Roy and Krishnan, 2005). Mangroves mostly grow in the intertidal zone and each mangrove appears in certain specific tide levels. Sea level changes, past and present, influence the geomorphic character of a locality and every locality has its own sea level history resulting from global and eustatic changes of sea level. Changes in sea level cause shifting of mangrove zone seaward or land-ward. The mangrove ecosystem itself gives some evidence of past sea level changes. In general Mangroves favour low-wave energy zones as higher waves can cause damage to them; hence they are found in sheltered areas. 31.3.10
Coral reefs
Corals act as natural breakwater and mangroves as natural shock absorbers. Protection of these natural barriers from avoidable destruction is central to the conservation of coastal ecosystems. In Andaman and Nicobar Islands, the Coral reefs have suffered two kinds of damages: (1) the reefs were either completely eroded and/or (2) sand, mud, detritus, was deposited on them. The giant tsunami waves smashed and crushed the reefs, while the backwash deposited the debris. The deposition of such material kills the live coral colonies. It has been estimated from satellite imagery (SAC, 2005) that the Andaman and Nicobar Islands lost reefs of about 23,000 and 17,000 ha, respectively during the 2004 tsunami. About 6740 ha of reef in the Andaman Islands and 6140 ha of reef in the Nicobar Islands is now covered by sand, mud, or detritus. The reefs of the Car Nicobar, Comorta, Nancowry, Trinkat, Katchal, Tilakchang, and Little Nicobar have
386
D.V. Subba Rao et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Table 31.2.
Impact of tsunami on mangrove stands of Andamans (based on Roy and Krishnan, 2005).
Mangrove stand (Study sites)
Dominant species
Mangrove observation
South Andaman Sipighat junction
Rhizophora mucronata
Minnie Bay
R. apiculata Avicenia marina R. mucronata
80% of Rhizophora spp. were affected in low saline microhabitat; death of mangroves in south. Andaman is due to continuous subduction. Thriving. 40% of Rhizophora spp. was affected and most of them died. R. apiculata continuous inundation. Thriving. 30% of Rizophora spp. was affected and most of them died due. Not affected.
Chouldari
Wandoor
Middle Andaman Uttara
Rangat Rangat Bay
North Andaman Mayabunder Karmatang, Danapur, Dobidehra, Baludehra Diglipur Durgapur Arial Bay
R. apiculata A. marina R. apiculata, R. mucronata A. marina and Sonneratia alba R. apiculata, R.mucronata A. marina
30% of Rizophora spp. was affected in low saline microhabitat and most of them died. Not affected.
R. apiculata C. tagal A. marina R. mucronata A. marina A. officinalis Exoecaria agallocha A. marina R. apiculata R. mucronata
In middle Andamans all mangroves are healthy and unaffected.
R. apiculata R. mucronata
In north Andamans all mangroves are healthy and unaffected.
A. marina A. marina R. apiculata R. mucronata R. apiculata R. mucronata A. marina
been totally lost while those of the Gulf of Mannar have been affected marginally. On the day of the tsunami and after, extensive silt-laden turbid waters were seen all over the reef area; these turbid waters were mapped using satellite imageries and they were observed to cover an area of 401 ha in Nancowry and about 552 ha in Trinkat. The effect decreased after 10 days, but silt and mud were found to be deposited on the reef area (SAC, 2005; Ramachandran et al., 2005). The increase in the extent of sandy beaches after the tsunami was 18.7 ha in Trinkat and 1242 ha in Katchal, whereas decrease in sand cover was witnessed in Camorta 369 ha (103.43%) and Nancowry 79 ha (31%) (SAC, 2005). The underwater surveys carried out by the Center for Marine and Coastal Studies of Madurai Kamaraj University, using Line Intercept Transect (LIT) method revealed no appreciable changes
Tsunamis and marine life Table 31.3.
Summary of observations made by dives on coral reefs. Date December 2004
Dive site
Damaged
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Thai Island 3 5 7
30 29 27–29
8
27
9
29 27
Koh Bon Koh Tachai Richelieu Rock Mergui Archipelago Phi Phi Island Thailand Maldives Sri Lanka
387
29 28 Westridge Whole area Whole area
Shark Fin Reef Reef East of Eden Atlantis wreck Beacon Reef Elephant Head Breakfast Bend Snapper Alley SE Christmas Point North Point Westridge 28 31 Continual
Not damaged X
5–10% in the north More damage in the south to hard corals 15% damaged Most damage Major damage 80% hard corals and sea fans 20% hard corals Some damage
X
X
Minimal damage Unchanged Unchanged Loads of fish
Alistair Beveridge Debris could damage marine life
Intact coral systems Dead fish strewn on Mangroves protected the beach, coral reef coral reefs
in the bio-physical status of corals in the Gulf of Mannar. Studies carried out by the SDMRI, Tuticorin, India have also revealed that in the Gulf of Mannar, there was no major impact on the fish community and common reef fishes were abundant (SDMRI, 2005). Fragments of seaweed and sea grass had been washed ashore by the strong waves, and in some areas (Keezhakkarai group) the fragments were entangled with branching corals (Acropora sp.). There was, however, no deposition of sand and debris on the seaweed and sea grass beds (SDMRI, 2005). Further north there was no impact on Pulicat Lake ecosystem. The estuaries such as Vellar (Cuddalore), Manakudi (Kanyakumari) suffered sand deposition. The mouth of Manakudi estuary was completely closed by sediment deposition. Other estuaries along the east coast such as Cauvery, Coleroon, Muthupet salt swamp, Vedaraniyam backwaters, Vellar estuary, Killai backwaters, Kaazuveli backwaters, Palar estuary, Kovelong backwaters, Adyar estuary, Cooum, Ennore creek, and Pulicat lake areas remained open immediately after tsunami (SDMRI, 2005; DOD, 2005) but they became heavily silted and closed as a result of the high rate of sedimentation following the tsunami. In the absence of even normal south-west monsoon rains, there is every possibility that they will not have any connection with Bay of Bengal till the onset of north-east monsoon. The absence of such a connection to these brackish water systems with the Bay of Bengal will impair the recruitment of anadromous shrimp and fish species which depend on these ecosystems for breeding, feeding, and further growth. Observations made by divers (Table 31.3) revealed that the impact of the tsunami varied; for example the damage to reefs off Thailand, that is Shark Fin reef, Atlantis wreck, Beacon
D.V. Subba Rao et al.
388
Table 31.4.
Impact of tsunami on aquaculture in peninsular India.
Location
Impact
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Andhra Cuddalore Villupuram Nagapattinam Pudukottai Ramanathapuram Tuticorin Kanyakumari Pondicherry Karaikal Thiruvananthapuram Kollam Alleppey Ernakulam Mallapuram Kannur Kasargode
4229 shrimp farms + 53 hatcheries, 4 feed mills, 4 processing plants and 1 integrated unit – unaffected 10 hectares affected 6 shrimp hatcheries damaged Of the 996 farms the total loss Rs 60 million No damage Minor damages 16 hectares shrimp farm suffered minor damages Government shrimp farm severely damaged No serious damage A few farms reported damaged No damage Pump-house damaged; death of larvae loss Rs 3 million 3 shrimp hatcheries damaged loss Rs 5.0 million 24 shrimp farms damaged. Rs 8.73 million No damage One hatchery damaged loss Rs 6 million Another suffered a loss of Rs 9.10 million Rope culture of mussel. 100 units suffered loss
Note: Rs one million is ∼ US $ 20,000.
Reef was negligible while reefs off Breakfast Bend, Snapper Alley SE suffered very extensive damages comparable to those in Andaman Nicobar Islands. Reefs in Maldives and Sri Lanka were protected. 31.4 AQUACULTURE The aquaculture units that lie close to the shore zone, were badly affected by the tsunami: the farms were damaged significantly; dykes collapsed; sluice gates sank; pipelines (drains and feeders) were covered with silt and sand; cultured shrimp (whatever present) were either dead or drained into the sea (Table 31.4). Food and Agriculture Organization (FAO) estimated that marine exports from India may decline by as much as 30%, as a consequence of the impact of tsunami in south-east Asian region. The fishermen living along the 1000 km of Indian coastline were the worst hit by tsunami (FAO, 2005). 31.5 31.5.1
OTHER DAMAGES Short-term damage
Flooding with flow velocities approximating 10 m/s cause potentially irreversible damages. Consistent with Plafker’s rule (Borrero, 2005) which states that usually the maximum run-up is less than twice the maximum seafloor displacement at the southern end of Lhoknga, Banda Aceh, runup reached 31 m; the wave scour and subsidence shifted the shoreline 1.5 km inland, and flooded 65 km2 of land between Banda Aceh and Lhoknga (Borrero, 2005). The resulting damages include: • Reduction to rubble resulting from mechanical damage.
Tsunamis and marine life
389
• Increase in turbidity – smothers plant growth and suffocated animals. • Exposure to receding sea level causes mortality of fish inhabiting coral reefs. 31.5.2
Long-term damages
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Damages under this category are more extensive and need a long recovery time. They include: • Shorelines may have changed. • Mangroves forests may have been wiped out. • Livelihood of coastal populations and ecotourism industry affected. For example in Yala, Sri Lanka, following removal of natural coastal protection, an 8 m tsunami run-up destroyed a hotel and affected tourism (Liu et al., 2005). • Funnelling of water inland, leading to floods as in Hambanthota town in southern Sri Lanka. • Coastal industrial establishments such as Nuclear plants (Jayaraman 2005), desalination plants, thermal power plants, and sewage plants may be affected. 31.6
RECONSTRUCTION AND MANAGEMENT
To some extent mangrove forests and coastal green belts reduced the damage to coast lines. Accordingly afforestation through conservation and replanting coastal mangrove vegetation is advocated to provide “bioshields” to lessen the tsunami impacts. These, besides acting as a better shield to attenuate the impact of tsunamis than would seawalls, would also provide a habitat for migratory birds, turtles, manatees, dolphins, and fish. The desirability of promoting partnerships with other regional governments to set up technical coordination for effective prognostication of tsunamis and periodic environmental assessment is emphasized. Institution of a global multihazard tsunami warning system that is fully integrated with an operational ocean-observing system along the lines of Global Sea Level Observing System (GLOSS) and its component Global Ocean Observing System (GOOS) would provide an early warning system and should be the topmost priority. ACKNOWLEDGEMENTS We are grateful to Dr James E. Stewart, Bedford Institute of Oceanography for constructive comments on the manuscript. We thank Drs T.S. Murty and U. Aswathnarayana, Editors of this volume for inviting us to contribute this chapter. The present work was supported by the National Natural Science Foundation of China (NNSFC2006), Chinese Academy of Sciences (CAS) Research Fund (KZCX3-SW-227-3), and “One Hundred Talents Program” of CAS awarded to Prof. DL TANG. A post-doctoral research fund allowed Dr B. Satyanarayana to contribute to this work. Mr Zhang MG and Lv JH of SCSIO, CAS, have contributed to some aspects of this study. We also thank Dr S.R. Shetye, Director of NIO, Goa, India and Dr P.A. Lokh Bharathi (NIO) for their encouragement in preparation of this manuscript. For formatting and graphics we thank Mr Art Cosgrove and Mrs Bala T. Durvasula. REFERENCES Altaff, K., Sugumaran, J., and Naveed, Md.S. (2005). Impact of tsunami on meiofauna of Marine beach, Chennai, India. Curr. Sci., 89(1), 34–38. Bagnis, R. (1994). Natural versus anthropogenic disturbances to coral reefs-comparison in epidemiological patterns of ciguatera. Mem. Queensl. Mus., 34, 455–460. Barnes, D.K.A. (2002). Invasions by marine life on plastic debris. Nature, 416, 808–809.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
390
D.V. Subba Rao et al.
Borrero, J.C. (2005). Field data and satellite imagery of tsunami effects in Banda Aceh. Science, 308, 1596. Chadha, R.K., Latha, G., Yeh, H., Peterson, C., and Katada, T. (2005). The tsunami of the great Sumatra earthquake of M 9.0 on 26 December 2004 – impact on the east coast of India. Curr. Sci., 88(8), 1297–1301. Chaturvedi, N., and Narain, A. (2003). Chlorophyll distribution pattern in the Arabia Sea: seasonal and regional variability, as observed from SeaWiFS data. Int. J. Remote Sens., 24, 511–518. CIBA (2005). Report on Assessment of Loss Due to Tsunami to Brackish water Aquaculture and Fisheries Sectors in Coastal States of Andhra Pradesh, Tamil Nadu and Kerala. 37 pp. Dey, S., and Singh, R.P. (2003). Surface latent heat flux as an earthquake precursor. Nat. Hazards Earth System Sci., 3, 1–7. Dey, S., Sarkar, S., and Singh, R.P. (2004). Anomalous changes in column water vapour after Gujarat earthquake. Adv. Space Res., 33, 274–278. DOD (2005). Preliminary Assessment of Impact of Tsunami in Selected Coastal Areas of India. Report, 42 pp., Compiled by Department of Ocean Development (DOD), ICMAM Project Directorate, Chennai, India. FAO (2005). Food and Agricultural Organization 2005 (http://www.fao.org). Ittekkot, V., Nair, R.R., Honjo, S., Ramaswamy, V., Barstch M., Manganini, S., and Desai, B.N. (1991). Enhanced particle fluxes in Bay of Bengal induced by injection of fresh water. Nature, 351, 385–387. Jayaraman, K.S. (2005). India’s nuclear debate hots up after tsunami floods reactor. Nature, 433, 675. Krishna, S.K. (2005). Science plan for coastal hazard preparations. Curr. Sci., 89(8), 1339–1347. Kumar, S.P., Muraleedharan, P.M., Prasad, T.G., Gauns, M., Ramaiah, N., de Souza, S.N., Sardesai, S., and Madhupratap, M. (2002). Why is the Bay of Bengal less productive during summer monsoon compared to the Arabian Sea? Geophys. Res. Lett., 29(D24), 2235, doi: 10.1029/2002GL016013. Liu, P.L.F., Lynette, P., Fernando, H., Jaffe, B.E., Fritz, H., Higman, B., Morton, R., Goff, J., and Synolakis. C. (2005). Observations by the International Tsunami Survey Team in Sri Lanka. Science, 308, 1595. Murthy, K.S.R. (2005). First Oceanographic expedition to survey the impact of the Sumatra earthquake and the tsunami of 26 December 2004. Curr. Sci., 88(10), 1038–1039. Naidu, P.D., Ramesh Kumar, M.R., and Ramesh Babu, V. (1999). Time and space variations of monsoonal upwelling along the west and east coasts of India. Cont. Shelf Res., 19, 559–572. NRSA (2005). Rapid Assessment of Damage due to Tsunami in Nancowry group of Islands (Andaman & Nicobar Islands) using Satellite data by National Remote Sensing Agency, Hyderabad. 5 pp (personal communication). Ouzounov, D. and Freund, F. (2004). Mid-infrared emission prior to strong earthquakes analyzed by remote sensing data. Adv. Space Res., 33, 268–273. Ram Mohan, V. (2005). Mapping of areas of inundation – Marakkanam to Kovalam. pp. 4–8. In: N. Chandrasekar. (ed.) 2005 Mid term review of the tsunami projects. Ramachandran, S., Anitha, S., Balamurugan, V., Dharanirajan, K., Ezhil Vendhan, K., Divien, M.I.P., Vel, A.S., Hussain I.S. and Udayaraj, A. (2005). Ecological impact of tsunami on Nicobar Islands (Camorta, Katchal, Nancowry and Trinkat). Curr. Sci., 89(1), 195–200. Ramanamurthy, M.V., Sundaramurthy, S., Pari, Y., Ranga Rao, V., Mishra, P. Bhat, M. Tune Usha, Venkatesan, R., and Subramanian, B.R. (2005). Inundation of sea water in Andaman and Nicobar Islands and parts of Tamil Nadu coast during 2004 Sumatra tsunami. Curr. Sci., 88(11), 1736–1740. Roy, D.S. and Krishnan, P. (2005). Mangrove stands of Andamans vis-à-vis tsunami. Curr. Sci., 89(11), 1800–1804. SAC (2005). Assessment of damages to coastal ecosystems due to the recent tsunami – Summary report – Prepared by SAC, Ahmedabad for MOEF April 2005. 36 pp. SDMRI (2005). Rapid environmental impact assessment after tsunami in the inter-tidal and sub-tidal and coastal areas including water bodies and lakes along Tamil Nadu coast. Final Report. (Funded by Department of Environment, Tamil Nadu Government.) 30 pp. Singh, R.P., Bhoi, S., and Sahoo, A.K. (2002). Changes observed on land and ocean after Gujarat earthquake 26 January 2001 using IRS data. Int. J. Remote Sen., 23, 3123–3128. Solem, I.O., Solem, T., Aagaard K., and Hansen, K. (1997). Colonization and evolution of lakes on the central Norwegian coast following deglaciation and land uplift 9500 to 7800 years B.P. J. Paleolimnol., 18, 269–281.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunamis and marine life
391
Stein, S. and Okal, E.A. (2005). Ultra-long period seismic moment of the great December 26, 2004, Sumatra earthquake and implications for the slip process. Unpublished report. Subba Rao, D.V. (2005). Comprehensive review of the records of the biota of the Indian Seas and introduction of non-indigenous species. Aquatic Conserv: Mar. Freshw.Ecosyst., 15, 117–146. Tang, D.L., Kawamura, H., and Luis, A.J. (2002). Short-term variability of phytoplankton blooms associated with a cold eddy on the north-western Arabian Sea. Remote Sens. Environ., 81, 81–89. Tang, D.L., Kawamura, H., Lee, M.A., and Dien, T.V. (2003). Seasonal and spatial distribution of chlorophyll-a concentrations and water conditions in the Gulf of Tonkin, South China Sea. Remote Sens. Environ., 85, 475–483. Tang, D.L., Kawamura, H., Nhu, H.D., and Takahashi, W. (2004). Remote sensing oceanography of a harmful algal bloom (HAB) off the coast of south-eastern Vietnam. J. Geophys. Res., 109(C3), C03014, doi:10.1029/2003jc002045. USGS (2004). United States Geological Survey – Earthquake Hazards Program, Available at http://earthquake.usgs.gov/eqinthenews/2004/usslav USGS (2005). United States Geological Survey – Earthquake Hazards Program, Available at http://earthquake.usgs.gov/eqinthenews/2005/usweax
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 32
Tsunami Impact on Coastal Habitats of India
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
P.N. Sridhar, A. Surendran, S. Jain and B. Veera Narayan Oceanography Division, National Remote Sensing Agency, Hyderabad, Andhra Pradesh, India
32.1
INTRODUCTION
On 26 December 2004, following an earthquake of 8.9 Richter scale off Sumatra coast in Indian Ocean, large seismic sea waves had hit Andaman–Nicobar Islands and east coast of southern India (Figure 32.1). These seismic sea waves are popularly known as Tsunami waves. Tsunami waves traveled a distance of 3000 km from Sumatra to east coast of India within few hours at a speed of about 160–250 m/s (600–900 km/h). In deep sea, these tsunami waves were unnoticed because they are only few centimeters to a meter high. But they have long wavelength of hundreds of kilometers and wave periods of 100–1000 s. In shallow sea, tsunami waves travel at a speed of 10 m/s but grow 10 times taller than deep-water waves. Therefore, large wavelength, long wave period and high velocity are accounted for high-destructive nature of the tsunami waves (Ward, 2002). As the result, the shallow water tsunami waves can modify the coastal topography, inundate coastal areas and destroy coastal habitats. But in open coasts such as bay, estuaries, creeks and lagoons, the effect of tsunami wave is marginal as they provided conduit to the tsunami waves runoff. During recent tsunami, several coastal marine habitats have been affected by largescale sedimentation and seawater inundation. Among them, the Andaman and Nicobar Islands, Pichavaram Mangroves and Pulicat Lagoon are the most significant and socio-economically important coastal marine ecosystems. Already these habitats are under severe anthropogenic stress; no doubt, the effects of recent tsunami have long-term impacts to jeopardize livelihood of sizable coastal population living in the coastal areas. Therefore quick review of recent tsunami, its impacts and implication to the coastal community has been carried out in the following sections.
32.2
POST-TSUNAMI OBSERVATIONS IN INDIAN COASTAL HABITATS USING SATELLITE DATA
The post-tsunami scenario of Indian coastal zone within few hours of tsunami attack was observed from Indian Remote Sensing Satellite–Advance Wide Instantaneous Field of View Sensor (IRS– AWIFS) data. It revealed a widespread inundation in Andaman, Nicobar Islands (Figure 32.2) and Tamil Nadu coasts. High-resolution satellite data viz. Quick Bird, IRS- PAN (Panchromatic) and LISS III indicated changes in the shoreline morphology caused by breaching and sedimentation of river mouths and tidal inlets due to tsunami wave runoff (Figures 32.3 and 32.4). OceansatOcean Colour Monitor (OCM) data showed high turbidity and suspended sediment concentration in Andaman–Nicobar coral reefs and southern India coastal waters (Figures 32.5–32.7). Tropical Rainfall Measuring Mission-Microwave Imager (TMI) data showed a marginal fall in sea surface temperature of 1◦ C in the reef areas. 393
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
394
P.N. Sridhar et al.
Figure 32.1.
(a)
Circled areas indicate the severely affected coastal areas in (1). Andaman–Nicobar Islands and (2) southern east coast of India and north coast of Sri Lanka.
(b)
Figure 32.2. AWiFs data showing pre- (A) and post-tsunami (B) scenario in Nicobar Islands (a) Katchal, (b) Kamorta and (c) Trankati.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunami impact on coastal habitats of India
Figure 32.3.
IRS-LISS III data showing north Chennai coast affected by tsunami waves.
Figure 32.4.
Quick Bird data showing inundation of tsunami devastated Nagapattanam.
395
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
396
P.N. Sridhar et al.
Figure 32.5.
False color image showing high turbidity in the coastal waters of Andaman Islands on 27th December 2004.
Figure 32.6. The pre- and post-tsunami ocean color data showing increase in suspended particulate in coral reef areas of Andaman Islands on 27 December 2004.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunami impact on coastal habitats of India
25 December 2004
Figure 32.7.
32.3
27 December 2004
397
31 December 2004
Oceansat-OCM derived suspended sediment concentration along Pulicat Lake, Chennai coast during pre-tsunami (25 December 2004) and post-tsunami (27 and 31 December 2004).
MARINE AND COASTAL ECOSYSTEMS
In oceanic environment, marine ecosystem is a unit of unique biological communities. It is a sum of all coexistent flora and fauna species, their actions and interactions with other living and nonliving resources (Chivian, 2001). As a whole, a marine ecosystem plays a major role in regulating atmospheric oxygen, carbon dioxide and water vapor, global temperature and precipitation and filtering pollutants in the water (Cohen and Tilman, 1996; Costanza et al., 1997; Daily, 1997). The coast is an intermediate region between sea and terrestrial environment with unique intertidal floras and faunas. Coastal habitats provide livelihood and health to a large population by means of fishery, aquaculture, transportation, tourism, recreation and housing, etc. Therefore, any changes in the coastal marine ecosystem impacts on the livelihood, health and socio-economic aspects of mankind. 32.3.1
Corals of Andaman and Nicobar Islands
The coral reefs are the limestone structures, built by soft corals and other calcifying algae through secretion of calcium carbonate. Currently, coral reefs act as a sink for 111 million tons of carbon each year, “the equivalent of 2% of present output of anthropogenic CO2 ” (Kinsey and Hopley, 1991). There reef building process also helps regulating CO2 and linked to global climate changes. Andaman–Nicobar Islands and Gulf of Mannar are the two major coral reef systems in the Bay of Bengal. The Andaman and Nicobar Islands are the India’s largest coral Island systems covering 11,000 km2 and 2700 km2 of area, respectively (Gopinadhan, 1997), spread over 38 islands supporting a human population of 88,741 as per 1981 census. Several faunal assemblages such as dugong, a sea mammal and several endangered species like hawksbill, Olive Ridley and great leatherback turtles with numerous species of fish, algae, anemones and mollusks are found in
398
P.N. Sridhar et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
this region. Hard corals protect shoreline as barrier from high waves and currents and supply raw material for life saving drugs (Krishnakumar, 1997).
32.3.1.1 Tsunami related damage to corals World corals are under constant threat from bleaching, sedimentation pollution and diseases (Pamela et al., 2004). The recent earthquake in Indian Ocean region devastated several coral reefs within 500 km radius from the epicenter followed by tsunami wave attack (Chandha et al., 2005). In Andaman and Nicobar region tsunami waves damaged island and reef area by the raising waves of 2–10 m. In north Andaman Islands, large tidal and sub-tidal reefs areas were exposed during this event. As a result corals and coral reef organisms were dislocated and washed ashore. Further many stag horn corals were broken off by the force of tsunami waves and large boulders of corals had been seen moved towards the sea. The coral reefs were split and carried away. In southern Andaman and Nicobar Islands corals have survived from the tsunami waves but large volume of sand and debris were swept by seawater and had been deposited over corals. This event smothered and choked corals and other marine species (Sarang, 2005). A sudden spurt turbidity and suspended load is evident from the satellite data. Coral succumb to sustained sediment load of 0.2 kg/m2 /day (Hodgson, 1997) and many coral species are reported to have withstood certain concentration of periodic sediment loads (Solandt et al., 2001). But heavy sedimentation is capable of triggering expulsion of symbiotic algae and enriches nutrient level to high productivity and eutrophication. The eutrophication tends to lower the resistance of corals to diseases (Szmant, 2002). On the other hand, huge sand and coral debris deposited on the seagrass and seaweeds have had damaging effects on corals and other organisms. This may imbalance the intricate relationship among dependent marine species in the coral ecosystem. In Nicobar Islands, in the nesting grounds, several hundreds of sea turtles are jeopardize by damage caused to shoreline. The coral reefs of Gulf of Mannar have been protected by Island of Sri Lanka from direct impact of tsunami waves.
32.3.2
Coastal wetlands
Coastal wetlands are a sanctuary to several migrating and resident birds. Estuarine mangroves control flow velocity of river, trap the sediment and control sediment erosion in the deltaic region. Mangroves provide buffer zone against coastal erosion (Khalil, 1999). Estuarine mangroves are the breeding and nursing ground for crabs, shrimp, fish and other invertebrates (MEWM, 1996). Many other little known species like bream, mullet, milkfish, mojarras, snooks, barramundi, sea trout, snapper, drum, croaker, grouper and tarpon are important mangrove dependent commercial species (HMAM, 1984). Pichavaram Mangrove forest in the Tamil Nadu coast of Indian spread over 900 hectares, is providing livelihood to three thousand fishermen. Annually 230 tons of prawn fish, crabs are harvested and 74% penaeid prawns caught in adjacent coastal waters are nurtured in this mangrove swamp (Krishnamurthy, 1980). During recent tsunami the Pichavaram Mangroves have protected the hinterland from surging waves. However, several changes in the shoreline topography and coastal morphology have been recorded in satellite data and in the field. In particular, Vellar and Kollidam Estuaries morphology have been phenomenally changed by the tsunami waves (Figure 32.8). Significant increases in turbidity and suspended sediment have been observed in the Modasalodai, Vellar and Kollidam. The erosion of inlet mouth sand bars and coastal dunes (Figures 32.9 and 32.10) by tsunami wave have resulted in sedimentation and shoaling of the inlet channel. Consequently, the estuaries are limited to free exchange of sea and fresh water may lead to poor water quality and fall in fish production. This will cause severe implications on the socio-economic local fishing communities.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunami impact on coastal habitats of India
Figure 32.8.
399
IRS-LISS III data showing the opening of Pichavaram inlet (2 and 4) and reaching of Vellar and Kollidam mouths (1 and 3) during pre- and post-tsunami.
Figure 32.9. The field photography showing damage to the coastal geomorphology and dunes erosion in MGR Tittu in Pichavaram due to tsunami wave runoff.
32.3.3
Lagoons and creeks
Pulicat Lake, situated in the north of Chennai, is the second largest brackish water lagoon in the east coast of India and is a nesting ground for many migrating birds across sea. It is also a bird
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
400
P.N. Sridhar et al.
Figure 32.10.
(A) Fishermen village houses damaged by tsunami waves in Pichavaram, (B) Tiruchendur, a temple town experienced retreat of sea during tsunami, (C and D) Along the coast of Mannakkudi (Kanyakumari District), the church and the bridges suffered damage where there was no seawall to protect them from tsunami waves.
sanctuary, fishing ground and a tourism attraction, providing livelihood to thousands of coastal fishing communities living in the periphery. It is a major concern for several years that the lake is subjected to environmental stress from various anthropogenic causes (Nanda Kumar et al., 2001). Post-tsunami satellite observations show marginal inundation and shoreline changes in the Pulicat Inlet. After the tsunami widening of Pulicat tidal inlet mouth bars and several other inlets in these regions has been reported (Gupta et al., 2005). Yet, returning of migratory birds to their original habitat from this lagoon had been reported due to lack of seawater flow at the lake entrance (Anonymous, 2005). Oceansat-OCM data of post-tsunami showed heavy suspended sediment flux along Pulicat Inlet mouth, Cooum, Adyar and Palar river mouths (Figure 32.7). It suggests that the nearshore profile of this coastal stretch have been modified due to sedimentation by retreating tsunami waves. 32.4
COASTAL INUNDATIONS AND SHORELINE EROSION
In the east coast of India, several densely populated human settlements along 2260 km of shoreline were affected by tsunami waves. Satellite and field data observations showed human settlements along certain stretches of the Tamil Nadu coast was most affected by the tsunami due to the shoreline morphology. The bathymetric profile of the Chennai–Mahabalipuram coast showed moderate to gentle shoreward slope and the tsunami inundation cause less damage to the densely populated Chennai city and their suburbs. In contrast, low to gentle profile Kalpakkam– Nagapattanam coast had suffered severe damage and loss of lives due to inundated by the tsunami waves. Further in the south, the Pichavaram Mangroves provided shielding effect to the small stretch of coast, however low-profile estuarine, and wetland in area undergone considerable
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunami impact on coastal habitats of India
Figure 32.11.
