Geolooical Criteria for Evaluatin9 Seismicity Revisited
-------------------
Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes
•
THE GEOLOGICAL SOCIETY OF AMERICA®
Special Paper 479
Edited by Franck A. Audemard M., Alessandro Maria Michetti, and James P. McCalpin
Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes
edited by
Franck A. Audemard M. Fundación Venezolana de Investigaciones Sismológicas FUNVISIS El Llanito, Caracas 1073 Venezuela Alessandro Maria Michetti Dipartimento di Scienze Chimiche e Ambientali Università dell’Insubria Via Valleggio, 11 22100 Como Italy James P. McCalpin GEO-HAZ Consulting Box 837 Crestone, Colorado 81131 USA
Special Paper 479 3300 Penrose Place, P.O. Box 9140
Boulder, Colorado 80301-9140, USA
2011
Copyright © 2011, The Geological Society of America (GSA), Inc. All rights reserved. GSA grants permission to individual scientists to make unlimited photocopies of one or more items from this volume for noncommercial purposes advancing science or education, including classroom use. In addition, an author has the right to use his or her article or a portion of the article in a thesis or dissertation without requesting permission from GSA, provided the bibliographic citation and the GSA