401
Post-tsunami vulnerability ranking of coastal stretch of Tamil Nadu based on the slope, type of coast, presence of barriers and relative elevation, etc.
topographic changes (Figure 32.8). However, other coastal vegetation like coconut and casuarinas trees, sand dunes are not effective in protecting the coast. The beach morphology of the MGR Tittu (Island), a fisherman hamlet in Pichavaram, was totally modified by the tsunami waves (Figures 32.9 and 32.10). Vedaranyam (Point Calimer) to Gulf of Mannar coast, where the damage was least, the presence of barrier-island (Sri Lanka) supposed to have protection against tsunami waves. The intensity of the damage varied place-to-place depending on the type of coast, shoreline orientation, shoreline topography, coastal morphology, land cover and land use, etc. Steep rocky coast of Kanyakumari district had suffered more loss of lives and damage, wherever the protection structures are absent. In open coasts like tidal inlet, creeks, estuaries and bays inlet channels have provided conduit to the tsunami runoff. Natural barriers like mangrove vegetations, sand dunes, barrier islands and reefs are more effective than costal structures like seawalls and groins in protecting the coast. The unprotected coasts of Mannakkudi and Colachal villages of Kanyakumari district had suffered worse damage than that in protected coasts (Figure 32.10). The post-tsunami vulnerability ranking of Tamil Nadu coast (Figure 32.11) showed different intensities of tsunami
402
P.N. Sridhar et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
wave calamity due to difference in type (open and closed), nature (rock, sand and clay), relative height (shore to coast), slope (shoreward and landward) and orientation of the coasts. On the other hand, from the socio-economic perspective, coastal population suffered severe blow in terms of mental, physical and economic aspects. The effects of tsunami are immensely felt in two major revenue earning sectors viz. fisheries and tourism. Those small and marginal fishermen, who have survived from tsunami, have lost their boats and fishing gears. After traumatic loss of kith and kin of people, counseling and rehabilitation was needed. Gradually ecotourism was becoming major revenue in coastal sectors; sharp fall in tourist influx during January to April 2005 had affected the tour operators and small time vendors in Kanyakumari. The fear of recurrence of tsunami has jeopardized livelihood of many fishing communities in the coastal areas. 32.5
CONCLUSION
Satellite data of different ground resolutions supplemented by ground data, prior knowledge and field observations enabled quick and detailed review of post-tsunami scenario of Indian coastal and marine habitats. The evaluation of post-tsunami effects on different coastal habitats has provided knowledge and experience on the role of shoreline topography, coastal morphology, nature and type of coast in controlling the damage due to tsunami waves. This study helped in the understating of restoration of the affected habitats, socio-economic implications to coastal communities from tsunami impacts, planning rehabilitation measures needed by local communities and also future preparedness. ACKNOWLEDGEMENTS Authors express our sincere thanks to Director, Deputy Director and Group Director of NRSA and my colleagues who encouraged this work. REFERENCES Anonymous (2005). Tsunami caused major changes in shoreline study. The Hindu, April 25, 2004. Chadha, R.K., Latha, G., Yeh, H., Curt, P., and Toshitama, K. (2005). The tsunami of the great Sumatra earthquake of M 9.0 on 26 December 2004 – impact on the east coast of India. Curr. Sci., 88(8), 1297–1301. Chivian, E. (2001). Environment and health: species loss and ecosystem disruption – the implications for human health. CMAJ, 164(1), 66–69. Cohen, J.E. and Tilman, D. (1996). Biosphere and biodiversity: the lessons so far. Science, 74, 1150–1151. Costanza, R., d’Arge, R., de Groot, R., Farber, S., Grasso, M., and Hannoh, B. (1997). The value of the world’s ecosystem services and natural capital. Nature, 387, 253–260. Daily, G.C. (1997). Societal dependence on natural ecosystems. Nature’s Services, Washington DC, Island Press. Gopinadhan Pillai, C.S. (1997). A brief resume of research and understanding of the reef corals and coral reefs around India. In: Proceedings of the Regional Workshop on the Conser. Sustain. Manag. Coral Reefs Ecosystem (ed. Vineeta Hoon). M.S. Swaminathan Research Foundation and BOBP of FAO/UN, pp. 13–21. Gupta, G.V.M., Ramanamurthy M.V., and Subramanian B.R. (2005). Preliminary Assessment of Impact of Tsunami in selected coastal areas of India. ICMAM, DOD, p. 13. HMAM (Handbook for Mangrove Area Management) (1984). In: L.S. Hamilton, and S.C. Snedaker (eds.), Environment and Policy Institute, East West Center, IUCN/UNESCO/UNEP, Honolulu.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Tsunami impact on coastal habitats of India
403
Hodgson, G. (1997). Resource use conflicts and solutions. In: C. Brikeland (ed.), Life and Death of Coral Reefs, London, Chapman and Hall, pp. 536. Khalil, S. (1999). The economic value of the environment, cases from South Asia. Economic Valuation of the Mangrove Ecosystem Along the Karachi Coastal Asia. Applied Economic Research Institute, IUCN, Karachi, pp. 1–11. Kinsey, D.W. and Hopley, D. (1991). The significance of coral reefs as global carbonsink – Response to greenhouse. Palaeogeogr. Palaeocli. Palaeoecol., 89, 363–377. Krishnakumar, (1997). The coral reef ecosystem of the Andaman and Nicobar islands. Problems and prospects and the World Wide Fund for Nature – India initiatives for its conservation. In: Proceedings of the Regional Workshop on the Conser. Sustain. Manag. Coral Reefs Ecosystem (ed. Vineeta Hoon). M.S. Swaminathan Research Foundation and BOBP of FAO/UN, pp. 29–47. Krishnamurthy, K. (1980). Humans’ Impact on the Pichavaram Mangrove Ecosystem: A Case Study from southern India. Proceedings of the Asian Symposium on Mangrove Environment, Research and Management, Kuala Lumpur 25–29 August, pp. 624–632. MEWM (Mangrove Ecology Workshop Manual) (1996). In: I.C. Feller, and S. Marsha, (eds.), A Field Manual Focused on the Biocomplexity on Mangrove Ecosystems, Washington DC Smithsonian Institution. Nanda Kumar, N.V., Saritha, K., Rajasekhar M., and Ameer Basha, S. (2001). Aquaculture effluent effect on physico chemical characteristics and zooplankton of Pulicat lake bird sanctuary. Eco. Env. Cons., 7(1), 25–29. Div. Environ Bio, SVU, Tirupati-517502. Pamela, H., K’wasi, B., and Fisher, E.M. (2004). Coral reef risk assessment. Environ. Micropaleontol. Microbiol. Meiobenthol. 1, 11–39. Sarang, K. (2005). Tsunami Impact Assessment of Coral Reefs in the Andaman and Nicobar. CARDIO News, Interim Report, pp. 1–6. Solandt, J.L., Goodwin, L., Beger, M., and Harborne, A.R. (2001). Sedimentation and related habitat characteristics in the vicinity of Danjugan Island, Negros Occidental. Danjugan Island Survey, Summary Report 5, p. 5. Szmant, A.M. (2002). Nutrient enrichment on coral reefs: is it a major cause of coral reef decline? Estuaries, 25(4b), 743–766. Ward, S.N. (2002). In: R.A. Meyers (ed.), Tsunamis in Encyclopedia of Physical Sciences and Technology, 3rd edn. New York, Academic Press.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 33
Overview and Integration of Part 4
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
U. Aswathanarayana Mahadevan International centre for Water Resources Management, Hyderabad, AP, India
33.1
BIOPHYSICAL AND SOCIO-ECONOMIC DIMENSIONS OF TSUNAMI DAMAGE
Raison d’ etre: This part gives an account of the different dimensions of the tsunami damage. In the case of Indonesia and Thailand, an assessment is made of damage to engineering structures. In the case of India, the variation in the severity of damage at various sites is shown to be determined by the geophysical, ecological, and socio-economic ambiences at these sites. The relative information is made use of to figure out strategies for minimizing the vulnerability of structures and sites to tsunami damage. In Chapter 26, Murat Saatcioglu et al. drew the following conclusions from the reconnaissance conducted in Thailand and Indonesia to assess the engineering significance of the December 26, 2004 tsunami in respect of buildings, coastal structure, bridges, and other physical infrastructure. In the case of Phuket Island as well as the coastline extending north of Phuket, it was observed that non-engineered reinforced concrete buildings, low-rise timber frames, and unreinforced masonry walls suffered extensive damage from hydrodynamic pressures generated by the tsunami. No seismic damage was observed in Thailand as damage occurred only as a result of the tsunami wave impact. On the other hand, the infrastructure in place along the northern and southern coastal regions of Banda Aceh was completely devastated by both tsunami wave forces and seismic ground excitations. The damaging effects of the tsunami were most pronounced in unreinforced masonry walls, non-engineered reinforced concrete buildings, and low-rise timber framed buildings. Engineered construction survived the tsunami pressure, but many suffered extensive damage due to seismic forces. The majority of seismic damage was attributed to poor design and detailing of non-ductile buildings. There was very little damage observed in structural components of well-designed concrete buildings. Bridge infrastructure was devastated by tsunami forces. Many bridges were swept away from their supports, disabling the transportation network. Storage tanks and other light structures should be well anchored to their foundations to resist tsunami pressures. Many steel storage tanks, as well as other unanchored structures floated away long distances due to uplift pressures generated by the tsunami. In Chapter 27, Kurian et al. gave a detailed account of the tsunami damage on the Kerala coast. The devastation has been found to be most severe along the Neendakara–Arattupuzha sector, resulting in significant changes in the coastal geomorphologic setting along this region. The run-up level distribution along the Kerala coast shows wide variation, with maximum of 5 m at Kayamkulam Inlet. The interactions of the tsunami with tides, currents, and waves could have played a role in the tsunami intensity variation along the coast. The arrival of tsunami waves at high tide is a factor that compounded the inundation, leading to higher intensity of damage around Kayamkulam Inlet. In the same way, the low tide minimized the effect, as observed in the northern tracts of the coast, where the tsunami waves arrived in the afternoon. The highest 405
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
406
U. Aswathanarayana
waves along the northern Kerala coast occurred in the midnight, coinciding with the next high tide and the occurrence of two major waves at that time. The data on the shelf and beach profiles before and after the tsunami has been made use of to quantify the extent of erosion brought about by the tsunami sea level. It is quite possible that the tsunami waves might have stirred the bottom and taken ashore the finer sediments leading to dominance of sandy sediments. The run-up distances of seawater to the inland region is directly related to the beach slope, land elevation, settlement pattern, infrastructure facilities, etc. Tsunami inundation into the hinterland regions is relatively less in Kerala due to the occurrences of coast parallel backwaters/lagoons. It can be surmised that the sediment resuspended in the innershelf by the tsunami is transported on to the barrier beach resulting in its deposition inland. This accounts for the huge deposits of heavy minerals in the hinterland areas and siltation seen in the TS canal. In Chapter 28, C.S.P. Iyer investigated the ecological impact of the Indian Ocean Tsunami. The immediate impact of tsunami on the southwest coast of India was the draining of nutrients with the consequent lowering in chlorophyll concentration and thereby of primary productivity. Naturally, this affected the food web and resulted in depletion in fish. With time, there was perceptible recovery in the marine system, as evident from the monitoring in May 2005. It is also seen that the benthos being bottom dwelling, suffered maximum from tsunami. Naturally, it may take more time to recover. The maximum impact of tsunami was on those coastal regions, which had inland basins as Muttam, Vizhinjam, and Valiyazhikkal. On the other hand, the impact was least from Veli to Quilon, due to long stretches of plains. That the marine environment is slowly recovering from the impact of tsunami is evident from the improvement in biological productivity of this coastal stretch. In Chapter 29, Chandrasekar and Ramesh made a detailed study of the biophysical and socioeconomic damage caused by the tsunami on the southwest coast of India. The inundated seawater rendered the groundwater highly saline. In some of the coasts like Mankudy and Colachel, the pH level went up to 8.7 and the total dissolved solids varied between 2000 and 7000 mg/l. There was negligible damage to the mangroves of Manakudy Estuary and to the salt marshes of some of the beaches like Colachel, Chothavilai. On the basis of the number of casualties, and estimated damage to buildings, boats and fishing nets, and assigning a weightage to them, the vulnerability of a given segment of the coast was categorized into Highly Vulnerable, Medium Vulnerable, and Less Vulnerable. On this basis, a vulnerability map was prepared for the coast. Thus, the coasts of Kanyakumari, Covalam, Manakudy, Pallam, Azhikkal, Muttam, Kadiapatanam, Kottilpadu, and Colachel are identified as highly vulnerable in the event of a tsunami occurring. Evidently, the most vulnerable areas of the coast need to be provided greater protection when planning for preparedness. In Chapter 30, Sadhuram et al. investigated the hydrophysical manifestations (such as, the salinity, temperature, coastal currents, internal waves, etc.) of the tsunami. Cumulative trauma disorder profiles taken before and after the tsunami at roughly the same location off Visakhapatnam, have been compared. Cooling of temperature could be seen throughout the water column with a maximum of 1.5◦ C at 40 m depth. Salinity after tsunami increased by 1 psu at surface while at bottom it decreased by more than 1.5 psu compared with that before tsunami. There is evidence to show that the coastal current was strengthened by the tsunami waves which hit east coast of India. Spectral analysis of temperature data (2 min interval) collected at 10 m and 60 m depths indicate high-frequency internal waves during the observational period. Temperature and salinity data obtained from an Argo float in the Indian Ocean about 8 h after the tsunami hit the east coast of India was compared with comparable data available near the same location before the tsunami. The cooler water near the coast as a result of tsunami waves might have advected to this location, which may be possible in 8 h. The high speed of current generated by the tsunami waves (>3.0 m/sec ?) might have caused cooling in temperature and changes in the salinity of top 20 m layer. The vulnerability of the Indian coast for damage due to tsunami closely follows
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Overview and integration of Part 4
407
the vulnerability of the coast to sealevel rise due to global warming, and is strongly dependent on nearshore bathymetry and topography. It may be inferred that on the east coast, 10–12◦ N (south Tamil Nadu coast), 14–16◦ N (south Andhra coast) and >20◦ N (West Bengal) , and on the west coast; 9–10◦ N (Kerala) and 21–24◦ N (Gujarat) appear to be more vulnerable for the damages due to tsunamis/storm surges/sea level rise. In Chapter 31, Subba Rao et al. made a wide-ranging synthesis of the impact of the tsunami on marine life in the Indian Ocean, ranging from Indonesia to India. Geomorphologic changes, inundation by oceanic water, submarine land slides resulting from tsunamis impact the shoreline, pelagic, and sea bed habitats. Populations ranging from meiofauna to large aquatic mammals inhabiting the backwaters, salt marshes, tidal waters, mangroves, coral reefs, sandy shores, rocky shores, mud flats, and shoreline structures will be adversely affected. For instance, in the coastal waters off Pondicherry, that the sewage mix extended even up to 10 km is evident from growth of E. coli, faecal coliforms and Salmonella. Concomitantly there were changes in the biodiversity. Harpacticoids, nematodes, oligochaetes, gastrotrichs, and polychaetes were dominant in the meiofauna prior to tsunami but by the third and fourth day after the tsunami changed to foraminiferans, cnidarians, nemertines, gastrotrichs, rotifers, kinorhynchs, ostracods, isopods, halacarids, and insects. The tsunami impacted both the oceanic waters and the nearshore waters. The massive dislocation of sub-surface deep waters was similar to an upwelling, and was characterized by a decrease in sea surface temperature by about 1◦ C, increase in suspended particles, and increased nutrients which probably caused an increase in phytoplankton biomass to the northeast of Sumatra, and off Chennai. The time series data for chlorophyll a compiled for 2000–2005 showed an increase in phytoplankton biomass (>0.35 chl a µg/l) between mid-January and February 2005 soon after the tsunami. Two weeks after the tsunami, in January 2005 a phytoplankton bloom developed with chlorophyll a (>0.5 µg/l) in a 300 × 300 km area to the southeast of Sri Lanka and north of Aceh Province of Indonesia in the Andaman Sea. Similarly, in the nearshore waters near Chennai, a bloom dominated by the diatom Lauderia annulata developed and could be attributed to nutrient enrichment. Many coral reefs have lost both their structure and biota, and are now reduced to rubble due to mechanical damage. The force of the tsunami can move enormous boulders and sections of reef, as well as thousands of tons of smaller fragments, sand and silt, which dislodge, crush, and kill marine biota. Further, there is significant contamination by run-off from land, with large quantities of wastes and pollutants, debris, soil, and organic matter. This will further increase the damage and hamper recovery. Similarly, devastation of vast stretches of coral reef beds, mangroves, seagrass, and seaweed beds which act as feeding and nursery grounds for a myriad species of finfish and shellfish species would affect the capture and culture fishery for long. The aquaculture units that lie close to the shore zone, were badly affected by the tsunami: the farms were damaged significantly; dykes collapsed, sluice gates sank; pipelines (drains and feeders) were covered with silt and sand; cultured shrimp (whatever present) were either dead or drained into the sea. Food and Agriculture Organization (FAO) estimated that marine exports from India may decline by as much as 30%, as a consequence of the impact of tsunami in SouthEast Asian region. The fishermen living along the 1000 km of Indian coastline were the worst hit by tsunami. When a catstrophic natural disaster struck the Norwegian coast at Kvennavatnet, it took between 80 and 120 years before a comparatively diverse invertebrate community developed. But in the tropics where the biological wheels turn faster unlike in the sub-arctic, recovery of the marine environment from the tsunami impacts could be faster and this needs to be monitored. In Chapter 32, according to Sridhar et al. a few hours after the tsunami struck the Indian coast, the Indian Remote Sensing Satellite–Advance Wide Instantaneous Field of View Sensor (IRS–AWIFS) data could provide information on the widespread inundation in Andaman, Nicobar Islands, and Tamil Nadu coasts. High-resolution satellite data viz. Quick Bird, IRS- PAN
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
408
U. Aswathanarayana
(Panchromatic), and LISS III indicated changes in the shoreline morphology caused by breaching and sedimentation of river mouths and tidal inlets due to tsunami wave runoff. Oceansat-Ocean Colour Monitor (OCM) data showed high turbidity and suspended sediment concentration in Andaman–Nicobar corals reefs and southern India coastal waters. Tropical Rainfall Measuring Mission-Microwave Imager (TMI) data showed a marginal fall in sea surface temperature of 1◦ C in the reef areas. The Andaman and Nicobar Islands are the India’s largest coral Island systems. The tsunami event smothered and choked corals and other marine species. A sudden spurt of turbidity and suspended load is evident from the satellite data. Heavy sedimentation is capable of triggering expulsion of symbiotic algae, thereby enriching nutrient level to high productivity and eutrophication. The coral reefs of Gulf of Mannar have been protected by Island of Sri Lanka from direct impact of tsunami wave. Oceansat-OCM data of post-tsunami showed heavy suspended sediment flux along Pulicat Inlet mouth, Cooum, Adyar, and Palar river mouths. It suggests that the nearshore profile of this coastal stretch have been modified due to sedimentation by retreating tsunami waves. The evaluation of post-tsunami effects on different coastal habitats has been found useful in planning the restoration of the affected habitats, understanding the socio-economic implications to coastal communities, planning rehabilitation measures needed by local communities and also future preparedness.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Part 5
Quo Vadis
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 34
Protection Measures Against Tsunami-type Hazards for the Coast of Tamil Nadu, India
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
V. Sundar Department of Ocean Engineering, Indian Institute of Technology, Chennai, India
34.1
INTRODUCTION
The new approach towards the mitigation of major disasters, such as tsunami, is to promote dual-use technologies to improve the resiliency of the biophysical and socioeconomic systems. This chapter explains how the groins that have been designed for coastal protection have proved beneficial in the mitigation of the adverse consequences of the tsunami. This results of the posttsunami survey clearly indicate that the coastal protection structures (engineering structures, vegetation, etc.) helped in the mitigation of the adverse consequences of the tsunamis. They would be equally protective against extreme weather events, such as tidal waves and hurricanes when they happen. Details are given as to how at specific sites along the Tamil Nadu coast, groins gave protection against the tsunami, and how the structures that are planned will serve the double purpose of giving protection against continuous processes such as erosion, and episodic processes such as sea surges, tsunamis, etc.
34.2
BEHAVIOR OF SHORELINE IN-BETWEEN GROIN FIELD
The rate of erosion along the Tamil Nadu coast is given in Table 34.1. Ever since the construction of the harbor of Chennai port with breakwaters, the coast on its north has been subjected to erosion at a rate of about 6.6 m/year for the last four decades due to the predominant northerly drift. A part of the existing National Highway and the residential area nearer to this coastline have already been eroded. In spite of the provision of a seawall, the erosion continued along few pockets of the northern coast. The solution for the said coastal erosion problem was divided into two categories: a temporary strengthening of the existing seawall and a permanent remedial measure by providing suitable groins. In the first phase, a detailed bathymetry survey was made for the measurement of existing cross section of seawall and its status in order to assess its adequacy for the design wave regime. The wave characteristics such as average wave height, wave period and wave direction from which the average breaking wave characteristics were derived, were drawn from the wave atlas prepared for the Indian coast by National Institute of Oceanography (NIO). The monthly sediment transport has been estimated based on Energy Flux method of Coastal Engineering Research Centre (SPM, 1984), and the method of Komar (1969) which involves integrating the distribution of sediments within the surf zone. The net sediment drift along the Chennai coast is observed to be about 1.2 × 106 m3 /year towards the north. In the second phase, as a permanent solution for the coastal erosion problem, ten numbers of shore-connected straight rubble mound groins in the two severely 411
412 V. Sundar Table 34.1.
Sl.No.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
1 2 3 4 5 6 7 8 9 10 11 11a 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Rate of erosion along the Tamil Nadu coast (Public Work Department, Tamil Nadu, 2002). Location Pulicate Ennore Royapuram Marina Foreshore Elliot/Astalakshmi temple site Kanathur Kovalam Mahabalipuram Pondicherry Cuddalore (north) Cuddalore (south) Poombuhar Tranquebar Nagapattinam Point Calimere Ammapattinam Keelakarai Mandapam Rameswaram Tiruchendur Manappadu Uvari Kanyakumari Manakkudi Pallam Muttom Manavalakurichi Colachel Midalam
Length in m
Accretion/ erosion
Rate in m/year
0.71 3.27 5.38 2.97 2.3 2.08
– – – – – –
3.20 1.30 6.60 1.70 1.09 1.28
0.24 3.15 5.45 1.19 1.538 0.483 1.905 0.76 4.27 0.966 3.6 2.9 2.19 3.3 1.53 1.6 2.6 0.7 3.65 2.6 3.0 3.5 1.75 2.5
– – – – – – – – – – – – – – – – – – – – – – – –
1.4 0.81 0.25 0.15 8.00 2.98 0.65 1.80 0.11 3.40 0.72 0.29 0.25 0.06 0.33 1.10 0.86 1.74 0.57 0.93 0.17 0.60 1.20 0.84
affected stretches, were proposed (Figure 34.1). The length and the spacing between groins were designed based on the recommendations of Shore Protection Manual (SPM, 1984). Mathematical modeling to evaluate the shoreline changes due to the proposed groin field was carried out. The mathematical modeling of shoreline evolution essentially relates the change in the beach volume to the rate of material transported from the beach. The methodology for the present numerical model is based on the numerical scheme proposed by Janardanan and Sundar (1994) and the one line model was solved by using Crank Nicholson implicit finite difference method. The seabed bathymetry of the proposed location, length of the groins, height of the berm, grain size of the sediments required for the calculation of active depth of the sediment transport and water depth at the tip of the structure are used as inputs for the model. The model predicted a significant advancement of beach over a period of 15 years. The construction of the proposed groin field started in May 2004. That there has been immediate shoreline advancement on the south of the executed groin has convincingly demonstrated the viability of the design of the suggested remedial measure. Since then all the groins have been
Protection measures against tsunami-type hazards for the coast of Tamil Nadu, India
413
Shoreline
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Satellite Port
En
ek
re
eC
r no
Kathivakkam
Ennore E.I.D Parry
ICI Ltd.
A New 500 m long groin
Sandtrap
Ernavoor Ennore Expressway Ernavoor bridge Stretch I (2.0 km) Proposed 6 groins by IIT Madras (Not to scale)
Thiruvotriyur Highroad
Eveready&co Tollgate
Stretch II (1.5 km) Proposed 4 groins by IIT Madras (Not to scale)
Tondiarpet Royapuram
0
0.5 1.0 1.5 2.0 km
Harbor
Figure 34.1.
System of groins in the Chennai area.
completed and the road has been saved from any further damage. This is evident from the fact that the groin field not only withstood the onslaught of the recent tsunami, but also has helped to a very great extent in reducing the inundation and damage on the landward side of this stretch of coast. The shoreline advancement due to the groin field (6 groins) in stretch one after the tsunami in December 2004 proves the effectiveness of the proposed groin field not only in preventing further erosion, but also in enhancing the formation of beach. The post-tsunami shoreline survey
414 V. Sundar Table 34.2. Area of beach in-between groins 5 and 6. Date of measurement
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Work commenced in May 2004 13 August 2004 25 August 2004 14 September 2004 Post-tsunami 06 January 2005 21 January 2005
Figure 34.2.
Area in m2 3700 6970 8800 4660 10,450
Satellite imagery showing the shoreline advancement due to groin field in Kanyakumari district.
on 6 January 2005 showed that nearly 50% of the beach was lost due to the propagation of the tsunami over the continental shelf in between the groins 6 and 5. However, the survey carried out on 21 January 2005 has revealed that a quantity slightly more than that removed has been redeposited. The above survey brings out clearly the effect of tsunami on the behavior of shoreline in between the groins. The area of beach formed in-between groins 5 and 6 are projected in Table 34.2. Another success with groin field is that for a stretch of 3 km covering Kurumbanai, Vaniyakudi and Simon colony villages in the west coast of Tamil Nadu. The geographical position of the
Protection measures against tsunami-type hazards for the coast of Tamil Nadu, India
415
study area (8◦ 11.5 , 77◦ 13.5 E and 8◦ 10.3 , 77◦ 14.8 E) is shown in Figure 34.2. The groin field enhanced the beach formation in the originally eroding stretches, and acted as buffers in reducing the inundation of seawater due to tsunami. The shoreline advancement could be seen in satellite imagery for the said study area. Creating beaches to combat seawater ingress into land as well as to mitigate the usual erosion problem constitute suitable protection measures for tsunami as well. Case histories in respect of Chennai, Trichy and Madurai areas are described as follows.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
34.3
PROPOSED COASTAL PROTECTION MEASURES
34.3.1 Chennai region Ennore (N 13◦ 13 56.9 E 80◦ 19 51.7 to Royapuram (N 11◦ 54 59.03 E 79◦ 49 51.7 ) A number of fishing hamlets are located in the stretch of about 15 km from Ennore towards south up to Royapuram. Even though the stretch from Chinna Kuppam (about 3 km from south of Ennore creek mouth) to Ennore mouth has been protected by a seawall, this stretch is liable to be eroded in future. Hence, this should be strengthened by a groin field, by which additional beach width can be gained. An added benefit will be the reduction in the quantity of sand entering the Ennore river mouth and harbor, leading to lesser quantity of maintenance dredging of the approach that need to be carried out by the Ennore port. The number of groins for this stretch will be about 10, with the average length of the groin being 150 m. The coast north of Chennai harbor for a distance of 9 km has already been protected by groin fields designed by IITM, which has served as an effective measure against coastal erosion (Sundar, 2005). Further strengthening is planed with seawall to reduce the run-up during the new moon and full moon days, phenomena which is experienced in the very recent past after the occurrence of the tsunami. 34.3.1.1
Stretch between Chennai port (N 13◦ 02 24.9 E 80◦ 16 47.5 ) to Foreshore estate (N 13◦ 02 04.9 E 80◦ 16 35.3 ) This stretch of the coast includes the Marina beach which is one of the world’s longest beaches. The net littoral drift being towards north has resulted in the sand bar formations at the mouth of rivers, Cooum and Adyar. A pair of groins as training walls is suggested for training of the mouth of Cooum river – one on the south of the Cooum mouth extending up to a water depth of about 6 m, and the second one in the north extending up to a water depth of about 4 m. The reason for longer groins is to trap the sediments for 1.5 to 2 times the surf width. In the case of Adyar, regular maintenance dredging of the mouth is recommended. The stretch in between these two river mouths is best suited for plantations. 34.3.1.2
34.3.1.3 Meyyur Kuppam (N 12◦ 31 44.5 E 80◦ 10 00.8 ) and Sadras kuppam This village lies south of Kalpakam Atomic Power Station. This stretch is highly vulnerable to tsunami. In the immediate south of this village, the presence of coconut plantations had given some protection against the tsunami. The tsunami wave rose to a height of 3 m, and caused inundation. Further south, the flatness of the beach and the area of low elevation has enabled the tsunami to cause much damage. Based on the behavior of the tsunami and taking into consideration the low elevation of ground, it is recommended that this stretch of 1.5 km be protected with a combination of groin field (about six with an average length of about 200m) and seawall. The crest of the seawall should be fixed 2 m higher than the ground level in order to protect the coast and also to serve as a buffer against natural hazards in future. A similar measure is suggested for Oyyalikuppam, which is south of Meyyur kuppam. Bioshields constitute a long-term solution.
416 V. Sundar
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Plantations
Buffer blocks
Shoreline
Figure 34.3.
Suggested protection measure for the stretch of the coast at Pudukuppam in Parangipettai.
34.3.1.4 Pudukuppam (N 11◦ 31 34.9 E 79◦ 45 47.8 ) During the tsunami onslaught, water penetrated for about 1 km landward, and damaged the houses that are located at a distance of about 500 m from the shoreline. The barren land in this stretch of the coast has not offered any protection from the tsunami run-up. Hence, dense plantations are proposed. The dwelling units are located mostly about 200 m away from the shoreline and the seabed is quite flat, without much shoreline oscillations. For these reasons, conventional hard measures such as seawalls or groin field would not be effective as coastal protection measure. In the event of a tsunami or storm surge, the aim of the protection measure should be to reduce its speed. This purpose is best served by the construction of two rows of masonry blocks (4 m × 0.5 m × 0.5 m) at a distance of 200 m from the shoreline and in between these blocks and shoreline, trees may be planted. These could act as front line soldiers to reduce the speed. The priority in this case should be plantations followed by the construction of the buffer blocks. The details of the concept are projected in Figure 34.3. 34.3.2
Madurai region
A few stretches of the coast of Tamil Nadu has been protected with groin field and seawalls designed by IITM. They have withstood the onslaught of the tsunami and have also been very effective in reducing the inundation, particularly due to beach formation in between the groins. The most affected stretch of the coast under this coastal region are 34.3.2.1 Colachel jetty (N 8◦ 10 18.4 E 77◦ 15 18.2 ) The beach is found to be very flat on either side of the jetty and can be used for plantations. A pair of groins with a crest elevation of about 6.0 m from mean sea level (MSL) (locally called as thoondil valaivu) can serve as a protection measure against high waves particularly during
Protection measures against tsunami-type hazards for the coast of Tamil Nadu, India
417
Break waters
Appr
oach
chan
nel
800 m
Existing jetty
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
600 m
500 m
Figure 34.4.
500 m
Proposed groins (thoondil valaivu) at Colachel jetty.
cyclones. The proposed structure could also serve as a landing facility for boats and catamarans (vide concept projected in Figure 34.4). The existing jetty can be used more effectively if the above proposal is implemented. The length of the coast that should be covered under this proposal will be about 1 km. 34.3.3 Trichy region The tsunami wave with a height of about 8 m rushed inland with a great velocity and penetrated up to a distance of about 1 km. This is mainly because of the flat beach that exists in this stretch of the Nagapattinam coast. The barren land north of Nagapattinam port is ideally suited for dense plantations. 34.3.3.1 Keechankuppam (N 10◦ 45 16 E 79◦ 50 57.8 ) This is the worst affected area due to tsunami. The tsunami destroyed several bridges and houses along this stretch. Dense plantations for a distance of about 9 km (Nagore to Keechankuppam) would protect about six villages in this stretch of the coast. The recommendation is made for the two stretches namely, between Kallar and Kaduvayar river mouths and Nagore to Kaduvayar river mouth. As littoral drift in this stretch is more towards the north and the coast is of sandy type, T-shaped groins would certainly trap the sediments and also retain the same within them. A seawall is recommended to minimize the inundation of seawater landward. As the coast is just 0.6 m above MSL, a seawall is also recommended for the entire stretch of the coast. The details of the proposed layout for the stretch between the rivers Kaduvayar and Kallar to protect the stretch of the coast in Keechankuppam are shown in Figure 34.5. 34.3.3.2 Tharangampadi (Tranquebar) (N 11◦ 01 32.4 E 79◦ 51 23.1 ) This stretch of the coast at Tharangampadi involves the protection of monuments and places of National Heritage. The existing old groins are ineffective in trapping the sediments. It is recommended to rehabilitate the groin A–A with a proper head with a top elevation of +3.35 m. Also, an extra groin of length 70 m at a distance of 50 m south of A–A is recommended. The existing two groins south of A–A should be rehabilitated and the length should protrude to a distance of 50–60 m from shoreline with a top level of +3.35 m. Plantations on the leeside of the existing seawall is recommended as a long-term measure. The proposed scheme
418 V. Sundar
N
Kaduvaiyar river
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Seawall
5 T-groins of length 200 m (average)
3 Km
Kallar river
Figure 34.5.
Proposed layout of groins from Nagore to Keechankuppam.
is shown in Figure 34.6. The village Sathankudi (N 11◦ 01 52.7 E 79◦ 51 19.6 ), located north of the fort has suffered huge loss of life and dwelling units. The water has penetrated about a distance of about 750 m from the shoreline. The Public Work Department (PWD) has a proposal for construction of a seawall for a distance of about 850 m from the existing seawall. Construction of a seawall with the crest level at +4.35 m, is recommended. In addition to the seawall, a groin field consisting of 5 transition groins of average length of 100 m,
Protection measures against tsunami-type hazards for the coast of Tamil Nadu, India
Plantations
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
N
Existing groins
Temple
419
Remains of an old existing building A A
A A
Collector’s bungalow
Additional Groin Dutch fort
Figure 34.6.
Rehabilitation of groins
Rehabilitation of existing groins at Tranquebar coast.
Plantations
+6.0 MSL
1:5 to 1:6
1:2.5
Plantations
Sand Dune Ditch
Figure 34.7.
Proposed shape of the sand dune at Palayur.
with 1 or 2 groins bent to be formed as “Thoondilvalaivu” as it is called by the locals, is recommended. 34.3.3.3 Palayur (N 11◦ 21 14.6 E 79◦ 49 44.5 ) A number of casualties and damage to the property have taken place in this stretch of coast during the tsunami. As the village is right on the banks of river Coleroon, one suggestion is to retain the dunes already constructed by the local people. The top level of the dune may be further raised. The ditch in front of the dune should be shifted to the rear side of the dune. The dune should take the shape as shown in Figure 34.7 for a distance of about 1 km. Plantations on the seaside and on the dune are recommended. As a long-term measure, the dunes can be converted
420 V. Sundar
Kodiyam Palayam Ri
ve
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
fish ting etty s i x j E ing load
Sand bar
rC
ole
N
ro
on Spurs loc Ne atio w no Ha f rb or
Short
spurs
Sand D
une
Ditch
Bay
Figure 34.8.
of B
eng
al
Proposed coastal protection measure at Palayur.
to revetments or with geo-tubes with its top level of +6.0 m above MSL. For this purpose, the shallow regions can be dredged and the dredged spoil can be used for the creation of the dune. A portion of the bank can also be planned for landing jetty in future after the protection of the riverbank with spurs. The details of the proposed scheme are shown in Figure 34.8. 34.3.3.4 Thirumalaivasal (N 11◦ 14 31.5 E 79◦ 50 37.9 ) This stretch of the coast is at the confluence point of the river Vellapallam Uppanar. The entire stretch needs to be dredged and a bund has to be created using this dredged spoil for a distance of about 1 km from the mouth. Two training walls, at the mouth of the river Vellapallam Uppanar, are recommended (vide Figure 34.9). A few spurs along the banks of this river need to be provided in order to divert the flow into the ocean. Plantation along the banks of the river is recommended.
34.4
SUMMARY
A general post-tsunami survey of the coast of Tamil Nadu was made during February–March 2005 in order to assess the vulnerable areas being affected by the perennial problem of erosion. The effect of the recent tsunami was considered in the said exercise. A “mix” of different coastal protection measures, such as, seawall, groin field, combination of seawall and groins, training walls, plantations, buffer blocks, curved groins (thoondil valaivu), geo-tubes, need to be customdesigned for the three coastal regions, namely, Chennai, Madurai and Trichy. The protection measures are site specific and are dictated by the direction and magnitude of the littoral drift. A host of parameters like the beach profile, bathymetry, shoreline changes over the past few years, behavior of already existing protection measures, etc. which control the magnitude and quantity of littoral drift, need to be investigated before the recommendations made in this chapter, are implemented on the ground.
Protection measures against tsunami-type hazards for the coast of Tamil Nadu, India
421
N Dredging the shallow patches Training walls Spurs
Bay of Bengal
To be formed as Bund
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Plantations
Existing 2 No.s Training walls Dune – 150 m Sand dune – 1 Km 4 Spurs – 70 m Vellapallam Uppanar River
Figure 34.9.
Proposed coastal protection measure at Thirumalaivasal.
ACKNOWLEDGEMENT The author wishes to record his thanks to the Public Work Department, Government of Tamil Nadu for entrusting the responsibility for carrying out the study as well as to permit the author to present the salient parts of the comprehensive report in the form of a publication. The support of his colleagues and students as well as PWD engineers is gratefully acknowledged. REFERENCES Anonymous (1984). Shore Protection Manual (SPM), Volume 1 and 2. US Army Corps of Engineers, Vicksberg. Janardanan, K. and Sundar, V. (1994). Effect of uncertainties in wave characteristics on Shoreline Evolution. J. Coastal Res., 13(1), 88–95. Komar, P.D. (1969). Longshore transport of sand on beaches. PhD Thesis, University of California, Sandiego. Sundar, V. (2005). Behaviour of shoreline between groin field and its effect on the tsunami propagation. Proceedings of Solutions to Coastal Disasters Conference of ASCE, 8–11 May, Charleston, South Carolina, USA.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 35
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Protective Role of Coastal Ecosystems in the Context of the Tsunami in Tamil Nadu Coast, India: Implications for Hazard Preparedness A. Mascarenhas and S. Jayakumar Geology Division, National Institute of Oceanography, Dona Paula, Goa, India
35.1
INTRODUCTION
To assess the damage caused by the Indian Ocean Tsunami of 26 December 2004, post-tsunami field surveys were undertaken during January and April, 2005 from Chennai to Velankanni along the Tamil Nadu coast. Beach profiles were measured at 24 sites (Figure 35.1). Tsunami-induced landscape changes, identified in April 2005, have been compared with earlier observations conducted along the Nagore–Velankanni sector in November 1998. The success of sustainable coastal development depends on the resilience towards natural hazards and reduction of disasters. With this objective, this paper highlights the natural buffer capacity of littoral sand dunes and coastal forests, as proved during the December 2004 tsunami. Some coastal management imperatives are explained and the lack of appropriate coastal hazard policies as a cause of human loss is emphasized. 35.2
EXTREME OCEANOGRAPHIC EPISODES AND THEIR AFTERMATH
During the period 1891–2000, of the 215 severe cyclones in the Bay of Bengal, 55 crossed the coast of Tamil Nadu (Shrestha, 1998). Storm surges that accompany high intensity wind systems have resulted in extensive inundations of coastal lowlands. The Orissa super cyclone of October 1999 induced a storm surge of 7–9 m with consequent flooding up to 35 km from the coast around Paradip (Thapliyal et al., 2000). As a consequence of recurring hydrometeorological events, damage to coastal population and habitations, loss of property and livestock, damage to infrastructure, breakdown of communications, changes in land use patterns, modifications of landforms, alterations of vegetation and effects on ports, etc. have occurred (Raghavan and Sen Sarma, 2000; Paul, 2001; Mascarenhas, 2004). The coast from Nagore to Velankanni (Figure 35.2) has been chosen a case to highlight impacts of the December 26 tsunami event on coastal landforms, assets and humankind. From north to south, the sea front can be subdivided into various sectors based on the type and state of coastal landforms as observed in November 1998. These observations have been compared with post-tsunami studies conducted in April 2005 (Table 35.1). Intense traditional fishing activity marks the coast from Nagore to Nagapattinam lighthouse. The area is characterized by linear sandy beaches. The sand dunes in the back of the beaches may be thickly or sparsely vegetated. Fisherfolk live in the huts built on these sand dunes (Figure 35.3(a)). A narrow creek, congested with fishing boats, lies across the linear beach opposite the lighthouse where high value storage tanks are located along the shore, near the port. The area 423
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
424 A. Mascarenhas and S. Jayakumar
Tamil Nadu
Figure 35.1.
Location of 24 stations alongTamil Nadu coast where post-tsunami surveys and beach profiles were carried out.
between Keechankuppam and Akaripettai constitutes a packed coastal strip with innumerable permanent houses, which were evidently built on palaeo-dunes. Further south, the stretch up to Puatadi formed a fairly natural coast, except at Velankanni where several make-shift shops are located on the beach front. It is these shops that suffered most damage (Figure 35.3(c)). The pounding the sea front by the tsunami was a result of powerful waves that rose 0.7–6.5 m above sea level causing wave up-rush from 31 to 862 m inland (Chadha et al., 2005; Jayakumar
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Protective role of coastal ecosystems
Figure 35.2.
425
Map of Nagore–Velankanni stretch showing some of the coastal geomorphic and vegetal features.
et al., 2005; Ramanamurthy et al., 2005). The tsunami drastically transformed the coastal landscape of Tamil Nadu (see Table 35.1 for details). Our post-tsunami field surveys in April 2005 confirmed that the sand dunes bore the brunt of high waves; impacts are indicated by various degrees of over wash, breaching, erosion and even flattening of dune complexes (Figure 33.3(b)). New features in the form of lagoons that have formed in the low-lying areas could be seen at Silladi and Kallar. Exposure of black mineral rich sands, as in Nagore and Karaikal, is a new sedimentary feature. All the frontal houses, dwellings and huts are leveled, and Keechankuppam has turned into a disaster zone. Powerful waves over-ran sea walls at Nagapattinam port and tossed the boats inland. The pilgrims at Velankanni and the make-shift structures along the beach were washed off in totality (Figure 35.3(d)). It has been observed that at several points along the Tamil Nadu coast, including the area behind the Nagapattinam lighthouse, roads perpendicular to the coast appear to have facilitated the tsunami run-up in invading the hinterland. The most significant finding, however, is that casuarina plantations and coconut groves remained virtually intact. 35.3
NATURAL HAZARDS AND COASTAL MANAGEMENT
The kind of human suffering witnessed every year following severe cyclones (Table 35.2), and in the aftermath of the tsunami that razed villages such as Keechankuppam and Akaripettai in Tamil Nadu (Table 35.1), constitutes a grim reminder of what happens when our coasts are developed and colonized improperly. The damage inflicted by recurring extreme events over the last 50 years involves a colossal economic loss (Table 35.2). Independent estimates reveal that the financial loss incurred by the super cyclone of October 1999 in Orissa was over 2750 crore rupees (Priyadarshini, 1999), whereas the tsunami of December 2004 resulted in a monetary loss of 512 and 2730 crore rupees in Pondicherry and Tamil Nadu respectively (Datta, 2005) (one crore = 10 million; Ind. Rs. (INR) one crore = ∼USD 0.23 million). The figures shown in
426 A. Mascarenhas and S. Jayakumar
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Table 35.1.
State of coastal landforms along Nagore–Velankanni sea front of Tamil Nadu in November 1998, and landscape changes and impacts recorded after the tsunami of December 2004.
Coast
Geological setting and state of coastal landforms (November 1998)
Landscape changes and impacts after the tsunami (April 2005)
Nagore
Broad sandy beach backed by sand dunes with sparse vegetation, occupied by dwellings and packed huts along high tide line; dunes often obliterated by human actions; active traditional fishing activity Flat sandy beach; prominent high sand dunes seen on either side of entry point; dunes are vegetated; thick casuarina plantations seen; a mosque is located backshore; a road terminates at sea front; a fairly natural coast, used for recreation
Sand dunes flattened; dune vegetation uprooted; heavy mineral layer exposed on beach, now being covered by wind blown sand; natural dune rebuilding observed; only frontal coconut trees damaged; huts washed off A new lagoon is formed; frontal high dunes show differential damage; flattening, breaches and over wash observed; no evidence of frontal dune vegetation; black sands are exposed; only frontal casuarinas trees bent, now bare; road to beach cut off by new lagoon; road served as a pathway for tsunami to travel far inland Sand dunes eroded; a 30-m long single dune remains, but breached; exposed heavy mineral carpet now being covered by wind blown sand; casuarinas only partly affected; all houses on dunes smashed Sand dunes eroded, some breached and collapsed; dune vegetation removed; wooden structures along sea front disappeared; all houses and huts over a large beach tract demolished Beach shows severe erosion; dunes leveled; a new inlet was identified; black sand lies exposed; plenty of wind blown sand accumulations seen; dunes showed signs of restoration; houses,
Silladi
Samanthanapettai
Wide sandy beach backed by conspicuous high dunes; dune plants common; houses and huts are located on dune crests; intense fishing activity seen; road ends at dune line
Nambiar Nagar
Steep, narrow, sandy beach; high dunes present, houses on dunes, crowded fishing settlements noted; extensive traditional fishing observed; road ends on a high dune
Nagapattinam lighthouse
Beach with low vegetated dunes in northern part; lighthouse located within a high compound wall; huge oil tanks located along sea front; dense colonies of
Activities/reasons responsible for impacts/human tragedy Huts along coast too close to waterline; lacked setback and natural protection; mostly human influenced impacts Natural impacts of an extreme event
Houses and huts in particular built on the frontal dune without any buffer; affected by direct wave attack
Inappropriate location of numerous small structures even on the upper beach
Lighthouse wall and oil storage tanks did not have any natural protection
Continued
Protective role of coastal ecosystems
427
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Table 35.1. Continued huts present; a tourist recreational area
huts, 100 m of compound wall destroyed; tower crashed; foundations of oil tanks undermined, eroded
Nagapattinam port
Beach interrupted by an inlet; sea facing narrow creek served as a harbor for trawlers and boats; a sea wall borders the creek; a bridge spans this creek
Inlet promoted tsunami movement; sea wall overrun by high waves; boats tossed, lying scattered on land
Fishing harbor was not a sheltered one
Keechankuppam
Beach with dense dwellings; sand dunes razed for construction of houses
All houses in ruins; road under a layer of sand; resembles an abandoned village
Dense houses too close to waterline; no suitable setback; affected by direct run-up
Akaripettai
Beach, densely populated strip; no evidence of sand dunes, except at southern end
Huts, houses and infrastructure destroyed; a patch of casuarinas are still standing; a deserted coastal strip
Crowded hamlets along the sea front; no suitable setback; impacted by direct wave attack; a manmade disaster
Kallar
A wide stretch of very low, flat, flood-prone sandy area, >300 m long; a narrow motorable N–S road crosses this zone
A new, persistent water body occupies the entire area; a narrow inlet is a new feature; the lagoon has linked itself with the existing river inland; the road is fully submerged; columns of new bridge (under construction) dislocated
Natural impacts of an extreme event
Velankanni
A large, wide, flat beach, fully occupied by thatched, make-shift restaurants, shops, shacks up to the water line; a markedly degraded strip; a shrine is located backshore
An elongated lagoon linked by an inlet was noticed towards south; Improvised shops and structures washed off in totality; a wide, flat beach now lies bare
Absence of a buffer zone; flouting of coastal regulations; unplanned shopping area; lack of awareness of natural hazards
Puatadi
A pristine coast with well developed, high, vegetated, sand dunes; thriving casuarina forests
Dune erosion, breaching due to over wash; casuarina trees unaffected
Natural impacts of an extreme event
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
428 A. Mascarenhas and S. Jayakumar
Figure 35.3.
Landscape changes before (November 1998) and after (April 2005) the tsunami: (a), Human occupation of dunes at Nagore (November/1998). (b) All huts and shacks flattened by the tsunami; only coconut groves survived (April/2005). (c), Crowded improvised structures on the beach opposite the shrine at Velankanni (November/1998). (d) Make-shift structures washed off in totality by violent waves (April/2005).
Table 35.2 are only estimates, as the overall losses may be of the order of tens of thousands of crores (billions of USD). The coastal residents of eastern India frequently face extreme oceanographic events in the form of storm surges that devastate large areas. Now the December 26 tsunami destroyed entire coastal villages within an hour. Every such episode derails the economic activity of the region. Unfortunately, most of the people that live in the low-lying coastal areas comprise farmers or fisher folk who live in mud houses or thatched huts. This section of society bears the recurring loss of humans, livestock and property. Prudent development of coasts is the obvious solution. The inherent benefits offered by a functional sea front have to be considered for a sustainable management of hazard-prone coasts. Knowledge of coastal geological processes is gaining importance mainly because coasts offer a natural buffer protection against oceanic forces (Nordstrom, 2000; Pilkey et al., 2000). When coastal processes are ignored and natural protection entities (such as, mangroves and coral reefs) are destroyed, the vulnerability to natural hazards increases. The oceanographic significance of coastal buffer zones is based on the fact that the shore front is a site where extraordinary energy is at play. Structures that come directly in the path of oceanic forces are severely affected. As a result, hard structures along coastal strips alter, enhance or suppress coastal geological processes, and impacts are thus multiplied (Clark, 1996; Pilkey et al., 2000). That is why sustainable coastal development has to take into account the strength
Protective role of coastal ecosystems Table 35.2.
Monetary loss (crops, property and infrastructure) incurred and deaths reported as a consequence of some of the extreme events along the Indian coasts during the last 53 years.
Year of event
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
429
Place
November 1952 October 1955 December 1964
Nagapattinam Kalingapatnam Rameshwaram
November 1969 December 1972 November 1977 November 1977 October 1983 October 1987 November 1988
Kakinada Cuddalore Nagapattinam Nizampatnam Bheemunipatnam Ongole 24-Parganas (N) 24-Parganas (S) Medinipur 24-Parganas (N) 24-Parganas (S) Medinipur Machilipatnam Cuddalore 24-Parganas (N) 24-Parganas (S) Medinipur Kakinada Digha Khajuri Hugli estuary Balasore–Paradip Kakinada Pondicherry Karaikal
May 1989 May 1990 November 1991 May 1995 November 1996 August 1997 October 1999 August 2000 December 2004 (tsunami)
Cuddalore Nagore Samanthanapettai Nagapattinam Velankanni
Damage to embankments (km)
Overall monetary loss (crores, INR)
Human loss
– – (Train swept off) – – – – – – 2 – 20 22 160 59 – – 77 78 12 – <1 – – – – – (Sea wall uprooted; bridge column broken) – – (Rail tracks breached) –
6.00 1.00 8.00
400 – 900
110.00 40.00 155.00 350.00 520.00 34.00 02.10 00.69 06.36 15.63 41.06 02.37 2248.00 323.00 11.17 33.00 23.86 150.00 75.12 00.05 14.00 >2750.00 776.75 512.00
∼900 80 560 10,000 120 48 – – – 485 – – 967 201 – – – 2000 400 – – 9885 131 107 484
2730.00
606 6629 ∼900
(Source: Shrestha, 1998; Priyadarshini, 1999; Raghavan and Sen Sarma, 2000; Thapliyal et al., 2000; Paul, 2001; Mascarenhas, 2004; Datta, 2005).
of natural energy operating at the shore. Coastal systems need spaces and liberty to function and evolve. Therefore, adopting adequate buffer zones by retreating inland is a sensible long-term sustainable management option for low-lying coasts prone to extreme events (Pilkey et al., 2000; Valdiya, 2001; Mascarenhas, 2002, 2004).
430 A. Mascarenhas and S. Jayakumar
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
35.4
COASTAL ECOSYSTEMS AS “NATURAL SHOCK ABSORBERS”
Fundamentally, coasts dissipate oceanic energy on their mudflats, beaches, dunes, marshes or mangroves. A natural sustainable coast is the one that reduces the rate and intensity of flooding and erosion. This is a basic function that must be recognized and promoted. Importantly, so as to defuse energy, coasts need space to operate. But simultaneously, human use also requires space. It is this state of affairs that results in conflicts that ultimately end up in the unsustainable use of the coastal zone. Therefore, the central issue in sustainable coastal management is that coasts need buffer zones to function naturally, and shore front development should be harmonized with it. Natural coastal ecosystems such as sand dunes, coastal forests and wetlands are capable of attenuating the forces of extraordinary oceanic events. Each geomorphic and ecological unit has a specific role to play and functions to perform. As such, coasts protect mankind. 35.4.1
Coastal sand dunes
Sand dune complexes act as Nature’s natural protection, and are called the last line of defense between oceanic forces and inland property. As natural geomorphic barriers, dune ridges protect the hinterland from wave run-up due to extreme oceanographic conditions. Wide beaches and high dunes act as efficient dissipaters of wave energy. Sand dunes therefore serve as stores that strong waves draw on during extreme events. Society at large is immensely benefitted. The role played by sand dunes and beach gradients during the tsunami run-up is illustrated in Figure 35.4. Nanjalingampettai and Samanthanapettai are characterized by high dunes and steep seaward gradients. The wave up-rush of 3.7 and 3.2 m respectively induced inundations up to 150 m (Jayakumar et al., 2005). At Samanthanapettai, the over wash was nominal. It is pertinent to note that all the assets behind dunes remained intact due to the natural buffer protection offered by sand dunes. Similarly, at Urur Alcot Kuppam near Chennai, although the beach is relatively flat, a 2-m high compound wall was spared despite a 5.0 m run-up, owing to a steep gradient with thick bushes in front of it. In comparison, Tarangambadi and Velankanni shores were occupied by rows of dense dwellings and make-shift structures respectively. Sand dunes were flattened and colonized. The wave run-up of 2.4 and 3.9 m bypassed the flat beach thus razing whatever came its way. Inland inundation was 401 and 325 m respectively. Roads over dunes, which are perpendicular to the coast, also contributed towards the invasion of tsunami waters further inland. As they lacked natural protection, all the beach front houses were washed off totally. Globally, dune fields have offered protection in the wake of extreme events (Clark, 1996; Nordstrom, 2000; Pilkey et al., 2000). The Digha coast of West Bengal is a typical example where urbanization of sandy coasts resulted in leveling, and hence large-scale removal of sand dunes. A consequent sediment deficit over time resulted in man-induced erosion and shoreline recession. Inundations and destruction of habitations, dwellings and resorts during extreme events, as a result of human interference on sand dunes in West Bengal, is well documented (Paul, 2001). Obviously therefore, sand dunes need to be conserved and nourished for which dunes should be allowed and be able to migrate, develop and evolve freely and naturally in form and space. 35.4.2
Coastal vegetation
Published research indicates that coastal vegetation plays a distinctive role in shore front dynamics (Clark, 1996; Ali and Chowdhury, 1997; Pilkey et al., 2000; Nayak et al., 2001). Sand deposition is induced by beach and dune grasses that baffle winds, creating higher, wider and laterally continuous dunes. Marsh grass also traps sediments allowing the marsh to build upwards. Absence of vegetation leads to unstable coastal sedimentary landforms (Clark, 1996). Our post-tsunami surveys confirmed that the damage to casuarina plantations (Figure 35.5) and coconut groves (Figure 35.3(b)) by the tsunami onslaught was minimal. The frontal casuarina
Protective role of coastal ecosystems
431
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Uroor Alcott Kuppam (CHENNAI)
Nanjalingampettai
Tarangambadi
Samanthanpettai
Velankanni
Figure 35.4.
Profiles of the beach at five stations along Tamil Nadu coast. The arrow indicates run-up heights at each location.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
432 A. Mascarenhas and S. Jayakumar
Figure 35.5.
Casuarina plantations served as excellent buffers against the tsunami onslaught at Nanjalingampettai (a) and at Karaikal (b); these trees remained intact all along the Tamil Nadu coast (April 2005).
strips, around 10 m in width, were attacked, bent and stripped of their leaves by wave up-rush. These vegetal species are largely intact and healthy. This phenomenon was verified in Puatadi, Samanthanapettai, Silladi, Karaikal, Nanjaligampettai and Mahabalipuram along Tamil Nadu coast, and Karaikal and Periyakalapet in Pondicherry. Sea water intrusion was negligible in areas covered with thick vegetation. Similarly, tsunami damage was minimal on shorelines fringed by mangroves (Chada et al., 2005) as compared to bare lands of Andamans (Jain et al., 2005). It is known that Sunderban mangrove areas of Bengal suffer less from wind and surges (Ali and Chowdhury, 1997). Casuarina trees survive wind speeds of 100 km/h, and the protective role of mangrove during cyclones has been demonstrated by post-cyclone imageries of Orissa coast (Nayak et al., 2001). This evidence supports bio-shields as energy dissipaters during powerful oceanographic events. The efficiency of protective bio-shields depends on a progression of species landward from the shore (Clark, 1996). Herbs – shrubs – bushes – trees form a gradation and a natural slope wherein winds get deflected upwards and onrushing waves decay quickly. Bangladesh has planted fringing mangroves to protect the occupied coastline (Clark, 1996; Ali and Chowdhury, 1997). Casuarina plantations along Karaikal, Nagore, Nagapattinam, Puatadi coast (Mascarenhas, 2004), the East coast highway, Kakinada–Uppada sector constitute coastal forests that offer protection as shelter belts. Bio-shields are useful for various reasons: (a) Shrubs control erosion and stabilize the shore; (b) Green belts significantly alleviate wind energy thus protecting the hinterland from oceanic forces; (c) A green belt of trees effectively reduces the force of devastating storm surges and waves; (d) Trees are beneficial for biodiversity and can induce habitats for wild life (Clark, 1996); (e) People inhabiting hazard-prone coasts would benefit from green belts in terms of security, access to food, materials, shelter and income (Clark, 1996); and (f ) Strips behind the green belts serve as areas of peace and tranquility. In brief, the physical and geological processes are more intense along the open ocean. Manmade modifications of the coast thus experience extreme processes due to lack of natural protection. Such sites become vulnerable to natural hazards (Pilkey et al., 2000). In comparison, the forested hinterland is sufficiently protected from physical forces as the gradation of vegetal species from the coast towards the hinterland forms a natural slope that forces winds to deflect upwards and onrushing waters to absorb energy. Therefore, protective vegetation (and elevation) is the only coastal environment where risks of impacts due to extreme events are modest.
Protective role of coastal ecosystems
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
35.5
433
NEED FOR COASTAL HAZARD POLICIES: THE INDIAN CONTEXT
In 1991, the Government of India declared coastal stretches influenced by tidal action as Coastal Regulation Zone (CRZ) (MEF, 1991). Of the four categories, CRZ I comprises areas that are ecologically sensitive and important, such as national parks/marine parks, sanctuaries, reserve forests, wildlife habitats, mangroves, corals/coral reefs, areas close to breeding and spawning grounds of fish and other marine life, areas of outstanding natural beauty/historical/heritage areas, areas rich in genetic diversity, areas likely to be inundated due to rise in sea level consequent upon global warming. The main purpose of the Notification is to protect fragile ecosystems from unplanned human interference. Accordingly, no new constructions are permitted up to 500 m in CRZ I and up to 200 m in CRZ III. This notification is the prevailing law that governs developmental activities along the coasts, rivers and backwaters of India. Subsequently, the Science and Technology Policy (STP) 2003 was unveiled with the concurrence of scientists, technologists, politicians and citizens (Sen, 2003). One of the key points of the Policy involves the development of technologies for the mitigation and management of natural hazards wherein a concerted plan for enhancing predictive capabilities and meeting emergencies in natural hazards would be made. Special emphasis would be placed on forecasting, prevention and mitigation of natural hazards. Our post-tsunami field data confirms our earlier view that the existing instruments are deficient in the wake of inundations following extreme oceanographic events (Mascarenhas, 2002, 2004). Except for CRZ 1991 that regulates development, the country does not have a coastal hazard policy. Such anomalies merit attention and call for a dialog particularly in view of the recent tsunami attack that added yet another dimension to the issue of natural hazards on the Indian coasts. Herein lies the need to examine whether new regulations have to be formulated, or whether enforcement mechanisms or implementation strategies need a revamp. The inhabitants of coastal Tamil Nadu lost their lives because they were occupying landforms that they were not supposed to alter (Table 35.1). The prevailing CRZ laws prohibit indiscriminate use of sea side spaces, and sand dunes are fully protected from human interference. Loss of life due to the tsunami invasion occurred exclusively up to 50–80 m from the frontal dune. This strip contained most of the destroyed dwellings inhabited largely by the fishing community. Since CRZ is routinely violated, these constructions lacked designated setback and consistent natural protection. As such, complete destruction of sea front houses and make-shift structures occurred at Velankanni, Keechankuppam and Akaripettai – in particular, Nagapattinam, Samanthanapettai, Nagore, Karaikal, Tarangambadi, Periyakalapet, Poompuhar and Besant Nagar at Chennai. Coastal response against natural hazards comprises several options of adaptation (Klein et al., 2001). In the Indian situation, adaptation in the form of managed retreat is a viable solution to counteract hazards. Such a strategy has to consider increasing setback zones, shifting of buildings, no development in susceptible areas, relocation inland, realignment, creating forested buffer zones, hazard insurance, appropriate land use, regulation of hazard zones and improved drainage. In view of recurring natural hazards along Indian coasts, Valdiya (2001) advocated a public instrument for hazard management. Policy plans include enforcing laws, restrictions and regulations for preventing unplanned development in lands prone to natural hazards. An effective way to protect hazard-prone tracts would be to impose a series of disincentives. Alternatively, governments can acquire and convert sea side strips into recreation parks, forests, gardens, agriculture farms, wild life sanctuaries, by forbidding constructions. Therefore, productive use of vulnerable coastal land can be made while simultaneously reducing the degree of risk. Further, there is need for legislation enforcing disincentives, for which the setting of a National Commission of Hazards Management is proposed (Valdiya, 2001). Coastal hazards do not figure adequately in the Indian environment Acts. The CRZ which includes areas likely to be flooded by sea level rise should necessarily include protection to
434 A. Mascarenhas and S. Jayakumar
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
lowlands prone to inundations due to extreme oceanographic episodes. Natural processes active in the coastal zone need to be taken into account in formulating legislation.. Moreover, the benefits of soft protection are gaining international prominence. Therefore, an ideal coastal hazard policy for India is the one that (a) encourages the most ideal low-risk development, and (b) recognizes coastal geological processes, preserves natural coastal landforms and promotes afforestation. In summary, a human disaster of such a magnitude as the one at Nagapattinam could have been averted, or at least minimized, had the inherent natural protective value of geomorphic features and forested shelter belts been understood, had appropriate policies for natural coastal hazards been put in place and had the CRZ been enforced by locating the coastal inhabitants and dwellings further inland. ACKNOWLEDGEMENT The authors are grateful to the Director, NIO, Goa, for permission to publish this paper. Reviews by Drs. P. Vethamony and R. Mukhopadhyay, NIO, Goa, improved the manuscript. Mr. R. Gowthaman helped in the field. Mr. K. Chitari generated computer figures. NIO Contribution no. 4166. REFERENCES Ali, A. and Chowdhury, J.U. (1997). Tropical cyclone risk assessment with special reference to Bangladesh. Mausam, 48, 305–322. Chadha, R.K., Latha, G., Yeh, H., Peterson, C. and Katada, T. (2005). The Tsunami of the Great Sumatra Earthquake of M 9.0 on 26 December 2004 – Impact on the East Coast of India. Curr. Sci., 88, 1297–1301. Clark, J.R. (1996). Coastal zone management handbook. Lewis Publication, USA, p. 694. Datta, S. (2005). Little men, big water. Outlook, the weekly news magazine, 17 January, 2005, 28–35. Jain, S.K., Murty, C.V.R., Rai, D.C., Malik, J.N., Sheth, A. and Jaiswal, A. (2005). Effects of M9 Sumatra Earthquake and Tsunami of 26 December 2004. Curr. Sci., 88, 357–359. Jayakumar, S., Ilangovan, D., Naik, K.A., Gowthaman, R., Tirodkar, G., Naik, G.N., Ganeshan, P., Mani Murali, R., Michael, G.S., Ramana, M.V. and Bhattacharya, G.C. (2005). Run-up and inundation limits along southeast coast of India during the 26 December 2004 Indian Ocean Tsunami. Curr. Sci., 88, 1741–1743. Klein, R.J.T., Nicholls, R.J., Ragoonaden, S., Capobianco, M., Aston, J. and Buckley, E.N. (2001). Technological options for adaptation to climate change in coastal zones. J. Coast. Res., 17, 531–543. Mascarenhas, A. (2002). Need for setback lines in coastal zone management: a meteorological point of view. In: Tropmet 2001, Meteorology for sustainable development, Indian Meteorological Society, Mumbai, 564–568. Mascarenhas, A. (2004). Oceanographic validity of buffer zones for the east coast of India: a hydrometeorological perspective. Curr. Sci., 86, 399–406. Ministry of Environment and Forests (MEF) (1991). Declaration of Coastal Stretches as Coastal Regulation Zone (CRZ), Notification, S.O. 114(E), 19 February 1991, 14 pp. Nayak, S.R., Sarangi, R.K. and Rajawat, A.S. (2001). Application of IRS-P4 OCM Data to Study the Impact of Cyclone on Coastal Environment of Orissa. Curr. Sci., 80, 1208–1213. Nordstrom, K.F. (2000). Beaches and Dunes of Developed Coasts. Cambridge University Press, Cambridge, UK, 338 pp. Paul, A.K. (2001). Cyclonic storms and their impacts on West Bengal coast. In: G.V. Rajamanickam, and M.J. Tooley, (eds.), Proceedings of the International Seminar on Quaternary Sea Level Variation, Shoreline Displacement and Coastal Environment, New Academic Publishers, Delhi, pp. 8–31. Pilkey, O.H., Bush, D.M. and Neal, W.J. (2000). Storms and the coast. In: R. Pielke, and R. Pielke, (Eds.) Storms Vol. I. Routledge Hazards and Disaster Series, New York, 427–448. Priyadarshini, S. (1999). Orissa cyclone losses seen at over Rs 2,750 cr. Financial Express, 8 November 1999.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Protective role of coastal ecosystems
435
Raghavan, S. and Sen Sarma, A.K. (2000). Tropical cyclone impacts in India and neighbourhood. In: R. Pielke and R. Pielke, (Eds.) Storms (Vol. I). Routledge Hazards and Disaster Series, New York, pp. 339–356. Ramanamurthy, M.V., Sundaramoorthy, S., Pari,Y., Rao, V.R., Mishra, P., Bhat, M., Usha, T., Venkatesan, R. and Subramanian, B.R. (2005). Inundation of seawater in Andaman and Nicobar Islands and parts of Tamil Nadu coast during 2004 Sumatra Tsunami. Curr. Sci., 88, 1736–1740. Sen, N. (2003). Science and Technology policy – 2003. Curr. Sci., 84, 13. Shrestha, M.L. (Ed.) (1998). The Impact of Tropical Cyclones on the Coastal Regions of SAARC Countries and Their Influence in the Region. SAARC Meteorological Research Centre, Agargaon, Bangladesh, 329 pp. Thapliyal, V., Desai, D.S. and Krishnan, V. (2000). Cyclones and depressions over North Indian Ocean during 1999. Mausam, 51, 215–224. Valdiya, K.S. (2001). Public Policy for natural hazard management. Curr. Sci., 80, 486–487.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 36
Integrated Preparedness Systems
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
U. Aswathanarayana Mahadevan International Centre for Water Resources Management, Hyderabad, Andhra Pradesh, India
36.1
INTRODUCTION
The devastating impact of the December 26, 2004 Indian Ocean Tsunami (also known as the Asian Tsunami, Sumatran Tsunami, Aceh–Andaman Tsunami) on coastal communities has been widely reported in the media. The following table summarizes the enormity of the human tragedy involved (source: Science, 309, 1032, 2005). The data in the Table 36.1 indicates that Indonesia suffered the greatest damage as the epicenter of the earthquake, and area of initiation of the tsunami are located there. There is a scientific explanation for the linear path of the tsunami – why it affected Sri Lanka and the Tamil Nadu coasts, but not Bangladesh and Orissa coasts. Singh (2005) showed that as a consequence of the existence of a lithosphere-scale boundary around the Simeulue Island which continues up to the east of Nicobar Island, the December 26 earthquake rupture which seems to have been initiated west of this boundary, did not cross the boundary to the east, but got propagated northwards up to the Andaman Islands. The same boundary also explains why the after shocks of Mw 9.3 earthquake of December 26, 2004 and Mw 8.6 earthquake of March 28, 2005 did not overlap. 36.2
FRAMEWORK OF PREPAREDNESS SYSTEMS
Population densities tend to be markedly high in coastal belts, due to habitations, industries (particularly oil refineries), fisheries, tourism, naval establishments, harbors, etc. More important, this trend is becoming accentuated, and in the future more people and more installations will be at risk. Currently 23% (1.2 billion people) of the world’s population live within 100 km of the coast. It has been projected that by 2030, half of the world’s population will be living in this zone. Table 36.1.
Populations affected by the tsunami.
Country Indonesia Sri Lanka India Thailand Burma (Myanmar) Maladives Malaysia Bangladesh Total
Number dead
Number missing
128,645 31,229 10,749 5395 90 82 68 2 176,260
37,063 4100 5640 2845
532,898 516,150 647,559
26 8
21,663 8000
49,682
1,726,270
437
Number displaced
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
438
U. Aswathanarayana
Apart from risks shared with others, the populations in this zone are exposed to coastal flooding, tsunamis, and marine-related infectious diseases. For obvious reasons, beach-based tourism will not move inland, whatever danger it faces from tsunamis. Simply put, the design of the preparedness systems for the tsunami is based on our understanding of what happened and why it happened that way. Aswathanarayana (2005) gave an account of the various possible preparedness and mitigation systems. An important component of preparedness systems is the application of dual-use technologies and practices. This can be illustrated with a simple example. A seawall to protect the coast against the tsunami has no other use except providing protection against a tsunami. But a bioshield involving (say) mangrove species or salt-tolerant plants, such as, Salicornia atriplex and casuarina, not only give protection against the tsunami, but are also socially, economically and environmentally beneficial, even if no tsunami occurs. Civil engineering structures to reduce the impact of tsunami include specially designed ditches, slopes and berms to slow down the tsunami, strategically placed angled walls, ditches and paved surfaces to steer the tsunami wave away from high-value sites, and hardened structures, such as walls and compacted structures to block the tsunami. These are no doubt expensive, but may be needed in some situations. For instance, it has been reported that the town of Yoshihima in Japan which has been destroyed by tsunami in 1896, built a 800 m long and 36 m high wall to protect itself. Sundar (2006, this volume) reports that the groins built to protect against severe beach erosion in the Chennai–Ennore belt in Tamil Nadu, India, markedly reduced the impact of the December 26, tsunami. The extensive damage suffered by the Nagapattinam–Cuddalore part of the Tamil Nadu coast is attributed to the narrow and gentle shelf and the fault-controlled concave nature of the basin (Murthy et al., 2006, this volume). 36.3
DUAL-USE TECHNOLOGIES AND PRACTICES
Resiliency has been defined as the “capability of a system to maintain its function and structure in the face of internal and external danger. . .”. Allenby and Fisk (2005) recommend the development and implementation of dual-use technologies that provide substantial economic benefits in addition to resiliency. Urban systems routinely use dual use technologies, practices and systems, even if no negative events occur. The developed countries are already highly urbanized, and the developing countries inAsia, Africa and LatinAmerica are getting rapidly urbanized. It has been estimated that by 2030, 60% of the world’s population will live in cities. Consequently, information-dense urban structures are coming into vogue. Instead of rigid systems (e.g. land phone), more fluid and responsive, network-centric organizational patterns are emerging (e.g. Internet phones). A network-centric society is more equitable (e.g. employment of handicapped persons, housewives), more productive (less commuting) and less fragile. The urban systems have necessitated the development of tools that aggregate and display complex urban systems data. The availability of such a system can serve as a training tool for disaster management, besides being useful in the coordination of disaster management (Allenby and Fisk, 2005). A good example of dual-use technology is “Mission 2007: Every Village a Knowledge Centre”, launched by the Government of India. The Mission consists of an ambitious plan for establishing Information and Communication Technology based Village Knowledge Centers in each of the 100,000 villages in India. At least two persons, one woman and one man, will be trained to run each of these centers, which will serve a variety of developmental purposes, such as, health, drinking water, agriculture, hazards, guaranteed employment for 100 days for one member of the families which are below the poverty line (one USD per day), etc. The broad-band connectivity of the
Integrated preparedness systems
439
Village Knowledge Centre allows these centres to be used effectively in all hazard preparedness, public education, public warning, etc. campaigns.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
36.4
RESILIENCY LINKED TO SOCIAL–ECOLOGICAL SYSTEMS
Resilience to coastal disasters is related to the capacity of linked social–ecological systems to absorb recurrent disturbances, such as hurricanes, tidal waves, tsunami, etc. so as to retain essential structures, processes and feed-backs. Just as a healthy man is less vulnerable to disease, and more capable of recuperation after he is affected by a disease, the healthier an ecosystem is the more capable it is of regenerating itself and continuing to deliver resources and ecosystem services that are essential for human livelihoods and societal development (Adger et al., 2005). In coastal Asia, the resilience of the ecosystems have been degraded by activities, such as, deforestation of mangroves for intensive shrimp farming, overfishing, coral mining, land clearing, etc. Consequently, the adverse consequences of the tsunami got accentuated in some places in Sri Lanka and Tamil Nadu coast. Similar considerations hold good for social systems. Thanks to the inherited knowledge of tsunamis, and institutional preparedness for disasters, fishing communities in Simeulue Island, west of Sumatra, close to the epicenter of the December 26, earthquake, survived the tsunami better than the fishing communities (say) in the Tamil Nadu coast which have no such knowledge or tradition. The Cayman Islands in the Carribean which suffered three devastating hurricanes during 1988–2000, has put in place specific rules and governance of hurricane risk, dedicated organizations, establishment of early warning systems, promotion of self-mobilization in civil society and private companies, public infrastructure, coral reef management, etc. This effort paid off when the Cayman Islands was hit by Hurricane Evan in 2004 (Adger et al., 2005). Mangroves of Pichavaram along the Tamil Nadu coast saved the lives and property of the people living in seven hamlets. Some mangroves on the coast were uprooted, but beyond that, there was hardly any damage. Apparently, the velocity of the tsunami wave got drastically reduced due to friction with mangrove forest, and the tsunami water got dispersed into the creeks and canals that crisscross the mangrove forest. A fisherman said, “ we saved the mangroves, and the mangroves saved us” (MSSRF note). Similar observation has been made in Sri Lanka about the protection afforded by coastal ecosystems, such as mangroves and corals, against the tsunami. 36.5 TSUNAMI RISK MANAGEMENT THROUGH SECURITIZATION World Bank studies have shown that natural catastrophes have a strong adverse impact on the economic development of the low-income countries, and can offset poverty reduction activities. Raising taxes by a government is never popular, and people who will never be affected by a tsunami, will be totally unwilling to pay taxes to provide tsunami protection. The main purpose of the tsunami risk management is to minimize disaster losses in hazardprone areas. These management alternatives may be structural (such as, seawalls, groins, bioshields) and nonstructural (such as, Catastrophe Bonds or cat bonds). Such actions are taken before or in advance of the event (ex ante), and are, therefore, proactive measures. A comprehensive tsunami risk management program is usually a combination of structural, nonstructural and reactive actions. Reactive actions are those taken once the event has occurred, such as evacuation and relief. Pizarro and Lall (2005) gave a detailed account of the use of innovative insurance systems, such as cat bonds, for unpredictable, disastrous events. Figure 36.1 shows how the risk premium amounts are determined depending upon the function of the size of the risk community. Stipple (1998) gave a lucid account of the methodology of securitization, through the use of catastrophe or cat bonds. The institutional structure for a catastrophe bond is shown in Figure 36.2. To date,
U. Aswathanarayana
3%
2%
Hazard: Affected every:
100– 200 years
200– 300 years
Risk premium per zone
3.5‰
1.6 ‰
Risk premium when all property in relevant zone is insured for the same premium
1%
300– 400 years
400– 500 years
500– 1000 years
1000 years
0.5 ‰
0.2 ‰
0.05 ‰
0.8 ‰
Compulsory insurance
Risk adequate premiums as a function of the size of the risk community. (Source: Menzinger and Brauner, 2002.) Supplemental interest expense
Government or Corporation
Reimbursement contract
Special Purpose Finance Vehicle, SPFV
Catastrophe bond
or
Interest and principal
Special Purpose Reinsurance Vehicle, SPRV
Investors
Proceeds
Interest
Figure 36.1.
92%
1%
1%
3.5 ‰ 2.8 ‰ 2.4 ‰ 2.2 ‰ 1.9 ‰ 0.2 ‰
Proceeds
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Property distribution
Principal
440
Trust fund
Figure 36.2.
Institutional structure for a catastrophe bond (cat bond). (Source: Stipple, 1998.)
cat bonds have been issued for earthquakes in California and Japan, hurricanes in the coast of the USA, typhoons in Japan, and windstorms in Europe. Presently, the volume of cat bonds is about USD 6 billion. Mills (2005) proposed an insurance mechanism, based on the globalization of risk, to cover the disasters arising from climate change. A similar system may be developed to take care of the tsunami risk. Insurance Linked Securities (ILS) are the most common form of cat bonds. A “Special Purpose Financing Vehicle” (SPFV), or “Special Purpose Reinsurance Vehicle”( SPRV) is created for the purpose. There are two kinds of bonds, namely, “principal at risk” or “interest at risk”. If a catastrophe occurs, the investors will lose their principal (the invested capital), or part of it, or only the interest, depending upon the type of cat bond purchased.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Integrated preparedness systems
441
It has been estimated that out of the global economic costs of USD 1400 billion due to weatherrelated disasters during the period 1980–2004, only USD 340 billion were insured. Climate change is exacerbating the economic losses. Some forward-looking insurance companies (like Swiss Re) are attempting to reduce their losses by promoting the protection of mangroves, reefs and wetlands that buffer storm surge and wave risks, and funding the development of energyefficient technologies. This is akin to life insurance companies promoting practices to increase life expectancy. In the aftermath of December 26, tsunami, the developing countries which were affected by it, sought financial assistance of the donor countries. Linnerooth-Bayer, Mechler and Pflug (2005) suggest novel insurance instruments to be adopted by the donor community for assisting the poor countries before the disasters happen. Under this arrangement, the donor community transfers the catastrophe risks to global financial markets. The proposed donor-supported risk transfer programme would on one hand reduce the disaster aid demands from the donor countries, while freeing the poor countries from the vagaries of post-disaster assistance, on the other. Recent advances in computer modeling have made it possible to better estimate and price low-probability extreme events for which there is limited historical data. The proposed scheme seeks to complement post-disaster humanitarian aid, with pre-disaster programmes of preparedness and risk transfer. 36.6
FUNDAMENTAL STUDIES
The total damage caused by a tsunami is a function of two kinds of biophysical factors. First is the cause, mode of generation, location, and geological environment in which a tsunami is generated; this would determine the how and with what energy and velocity the wave train would move. Second is the bathymetry and geomorphology of the coast concerned. A variety of fundamental studies need to be made to understand the tsunami-causing processes in the Indian Ocean region: (i) Identification and location of the possible causes and potential sites where tsunamis could be triggered. The stress map of the region, downloadable from http://www.world.stress.map.org, could be a good starting point. (ii) Location of sites which are prone to submarine landslides, such as submarine canyons with high slopes and thick sediment cover. (iii) Probability studies of the temporal and spatial distribution and possible magnitudes, characteristics, dynamics, movement patterns, etc., of the potential tsunami. (iv) Coastal bathymetry and the shoaling effect. When the tsunami wave is moving through the deep ocean, it may have a wavelength of hundreds of kilometers but a wave height of a few tens of centimeters. When such a wave approaches the coast, it slows down depending upon the depth of the water, and increases greatly in height (to 10 m or more), without losing energy. (v) Land cover and land use of the coast. A rocky coast or a coast with extensive mangrove stands will suffer less damage than a sandy coast without mangroves. (vi) Modeling and simulation studies involving the above parameters. Three-dimensional modeling of the propagation of the tsunami is horrendously difficult, as the wavelength of the tsunami depends upon the water depth along the propagation path, which may not be known with sufficient accuracy. As regards the coastal bathymetry, almost all countries zealously guard detailed information about this on grounds of security. 36.7
MONITORING AND WARNING SYSTEMS
Unlike the Pacific Ocean where there are numerous sites from which tsunamis can be triggered, the Indian Ocean has only two belts of possible tsunamigenic sites – Sumatra–Andaman fault
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
442
U. Aswathanarayana
belt wherefrom the December 26, tsunami got originated, and the Makran fault off Pakistan. The Indian Ocean Tsunami Warning and Mitigation System (IOTWS) which is being planned involves the participation of Australia, India, Indonesia, Iran, Malaysia, Pakistan and Thailand, and envisages the eventual installation of about 20 DART (Deep Ocean Assessment and Reporting for Tsunamis) systems in such a way that there are no gaps. The protocols for data formats, transmission, analysis and storage, etc. are being standardized. A DART system comprises of pressure-sensitive sensors located on the ocean floor (which detects the greater pressure of a passing tsunami). This information is passed on to sondes mounted on floating buoys, which are linked to satellites and monitored continuously. India would also be putting in place 20–25 automated sea-level gauges. In future, it should be possible to identify the location and magnitude of the earthquake in less than 3 min of its occurrence, and provide tsunami warning in less than 30 min. Countries in the Indian Ocean region are in the process of developing their own national detection networks, their own risk assessment and preparedness plans and their own public education and awareness campaigns. According to Eddie Bernard of NOAA (EOS, 4 January 2005), the accuracy of detection of the currently available systems is 0.5 cm and the numerical models for tsunami forecasting have an accuracy of about 80%. Bernard cautions that while smaller tsunamis generally behave in a linear way, the behavior of larger tsunamis tends to be nonlinear. Monte Carlo calculations may be useful for this purpose. As is well known, the height of the tsunami waves increase sharply as it approaches the coast because of the sea becoming shallower near the coast. In situations where the sea floor has a gentle slope, it is possible that high-frequency shore-based radars which are used to monitor storm surges on the basis of the measurement of surface currents and wave characteristics, could provide the advance warnings of tsunami. It has been reported (Roder et al., 2005) that strong electrical signals corresponding to the M9.3 Great Sumatra Earthquake of December 26, which took place at 0058:50.7 UTC, were picked up by an electrostatic sensor in Tuscany, Italy, almost immediately after the event, as the electric signals travel at the speed of light. The p-waves from the same earthquake reached the station 740 s later. This suggests the possibility of increasing the hazard alert window. 36.8
GENERAL STRATEGY TO MEET THE TSUNAMI THREAT
As tsunamis are comparatively rare events, the expenditure on preparedness should be commensurate with the probability of risk. A tsunami cannot be prevented, but through an understanding of the processes involved in tsunami formation and propagation, the damage it can cause can be minimized. The design of a cost-effective strategy for preparedness for tsunamis has to take into account two attributes of a tsunami, namely, its genesis and its impact: a tsunami is triggered by earthquakes, volcanism, submarine landslides, asteroid impacts, etc., and a tsunami is similar to the manifestations of a tidal wave in its impact on the coasts. This consideration should be kept in mind while designing a cost-effective strategy for the preparedness for tsunamis. The strategy may have the following goals: • enhance the warning time as much as possible by the study of the electromagnetic, ionospheric, thermal, geochemical, geotechnical and geobiological signatures of earthquake precursors; • develop methodologies for the identification of earthquakes which are capable of triggering a tsunami, while discriminating them from false alarms; • incorporate the tsunami impact as an add-on to the existing scientific and administrative arrangements for the warning and mitigation of storm surges, which are quite common. The pre-hazard, during hazard, and post-hazard phases of the preparedness and mitigation systems for the triad, tsunami–earthquake–cyclone, have both science-based and people-based
Integrated preparedness systems Table 36.2.
Science-based and people-based preparedness systems. People-based preparedness and mitigation systems
Science-based preparedness and mitigation
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Prehazard
During hazard
Posthazard
443
1 Earthquake precursors may have electromagnetic, thermal, geotechnical, geobiological, geochemical, etc. signatures. R&D is needed to identify these signatures which have the potentiality to lead to a longer warning time (1–2 days?) for an earthquake, and the resulting tsunami. 2 Network of detection systems on the seabed, on the sea surface and on land. 3 Location of the possible sites of earthquakes, volcanism, landslides, etc. 4 Probability studies of the temporal and spatial distribution and possible magnitudes, characteristics, movement patterns of potential tsunami , based on the ocean depth, coastal bathymetry, land use and land cover of the coast, etc.). War games-type simulations, involving Monte Carlo calculations. 1 The detection of (say) ∼M7 submarine earthquake should automatically activate a tsunami-warning system, which will project, based on previously made desk studies, when, where and how the tsunami is going to hit. 2 Monitoring of the tsunami with GOOS satellites. 1 Damage assessment by remote sensing, involving the nature and extent of damage, how and when it occurred, prioritization based on the severity of damage. 2 Long-term planning involving the identification of rehabilitation sites, coastal zone regeneration, identification of sensitive coastal environments, etc.
1 2 3 4 5
Public education. Longer warning time. Warning communication systems. Evacuation drills. Protective bioshields, (mangroves, salttolerant trees) which are ecologically sustainable and employment-generating. 6 Knowledge-based coastal bio-villages. 7 Hazard zoning. 8 Securitization through Catastrophe bonds.
1 Activation of warning communication systems. 2 Evacuation drills. 3 Help to children, women and old people. 4 Safety of animals.
Public–private sector partnership through an integrated psychological, ecological, agronomic and livelihood rehabilitation.
components which need to be integrated (Table 36.2). Data from different spectral bands of the Earth observation satellites, communications satellites and satellite web are useful in addressing all phases of the hazard. 36.9
PUBLIC AWARENESS MEASURES
A significant component of preparedness is public education. Every schoolboy in Japan knows that a sudden recession of the sea is an indication of an imminent tsunami attack and that he should run inland as fast as possible, alerting others while doing so. That this kind of public consciousness is absent in countries like India is evidenced by the fact that in some areas both children and adults were washed away when they rushed to collect fish and crabs lying on the
444
U. Aswathanarayana
exposed seabed. Thus, public education at all levels (school, community, college, etc.) constitutes a critically important part of preparedness. Tsunami hazard zoning, along the lines of earthquake zoning, will be useful not only for regulating civil constructions, but also for designing warning communications systems. Evacuation drills need to be performed periodically to ensure that warning communications and evacuation protocols are viable.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
36.10
REHABILITATION MEASURES
An integrated plan for psychological, ecological, agronomic and livelihood generation in the affected areas is best effected through public–private partnership. A major challenge would be the restoration of the ecosystem and biodiversity, involving the desalinization of agricultural land, water wells, etc. Remote sensing is useful for damage assessment, covering the nature and extent of the damage, how and when it occurred, and prioritization based on the severity of damage, and for long-term planning involving the identification of rehabilitation sites, coastal zone regeneration and identification of sensitive coastal environments. REFERENCES Adger, W.N. et al. (2005). Social-ecological resilience to coastal disasters. Science, 309, 1036–1039. Allenby, B. and Fisk, J. (2005). Toward inherently secure and resilient societies. Science, 309, 1034–1036. Aswathanarayana, U. (2005). Preparedness and mitigation systems for asian tsunami-type hazards. EOS, 86(11), 2005. Linnerooth-Bayer, J., Mechler, R., and Pflug, G. (2005). Refocussing disaster aid. Science, 309, 1044–1046, 2005. Menzinger, I. and Brauner, C. (2002). Floods are insurable Swiss Re, Zurich. Mills, E. (2005). Insurance in a climate of change. Science, 309, 1040–1043, 2005. Pizarro, G. and Lall, U. (2005). Climate drivers, streamflow forecasting and flood risk management, In: U. Aswathanarayana (ed.), Advances in Water Science Methodologies, Taylor & Francis, UK, 134–155. Roder, H., et al. (2005). Great sumatra earthquake registers on electrostatic sensor. EOS, 86(45), 2005. Singh, S.C. (2005). Sumatra earthquake research indicates why rupture propagated northwards. EOS, 86(48), 2005. Stipple, I. (1998). Securitizing the Risks of Climate Change. IIASA, Laxenburg, Austria.
CHAPTER 37
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Social and Political Aspects of Tsunami Response, Recovery, and Preparedness Planning: A Transdisciplinary Approach from Canada C. Amaratunga Women’s Health Research Unit, Institute of Population Health, Ottawa, Canada H. Smith Fowler Department of Epidemiology and Community Medicine, University of Ottawa, Ottawa, Canada
37.1 THE SOCIOPOLITICAL NATURE OF DISASTER Disasters are not just meteorological or geo-thermal events. Implicit in the term is the perceived impact of these events in human as well as environmental terms. Gist and Lubin (1989) define disaster as “collective stress situations that . . . involve some degree of loss, interfere with the ongoing social life of a community, and are subject to human management.” While the physical force and scale of an earthquake, for example, can be measured on the Richter scale, it is the scale of its destructive force on the built and natural environments and its toll on human life that generally define it as a disaster. Moreover, the “felt” impact and consequence of these events occurs at a variety of levels – on individuals, families, communities, and countries, even globally. As such, disasters may also be seen as social phenomena, or more precisely, as the interaction of natural hazards with social structures and political communities (Drabek, 1986; Dynes et al., 1987; Dynes, 1998; Enarson and Morrow, 1998). That disasters have their foundation in social systems or structures (Quarantelli, 1989) has by now become an accepted paradigm in disaster research (Fothergill, 1998). Its application can be seen in the myriad fields that now encompass “disaster science,” as well as the variety of academic disciplines represented therein. In addition to the study of the physical and technical aspects of disasters at all phases (i.e., at their genesis, impact, reconstruction, and mitigation) by engineers, oceanographers, geologists, meteorologists, etc., the field now also includes sociologists, anthropologists, psychologists, and others who study social processes such as risk perception and assessment, risk management and preparedness, emergency response, psychological impacts, and community rehabilitation (as distinct from reconstruction) and development. The social characteristics of disasters are perhaps most obvious in terms of the ensuing response, which is invariably and of necessity collective and organizational (as well as individual) in scope. However, the impact of disasters is also largely influenced by social factors. As noted in a recent Oxfam bulletin on the 2004 Asian Tsunami, “disasters are profoundly discriminatory, even those that are ‘natural’ rather than man-made. Factors that were present before a disaster, such as poor social conditions, mean that some people will be more affected than others.” (Oxfam International, 2005) Likewise, Enarson and Morrow (1998) note, “vulnerabilities to disaster are not . . . equally distributed . . . . Exposure to environmental hazard and risk of 445
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
446
C. Amaratunga and H. Smith Fowler
catastrophic disaster, like other life chances, are shaped by overarching social structures of caste and class, race and ethnicity, age and physical ability, and sex and gender” (p. 2). Economic, race, age, and gender relations are key characteristics of social systems and key organizing principles of the social life of any society; they shape identity and social interaction, as well as social institutions. Insofar as some groups and communities have less access to fewer resources of quality – where housing is located, for example, or access to means of transportation and communication – then those groups face greater risk in terms of disaster exposure, impact and response, in both the short term and long term. Aguirre et al. (2005) call this a “conflict perspective” to the conceptualization of disaster, whereby “the vulnerabilities plaguing societies are said to reflect chronic and long-standing unequal social power and the accumulation of risks which make disasters inevitable” (p. 3). Indeed, Blaikie et al. (1994) have developed a “pressure model” that measures the root causes (i.e., limited access to power and resources), social lacks, and unsafe conditions of a social group to define the probabilities of its risk of disaster impacts. In the case of the Asian Tsunami, the majority of casualties in Sri Lanka and other tsunamiaffected nations were women, as well as children, and the elderly. In some areas, women accounted for 80% of victims. While biology played some part in this (i.e., less upper body strength to climb or pull oneself to safety), the reasons were primarily social, cultural, and political. Women and girls native to the tsunami-affected countries are not generally taught to swim or climb trees, and their long hair and traditional saris sometimes entangled them in thorn bushes and hampered their escape. Many of the women and children who died were on the beaches bathing or waiting to process the fish harvest, while the fishermen were able to ride out the waves in boats on the open sea. Many more women died trying to save the children and elderly in their care. Survivors have had further hurdles to face: reports have emerged of sexual assault and forced or early marriage for women in the settlement camps, and of the remaining women being further burdened with caring for even greater numbers of orphaned children and elderly persons, while the men get paid for reconstruction work. Women’s access to aid and material resources has been more difficult in the more traditional societies, where resources are scarce and men’s needs are prioritized. This inherent discrimination in disaster impact and response is not unique to the Asian Tsunami. Certainly the recent media coverage of Hurricane Katrina in the southern US has highlighted the disproportionate degree of loss amongst the impoverished residents of New Orleans, who, unlike their more affluent neighbors, could not easily evacuate the city without cars or credit cards at their disposal (e.g., Seager, 2005). As Enarson and Morrow state, regardless of the specific nature of the disaster, “when the dust clears or the waters recede, poor families around the world suffer the greatest losses and have access to the least public as well as private recovery assets” (p. 2). Insofar as women and female-headed households constitute the greatest number of the world’s poor in any society, it therefore holds that they also are the population most at risk (Schroeder, 1987; Vaughan, 1987; Blaikie et al., 1994; Ikeda, 1995). This is also true even after the immediate impact of a disaster. According to the UNDP report, The needs of women in disasters and emergencies (Wiest et al., 1994), “gender bias is . . . characteristic of assistance programs” (p. 5) of donor agencies and governments, due in part to expediency, the “comfort” of paternalism, and uninformed assessments of women’s roles and contributions within their local systems. Gender, then, along with the other key organizing factors of social systems of race, age, class, and economic status, is an important – if relatively ignored – feature of the social experience of disasters. These factors are crucial to understanding the public and private power structures and dynamics that operate before, during, and after a disaster. Moreover, attention paid to preexisting social inequalities is much more likely to yield successful interventions, whether it be for disaster and emergency preparedness, response and mitigation, reconstruction and rehabilitation, or long-term recovery.
Social and political aspects of tsunami
447
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
37.2 THE NATIONAL TSUNAMI FORUM, OTTAWA, 2005 It was out of concern that these key social features of disaster and long-term recovery were receiving insufficient emphasis after the Asian Tsunami of December 2004 that a series of fora were organized in Canada in early 2005. The largest of these was the National Tsunami Forum of April 21–22, 2005, organized by the University of Ottawa, the Ocean Management Research Network, and the Canadian Society for International Health. Its purpose was to build upon the tremendous international outpouring of compassion and aid for the survivors of the tsunami and the sense that more could and should be done by Canadians to assist with medium and long-term recovery efforts. Specifically, some members of the non-government organizations (NGO), academic, and public policy communities shared a sense that the disaster might provide an opportunity for the communities hardest hit to optimize their own recovery through the use of practical, evidence-based, and experiential knowledge; not just to rebuild, but to “build better,” based on the integration of international and local knowledge, research, and experience. Accordingly, the National Tsunami Forum in Ottawa brought together 75 experts from a number of disciplines and a variety of settings – academia, government, NGOs, the private sector, and foreign embassies – to exchange their knowledge of and experience with the tsunami in South and Southeast Asia, and to lay out the elements of a framework that could guide Canada’s involvement in medium and long-term reconstruction and rehabilitation efforts. In so doing, the Forum sought to explore and identify an intersection of diverse but related perspectives, including research, international development aid, disaster relief, community and economic development, engineering and infrastructure development, and public policy. The challenge was to articulate a framework for future action that would be broad enough to encompass the ecological, health, social, and economic dimensions of the rebuilding effort while at the same time focusing on the long-term impacts on individuals, communities, countries, and the region as a whole. In presentations and discussions in small groups and plenary, participants grappled with the unprecedented complexity and scale of this disaster. Several overarching systemic challenges were identified, including the lack of mechanisms for coordination of relief and recovery efforts, the shortage of regional organizations and little history of inter-country co-operation, and the challenges of working with formal power structures or institutions that are unstable or even corrupt. Likewise, it was recognized that traditional community development approaches would need to be reconsidered and adapted in areas where the community no longer exists, physically or socially. Forum participants also recognized that there was tremendous variation in the degree and type of devastation within the tsunami-affected countries. Whereas the tsunami hit only the southernmost region of India, for example, almost three-quarters of Sri Lanka’s coastline were directly impacted. Coastal areas of affected countries were hardest hit; in some areas, however, croplands and river systems as far as 5-km inland were destroyed from excessive salination and/or erosion. Moreover, while overall estimates suggest that 75–80% of the victims were women and children, aid agencies reported tremendous variation in the pattern of human casualties (such as the ratio of adult to child and male to female victims), even in communities in close proximity to one another. Pre-existing social conditions such as social and civil conflict, high levels of poverty, and inadequate health and social service infrastructure further exacerbated the impacts of the tsunami in some areas, and the capacity of specific countries to respond to their population’s needs. Likewise, participants recognized that some priorities for recovery and redevelopment were also country specific. Immediately following the tsunami, for example, the Indonesian government expressed that it – not international relief organizations – would take primary responsibility for relief efforts in that country. Likewise, Thailand expressed interest in receiving specialized technical support such as in seismology and ecosystem management, rather than traditional relief and reconstruction assistance. Nevertheless, participants felt that the tsunami-affected countries have more issues in common than differences. Governments in all affected countries are attempting to balance the multiple demands for reconstruction and rehabilitation, as well as
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
448
C. Amaratunga and H. Smith Fowler
meeting immediate needs. The challenges of restoring the economy and livelihoods to citizens, of redeveloping health and social service infrastructure to meet immediate and longer-term needs, and of rehabilitating and reconstructing the physical environments are faced by all countries in the region. Forum participants reached two different but complementary conclusions on the issue of country-specific challenges. One was that the “felt” impact of disaster is highly contextual, and recovery efforts need to account for, and be based on local needs, capacities, and resources. The second conclusion was that the most effective strategy for tsunami recovery would be one which takes a regional perspective, albeit one that is informed by local conditions. The degree of variation among localities, the commonality of challenges, the importance of regional equity and co-operation, and the opportunity for knowledge exchange – such as for a tsunami early warning system – were felt to be critical arguments for such a regional approach. There was strong consensus that Canada has a role, indeed a responsibility as a world leader, to contribute its expertise to alleviate suffering and devastation in the aftermath of the tsunami. The establishment of regional, relationship-based networks of scientists, academics, and policy decision-makers was recommended as a priority. Based on the collective development experience of those attending the Forum, a number of guiding principles were articulated by participants for Canada’s involvement in disaster rehabilitation, reconstruction, and research. These included focusing on risk reduction and prevention, especially for those most vulnerable; emphasizing resiliency, assets, and capacity development, not just meeting “needs”; ensuring that everyone affected by the tsunami – especially women – has an opportunity to be actively engaged in meaningful recovery efforts; and the need for an ecological approach that recognizes the inter-connectedness of life domains. The latter became known at the Forum as the “Pop-up Principle,” whereby actions to address a hazard in one area – establishing mangrove forests to stabilize and protect coastlines, for example – can unintentionally create a hazard in another, in this instance, by facilitating the spread of malaria. For these reasons, participants at the National Tsunami Forum agreed that once the immediate catastrophe is over, development and research efforts need to look to the broader, long-term environmental, social, economic and health impacts, and to frame future reconstruction and development planning within one conceptual framework. A common framework would help to ensure that all reconstruction efforts and research initiatives are undertaken with due diligence and consideration of systemic impacts, for example, the inter-relation among diverse domains, such as health, economic, and environmental factors. To the extent that a common framework addresses the redevelopment principles outlined above, the more useful and relevant it would be. National Forum participants also agreed that, to be useful, a framework needs to be forward looking, take a global perspective, and be innovative, creative, and flexible. It needs to be evidence based and holistic, and to strike a balance between human needs and those of the ecosystem we are part of and dependent upon. To this end, participants examined other relevant frameworks – in particular population health and ecosystem models – to determine the critical elements of an overarching conceptual framework. 37.3 A POPULATION HEALTH APPROACH The population health model was first developed by the Canadian Institute for Advancement in Research (CIAR), and has been singularly influential over the past decade in redefining health services planning and health policy and program development in Canada at the federal, provincial, and district levels. This model has been institutionalized in the research architecture of the Canadian Institutes for Health Research (CIHR), various university-based Institutes for Population Health, and the Canadian Institute for Health Information (CIHI). Successful Canadian toolkits and training modules related to the development and implementation of community-based
Social and political aspects of tsunami
449
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Determinants of population health Individual factors
Population Health status Determinants of health Personal health practices Social and economic environment
Individual capacity and coping Skills Physical environment
Collective factors Health services
Tools and Supports Research, Information and Public Policy
Foundations for action
Figure 37.1. The population health model.
population health initiatives and public health management, including resource allocation, have been developed and implemented. The population health model is unique in its recognition of the multidimensional influences on health and well-being, especially those influences that are not directly affected by the health system, such as income, literacy, nutrition, environment, and gender. It examines their distribution within a society, as well as the fair distribution of resources that address these socio-demographic, psychosocial, and cultural influences in relation to health outcomes (i.e., health equity) (Figure 37.1). 37.4 AN ECOSYSTEM APPROACH The ecosystem approach is a strategy derived from conservation biology for the integrated management of land, water, and living resources. It promotes the equitable conservation and sustainable use of these resources. It is based on “a collaboratively developed vision of desired future conditions that integrates ecological, economic, and social factors” with a goal of restoring and sustaining “the health, productivity, and biological diversity of ecosystems and . . . overall quality of life” (US Federal Interagency Ecosystem Management Task Force, June 1995). Like the population health model described above, an ecosystem approach is based on the concept of multidimensionality, and the enmeshed, inter-dependent relations among the various
450
C. Amaratunga and H. Smith Fowler
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
sub-systems. It recognizes that the health of any ecosystem (biological, social, or otherwise) is dependent upon the healthy functioning of these multidimensional parts; furthermore, that any action taken upon one component inevitably has consequences on the rest of the system, which are sometimes unpredictable. In addition to these fundamental concepts of multidimensionality and connectedness, there are additional key principles to the ecosystem approach. While these vary in number and degree of specificity depending on the source, they generally include the following: • Intergenerational sustainability is an overall goal. • Ecosystems should be managed for their intrinsic value as well as for human benefit, and that the latter are defined by societal choices; the full participation of indigenous and community populations is recommended. • Diversity and complexity – these strengthen ecosystems by enabling adaptation to change and overall resilience. • Dynamic character – “change and evolution are inherent in ecosystem sustainability” (Ecological Society of America, 1995). • Context and scale vary widely and a long-term view is important to achieve evolutionary potential. • Adaptability and accountability – our current knowledge is necessarily incomplete and must continually be revised and updated; all forms of relevant information should be considered, including scientific and indigenous and local knowledge, innovations, and practices (The Framework Convention on Biological Diversity, 1992). • Management should be decentralized to the lowest appropriate level. • Need to consider the economic context especially by reducing market distortions that affect diversity, aligning incentives to promote conservation and sustainable use, and internalizing costs and benefits within a given ecosystem (The Framework Convention on Biological Diversity, 1992).
37.5 AN ECO-HEALTH APPROACH An eco-health approach essentially applies an ecosystem approach to population health, and in so doing, utilizes the fundamental concepts of both approaches. It recognizes that “the economy, the environment, and the needs of the community all have an impact on the health of an ecosystem and, consequently, the people living within that ecosystem” (IDRC, 2005). Likewise, it is based on the assumption that human health – whether that of individuals or populations – closely mirrors the health of our surroundings, that “the inextricable links between humans and their biophysical, social, and economic environments . . . are reflected in an individual’s health” (Lebel, 2003) (Figure 37.2). By focusing on the context, the eco-health approach presupposes that health is directly linked to our power to control, develop, and use our environment in a sustainable way, or to abuse it. Accordingly, it encourages positive environmental action to enhance health and welfare at the community level, as part of the movement for global sustainable development. When applied to development research, Lebel (2003) identifies the three methodological pillars of the eco-health approach as follows: 1 Transdisciplinarity: This “inclusive vision” requires the full participation of researchers/ specialists, community members, and decision-makers, including government representatives but also “those with informal influence based on their knowledge, experience, and reputation.” 2 Participation: Aims to achieve consensus and co-operation within and among these three groups.
Social and political aspects of tsunami
451
Environment
Health Community
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Economy
Population health
Research, information and policy as tools and supports for action Proven, effective intervention strategies
Scaling up of intervention
Figure 37.2.
Suggested schemata for the application of an eco-health approach to development.
3 Equity: This dimension involves analyzing gender and social roles, recognizing that men and women have different responsibilities, different degrees of influence, and differential access to resources; likewise, “various castes, ethnic groups, and social classes often live in completely separate worlds: this isolation has its own repercussions on health and access to resources.” It can be argued that incorporating these three eco-health principles leads to a more “robust science” that ensures research findings are valid. Results can typically be implemented in a more spontaneous and natural way, since the community, decision-makers, and scientists are directly involved in defining the problem and identifying solutions. Moreover, by investing in long-term sustainable development that is both socially and economically effective, societies and leaders can help citizens to reap tangible health benefits as well (IDRC, 2005). After discussion and reflection, the National Tsunami Forum participants reached agreement that the eco-health approach is the best “fit” with respect to the principles and issues identified earlier in the day for post-tsunami reconstruction and rehabilitation. While the applicability and
452
C. Amaratunga and H. Smith Fowler Socio-cultural factors
opportunities for chan ge ntify Ide Knowledge
Behavioural factors
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Global change
Understanding the determinants of ecosystem and human health
Improved health
Natural resource management
Societal response to health and human development needs
Economic factors
Human resource management Environmental factors
Figure 37.3.
Development of policies
Empowerment
Op
ti m i
z e inte
r v e n ti o n s f o r
cha
nge
Iterative research strategy for improving human health using a participatory and transdisciplinary approach (Forget, 1997).
portability of the eco-health research framework to the real-world context was not immediately clear to everyone, there was a strong commitment to see the common analysis and suggestions develop into tangible action. Accordingly, a Call to Action was developed which captures the recommendations of the day. This was submitted to senior officials in the Canadian government, academic, and NGO sectors. It represents the founding principles of a strategy for long-term development assistance and post-tsunami reconstruction and rehabilitation in South and Southeast Asia (Figure 37.3).
37.6 A CALL TO ACTION Whereas the participants at the first National Tsunami Forum in Halifax, January 21, 2005 and the second National Tsunami Forum, Ottawa, April 21–22, 2005 collectively represented a national gathering of over 100 experts and key decision-makers from the government, NGO, academic, and private sectors in the area of international development and disaster relief, participants are calling on the federal government to . . . • Assist with the coordination of relief and recovery efforts among the various levels of government, NGO, academia, and the private sector, such as through the development of a National Clearinghouse or Databank of Tsunami Redevelopment Initiatives. • Adopt a long-term perspective for current and future reconstruction and rehabilitation activities, so that short-term gains do not create long-term problems. • Integrate research as a fundamental component to redevelopment, both as part of specific development initiatives and in its own right, drawing on international, evidence-based solutions as well as local, experiential knowledge. • Encourage the adoption of an Eco-Health Framework for redevelopment research initiatives, in order to help all stakeholders – including citizens, communities, countries, the region and the Canadian people, and their government – obtain the greatest long term, sustainable benefits from recovery.
Social and political aspects of tsunami
453
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
• As part of an eco-health model, mandate requirements for transdisciplinarity, participation, and gender equity on the part of all groups engaging in redevelopment activities under the purview of the federal government. • Adopt a regional networking strategy for South and Southeast Asia to achieve maximum equity, inter-country knowledge exchange and co-operation within the region and with government, academic, and NGO Canadian partners. • Develop infrastructure for disaster mitigation and management that builds on existing research capacity and networks (e.g., a Centre of Excellence on Disaster Management) that can act as a key resource internationally as well as help to prepare for disasters in Canada. 37.7
CONCLUSION
Implicit in the Eco-Health Framework and the Call to Action developed by participants at the National Tsunami Forum is recognition of the social – as well as economic, cultural, and political – consequences of the Asian tsunami, and a holistic approach to their remediation. They emphasize a transdisciplinary approach that balances reconstruction with redevelopment, the technical with the social. Both the Framework and the Call to Action are based on a participatory, even emancipatory approach to recovery and redevelopment, one that values individual and community empowerment and supports the resiliency and capacity of survivors to actively engage in the recovery of their own communities. Moreover, this approach recognizes the differential impact of the tsunami on women, and promotes their inclusion as critical to the recovery process. As Enarson and Morrow (1998) conclude, “understanding the gendered terrain of disaster . . . will help as we strive toward more democratic, participatory, and disaster-resilient communities” (p. 231). Canadians will hold their government and NGOs accountable for the medium- to longterm recovery effort, and monitor the efficacy of our interventions, especially in terms of the quality, the quantity of aid delivered, and also the effectiveness and timeliness of its delivery. REFERENCES Aguirre, B., Dynes, R.R., Kendra, J., and Connell, R. (2005). Institutional resilience and disaster planning for new hazards: insights from hospitals. J. Homel. Secur. Emerg. Manag., 2(1), 1–16. Blaikie, P., Cannon, T., Davis, I., and Wisner, B. (1994). At Risk: Natural Hazards, People’s Vulnerability, and Disasters. London, Routledge. Drabek, T.E. (1986). Human System Responses to Disaster: An Inventory of Sociological Findings. New York, Springer-Verlag. Dynes, R.R. (1998). Coming to terms with community disasters. In: L. Quarantelli (ed.), What is a Disaster? New York, Routledge. Dynes, R.R., DeMarchi, B., and Pelanda, C. (eds.) (1987). Sociology of Disasters; Contributions of Sociology to Disaster Research. Milan, Franco Angeli. Ecological Society of America (1995). The Scientific Basis for Ecosystem Management. Retrieved from http://www.nres.uiuc.edu/outreach/esm_il_lo/esm_defs.htm Enarson, E. and Morrow, B.H. (eds.) (1998). The Gendered Terrain of Disaster: Through Women’s Eyes. Miami, Praeger. Forget, G. (1997). From environmental health to health and the environment: research that focuses on people. In: G.S. Shahi, B.S. Levy, A. Binger, T. Kjellstrom, and R. Lawrence, (eds.), International Perspectives on Environment, Development and Health: Towards a Sustainable World. New York, Springer, pp. 644–659. Fothergill, A. (1998). The neglect of gender in disaster work: An overview of the literature. In: E. Enarson and B.H. Morrow (eds.), The Gendered Terrain of Disaster: Through Women’s Eyes. Miami, Praeger, pp. 11–29. Gist, R. and Lubin, B. (eds.) (1989). Psychosocial Aspects of Disaster. New York, Wiley.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
454
C. Amaratunga and H. Smith Fowler
Ikeda, K. (1995). Gender differences in human loss and vulnerability in natural disasters: a case study from Bangladesh. Indian J. Gender Stud., 2(2), 171–193. International Development Research Council (2005). The Ecohealth Approach. Retrieved from http://web.idrc.ca/en/ev-68491-201-1-DO_TOPIC.html Lebel, J. (2003). In_Focus: HEALTH An Ecosystem Approach. International Development Research Council. ISBN 1-55250-012-8; 100 p. Retrieved from http://web.idrc.ca/en/ev-29009-201-1-DO_TOPIC.html Oxfam International (2005). The tsunami’s impact on women. Oxfam Briefing Note, March 2005. Quarantelli, E.L. (1989). Conceptualizing disaster from a sociological perspective. Int. J. Mass Emerg. Disas., 7(3), 243–252. Schroeder, R.A. (1987). Gender Vulnerability to Drought: A Case Study of the Hausa Social Environment. Boulder, Institute of Behavioural Science, University of Colorado. Seager, J. (2005). Noticing gender (or not) in disasters. Chicago Tribune, September 14, 2005. The Framework Convention on Biological Diversity (1992). Retrieved from http://www.biodiv.org/ programmes/cross-cutting/ecosystem/principles.asp US Federal Interagency Ecosystem Management Task Force (June 1995). The Ecosystem Approach: Healthy Ecosystems and Sustainable Economies. Washington, DC: June 1995, 3 volumes. Vaughan, M. (1987). The Story of an African Famine: Gender and Famine in Twentieth Century Malawi. Cambridge, UK, Cambridge University Press. Wiest, R., Mocellin J., and Motsisi, T. (1994). The needs of women in disasters and emergencies. Technical Report for the United Nations Disaster Management Training Programme. Winnipeg, MAN, University of Manitoba.
CHAPTER 38
An Ideal Conceptual Tsunami Warning System for the Indian Ocean
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
T.S. Murty Department of Civil Engineering, University of Ottawa, Ottawa, Canada N. Nirupama Atkinson School of Administrative Studies, York University, Toronto, Canada A.D. Rao Centre for Atmospheric Sciences, Indian Institute of Technology, New Delhi, India I. Nistor Department of Civil Engineering, University of Ottawa, Ottawa, Canada
38.1
INTRODUCTION
Even though ideally one would like to have a global tsunami warning system, somewhat similar to a global weather forecasting system, in practice this is not achievable, at least for the following reasons. Whereas the atmosphere is continuous, the oceans are not, they are separated by continents (Figure 38.1). The Pacific, Atlantic and Indian oceans are connected in the south through,
Arctic Ocean
Atlantic Ocean Pacific Ocean
Indian Ocean
Figure 38.1. The four global oceans.
455
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
456 T.S. Murty et al.
Figure 38.2. The countries in and around the Indian Ocean.
what is termed as the Southern Ocean. The Pacific and Atlantic oceans are connected to the Arctic Ocean in the north. The Indian and Arctic Ocean are not directly connected. In addition to these geographical considerations, since different countries border on different oceans, it makes perfect sense for each ocean to have its own tsunami warning system. There are some 37 countries, which have a coastline on the Indian Ocean (Figure 38.2). It should be noted that the Persian (Arabian) Gulf and the Red Sea are the part of the Arabian Sea system, and hence, the nations bordering these inland seas are also part of the Indian Ocean rim countries. Ideally there should be only one tsunami warning center for the whole Indian Ocean (Indian Ocean Tsunami Warning Centre, IOTWC) which is a 24 hours per day, 365 days per year in operation. In addition to this, each country should have a national tsunami warning center closely linked in real time to the IOTWC. These national centers need not be a 24 hours per day, 7 days per week operation, however, these should be able to operate making use of personnel with pagers, at least three people identified as responsible for a defined period, such as a week or a month. As with the Pacific tsunami warning system, ideally all the Indian Ocean rim countries should provide their seismic and sea level data in real time to the IOTWC, under the auspices of the Intergovernmental Oceanographic Commission (IOC) of UNESCO in Paris. Some cost sharing may be necessary for the operations of the IOTWC among all member nations. The data archiving and non-real time information sharing, as well as training, capacity building, activities related to increasing public awareness and education can be handled jointly on behalf of all the member nations, by the International Tsunami Information Center (ITIC) in Honolulu, Hawaii, USA.
An ideal conceptual tsunami warning system for the Indian Ocean 15
30
45
60
75
90
105
120
135
150
90
105
120
135
150
30
30
45
45
60
60
0
0
0
15
15
Recent 26 December, 2004 Indian Ocean Tsunami
15
15
30
30
45
45
60
60
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
457
0
Figure 38.3.
15
30
45
60
75
250 locations in the Indian Ocean and adjoining region for which tsunami travel time charts have been prepared (Bhaskaran et al., 2005).
For an in-depth analysis and study of the Indian Ocean Tsunami of 26 December 2004, see Kowalic (2005a, b); Murty et al. (2005a–e, 2006); Nirupama et al. (2005, 2006).
38.2 TSUNAMI TRAVEL TIME CHARTS The most basic information the IOTWC will require, once a tsunami event has been identified, is travel times to selected locations in various countries around the Indian Ocean rim. It is prudent to produce ahead tsunami travel time charts to various locations, which can be done, because, the travel times, to a first order of approximation depend only upon the water depth. Contrary to a misconception in the public and the media, tsunami travel times do not depend upon the magnitude of the earthquake. The tsunami travel time charts in use for the Pacific Ocean, are supposed to be accurate to ±1 min, for each hour of travel time. There is no reason not to expect the same precision for the Indian Ocean tsunami travel charts, provided, of course, there is good bathymetric data available. Figure 38.3 shows the locations around the Indian Ocean rim for which tsunami travel time charts have been prepared (Bhaskaran et al., 2005). Figures 38.4 and 38.5 show two typical tsunami travel time charts in the Indian Ocean.
458 T.S. Murty et al. 60°
90°
120°
150°
16
15
11
12
11
12
15
14
16
15
16
17
13
18
17
17
16
60°
17
60° 18
17
16
17
18
16
16
31
14
9 10 10
16
17
16
16
30°
9
7 8
8 9
15
16
16
16
17
16
16
16
15
15
15
15
16
20
5 7
14
12
16
16
15
15
15
15
15
15
18
15
14
15
14
14
30°
1516
14
14
14
14
15
18
13
14
15
0°
13
13 13
13
13
13
12
12
12
12
14
15 14
14
15
15
12
13
11 11
11
11
12
12
13
14
14
15
60°
12
13
10
10
10
11
17
13
14
15
12
13
14
11
11
12
9
9
10
17
15
11
10
18
13
15
10
13
15
30°
8
8 9
19
18
16
13
15
10
7
11
8
30°
19
17
19 19 2201
8
9 9
9
6
7
8
11
6
6
7
10
7
8
6
5
6 8
8
4
5
0°
9
12
3
4
18
1920
3
19
2120
4
90°
120°
150°
Figure 38.4. Tsunami travel time chart for Rann of Kutch, India. Latitude 23.583◦ N; Longitude 68.367◦ E (from Bhaskaran et al., 2005). 90°
120° 15
4 131 14
5
3
2
5 8
11
11
11
10
12
13
60°
90°
11
10
11 11
11
11
11
12
12
12
11
10
11
10
12 12
30°
5 10
10
10 10
12
11
11
9
60°
12
11
10
11
7
2
3
10
11
12
12 12
13
10
9
4
5
9 10
8
11
9
9
9
9
10
10
12
11
9
10
11
12
13
10
11
12
10
9
30°
10
11
9
9
8
8
8
8
25
12
9
7
8
8
12
8
7
7
7
10
9
11
10 11
7
8
9
6
6
6 6
6
9
7
3
7
9
9 10
6
7
89
6
5
5
5
11
7
4
5
4 4
0°
8
6
5
4
5
4
8
6
3
3
6
8
11
3
4
13
3
4
6
7
30°
2
5
9
12
2
4 5
6
7
13 15 17
2
4
7
6
10
14
1
5
0°
13
3
1
3
12
11
12
3
2
4
30°
13
14 13 12 12
13
5
5
60°
150° 13
18
60°
12
30° 30°
6
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
19
16
30° 30°
120°
12
13
150°
Figure 38.5. Tsunami travel time chart for a location off Sri Lanka: Latitude 8.570◦ N; Longitude 81.230◦ E (from Bhaskaran et al., 2005).
An ideal conceptual tsunami warning system for the Indian Ocean
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
38.3
459
INSTRUMENTATION FOR THE TSUNAMI WARNING SYSTEM
The two most basic instrumentation that is required for the tsunami warning system are a network of seismographs to record the earthquakes and another network of tide gauges to register the water levels associated with the tsunami waves. It is expected that the existing networks will be expanded and augmented with even more modern instrumentation over time. It should be noted that, the tsunami warning system, as presently envisaged is only for under ocean earthquake generated tsunamis, unless of course, under water volcanic eruptions generate a strong enough seismic signal. Otherwise the tsunami warning system will not be able to detect a tsunami generated by under water volcanic eruption, until the tsunami reaches at least one tide gauge. Same situation exists with tsunamis generated by submarine landslides. Either due to an earthquake or completely independent of an earthquake, submarine landslides can occur and they can also generate tsunamis. However, usually these tsunamis have only local effects and they cannot propagate over trans-oceanic distances. Some technologies for tsunami detection in real time (DART systems) are described elsewhere in this book. 38.4
NUMERICAL MODELS
Three sets of numerical models are required for the Indian Ocean Tsunami Warning System (IOTWS), second set for tsunami generation, second for propagation, and the third for coastal inundation. In principal at least, the numerical models for tsunami generation and propagation can be developed at one research center, for the whole Indian Ocean, making use of finite difference models with rectangular grids. Since the duration of most earthquakes is only a few seconds, these models cannot be run in real time, they all have to be developed ahead and the synthesized results be provided to the IOTWC in electronic format. Because all the models have to be run ahead, all kinds of permutations and combinations of all possibilities should be considered and this will require dozens of scenarios, involving locations of epicenters, magnitude of earthquakes, types of earthquakes (dip-slip or thrust-type involving vertical movement of the ocean bottom, as opposed to strike-slip type which generate horizontal motion at tectonic plate boundaries, but no significant tsunamis, hypo central depth (depth under the ocean bottom where the earthquake occurred, only shallow focus earthquakes, i.e., with focal depth of up to 30 km usually, generate major tsunami), and rupture parameters (rupture length, width, and vertical uplift) which determine the volume of the initial mound of water at the ocean surface, which spreads as tsunami waves. When an earthquake occurs that has the potential to generate a tsunami, the IOTWC should be able to punch in a few earthquakes parameters and retrieve the closest match from scenarios, to be used for warning purposes. The situation with coastal inundation models is completely different, one needs to know the horizontal extent of coastal inundation, as well as the tsunami amplitudes at the coast. Since one has to resolve the coastline geometry, shallow water ocean bathymetry and coastal topography to the finest detail, finite difference methods with rectangular grids cannot be used for coastal inundation models. Irregular triangular grids, such as those shown in Figures 38.6 and 38.7 should be used with finite element models. As is shown below, the Indian Ocean should be modeled as an elliptic system, in which boundary reflections play an important role. This is different from the Pacific and Atlantic oceans, which are respectively hyperbolic and parabolic systems. It is not realistic to expect the IOTWC to have the coastal inundation models. These models should be developed by each nation separately and the results should be stored in electronic format in their national tsunami warning centers.
460 T.S. Murty et al.
Gulf Island Grid
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Canada
United States
Figure 38.6.
Irregular triangular grid for a finite element model for a part of the Pacific coast of Canada and USA.
100 km
Figure 38.7.
Irregular triangular grid for a finite element model of the Queen Charlotte Islands of Canada.
Since tsunami is a rare event in the Indian Ocean, it is not economic to have a national tsunami warning center as a stand alone center. It should be part of a multi-hazard warning center, which also deals with more frequent hazards, such as cyclones, storm surges, river, floods, monsoons, and droughts.
An ideal conceptual tsunami warning system for the Indian Ocean
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
38.5
461
EQUATIONS FOR MODELLING TSUNAMI PROPAGATION
Since tsunamis propagate over trans-oceanic distances, the curvature of the earth must be taken into account in the numerical models for tsunami propagation. Hence the governing equations should be written in a spherical polar coordinate system, rather than in the traditional Cartesian coordinates. Following Kowalik and Murty (1993) the basic equations can be written as follows. Whenever fluid motions is considered along large distances on the globe the equation of motion and continuity can be written in the spherical polar coordinates λ, φ and R, defined as longitude, latitude and distance from the Earth’s center. If the origin of the system is located on the ocean surface, it is more suitable to introduce a vertical coordinate z = R − R0 . Here R0 is the radius of Earth and is equal 6370 km. Because Earth does not exactly has a spherical shape, the equation given below will better describe the large scale motion relative to the geopotential and not the spherical surfaces. The equations of motion in the spherical system are: Du u 1 ∂p (v sin φ − w cos φ) = − (38.1) − 2 + + Aλ Dt R cos φ ρR cos φ ∂λ Dv wv u 1 ρp + + 2 + (38.2) u sin φ = − + Aφ Dt R R cos φ ρR ρφ Dw v2 u 1 ∂p − − 2 + u cos φ = − − g + Az Dt R R cos φ ρ ∂z
(38.3)
where Aλ , Aφ , and Az are the components of the viscous force, and the time operator expressed as: D ∂ u ∂ v ∂ ∂ = + + +w Dt ∂t R cos φ ∂λ R ∂φ ∂z The frictional forces are written in a somewhat complicated form: Nh ∂ Aλ = A1 u − 2 (v sin φ − w cos φ) u + 2 R cos2 φ ∂λ v 2 sin φ ∂u 2 ∂w Aφ = A1 v + Nh − 2 + + R cos2 φ R2 cos2 φ ∂λ R2 ∂φ 2w 2 ∂u 2 ∂ Az = A1 w + Nh − 2 − 2 − (v cos φ) R R cos2 φ ∂λ R2 cos2 φ ∂λ where the operator A1 has the following form, 1 ∂ 1 ∂ ∂ ∂2 ∂ A1 = Nh + 2 cos φ + Nz R2 cos2 φ ∂λ2 R cos2 φ ∂φ ∂φ ∂z ∂z
(38.4)
(38.5) (38.6) (38.7)
(38.8)
The equation of continuity in the spherical system is: 1 ∂u 1 ∂ ∂w + (v cos φ) + =0 R cos φ ∂λ R cos φ ∂φ ∂z
(38.9)
462 T.S. Murty et al. Finally the equation for diffusion can be expressed in the following way: Dc = A2 c Dt
(38.10)
where operator A2 is:
1 ∂ ∂ 1 ∂2 ∂ ∂ cos φ D + + z R2 cos2 φ ∂λ2 R2 cos φ ∂φ ∂z ∂z ∂φ
(38.11)
Variable c stands for concentration, and it can denote salinity, temperature, or any passive admixture in the sea water. The temperature and density fields become relevant if internal waves play a role in the coastal effects of the tsunami. It has been suggested that, some of the coastal amplification of the 26 December 2004 tsunami in the Indian Ocean, could have been due to the interaction between the tsunami waves and the internal waves. 38.6
METHOD OF CHARACTERISTICS AND DIFFERENT APPROACHES TO TSUNAMI MODELLING
One can distinguish between propagation (or marching) problems (Figure 38.8) and equilibrium (or jury) problems (Figure 38.9) (Crandall, 1956).
ditions
Solution domain
s
ition
C
C Solution morches out
ary con
ond
ry c
P
Bound
nda
Bou
Initial conditions
Figure 38.8. The Marching problem (Hyperbolic) (from Crandall, 1956).
ed ib
y conditions pr es dar cr un o B
Solution domain A lo n
g
en
t i re boun da ry
of
clo
in
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
A2 = Dh
sed d o m
Figure 38.9. The Jury problem (Elliptic) (from Crandall, 1956).
a
An ideal conceptual tsunami warning system for the Indian Ocean
463
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Initial conditions must be specified throughout the domain, and boundary conditions must be specified at the open boundary at all times. (Generally speaking, boundary conditions must also be specified at the closed boundary at all time.) It will be shown below that the governing equations for propagation problems are either hyperbolic or parabolic whereas for the equilibrium problems, they are elliptic. Crandall (1956, p. 352) formulated the general propagation problem in the following manner. The problem is to march out the solution of a governing system of partial differential equations of hyperbolic or parabolic type from prescribed conditions on an open boundary. The following is a pair of simultaneous first-order differential equations: A1
∂u ∂u ∂v ∂v + B1 + C1 + D1 =0 ∂x ∂y ∂x ∂y
(38.12)
A2
∂u ∂u ∂v ∂v + B2 + C2 + D2 =0 ∂x ∂y ∂x ∂y
(38.13)
Here, u and v are the dependent variables, x and y are the independent variables, and the coefficients A1 , A2 , B1 , B2 , C1 , C2 , D1 , and D2 are functions of u and v but not of x and y. Usually, the system of equations (38.12) and (38.13) is non-linear, but since the coefficients are not functions of x and y, one can make it linear by treating u and v as the independent variables and x and y as the dependent variables. Hence, the system of equations (38.12) and (38.13) is referred to as a “reducible system”. Assume that the system of equations (38.12) and (38.13) is being solved in the domain shown in Figure 38.6a and that the solution up to curve CPC is known. At P, the continuously differentiable values of u and v along CPC are known, as well as all of their derivatives in the directions pointing towards the interior of the curve. The following question is asked: is the behaviour of the solution above P completely determined by the solution below CPC, or is additional information at the boundaries of C required? To answer this, consider the following argument. Let S be a direction in which the distance is measured; then one can write: ∂u ∂u ∂x ∂u ∂y = + ∂S ∂x ∂S ∂y ∂S
(38.14)
∂v ∂v ∂x ∂v ∂y = + ∂S ∂x ∂S ∂y ∂S
(38.15)
The above question can be reformulated: for a solution of equations (38.12) and (38.13), do the values of u and v along CPC uniquely determine the derivatives? Here, S measures the distance along CPC. Define ∂u dS ∂S ∂v dv ≡ dS ∂S ∂x dx ≡ dS ∂S ∂y dy ≡ dS ∂S
du ≡
(38.16)
464 T.S. Murty et al. The set of equations (38.12–38.15) written at P can be expressed in the following compact form: ⎡
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
A1 ⎢ A2 ⎢ ⎣ dx 0
B1 B2 dy 0
C1 C2 0 dx
⎤ ⎡ ∂u/∂x ⎤ ⎡ 0 ⎤ D1 ⎢ ∂u/∂y ⎥ ⎢ 0 ⎥ D2 ⎥ ⎥ ⎢ ⎥ ⎥⎢ ⎢ ⎥=⎢ ⎥ 0 ⎦ ⎣ ∂v/∂x ⎦ ⎣ du ⎦ ∂v/∂y dv dy
(38.17)
Since u and v are known at P, the coefficients A1 , A2 , D1 , and D2 are also known. If the curve CPC is specified (i.e. its direction), then dx and dy are known. When u and v are known along CPC, then du and dv are also known. The system (equation (38.17)) constitutes a set of four simultaneous linear algebraic equations for the four unknowns ∂u/∂x, ∂u/∂y, ∂v/∂x, and ∂v/∂y. Two possibilities exist. If the determinant of matrix = 0, there is an indefinite set of solutions; there may be discontinuities in the solutions on either side of CPC. If the determinant = 0, there is a unique solution. To find out under what conditions the determinant of this matrix can be zero, it is expanded to give (A1 C2 − A2 C1 )(dy)2 − (A1 D2 − A2 D1 + B1 C2 − B2 C1 )dxdy + (B1 D2 − B2 D1 )(dx)2 = 0
(38.18)
One can consider this as a quadratic equation for the slope dy/dx. If the direction of CPC at P is such that it has a slope satisfying equation (38.18), then the derivatives of u and v are not uniquely determined by the values of u and v along CPC. Such a direction is called a characteristic direction. Let the discriminant (A1 D2 − A2 D1 + B1 C2 − B2 C1 )2 − 4(A1 C2 − A2 C1 )(B1 D2 − B2 D1 ) be denoted by D; then the following is true. If D is positive, equation (38.18) gives two real slopes; the system of equations (38.12) and (38.13) is hyperbolic (there are two real characteristic directions at P). If D = 0, equation (38.18) gives one real slope; the system of equations (38.12) and (38.13) is parabolic (there is only a single characteristics direction at P). If D is negative, equation (38.18) gives a pair of complex slopes; the system of equations (38.12) and (38.13) is elliptic (there are no real characteristic directions at P). Similar analysis can be made for the following single second-order quasilinear equation: a
∂2 ψ ∂2 ψ ∂2 ψ + b =f + c ∂x2 ∂x∂y ∂y2
(38.19)
In which a, b, c, and f are functions of x, y, ψ, ∂ψ/∂x, and ∂ψ/∂y. The characteristic directions are determined from the following quadratic equation: a(dy)2 − bdxdy + c(dx)2 = 0
(38.20)
Thus, if b2 − 4ac is positive, equation (38.19) is hyperbolic; if b2 − 4ac = 0, equation (38.19) is parabolic; if b2 − 4ac is negative, equation (38.19) is elliptic. A technique referred to as the method of characteristics, which will be dealt with in more detail in later sections, will be briefly outlined. Assume that the system given by equations (38.12) and (38.13) is hyperbolic in the domain under consideration. Thus, at every point there are
An ideal conceptual tsunami warning system for the Indian Ocean
465
a family
b tic ris te ac ar ch
h
P y
ic
ist
r cte ara
ac
b family
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
x
Figure 38.10.
Method of characteristics (from Crandall, 1956).
two roots, (dy/dx)α , and (dy/dx)β , to the quadratic equation (38.18). A curve having a slope (dy/dx)α at each of its points is an α-characteristic and a curve with a slope (dy/dx)β is a β-characteristic. There are thus two families of characteristics filling the domain as shown in Figure 38.10. It has been shown that the characteristics are loci of possible discontinuities in the derivatives of a solution. In equation (38.17), if a characteristic direction is considered such that the determinant is zero, then, when the right-hand column is substituted for any column on the left-hand side, the resulting determinant must also be zero. Thus, replacing the fourth column on the left with the column on the right and equating the determinant to zero results in the following: (A1 B2 − A2 B1 )du + [(A1 C2 − A2 C1 )dy/dx − (B1 C2 − B2 C1 )]dv = 0
38.7
(38.21)
HYPERBOLIC, PARABOLIC, AND ELLIPTIC METHODS
Table 38.1 compares various parameters for the four global oceans. Figures 38.11–38.13 respectively show the conceptual mathematical process through which the Pacific, Atlantic, and Indian oceans could be modelled. The most basic shape in the universe is the cone. By cutting the cone in different ways, we will get conic sections, a hyperbola, which is more or less an open system, a parabola, which is a partly open system and an ellipse (circle is a special case of an ellipse) which is completely closed system. The Pacific Ocean is very large, and boundary reflections from far off coastlines usually do not play a significant role in the overall tsunami amplitudes. Hence the Pacific Ocean can be modeled as a hyperbolic process, by prescribing initial conditions and radiation-type boundary conditions. The Atlantic Ocean does not have converging tectonic plates (the Pacific and the Indian oceans have these); what it has is the mid-Atlantic ridge, which is a diverging tectonic plate boundary that gives rise to new ocean floor, but not to major tsunami-genic earthquakes. Whatever tsunamis that occur in the Atlantic Ocean are mostly at the margins and since tsunamis propagate slowly in shallow water, their propagation into the deep ocean is more like a slow diffusion process, and hence can be modeled as a parabolic process. The Indian Ocean is the smallest of these three oceans and boundary reflections contribute significantly to the overall tsunami amplitudes. Hence the Indian Ocean has to be modeled as a
Table 38.1. Tsunami characteristics of the four oceans.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Parameter
Pacific Ocean
Atlantic Ocean
Indian Ocean
Arctic Ocean
Area in Km2 166,241,000 Average depth in meters 4188 Deepest point in meters Mariana Trench 11,033 Number of trenches 18 Unpopulated area (%) – ∼20 Southern Ocean Length scale available 6447 (Km). Half the side of an equivalent (in area) square Ocean-wide tsunamis Yes
86,557,000 3735 Puerto Rico Trench 8648 3 30
73,427,000 3872 Java Trench 7725 1 55
4642
Tectonic-converging plates producing tsunami-genic earthquakes Frequency of tsunami occurrence Frequency dispersion
Yes
No, at mid-Atlantic Ridge, the plates are diverging
4284 – If we 1540 omit the unpopulated area, it is about 2000 Yes Rare – extremely strong dissipative influence by ice cover Yes Some, but not as strong as in Pacific and Indian oceans
High
Rare
Rare
High
Amplitude dispersion (non-linear effects at the coast)
High
N/A – since no Low ocean-wide tsunamis High High
Tsunami travel times
Up to 23 h
Available warning time to most populated areas Importance of initial conditions Importance of boundary reflections
Generally sufficient
Relevance of boundary conditions Effect of Coriolis force Type of boundary conditions to be used Initial withdrawal of the ocean Which wave is the highest Nature of physical process
High Low
No
Only local tsunamis which travel within minutes Because of local tsunamis, only several minutes High
Up to 10 h to most populated area 1–4 h High
9,485,000 1038 Eurasian Basin 5450 None Almost all of it
Rare Small Moderate. Ice cover does not permit significant amplification of tsunami waves at the coast Local tsunamis. Several minutes to at most a couple of hours Small – mostly several minutes High
Low
Low, since tsunami High originate close to the coast Medium High
Moderate, ice cover does not allow significant refection of waves Medium
Noticeable
Highly noticeable
Noticeable
Radiation type
Not important since tsunamis are local Rare
High, since Coriolis force is strong Combination of radiative and reflective boundaries Not obvious due to ice cover Usually 1st or 2nd wave
Reflective boundaries Sometimes at some locations Usually 2nd wave Elliptic: tsunami Half way between behaviour parabolic and elliptic everywhere in the ocean including reflections at the boundaries is relevant everywhere else
Sometimes at some locations Among 3rd to 5th Hyperbolic (like astrology): everything is determined at birth
Usually 1st wave Parabolic: tsunamis travel slowly in shallow water near the coast. It is like a slow diffusion process
An ideal conceptual tsunami warning system for the Indian Ocean
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Mariana Trench
Figure 38.11.
467
Boundary reflections do not contribute significantly to the water levels during a tsunami event Hyperbolic method Except for local tsunamis, usually sufficient warning time Frequent tsunamis Convergent tectonic plate boundaries give rise to tsunamigenic earthquakes For tsunami
Schematic illustration of the tsunami numerical modelling concept for the Pacific Ocean.
Mainly local tsunamis Ocean-wide tsunamis almost non-existent The tsunamis get dissipated quickly in time and space
Tsunami waves aeneration
Parabolic method Tsunamis are rare Divergent tectonic plate boundaries at the mid-Atlantic Ridge do now give rise to tsunamigenic earth quakes For tsunami propagation as well as coastal inudation, small parts of the ocean can be modelled
Figure 38.12.
Schematic illustration of the tsunami numerical modelling concept for the Atlantic Ocean.
468 T.S. Murty et al.
Arabian Sea
Bay of Bengal
Boundary reflections are extremely important
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Elliptic method Available warning time is few hours Convergent tectonic plate boundaries give rise to tsunamigenic earthquakes which produce both oceanwide as well as local tsunamis Tsunamis are rare For tsunami propagation, whole ocean has to be modelled For coastal inundation modelling, small parts of the ocean can be modelled
Figure 38.13.
Schematic illustration of the tsunami numerical modelling concept for the Indian Ocean.
more or less closed system (i.e., the elliptic approach). In a numerical model, this can be achieved through using reflective boundaries, as opposed to radiative-type boundaries.
38.8
COASTAL INUNDATION
Since regular grids in a finite difference frame work cannot properly resolve the extreme complexity of the coastline geometry, a more appropriate approach is to use a finite element model with irregular triangular grids. Here we will briefly describe the so-called “ADCIRC” model, which is a finite element model for the computation of storm surges. With some modifications, this model can be adapted for the computation of coastal inundation from tsunamis. In a series of reports and papers (Blain, 1997; Cialone, 1991; Luettich et al., 1991, 1992; Mark and Scheffner, 1993; Scheffner et al., 1994; Westerink et al., 1992, 1993a, b) the socalled “ADCIRC” model of the US Army Corps of Engineers has been described. The following material is based on Westerink et al. (1993b). “ADCIRC” stands for “An advanced threedimensional circulation model for shelves, coasts, and estuaries”. The ADCIRC – 2DDI is a depth-integrated option of a system of two and three-dimensional hydrodynamic codes of “ADCIRC”. ADCIRC – 2DDI uses the depth-integrated equations of mass and momentum conservation, subject to the incompressibility, Boussinesq, and hydrostatic pressure approximations. Using the standard quadratic parameterization for bottom stress and neglecting baroclinic terms and lateral diffusion/dispersion effects leads to the following set of conservation statements in primitive
An ideal conceptual tsunami warning system for the Indian Ocean
469
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
non-conservative form expressed in a spherical coordinate system (Flather, 1988; Kolar et al., 1994):
∂ς 1 ∂UH ∂(VH cos ϕ) + + =0 ∂t R cos ϕ ∂λ ∂ϕ ∂U 1 ∂U 1 ∂U tan ϕ + U + V − U +f V ∂t R cos ϕ ∂λ R ∂ϕ R
1 τsλ ∂ ps =− + g(ς − η) + + −τ∗ U R cos ϕ ∂λ ρ0 ρ0 H ∂V 1 ∂V 1 ∂V tan ϕ + U + V + U +f U ∂t R cos ϕ ∂λ R ∂ϕ R
τsϕ 1 ∂ ps =− + g(ς − η) + + −τ∗ V R ∂ϕ ρ0 ρ0 H
(38.22)
(38.23)
(38.24)
where t = time λ, φ = degrees longitude (east of Greenwich positive) and degrees latitude (north of the equator positive) ζ = free surface elevation relative to the geoid U , V = depth-averaged horizontal velocities R = radius of the earth H = ζ + h = total water column h = bathymetric depth relative to the geoid f = 2 sin φ = Coriolis parameter = angular speed of the earth Ps = atmospheric pressure at the free surface g = acceleration due to gravity η = effective Newtonian equilibrium tide potential ρ0 = reference density of water τsλ , τsφ = applied free surface stress τ∗ = Cf
(U 2 + V 2 )1/2 H
Cf = bottom friction coefficient A practical expression for the effective Newtonian equilibrium tide potential as given by Reid (1990) is:
2π(t − t0 ) η(λ, ϕ, t) = αjn Cjn fjn (t0 )Lj (ϕ) cos (38.25) Tjn + jλ + Vjn (t0 ) n,j
where Cjn = constant characterizing the amplitude of tidal constituent n of species j αjn = effective earth elasticity factor for tidal constituent n of species j
470 T.S. Murty et al.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
fjn = time-dependent nodal factor Vjn = time-dependent astronomical argument J = 0,1,2 = tidal species (j = 0, declinational; j = 1, diurnal; j = 2, semidiurnal) L0 = 3 sin2 (φ – 1) L1 = sin(2φ) L2 = cos2 (φ) λ, φ = degrees longitude and latitude, respectively t0 = reference time Tjn = period of constituent n of species j Values for Cjn are presented by Reid (1990). The value for the effective earth elasticity factor is typically taken as 0.69 for all tidal constituents (Schwiderski, 1980; Hendershott, 1981) although its value has been shown to be slightly constituent-dependent. To facilitate an FE solution to equations (38.22) and (38.14), these equations are mapped from spherical form into a rectilinear coordinate system using a Carte Parallelogrammitque (CP) projection (Pearson, 1990): x = R(λ − λ0 ) cos ϕ0
(38.26)
y = Rϕ
(38.27)
where, λ0 , φ0 = center point of the projection. Applying the CP projection to equations (38.22), (38.24) gives the shallow-water equations in primitive non-conservative form expressed in the CP coordinate system: ∂ς 1 ∂(VH cos ϕ) cos ϕ0 ∂(UH ) + =0 + ∂t cos ϕ ∂x cos ϕ ∂y ∂U ∂U cos ϕ0 ∂U +V − + U ∂t cos ϕ ∂x ∂y =−
cos ϕ0 ∂ cos ϕ ∂x
tan ϕ U +f R
V
ps τsλ + g(ς − η) + − τ∗ U ρ0 ρ0 H
∂V ∂U cos ϕ0 ∂V + U +V − ∂t cos ϕ ∂x ∂y ∂ =− ∂y
(38.28)
tan ϕ U +f R
(38.29)
τsϕ ps + g(ς − η) + − τ∗ V ρ0 ρ0 H
U
(38.30)
Utilizing the FE method to resolve the spatial dependence in the shallow-water equations in their primitive form gives inaccurate solutions with severe artificial 2.x modes. However, reformulating the primitive equations into a GWCE (Generalized Wave Continuity Equation) form gives highly accurate, noise-free, FE-based solutions to the shallow-water equations (Lynch and Gray, 1979; Kinnmark, 1984). The GWCE is derived by combining a time-differentiated form of the primitive continuity equation and a spatially differentiated form of the primitive momentum equations recast into conservative form, reformulating the convective terms into non-conservative form, and adding the primitive form of the continuity equation multiplied by a constant in time
An ideal conceptual tsunami warning system for the Indian Ocean
471
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
and space, τ0 (Lynch and Gray, 1979; Luettich et al., 1992). The GWCE in the CP coordinate system is: ∂2 ∂ς cos ϕ0 ∂ ∂ς + τ0 + 2 ∂t ∂t cos ϕ ∂x ∂t ⎧ ⎫ cos ϕ0 ∂U tan ϕ ∂U ⎪ ⎪ ⎪ ⎪ U − − VH + UH U + f VH − ⎪ ⎪ ⎨ ⎬ cos ϕ ∂x ∂y R ×
⎪ cos ϕ0 ∂ ps τsλ ⎪ ⎪ ⎪ ⎪ ⎪ + g(ς − η) − (τ∗ − τ0 )UH + ⎩H ⎭ cos ϕ ∂x ρ0 ρ0 ⎧ ∂ς ⎫ ∂V cos ϕ0 ∂V ⎪ ⎪ V − VH − − UH ⎪ ⎪ ⎪ ∂t ⎪ ⎪ ⎪ cos ϕ ∂x ∂y ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎨ ⎬ tan ϕ ∂ U + f UH − + ⎪ R ∂y ⎪ ⎪ ⎪ ⎪ ⎪
⎪ ⎪ ⎪ ⎪ ⎪ τsϕ ⎪ ∂ ps ⎪ ⎪ ⎩H ⎭ + g(ς − η) − (τ∗ − τ0 )VH + ∂y ρ0 ρ0 ∂ − ∂t
tan ϕ VH R
+ −τ0
tan ϕ VH R
=0
(38.31)
The GWCE (equation (38.31)) is solved in conjunction with the primitive momentum equations in non-conservative form (equations (38.29) and (38.30)). The high accuracy of GWCE-based FE solutions is a result of their excellent numerical amplitude and phase propagation characteristics. In fact, Fourier analysis indicates that in constant depth water and using linear interpolation, a linear tidal wave resolved with 25 nodes√per wavelength is more than adequately resolved over the range of Courant numbers (C = ght/x ≤ 1.0) (Luettich et al., 1992). Furthermore, the monotonic dispersion behaviour of GWCE-based FE solutions avoids generating artificial near 2. x modes, which plague primitive-based FE solutions (Platzman, 1981; Foreman, 1983). The monotonic dispersion behaviour of GWCE-based FE solutions are very similar to that associated with staggered finite difference solutions to the primitive shallow-water equations allow for extremely flexible spatial discretizations, which result in a highly effective minimization of the discrete size of any problem (Foreman, 1988). The details of ADCIRC, the implementation of GWCE-based solution to the shallow-water equations, are described by Luettich et al. (1992). As most GWCE-based FE codes, ADCIRC applies three-noded linear triangles for surface elevation, velocity, and depth. Furthermore, the decoupling of the time and space discrete form of the GWCE and momentum equations, timeindependent and/or tri-diagonal system matrices, elimination of spatial integration procedures during time-stepping, and full vectorization of all major loops results in a highly efficient code. 38.9
SUMMARY
The basic components of an ideal tsunami warning system for the Indian Ocean are described. The functions of the Indian Ocean Tsunami warning center and the national centers of country bordering the Indian Ocean are identified for the whole systems to function efficiently, without the disadvantage of too many false alarms. It is essential that all the member nations working
472 T.S. Murty et al. together under the auspices of the united nations, should exchange seismic and sea level data in real time.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
REFERENCES Bhaskaran, P.K., Dube, S.K., Murty, T.S., Gangapadhyay, A., Chaudhuri, A. and Rao, A.D. (2005). Tsunami Travel Time Atlas for the Indian Ocean, Indian Institute of Technology Kharagpur, India, p. 279. Blain, C.A. (1997). Modeling methodologies for the prediction of hurricane storm surge in recent advances. In: N.K. Saxena (ed.), Mar. Sci. and Technol., PACON International, Honolulu, pp. 177–189. Cialone, M.A. (1991). Coastal Modelling System (CMS) user’s manual, instruction report CERC-91-1. Coastal Engineering Research Center, US Army Engineer Waterways Experiment Station, Vicksburg, MS. Crandall, S.H. (1956). Engineering Analysis – A Survey of Numerical Procedures, McGraw Hill, New York, pp. 417. Flather, R.A. (1988). A numerical model investigation of tides and diurnal-period continental shelf waves along vancouver Island. J. Phys. Oceangr., 18, 115–139. Foreman, M.G.G. (1983). An analysis of the “wave equation” model for finite element tidal computations. J. Comput. Phys. 52(2), 290–312. Foreman, M.G.G. (1988). A comparison of tidal models for the southwest coast of Vancouver Island. Proceedings of the 7th Internationl Conference on Computational Methods in Water Resources. Massachusetts Institute of Technology, Cambridge, MA, June 1988. Elsevier. Hendershott, M.C. (1981). Long waves and ocean tides. In: B.A. Warren, and C. Wunsch (eds.), Evolution of Physical Oceanography. MIT Press, Cambridge, MA, 292–346. Kinnmark, I.P.E. (1984). The shallow water wave equations: formulations, analysis and application. Ph.D. dissertation, Princeton University, Princeton, NJ. Kolar, R.L., Gray, W.J., Westerink, J.J. and Luettich, R.A. (1994). Shallow water modelling in spherical coordinates: equation formulation, numerical implementation, and application. J. Hydraulic. Res., 32, 3–24. Kowalik, Z. and Murty, T.S. (1993a). Numerical modeling of ocean dynamics. In: Advanced Series on Ocean Engineering, World Scientific Publishing, Singapore, 5, 481 pp. Kowalik, Z. and Murty, T.S. (1993b). Numerical simulation of two-dimensional tsunami runup, Marine Geodesy, 16, 87–100. Luettich, R.A., Westerink, J.J. and Scheffner, N.W. (1991). ADCIRC: an advanced three-dimensional circulation model for shelves, coasts and estuaries. Report 1: Theory and Methodology ofADCIRD-2DDI and ADCIRC-3DL, Coastal Engineering Research Center, US Army Engineer Waterways Experiment Station, Vicksburg, MS. Lynch, D.R. and Gray, W.G. (1979). A wave equation model for finite element tidal computation. Comput. Fluids, 7, 207–228. Mark, D.J., and Scheffner, N.W. (1993). Validation of a continental-scale storm surge model for the coast of Delaware. 3rd International Conference on Estuarine and Coastal Modelling, Chicago, IL, 8–10 September. Murty, T.S., Rao, A.D. and Nirupama, N. (2005a). Inconsistencies in travel times and amplitudes of the 26 December 2004 Tsunami. J. Mar. Med., 7(1), 4–11. Murty, T.S., Nirupama, N., Nistor, I. and Rao, A.D. (2005b). Conceptual differences between the Pacific, Atlantic and Arctic tsunami warning systems for Canada. Sci. Tsunami Hazards, 23(3), 39–51. Murty, T.S., Nirupama, N., Nistor, I. and Hamdi, S. (2005c). Why the Atlantic generally cannot generate trans-oceanic tsunamis. ISET J. Earthq. Technol., 42(4), 227–236. Murty, T.S., Nirupama, N., Nistor, I. and Hamdi, S. (2005d). Far field characteristics of the tsunami of 26 December 2004. ISET J. Earthq. Technol., 42(4), 213–217. Murty, T.S., Nirupama, N. and Rao, A.D. (2005e). Why the earthquakes of 26th December 2004 and the 27th March 2005 differed so drastically in their tsunami-genic potential. Newslett.: Voice Pacific, 21(2), 2–4. Murty, T.S., Rao, A.D., Nirupama, N. and Nistor, I. (2006). Numerical modelling concepts for the tsunami warning systems. Curre. Sci. 90(8), 1073–1081.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
An ideal conceptual tsunami warning system for the Indian Ocean
473
Nirupama, N., Murty, T.S., Rao, A.D. and Nistor, I. (2005). Numerical tsunami models for the Indian Ocean countries and states. Indian Ocean Survey, 2(1), 1–14. Nirupama, N., Murty, T.S., Nistor, I. and Rao, A.D. (2006). The energetics of the tsunami of 26 December 2004 in the Indian Ocean: a brief review Marine Geodesy, 29(1), 39–48. Pearson, F. (1990). Map projections: theory and applications. CRC Press, Boca Raton, FL. Platzman, G.W. (1981). Some response characteristics of finite element tidal models. J. Comput. Phys., 40, 36–63. Reid, R.O. (1990). Water level changes. In: J. Herbich (ed.), Handbook of Coastal and Ocean engineering. Gulf Publishing, Houston, TX, USA. Schwiderski, E.W. (1980). On chanting global ocean tides. Rev. Geophys. Space Phys., 18, 243–268. Westerink, J.J., Luettich, R.A., Baptista, A.M., Scheffner, N.W. and Farrar, P. (1992). Tide and storm surge predictions using a finite element model. J. Hydraulic Engg., 118, 1373–1390. Westerink, J.J., Luettich, R.A., Blain, C.A. and Scheffner, N.W. (1993a). ADCIRC: an advanced three-dimensional circulation model for shelves, coasts and estuaries. Report 2: Users manual for ADCIRD-2DDI, Technical Report DRP-92-6, Coastal Engineering Research Center, US Army Engineer Waterways Experiment Station, Vicksburg, MS. Westerink, J.J., Luettich, R.A. and Scheffner, N.W. (1993b). ADCIRC: an advanced three-dimensional circulation model for shelves, coasts and estuaries; Report 3: development of a tidal constituent database for the Western North Atlantic and Gulf of Mexico, Technical Report DRP-92-6, Coastal Engineering Research Center, US Army Engineer Waterways Experiment Station, Vicksburg, MS.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
CHAPTER 39
Overview and Integration of Part 5
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
N. Nirupama Atkinson School of Administrative Studies, York University, Toronto, Canada
39.1
QUO VADIS
In Chapter 34, Sundar discusses tsunami protection measures for the coast of Tamil Nadu in India. Even though the discussion was specifically for Tamil Nadu, the techniques elaborated upon are sufficiently general, so that they are adaptable, with some modifications, if required, to other tsunami-prone coastlines on the Indian Ocean, or indeed any other ocean. The author emphasises the use of dual-use technologies and mentions that groins built earlier for coastal protection helped to reduce the impact of the 26 December 2004 tsunami. He also suggests that both structural (seawalls, engineering structures) as well as non-structural measures, such as mangrove forests, etc. should be considered. Finally, the author mentions that tsunami protection measures will also offer protection against other marine hazards, such as storm surges and coastal erosion. In Chapter 35, Mascarenhas and Jayakumar discussed the protective role of coastal ecosystems against natural hazards. They made several suggestions for inclusion of natural marine hazards protection in the coastal zone, into the existing legislation. They also mention soft protection, such as afforestation in contrast to hard protection offered by civil engineering construction. They stated that, a human disaster of such magnitude, as the one at Nagapattinam, Tamil Nadu, India, could have been averted, or at least minimized, had the inherent natural protective value of geomorphic features and forested shelter belts been understood, had appropriate policies for natural coastal hazards been put in place, and had the Coastal Regulation Zone (CRZ) been enforced by locating the coastal inhabitants and dwellings further inland. In Chapter 36, Aswathanarayana described a framework for integrated preparedness systems for coastal marine hazards. At present 1.2 billion people out of the 6 billion inhabitants of this planet, live in the coastal zone, which is defined as the zone encompassing all the land up to a distance of 100 km from the coast. The present percentage of coastal population, which is 23%, is expected to increase to about 50% by the year 2030. The author points out that an important component of preparedness systems is the application of dual use technologies and practices. The authors mention that a seawall for protection against tsunamis serves only one purpose (the purpose it is intended to) and has no dual use. On the other hand, mangrove forests and vegetation belts offer not only some protection from tsunamis, but also yield economic and environmental benefits. However, the author does recognize that, in certain instances, civil engineering structures may be required to offer protection against tsunamis. The author also describes the practices of dual-use technologies, resiliency linked to social and ecological systems, tsunami risk management through securitization, fundamental studies needed in tsunami research, monitoring and warning systems, general strategy to meet the tsunami threat, how to increase public awareness and rehabilitation measures. In Chapter 37, Amaratunga and Fowler discussed the social and political aspects of the 26 December 2004 tsunami in the Indian Ocean, the authors pointed out that this tsunami had a much bigger impact on women than on men, and recognition of the gender element is required in 475
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
476
N. Nirupama
mitigation of all natural disasters. The authors also talk about the rehabilitation aspects, the roles of the Canadian Government and NGOs in the recovery and rehabilitation process. Implicit in the “Eco-Health Framework” and the “Call to Action” developed by participants at the National Tsunami Forum is recognition of the social as well as economic, cultural and political consequences of the Asian Tsunami, and a holistic approach to their remediation. They emphasize a transdisciplinary approach that balances reconstruction with redevelopment, and the technical with the social. Both the Framework and the Call to Action are based on a participatory, even emancipatory approach to recovery and redevelopment, one that values individual and community empowerment and supports the resiliency and capacity of survivors to actively engage in the recovery of their own communities. In Chapter 38, Murty et al. described an ideal conceptual tsunami warning system for the Indian Ocean, somewhat similar to the Pacific Tsunami Warning System, in the sense that all the nations in the Indian Ocean should cooperate and share data under the auspices of the Inter-Governmental Oceanographic Commission of UNESCO in Paris and the tsunami data and information being archived and handled by the International Tsunami Information Center in Honolulu. The tsunami data base for the Indian Ocean is part of the global tsunami data bases held by the National Geophysical Data Center in Boulder, Colorado, USA. The authors also point out some very important differences between the Pacific and Indian oceans, as far as the development of numerical models for the computation of coastal inundation is concerned. The Indian Ocean, being much smaller in geographical extent than the Pacific Ocean, has to be modelled to include the tsunami waves reflected from boundaries and interacting with the direct waves. Unless the boundary reflections are included, the total water levels associated with the tsunami could be underestimated at certain coastal locations.
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Author Index
Aagaard, K. 390 Aarup, T. 121, 257 Abe, K. 104, 120 Achache, J. 246 Achyuthan, H. 49 Adams, W.M. 121, 129, 272 Adger, W.N. 439, 444 Adomian, G. 128, 129 Afonin, V.V. 235, 245, 246 Agarvadekar, Y. 257 Agarwal, K. 162, 164, 172 Agarwal, N. 172 Agricole, W. 121 Aguirre, B. 446, 453 Akentieva, O.A. 245, 246 Aki, K. 104, 120, 171, 172 Ali, A. 16, 430, 432, 434 Alimov, O.A. 246 Allenby, B. 438, 444 Alonso, B. 55 Altaff, K. 384, 389 Amaratunga, C. 445, 475 Ambrasey, N. 353, 363 Ameer Basha, S. 403 Ammon, C.J. 37, 47, 121 Anderson, K. 89 Ando, M. 363 Andrade, C. 55 Anisimov, A.A. 138, 142 Anitha, S. 390 Annunziato, A. 165, 172 Anonymous 400, 402, 421 Apparao, C.H. 89 Arun Kumar, K. 49, 59 Artru, J. 246 Aster, R. 89 Aster, R.C. 47, 121 Aston, J. 434 Aswathanarayana, U. 47, 57, 405, 437, 438, 444, 475 Atwater, B.F. 52, 53, 54, 55 Baba, M. 323, 336, 337 Baba, T. 248, 256
Babeyko, A. 165, 172 Bagnis, R. 373, 389 Bailey, J. 55, 363 Baker, D.M. 268, 271 Balamurugan, V. 390 Bandibas, J. 363 Bapat, A. 17 Baptista, A.M. 363, 473 Baraza, J. 55 Barbier, P. 246 Barnes, D.K.A. 384, 389 Barstch, M. 390 Basset, A.B. 180, 182 Beck, S.L. 47, 121 Beger, M. 403 Bendick, R. 15, 17 Benioff, H. 266, 268, 271 Benoist, D. 246 Berman, A.E. 82, 88 Bernard, E.N. 262, 263, 271, 272 Bernard, P. 246 Berninghausen, W.H. 8, 14, 15, 16, 17, 323, 336 Berthelier, J.J. 236, 245, 246 Besana, G.M. 359, 363 Best, C. 165, 172 Bhaskaran, P.K. 273, 291, 294, 457, 458, 472 Bhat, M. 372, 390, 435 Bhattacharya, G.C. 372, 434 Bhattacharya, S. 235 Bhoi, S. 390 Bidoae, R. 336 Bilek, S.L. 47, 121 Bilham, R. 14, 15, 16, 17, 18, 38, 47 Binger, A. 453 Biswas, N.R. 31 Bjerknes, J. 144, 150 Bjerknes, V. 150 Black, K.P. 336 Blaikie, P. 446, 453 Blain Cheryl Ann 468, 472, 473 Blanc, E. 246 Blecki, J. 246 Bobrowski, P.T. 49, 53, 55
477
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
478 Author index Bodine, R.O. 101, 122 Bolt, B.A. 233 Borrero, J.C. 291, 388, 390 Boskova, J. 235, 246 Boudon, G. 55 Boyarchuk, K.A. 245, 246 Braddock, R.D. 178, 182 Brady, B.T. 218, 225 Braitseva, O.A. 52, 55 Brauner, C. 440, 444 Briggs, M. 291 Brochot, J.Y. 246 Brock, J.B. 363 Broer, L.J.F. 65, 68, 71 Brudzinski, M.R. 47, 121 Brundrit, G. 88, 121, 172 Bryant, E. 34, 47 Buckley, E.N. 434 Buckley, R.C. 363 Buland, R.P. 17 Bush, D.M. 434 Butler, R. 47, 121 Caldwell, P. 121 Canals, M. 55 Cannon, T. 453 Capobianco, M. 434 Carrier, G.F. 76, 79, 102, 120 Caruthers, C.G. 363 Catherine, J.K. 123, 129 CESS 332, 336 Chadha, R.K. 19, 22, 30, 33, 40, 43, 47, 58, 124, 129, 335, 336, 372, 381, 390, 402 Chandra, U. 30 Chandramohan, P. 129 Chandramouli, P. 372 Chandrasekar, N. 132, 142, 351, 358, 359, 363, 390, 406 Chandrasekhar, D.V. 31 Chang-gong, Dian 225 Chang-seng, D. 121 Chapuis, Y. 246 Chattopadyay, S. 336 Chaturvedi, N. 376, 390 Chaudhuri, A. 273, 291, 472 Cheminee, J.L. 55 Chen, J. 47 Chen, M. 63, 71 Cherkesov, L.V. 74, 79 Chittibabu, P. 159, 210 Chivian, E. 397, 402 Choi, B.H. 124, 129 Choudhury, S. 215, 218, 225 Chowdhury, J.U. 430, 432, 434 Chrystal, G. 144, 150 Chubarov, L.B. 124, 129
Cialone, M.A. 468, 472 CIBA 373, 390 Clague, D.A. 55 Clark, J.R. 428, 430, 432, 434 Clauge, J. 55 Cohen, J.E. 397, 402 Colin, F. 246 Connell, R. 453 Cooke, R.J.S. 51, 55 Costanza, R. 397, 402 Coutelier, J. 246 Cox, D.C. 55, 129 Crandall, S.H. 462, 463, 465, 472 Cros, A. 246 Cruikshank, K.M. 47 d’Arge, R. 402 Dabholkar, N. 257 Daily, G.C. 397, 402 Damodaran, V. 257 Daniels, F.B. 232, 233, 234 Daniels, G.M. 233 Darienzo, M.E. 52, 55 Das, J. 215 Das, P.K. 372 Das, S. 47 Dasgupta, S. 215, 218, 225 Datta, S. 425, 429, 434 Davies, K. 268, 271 Davis, I. 453 Dawson, A.G. 53, 55 DCRC Model 159, 166, 172 De Carvalho, D. 246 De Groot, R. 402 Defant, A. 144, 145, 150, 154, 156 Delft3D simulation package 124, 164, 172 DeMarchi, B. 453 Deplus, C. 51, 55 Desa, E. 257 Desa, E.S. 257 Desa, J.A.E. 257 Desai, B.N. 390 Desai, D.S. 435 Deschamps, A. 246 DeShon, H.R. 47, 121 de Souza, S.N. 390 Dey, S. 379, 390 Dharanirajan, K. 390 Dien, T.V. 391 Dimri, V.P. 123, 128, 129, 210 Diroky, D. 71 Divien, M.I.P. 390 DOD 336, 339, 377, 379, 380, 381, 387, 390, 402 Dominey-Howes, D. 363 Donn, W.L. 231, 233, 266, 267, 268, 270, 271 Drabek, T.E. 445, 453
Author index
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Dube, S.K. 273, 291, 372, 472 Dudley, W.C. 124, 129 Dungey, J.W. 230, 233 Dupuev, V.Kh. 246 Duxbury, A.B. 349 Duxbury, A.C. 349 Dynes, R.R. 445, 453 Eble, M.C. 271, 272 Ehrhardt, M. 349 Ekstrom, G. 47, 121 Elango, L. 124, 129 Elie, F. 246 Enarson, E. 445, 446, 453 Engdahl, E.R. 18, 47 Erdakos, G.B. 363 Ersoy, S. 350 ESA 450, 453 Ewing, M. 271 Ezhil Vendhan, K. 390 FAO 388, 390, 402, 403, 407 Farber, S. 402 Farrar, P. 473 Farre, R. 88, 121, 172, 251 FCBD 450, 454 Fejer, J.A. 230, 233 Feller, I.C. 403 Fengsha, Ge 225 Fergeau, P. 246 Fernando, H. 390 Fierro, J. 251, 256 Filloux, J.H. 248, 256 Firing, Y.L. 88, 121, 172 Fisher, E.M. 403 Fiske, R.S. 18 Flather, R.A. 101, 120, 469, 472 Folk, R.L. 342, 349 Forel, F.A. 144, 150 Foreman, M.G.G. 112, 120, 471, 472 Forget, G. 452, 453 Fortheringhum, D.G. 363 Fothergill, A. 445, 453 Francis, M. 47 Freeman, N.G. 154, 156 Freund, F. 218, 219, 220, 225, 379, 390 Friedman, G.M. 348, 349 Fritz, H. 390 Furumoto, A.S. 272 Gaboriaud, A. 246 Gahalaut, V.K. 129 Gaivoronskaya, T.V. 246 Ganeshan, P. 349, 372, 434 Gangloff, M. 246 Gangopadhyay, A. 273
Garrett, C.J.R. 154, 156 Gauns, M. 390 Gelfenbaum, G. 50, 55, 358 Geocities 2005 298, 321 George, Pararas-Carayannis 12, 13, 17, 47 Ghobarah, A. 297, 321 Gilbert, P. 246 Gill, A.E. 98, 112, 120 Girish, R. 129 Gist, R. 445, 453 Go, C.N. 11, 18 Goda, Y. 302 Godefroy, M. 246 Goff, J. 390 Gokhberg, M.B. 235, 245, 246 Goliavin, A.N. 246 Gonnert, G. 371, 372 Gonzalez, F.I. 121, 172, 262, 264, 271, 272 Goodwin, L. 403 Gopinadhan Pillai, C.S. 397, 402 Gorguis, A. 128, 129 Gorny, V.I. 218, 225 Goto, C. 101, 120 Gouveia, A.D. 257, 372 Gower, J.F.R. 263, 271 Gowthaman, R. 349, 372, 434 Grasshoff, K. 342, 349 Grasso, M. 402 Gray, W.G. 470, 471, 472 Gray, W.J. 472 Greenspan, H.P. 76, 79, 102, 120 Groot-Hedin, Catherine D.de 36, 47 Grygorczuk, J. 246 GSI 332, 336 Guilbault, J.P. 55 Guo, Muan-Hong 225 Gupta, G.V.M. 400, 402 Gupta, H.K. 19, 30 Gusiakov, V.K. 124, 129, 161, 172, 200 Gwal, A.K. 235, 293 Hamada, N. 121 Hamblin, P.F. 156 Hamdi, S. 71, 88, 95, 150, 157, 183, 208, 472 Hamilton, L.S. 402 Hamilton, T.S. 55 Hannoh, B. 402 Hansen, K. 390 Harbitz, C.B. 52, 55 Harborne, A.R. 403 Harford, C. 55 Harichandra Kumar, K.T. 372 Harivelo, F. 246 Harkrider, D.G. 233 Harris, A.K. 232, 233 Harry, Y. 30, 129
479
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
480 Author index Hasegawa, Y. 121 Hashimi, N.H. 335, 336 Hatori, T. 124, 129 Hayakawa, M. 235, 245, 246 Hayashi, Y. 121 Heaps, N.S. 101, 120 Hebenstreit, G.T. 336 Heck, N.H. 7, 8, 14, 15, 17 Heinitz, A.C. 363 Helmberger, D. 47 Hemphill-Haley, E. 55 Hendershott, M.C. 470, 472 Henry, R.F. 112, 120 Herron, T.J. 230, 232, 233, 234 Hidayat, R. 350 Higman, B. 390 Hindson, R. 54, 55 Hines, C.O. 227, 228, 230, 232, 233 Hirata, K. 120, 121, 256 Hiroi, I. 302, 321 Hjorleifsdottir, V. 47 Hobara, Y. 246 Hodgson, G. 398, 403 Hoffman, C. 248, 256 Holcomb, R.T. 55 Holgate, S. 257 Honjo, S. 390 Hopley, D. 397, 403 Horrillo, J.J. 102, 120, 121 Hough, S.S. 143, 150 HRV Harvard CMT Catalog 105, 121, 171 Hussain, I.S. 390 Hutchinson, I. 49, 55 Hwang, L.S. 71, 179, 182 Ichinose, G. 47 Idia, K. 124, 129 IDRC 450, 451, 454 Ikeda, K. 446, 454 Ilangovan, D. 349, 372, 434 Imamura, F. 91, 94, 95, 99, 100, 101, 121, 126, 129, 160, 363 Imamura, K. 55 Immanuel, J.L. 142, 359, 363 Ingole, B. 373 Ishii, H. 104, 120 Italy Model 159, 165, 172 Ittekkot, V. 376, 390 Iwasaki, S.I. 104, 121, 171, 172 Iwasaki, T. 120 Iyer, C.S.P. 339, 406 Jacquey, C. 246 Jaffe, B.E. 50, 55, 390 Jain, S. 393 Jain, S.K. 432, 434
Jaiswal, A. 434 Janardanan, K. 412, 421 Jardin, J. 88, 121 Jayakumar, S. 339, 349, 371, 372, 423, 424, 430, 434, 475 Jayaraman, K.S. 389, 390 Jeffreys, B.S. 177, 182 Jeffreys, H. 177, 182 Jestin, F. 47 Ji, C. 104, 105, 121, 171, 172 Jiricek, F. 246 Johnson, K.G. 348, 349 Johnson, R.W. 55 Joseph, A. 247, 251, 253, 256, 257, 293 Joseph, M. 336 Joseph, O. 257 K’wasi, B. 403 Kalairasan, P. 336 Kamphaus, A.R. 363 Kanamori, H. 47, 121 Kaneda, Y. 256 Kanishkan, B. 129 Karakus, H. 129 Katada, T. 47, 129, 336, 372, 390, 434 Kataoka, J. 55 Kato, H. 363 Kawamura, H. 391 Keating, B. 359, 363 Kelsey, H. 55 Kelvin, Lord W. 144, 150 Kelvin, W. 176, 182 Kendra, J. 453 Khalil, S. 398, 403 Kilonsky, B. 88, 121, 172, 251 Kim, K.O. 129 Kinnmark, I.P.E. 470, 472 Kinsey, D.W. 397, 403 Kitamura, N. 53, 55 Kjellstrom, T. 453 Klein, R.J.T. 433, 434 Klinge, D. 246 Knight, W. 71, 88, 95, 97, 121, 150, 172, 182, 208 Knowles, C.E. 263, 264, 265, 272 Kodama, T. 245 Kolar, R.L. 469, 472 Komar, P.D. 411, 421 Komorowsk, J.C. 55 Kon’no, E. 55 Kong, I. 88 Kong, Ling-Chang 121, 172, 225 Korellef, F. 342, 349 Koshimura, S. 159, 166, 172 Kotaka, T. 55 Kowalik, Z. 87, 88, 97, 102, 112, 121, 143, 150, 151, 156, 159, 169, 172, 185, 208, 461, 472
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Author index Kozacek, Z. 246 Kremling, K. 349 Krishna, S.K. 373, 374, 390 Krishna Kumari, A. 372 Krishnakumar 398, 403 Krishnamurthy, K. 398, 403 Krishnan, P. 385, 386, 390 Krishnan, V. 435 Krumbein, W.C. 348, 349 Kudela, K. 246 Kumanan, C.J. 124, 129, 336 Kumar, R. 172 Kumar, S.P. 377, 390 Kuragano, T. 121 Kuran, U. 173 Kurian, N.P. 323, 324, 326, 330, 331, 332, 333, 335, 336, 337, 405 Lagoutte, D. 245, 246 Lagrange, J.L. 144, 150 Lakshminarayana, S. 31 Lal, U. 444 Lamb, H. 64, 65, 71, 175, 176, 179, 180, 182 Lander, J.F. 124, 129 LaPlace, P.S. 143, 150 Latha, G. 30, 47, 129, 336, 372, 390, 402, 434 Lawrence, R. 453 Lay, T. 37, 47, 89, 120, 121 Le Friant, A. 55 Lebel, J. 450, 454 Leblanc, F. 246 LeBlond, P.H. 154, 156 Lebreton, J.P. 236, 245, 246 Lee, J.S. 129 Lee, M.A. 391 Lee, Y.K. 154, 156 Lefeuvre, F. 246 Legen’ka, A.D. 235, 246 Legendre, C. 246 LeMéhauté, B. 64, 71 Leveque, M. 246 Levy, B.S. 453 Lighthill, M.J. 66, 71 Lineykin, P.S. 157 Lingzhi, Li 225 Linnerooth-Bayer, J. 441, 444 Liperovskaya, E.V. 246 Liperovsky, V.A. 235, 246 Lipman, P.W. 55 Lisitzin, E. 13, 14, 17 Liu, C.H. 231, 233, 234 Liu, P. 121, 129 Liu, P.L.F. 102, 121, 122, 390 Lockridge, P.A. 129, 291 Logan, T. 71, 88, 95, 97, 121, 150, 172, 182, 208 Lomnitz, C. 82, 84, 85, 88
Longuet-Higgins, M.S. 112, 121 Loomis, H.G. 104,121 Lopez, G.I. 55 Lubin, B. 445, 453 Luettich, R.A. 468, 471, 472, 473 Luis, A.J. 391 Lynch, D.R. 470, 471, 472 Lynette, P. 390 Macdonald, G.A. 55 Macfarlane, A. 17 Macmurdo, Captain 14, 15, 17 Mader, C.L. 100, 121 Madhupratap, M. 390 Madrias, L. 246 Maggiolo, R. 246 Magori, C. 88, 121, 251 Mahadevan, R. 124, 129 Maksimovic, M. 246 Malik, J.N. 434 Malingre, M. 246 Malleswara Rao, M.M. 31 Manganini, S. 390 Manhong, Guo 225 Mani Murali, R. 372, 434 Manurung, P. 88, 121, 199 Marchuk, A.G. 138, 142 Mark, D.J. 468, 472 Marsha, S. 403 Mascarenhas, A. 423, 429, 432, 433, 434 Masson, D. 55 Mathur, S.M. 14, 15, 16, 17 Matsuyama, M. 102, 121 Mazhaeva, O.A. 246 McAdoo, B. 350 McCann, W.R. 7, 8, 9, 17 McCloskey, J. 19, 30, 36, 47 McCreery, C.S. 121, 260, 261, 264, 272 Mechler, R. 441, 444 MEF 433, 434 Mehra, P. 257 Mei, C. 111, 113, 121 Meinig, C. 271 Meister, C.V. 246 Melekestsev, I.V. 55 Menvielle, M. 246 Menzinger, I. 440, 444 Merian, J.R. 144, 150 Merikallio, S. 246 Merrifield, M.A. 88, 102, 104, 115, 117, 119, 121, 171, 172 Michael, G.S. 349, 372, 434 Miles, J.W. 144, 147, 148, 150, 153, 154, 156, 180, 182 Miller, G.R. 122 Mills, Evan 440, 444
481
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
482 Author index Min, Xu 225 Minoura, K. 49, 55, 363 Mirabueno, H.M. 363 Mishra, P. 372, 390, 435 Mitchell, W. 88, 121 Mocellin, J. 454 Mofjeld, H.O. 104, 121, 271, 272 Mohan Rao, K. 19, 31 Mohapatra, G.P. 31 Molchanov, O.A. 235, 245, 246 Moore, A.L. 51, 53, 54 Moore, J.G. 55 Moreau, T. 246 Morgounov, V.A. 246 Morrow, B.H. 445, 446, 453 Morton, R. 390 Motsisi, T. 454 Mruthyunjayareddy, K. 363 Munk, W.H. 74, 75, 79, 122, 147, 148, 150, 154, 156 Muraleedharan, G. 131, 135, 137, 138, 142, 210 Muraleedharan, P.M. 390 Murali Krishnan, B.T. 336 Murali, R.M. 349, 372, 434 Murthy, K.S.R. 19, 21, 23, 30, 31, 57, 365, 372, 375, 390 Murthy, P.S.N. 31 Murthy, T.V.R. 365, 369, 372 Murty, C.V.R. 434 Murty, G.P.S. 19, 21, 24, 26, 27, 29, 31 Murty, T.S. 15, 16, 17, 63, 71, 73, 74, 75, 79, 81, 84, 88, 89, 91, 95, 99, 101, 106, 121, 131, 142, 143, 150, 151, 156, 157, 159, 161, 172, 175, 182, 183, 185, 201, 208, 227, 233, 259, 262, 272, 273, 280, 281, 291, 335, 336, 371, 372, 376, 389, 455, 457, 461, 472, 473 Musumi-Rokkaku, S. 55 Myers, E.P. 363 Mysak, L.A. 154, 156 Nagarajan, K. 129 Nageswara Rao, A. 363 Nagvekar, S. 257 Naidu, P.D. 376, 390 Naik, G.N. 349, 372, 434 Naik, K.A. 349, 372, 434 Nair, R.R. 31, 335, 336, 390 Nair, M.M. 123, 129 Nakahara, S. 88, 121 Nakamura, K. 74, 79 Nakaya, S. 55, 363 Nalbant, S.S. 30, 47 Nanda Kumar, N.V. 400, 403 Narain, A. 376, 390 Narayana, A.C. 89, 323, 333, 336, 339, 349, 358, 363
Naveed, Md.S. 389 Nayak, S.R. 430, 432, 434 Neal, W.J. 434 NEIC 10, 16, 89, 104, 121, 171 Nekrasov, A.V. 104, 121 Nelson, Captain 14, 15, 17 Nettles, M. 47, 121 Newcomb, K.R. 7, 8, 9, 17 Newman, J.C. 121, 272 Nhu, H.D. 391 Ni, Sidao 47 Nicholls, R.J. 434 Nikitin, A.K. 182, 183 Nilsen-Hofseth, S. 82, 84, 85, 88 NIO 21, 89, 105, 122, 165, 171, 172, 251, 365, 372, 373, 384, 434 Nirupama, N. 63, 71, 73, 81, 88, 89, 91, 95, 121, 143, 150, 151, 156, 157, 175, 182, 183, 185, 208, 209, 210, 227, 259, 293, 455, 457, 472, 473, 475 Nistor, I. 63, 71, 73, 81, 88, 89, 91, 95, 121, 143, 150, 151, 157, 175, 183, 185, 208, 210, 227, 259, 297, 321, 455, 472, 473 Nkebi, E.K. 257 Nobuaki, K. 166, 172 Nordstrom, K.F. 428, 430, 434 Normark, W.R. 55 Northwestern University 321 NRSA 380, 382, 390, 402 Obayashi, T. 227, 233, 234 Oda, K. 122 Odametey, J.T. 257 Okada, Y. 104, 122, 161, 165, 166, 170, 172 Okal, E.A. 83, 89, 105, 122, 129, 171, 281, 291, 391 Oldham, T.A. 14, 15, 17 Orman, J. 46, 47 Ortiz, M. 14, 16, 18, 38, 47 Ouzounov, D. 218, 220, 225, 379, 390 Oxfam International 445, 454 Ozer, C. 129 Ozyurt, G. 129 Pai, Y. 372 Pamela, H. 398, 403 Pan, P. 229, 230, 231, 233, 234 Papadopoulos, G.A. 71, 95, 121, 150, 182, 363 Papathoma, M. 363 Parab, A. 257 Paras-Carayannis, G. 129 Pari, Y. 390, 435 Park, J. 82, 89 Parrot, M. 235, 236, 246 Pathak, M.C. 372 Paul, A.K. 423, 429, 430, 434
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Author index Pearson, F. 470, 473 Pelanda, C. 453 Pelinovski, E. 129 Pendse, C.G. 16, 18 Penou, E. 246 Pereira, A. 257 Perincex, D. 350 Peschard, D. 246 Peshwe, V.B. 257 Peters, B. 50, 55 Peterson, C.D. 30, 40, 47, 52, 55, 129, 336, 372, 390, 434 Pettijohn, F.J. 348, 349 Petukhov, V.K. 268, 269, 272 Pfeffer, R.L. 230, 234 Pflug, G. 441, 444 Pielke, R. 434, 435 Pilkey, O.H. 428, 429, 430, 432, 434 Pillai, A.P. 336 Pillay, S. 121 Pincon, J.L. 246 Pizarro, G. 439, 444 Plafker, G. 291 Planet, J.L. 246 Platzman, G.W. 153, 154, 157, 471, 473 Poirer, B. 246 Poirier, B. 246 Pokhotelov, O.A. 246 Polet, J. 47 Ponomareva, V.V. 55 Porter, F.-Y. 88, 121 Posmentier, E.S. 266, 267, 268, 270, 271 Potetyunko, E.N. 182, 183 Prabhudesai, R.G. 247, 251, 252, 254, 255, 256, 257, 293 Prabhudesai, S. 257 Prakash, T.N. 323, 332, 336, 372 Prasad, T.G. 17, 390 Presateya, S. 350 Press, F. 230, 234, 271 Preuss, P. 325, 336 Prithiviraj, M. 332, 336 Priyadarshini, S. 425, 429, 434 Proudman, J. 144, 150, 154, 157 Pulintes, S.A. 246 Quade, J. 17 Quarantelli, E.L. 445, 453, 454 Raad, P. 336 Rabinovich, A.B. 89, 102, 115, 119, 122, 257 Radhakrishna, B.P. 19, 31, 89 Raghavan, S. 423, 429, 435 Raghu, V. 363 Raghuram, G. 47 Ragoonaden, S. 434
483
Rai, D.C. 434 Raichlen, F. 151, 152, 153, 157 Rajamanickam, G.V. 142, 434 Rajamanickam, M. 142 Rajasekhar, M. 403 Rajawat, A.S. 434 Rajith, K. 336 Raju, Y.S.N. 31 Ram Mohan, V. 380, 390 Ramachandran, S. 380, 385, 386, 390 Ramachandrudu, G. 89 Ramaiah, N. 390 Ramalingeswara Rao, B. 124, 129 Ramana, M.V. 349, 372, 434 Ramana Murthy, M.V. 371, 372 Ramana Murthy, T.V. 365 Ramasamy, S.M. 124, 129 Ramaswamy, V. 390 Ramesh, R. 351, 406 Ramesh Babu, V. 390 Ramesh Kumar, M.R. 390 Ranga Rao, V. 372, 390 Rao, A.D. 63, 71, 73, 81, 88, 89, 91, 95, 121, 131, 142, 143, 150, 151, 156, 157, 175, 183, 208, 227, 259, 273, 291, 335, 336, 455 Rao, B.P. 365, 372 Rao, D.P. 372 Rao, K.M. 19, 31, 372 Rao, L.H.J. 31 Rao, M.M.M. 31, 372 Rao, T.C.S. 31 Rao, T.S. 31 Rao, V.P. 324, 337 Rao, V.R. 435 Rastogi, B.K. 3, 30, 31, 57 Raval, U. 23, 31, 351, 363 Reddy, D.R.S. 31 Reid, R.O. 101, 122, 263, 264, 265, 272, 469, 470, 473 Richards, P.G. 104, 120, 171, 172 Ritzwoller, M.H. 37, 38, 47 Robinson, D. 47 Roder, H. 442, 444 Romanova, N.N. 268, 269, 272 Rouzaud, J. 246 Rowell, G.A. 218, 225 Roy, D.S. 385, 386, 390 Saatcioglu, M. 297, 321, 405 Sadhuram, Y. 365, 370, 371, 372, 406 Sahoo, A.K. 390 Sahu, V.K. 129 Salaquarda, M. 246 Salman, A.G. 225 Saraf, A.K. 215, 218, 225, 293 Sarang, K. 398, 403
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
484 Author index Sarangi, R.K. 434 Saravanan, S. 142 Sardesai, S. 390 Saritha, K. 403 Sarkar, S. 235, 390 Sarma, K.V.L.N.S. 31 Sarma, V.V. 372 Satake, K. 47, 55, 71, 95, 121, 124, 129, 150, 166, 172, 182, 363 Satish Singh 19, 31 Satyanarayana, B. 373, 389 Sauvaud, J.A. 236, 246 Saxena, N.K. 291, 472 Sayabala, S.P. 17 Scheffner, N.W. 468, 472, 473 Schindele, F. 121 Schlichting, R.B. 50, 55 Schroeder, R.A. 446, 454 Schwiderski, E.W. 470, 473 Scott, R.H. 270, 272 SDMRI 382, 387, 390 Seager, J. 446, 454 Segoufin, J. 55 Sen, N. 433, 435 Sen Sarma, A.K. 423, 429, 435 Seran, E. 246 Seran, H.C. 246 Shahi, G.S. 453 Shahul Hameed, T.S. 324, 336, 337 Shankar, N. 49 Shillington, F. 88, 121 Shakdwipe, M. 89, 336, 349, 363 Shapiro, N.M. 47 Shastri, P.N.M. 372 Shaw, D.M. 231, 233 Shepard, F.P. 53, 55 Sheth, A. 434 Shetye, S.R. 30, 370, 371, 372, 389 Shi, S. 53, 55 Shibayama, T. 201, 311, 321 Shigematsu, T. 102, 122 Shilin, B.B. 225 Shillington, F. 88, 121 Short, A.D. 352, 363 Shrestha, M.L. 423, 429, 435 Shuto, N. 101, 120, 161, 172, 363 Siefert, W. 372 Sieh, K. 6, 19, 31 Simkin, T. 18 Simons, T.J. 154, 157 Simpson, D. 89 Singh, J.P. 47 Singh, R.P. 377, 379, 390 Singh, S.C. 37, 47, 437, 444 Sinha, M. 131, 142 Sipkin, S. 47, 121
Sivakumar, C. 124, 129 Slatkin, M.W. 179, 180, 183 Slominski, R. 246 Sloss, P.W. 291 Smilauer, J. 246 Smith, D. 257 Smith, D.E. 55 Smith, S. 363 Smith Fowler, H. 445, 475 Snedaker, S.C. 402 Snodgrass, F.E. 113, 122 Sobolev, S. 165, 172 SOE 325, 337 Solandt, J.L. 398, 403 Solberg, H. 150 Solem, I.O. 384, 390 Solem, T. 390 Soloviev, S.L. 11, 18, 124, 129 Soman, K. 323, 337 Sorkhabi, R.B. 17 Speilvogel, L.Q. 75, 76, 77, 78, 79, 209 SPM 411, 412, 421 Sridhar, P.N. 393, 407 Srinivasan, R. 129 Srivastava, Kirti 123 Srivastava, R.P. 128, 129, 210 Stalin, S. 271 Steacy, S. 30, 47 Stearns, H.T. 351, 363 Stein, S. 83, 89, 105, 122, 171, 391 Stipple, I. 439, 440, 444 Stoker, J.J. 177, 183 Strachey, R. 270, 271, 272 Stverak, S. 246 Subba Rao, A.V. 363 Subba Rao, D.V. 373, 391, 407 Subrahmanyam, A.S. 21, 23, 30, 31, 123 Subrahmanyam, C.S. 129 Subrahmanyam, M.V. 372 Subrahmanyam, V. 19, 21, 23 Subramanian, B.R. 19, 31, 372, 390, 402, 435 Sugumaran, J. 389 Sulerzhitsky, L.D. 55 Sundar, V. 411, 412, 415, 421, 422, 438, 475 Sundaramoorthy, S. 372, 435 Surendran, A. 393 Surya Prakash, S. 372 Suzuki, R.K. 272 Sverdrup, K.A. 342, 349 Synolakis, C. 121, 129, 172, 390 Szmant, A.M. 398, 403 Tadepalli, S. 73, 79 Takahashi, W. 391 Tanaka, H. 102, 121 Tang, D.L. 373, 376, 379, 389, 391
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Author index Tao, Liu 225 Tarits, P. 246 Tatavarti, R. 89, 336, 349, 363 Taymaz, T. 173 Tengali, S. 257 Testut, I. 121 Thacker, W.C. 102, 122 Thapliyal, V. 423, 429, 435 Thio, Hong-Kie 47 Thomas, K.V. 336 Tilman, D. 397, 402 Tirodkar, G. 349, 372, 434 Titov, V.V. 101, 121, 122, 161, 172, 173, 262, 272 Tolstoy, I. 229, 230, 231, 232, 233, 234 Tooley, M.J. 434 Torresan, M.E. 55 Toshctama, K. 30 Travnicek, P. 246 Treilhou, J.P. 246 Triska, P. 246 Tronin, A.A. 218, 225 Tsuji, Y. 55, 104, 122, 199, 201, 204 Tuck, E.O. 182 Turesky, N. 88 Uchida, M. 55, 363 Udayaraj, A. 390 Ueda, K. 55 UNESCO 160, 249, 251, 254, 259, 293, 342, 350 Urgeles, R. 51, 55 USFIEMTF 449, 454 USGS 10, 87, 169, 200, 374, 391 Usha, T. 372, 390, 435 Valdiya, K.S. 429, 433, 435 Van Cochran, J.R. 47 Van Den Driessche, P. 178, 182 Van der Hilst, R.D. 17 Van Dorn, W.G. 180, 183, 248, 257 Van Holland, G. 129 Van Hulsteyn, D.B. 229, 234 Varadachari, V.V.R. 24, 31 Vargese, T.I. 336 Vaughan, M. 446, 454 Veera Narayan, B. 393 Vel, A.S. 390 Venkatesan, R. 372, 390, 435 Venkateswarlu, K. 31 Venturato, A.J. 272 Vijay Kumar 251, 257 Villain, J.P. 246 Villemant, B. 55 Vinayak Prasad 17, 390 Vineeta Hoon 402, 403 Voyt, S.S. 154, 157
485
Wadia, D.N. 323, 337 Wald, D. 47 Walsh, T.J. 359, 363 Wang, Ge-Ping 225 Ward, S.N. 47, 121, 393, 403 Warita, K. 363 Warren, B.A. 472 Watanabe, H. 74, 79 Wazwaz, A.M. 128, 129 Weaver, P.F. 272 Webb, H.D. 232, 234 Weissel, J.K. 47 Wells, F.J. 230, 231, 233 Westerink, J.J. 468, 472, 473 Weston, V.H. 229, 234 Whelan, F. 363 Whitmore, P. 71, 88, 95, 97, 121, 150, 172, 182, 208 Wickersham Jr., A.F. 232, 233, 234 Wiest, R. 446, 454 Wijeratne, E.M.S. 88, 121 Wilson, B.W. 144, 150 Wilson, M. 82, 86, 89, 267 Wisner, B. 453 Witham, G.B. 128, 129 Woodworth, P. 257 Wronowski, S. 246 Wu, T.T. 120 Wu, T.R. 121 Wunsch, C. 472 Xie, K. 185 Yagi, Y. 89, 104, 122, 171, 173 Yalciner, A.C. 124, 126, 127, 128, 129, 160, 173, 339, 350 Yamaguchi, D.K. 52, 54, 55 Yamanaka 166, 173 Yanuma, T. 104, 122 Yeah, H. 121 Yeh, H. 40, 42, 47, 120, 129, 172, 281, 291, 336, 372, 390, 402, 434 Yeh, K.C. 231, 233, 234 Yong, Zhao 225 Yuen, P.C. 270, 272 Zaitsev, A. 173 Zamora, P. 246 Zarichny, J. 230, 234 Zhao, H. 373 Zhao, Yong 225 Zheng, Lan-Zhe 225 Zong, Y. 363 Zu-ji, Qiang 218, 225
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Subject Index
*Unless otherwise stated, all subject issues pertain to the Indian Ocean Tsunami (IOT) Andaman–Nicobar Islands Andaman–Sumatra tsunamigenic zone 4, 36–38, 46, 57, 58 impact on coral reefs 385–388, 396, 397–398 impact on mangroves 385, 386 inundation 352–355, 400–402, 468–471 land uplifts 380 run-up heights 201–205 sealevel retreat 380, 400 submergence 380, 385 subsidence 380 upwelling effect 379–380 arrival time 97–98, 115–117, 164, 171, 191–199, 280–281, 330–331 beach profiles 132, 137, 330–331, 356, 406 Devanaampatnam 42, 43, 44 Tamilnadu coast 423–434 biological impact 339 areas of high impact 349 areas of low impact 349 benthos community 349 biodiversity 384, 407, 432, 444 fish catch 349 nutrient concentrations 348 planktonic species diversity 349 primary productivity 342, 344, 345, 347, 348, 349, 406 biophysical dimensions 295, 405–408, 441, 450 Carlsberg ridge 7, 17, 57 Cauchy–Poisson problem 175 airy interface 176–178 application to tsunami generation 175–182 asymmetric sources 175, 178–180 Laplace and Hankel transforms 182 Stokes stream function 180 surface waves 175, 180 viscosity of fluid 180–182 Cauvery basin 20–30 bathymetry 24, 27, 29, 58
fault controlled 23, 438 Nagapattinam–Cuddalore shelf 22–24, 58 cellular-based instrumentation 247–256 sealevel gauges 247, 248, 249, 250, 251, 252, 255, 256, 442 Chagos ridge 3, 7, 14, 17 coastal ecosystems 382, 397, 423, 430, 475 casuarina plantations 401, 425, 430, 431–432, 438 coastal hazard policies 423, 433, 434, 475 coastal landforms 356, 362, 423, 426, 434 coastal sand dunes 398, 403 coastal vegetation 359, 430–433 natural shock absorbers 385, 430 coastal habitats of India 393 coastal inundation 400, 459, 468 coastal wetlands 398, 430 lagoons and creeks 399 marine and coastal ecosystems 397–400 shoreline erosion 385, 400–402 coastal morphology 19–30, 351–352 bathymetry 21, 22, 24, 27–29 eastern continental margin 20–21, 57, 58 fault control 23, 438 Nagapattinam–Cuddalore shelf 20, 21, 22, 23, 24, 58 seabed morphology 21, 58 submarine canyons 24, 441 tsunami surge 20, 22, 52, 358, 362, 372 coastal sealevel 247 coastal seismicity 24–27 coastal weak zones 24, 26, 58 coastal geophysical data 20, 24, 26, 30 gravity data 21, 23, 29, 57 magnetic data 23, 24, 27 DEMETER Satellite 235, 293 construction 236 electrostatic turbulence 239 mission objective 236–237 mode of operation 236
487
488
Subject index
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
DEMETER Satellite (Contd.) quick view 236 seismo-electromagnetic precursors 235 earthquakes 3 depth 19, 29, 34, 36, 37, 46, 58, 82, 84, 105, 123, 170, 215, 223, 237, 242, 459 energy dissipation 45 magnitude 15, 16, 19, 24, 50, 58, 87, 125, 215, 223, 242, 351 strike of the fault 45, 46 tsunamigenic earthquakes 3–13 eastern continental margin of India 20, 21, 22, 23, 57, 58 Cauvery basin 21, 22, 23, 24, 25, 26, 27, 29, 58 Krishna–Godavari basin 21, 22 Mahanadi basin 16, 21 seabed morphology 21, 24, 58, 416 slope characteristics 21, 23, 371, 401, 441, 442 ecological impact 339, 406 energetics of the tsunami 81, 209 directions of maximum energy 34, 45, 85, 114 directions of minimum energy 84, 85 energy of the tsunami 87, 151, 156, 189, 210 fast slip 83 normal modes of the earth 82–85, 143, 210 rupture process 82, 83 satellite detection of tsunami 85–86, 266, 271 slow slip 83 spheroidal normal mode 84, 85 total potential energy 87, 106 field observations 323, 353, 402 geostructural environment 57 global tsunami model 97, 119, 120 amplitude 109–113 energy flux vector 112, 114, 120 global distribution of maximum 109–113, 235 IOT in the global ocean 97, 102, 104, 106, 107, 108, 109, 111, 119 observations versus computation 117–119 residual maximum amplitude 111 time dependent propagation 113–115 tsunami travel time 115–117, 136, 141, 160, 273 Helmholtz mode 151–159 acoustic analogy 151–153 hydrodynamics 153–154 persistence of high water levels 151, 156, 188, 210 historical account 3 tsunamis during 326 BC–2005 AD 3, 13, 14, 57 tsunamis in the Indian Ocean 3–17, 281, 294, 381
tsunamis in the Pacific Ocean 12, 50–51, 135, 137, 140, 294 hydrophysical manifestations 365, 406 argo data 369, 370 coastal currents 324, 365, 368, 406 internal waves 91, 94, 187, 189, 209, 210, 227, 365, 406, 462 salinity and temperature structure 365–368, 369–370, 380, 406, 462 vulnerability of Indian coast 40, 370–372 Indian Ocean Andaman–Sumatra zone 4, 5, 6, 13, 36 geostructural features 38 Makran zone 7, 17, 36, 39, 40, 46, 59 tsunamigenic sources 3–7, 33–46 Indian Ocean tsunami propagation 123 Indonesia – Banda aceh bridges 317, 319 floating debris 311, 312, 313, 314, 320 precast concrete structures 318, 319 punching failures 305, 315 reinforced concrete bridges 318, 319 structural failures 311 timber-framed structures 314 initial withdrawal of the ocean 73 forerunner 73, 74, 75, 79, 98 period of tsunami 379 possible explanation 73–79 role of viscosity 73–75 solitary waves 69, 73 Spielvogel theory 75–79, 209 integrated preparedness systems 437, 475 dual-use technologies 411, 438–439, 475 ecosociological resilience 439 framework of preparedness systems 437–438 monitoring systems 213, 248, 249, 254, 293 public awareness measures 443–444 rehabilitation measures 444, 475 risk management 439–441 securitization 439–441, 443 warning systems 441–442 internal waves 91, 94, 187, 189, 209, 210, 227, 365, 406, 462 density of layers 94 depth of the layers 92, 93 Imamura and Imteaz model 91–94 ionosphere 227, 268–271, 293 IOT characteristics aftershocks 12, 19, 24, 37, 58, 104, 171, 215, 216, 223 average displacement 19 epicenter 29, 34, 36, 37, 39, 165, 209, 216, 218, 225, 237, 239, 242, 266, 268, 281–287 focal depth 19, 29, 82, 123, 223, 459
Subject index
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
magnitude 16, 33, 34, 36, 46, 54, 58, 88, 125, 351 rupture length 216, 459 stages of rupture 19, 63 tidal measurement 19 water column displaced 13, 37, 316, 318 Kanyakumari coast 351–352, 353, 356 coastal vegetation 353, 359, 362, 401, 430–432 coastal landforms 356, 362, 423, 426, 434 geomorphology 351–352, 353, 441 inundation 352–355 physical damage 41, 358, 360–361, 362 salinization of groundwater 54, 124 sedimentation 50, 356–358, 387 Kerala coast 323, 340, 405 beach profile variations 330–331, 356 continental shelf bathymetry 323 damage to life and property 327–329 Kayamkulam Inlet 323 Nearshore profiles 331, 332, 400 run-up levels 325–327, 330, 332, 405 sediment characteristics 332 siltation of backwaters 332–334 surficial sediments 324–325 time of arrival 187, 330–331 Krakatau volcanic eruption (1883) 12, 13, 15, 34, 35 Anak Krakatau (1928) 12 Krakatau eruption (416 AD) 13 related tsunami 12, 13, 15–17 tsunami wave attenuation 231, 232, 269 K-S-P waves 151–156, 210 long gravity waves 63, 151, 156, 186–187, 210, 280, 281 tidal water bodies 154 Makran coast 4, 14, 16, 35, 36, 38–39, 40, 210 geostructural environment 57–59 tsunami occurrences 3, 4, 7, 16, 17, 36, 38–39, 40, 46, 59 marine life 373 aquaculture 388, 407 biodiversity 384, 432, 444 chlorophyll a in the ocean 373, 374, 376–379, 380, 407 coastal ecosystems 397–400, 423 coral reefs 385–388, 397–398, 407 impact of IOT on marine life 339–349 mangrove forests 385, 386, 439 near shore studies 379 submergence 380 subsidence 380 temperature 379 time series analysis 377–378 upwelling effect 379–380
489
modeling of IOT basic equations and tools 98–101 bottom displacement 102–104, 165, 169 boundary conditions 101–102, 149, 179, 181, 463 domain 101–102 numerical grid 101–102 ocean bathymetry 101, 165, 166, 169, 189, 210, 280, 348 reflection from Maldives 45, 106, 107, 113, 335 reverse wave 106 sea level pattern 109 source function 104–106, 119, 169 trapped tsunami 108 tsunami amplitude 97, 107, 109, 110, 118, 119 wave reflection 103–104, 107, 165 Nagore–Velankanni stretch 423, 425, 426 Nagapattinam–Cuddalore shelf 21, 22–24, 58, 438 Ninety-East ridge 7, 17, 44, 57 natural hazards 411, 423 normal modes 82–85, 143–150, coastal behavior of tsunami 143, 149, 151, 411–415 coastal seiches 144–146 forced solutions 146–148 free and forced oscillations 146 free and forced seiches 148–149 free oscillations 143,144, 146, 149, 210 oscillations of first class 143–144, 154, 188, 210 oscillations of second class 143–144, 154, 188, 210 numerical models of tsunami 159 AIST (Satake) model 159, 166 Baird model 159, 168–169 DCRC – Tohoku model 159, 166 Delft model 159, 164 E.C. J.R.C. model 159 Italy model 159, 165 Koshimura model 159, 166–168 Kowalik et al model 159, 169–171 MOST model 161 NIO (India) model 159, 165–166 TOAST model 159, 162–164 Tsunami N2 model 126, 159, 160 Univ. of Frankfurt model 159, 165 Wakayama National College model 159, 166 paleo-tsunami deposits 49 Algarve tsunami deposits 53 Cascadia tsunami deposits 50, 52 characteristics 50, 59, 332 less contemporaneous reworking 49, 59 massive structure 49, 52, 332 Papua New Guinea deposits 52, 358
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
490
Subject index
paleo-tsunami deposits (Contd.) rapid deposition 49, 59 Reunion Island deposits 52 sedimentary context 50, 59 Storegga slides 52–53 performance of structures 297 buildings 297–301 column failures 305–306, 308–309, 320 coastal erosion 297–299, 411, 415, 475 concrete buildings 303, 304, 305, 308, 309, 310, 315, 318–320, 405 damage to buildings 300, 301, 406 debris flow 51, 54 engineered reinforced concrete 299, 306–310, 315–316, 317, 320–321, 405 impact loading 312–314 masonry walls 299, 300, 302, 303–305, 309, 315–316, 320, 405, 416 precast concrete bridges 310, 319 precast slab strips 309–310 reinforced concrete buildings 299–301, 302–303, 306–310, 315, 318, 319, 320 timber frame constructions 303, 304 tsunami loading 311–314 tsunami wave pressures 302–303, 320 unreinforced masonry walls 303–305 phase and amplitude dispersion application to tsunami Boussinesq equation 63, 67–68 celerity of solitary waves 64 Cnoidal waves 68–70 frequency and amplitude dispersion 66–67 Ursell parameter 63–66, 69 plate tectonic margins India–Australia plate 4–5 political aspects 445–453 possible amplification 91–94, 209, 227 possible detection 227–233 possible explanation 73–79, 209 predictive model 131, 142 maximum inundation distance 41, 42, 131, 132 maximum water height 132 tsunami beach run-up height 39, 40, 41, 124, 138, 311 tsunami travel time 136, 137–138, 141 pre-earthquake thermal anomaly 215, 218–220, 293 preparedness planning 445, 475 protection measures 411–421 Chennai region 415–416 coastal erosion 411–412 groin field 411–415 groins as dual-use technologies 413, 415 Madurai region 416 Trichy region 417
real-time reporting 247–256 remote sensing 217–218, 219, 340, 382, 393, 407, 408, 444 response to tsunami 445–453 satellite detection 85–86 sealevel gauges 247, 248, 249, 256 sea water turbidity 215, 221, 223–224, 384 seismic gap in the subduction zone 3, 4, 6, 57 seismic precursors 293 seismicity 19, 24–27, 30, 58 signals from earthquakes and tsunamis 227–233 social aspects 445–453 socio-economic dimensions 405–408 southeastern coast of India 351–362 southwestern coast of India 323–336 source zones of tsunamis 3–7, 57 Andaman–Sumatra 5, 6, 7, 36–38, 57, 210 Makran 3, 7, 16, 17, 36, 38–39, 40, 57, 210 Sunda trench 3, 34, 37, 39, 59, 82, 123, 215, 217, 223 storm surge deposits 49–51 surge monitoring 247, 248–249, 293 Thailand 297, 303, 309 damage to buildings 297–301 Khao Lak beach 204, 298, 299–301, 304, 307 Nai Thon beach 299–301, 307–308, 309–310 Phi Phi Island 301, 305, 306, 307 Phuket Island 297, 298 tsunami causes 3–4, 7, 15, 34, 46, 51, 53, 57, 218, 441 explosive volcanism 13, 34, 51, 52, 53 landslides 3, 13, 34, 293, 441 meteorite impact 34 submarine earthquakes 34, 179, 443 tectonic environment 4, 30, 35, 466 tsunami amplitude and travel times 115–117, 190 bathymetric features 106, 112–113, 123–124, 128 breaks in continental shelves 190, 211 coupling with internal waves 91–94, 187, 189, 209, 210 maximum tsunami amplitude in, 84, 97, 109, 160, 200–208 Banda Aceh 200, 201, 202, 203 India 207 Maldives 208 Srilanka 207 Thailand 204, 205, 206 OFC 143–144, 154, 188, 210 OSC 143–144, 154, 188, 210 Reynolds eddy stress 189, 211, 335 tidal current gradient 189, 190, 211 tsunami inundation 19, 50, 59, 332–333, 352–355, 400–402, 468–472
Downloaded by [INFLIBNET Centre] at 07:51 27 August 2012
Subject index Ursell parameter 63–65, 69, 71, 189 zones of convergence and divergence 190, 211, 335 tsunami beach run-up height 131–142, 201–205 bays and gulfs 186, 210 boundary reflection 187, 210 constructive interference 187, 210 east coast of India 15, 40, 41, 44, 45, 124, 127, 207, 371 harbours 186, 187, 210 interaction with tides 187, 188, 210 physical oceanographic processes 185–190, 210 run-up distributions 44, 126–127 tsunami coastal effects 143–150 tsunami damage 351–362, 405–408 tsunami detection systems 293 tsunami impact controlling factors 45 coastal topography 45 energy dissipation 45 strike of the fault 45 tsunami monitoring systems 247–256, 441–442 tsunami propagation 63–71, 73–75, 97, 113–115, 123–128, 175, 209, 461–465 discharge fluxes 126 elliptic modeling 128, 462 finite element method 210, 459, 460 fractals 128, 210 hyperbolic approach 128 long wave equations 126 sea state in the Indian Ocean 126, 127 tsunami magnitude scales 124 tsunami propagation models 123–128, 461–462 Voronoi tessellation 128 tsunami response 445 call to action 452–453 eco-health approach 450–452 ecosystem approach 449–450 integrated response strategy 442–448 sociopolitical nature of disasters 445–446 tsunami and storm surge detection 248–249, 293 cellular-based systems 253–254 cellular modem 250 communication performance evaluation 254 data communication options 249–250, 253
491
NIO sealevel gauge 251–253 sea-level measurement 252–253 tsunami surge heights 19, 371, 423 tsunami travel time atlas 273 hyperbolic, parabolic and elliptic methods 465–468 instrumentation for tsunami warning 459 numerical models for, 97, 459–460 Atlantic Ocean 467 Indian Ocean 97, 159, 168, 459, 468 Pacific Ocean 467 tsunami characteristics of four oceans 466 tsunami numerical modeling of, 97 coastal inundation 468–471 tsunami generation 160, 161, 162–164, 165, 170–171, 261–262 tsunami propagation 113–115, 160, 161, 162–164, 168–169, 169–170, 261–262 tsunami travel time charts 115–117, 274, 281, 287–291, 294, 457–458 tsunamigenic earthquakes 3–7, 33, 209, 281–287, frequency intervals 4, 57, tsunamigenic sources 3, 33 Andaman–Nicobar Islands 6, 36–38, 39, 46, 58, 59 Bangladesh–Myanmar coast 3, 4, 7, 15, 17 Chagos ridge 3, 7, 16, 17 Makran coast 7, 38–39, 40 Sunda arc 6, 7–13, 17, 57 thrust-type earthquakes 3, 459 volcanic eruptions 12, 52–53 Krakatau eruption 12, 13, 15, 34, 35 volcano collapse in Bismarck Sea 51–52 Ritter Island volcano collapse 51, 52 Santorini Island volcanic eruption 53 work–energy theorem 131–142, 210 maximum inundation distance 137, 140 maximum water height 137, 140 predictive model for beach run-up heights 133, 135, 137, 138 tsunami travel time 135, 138, 141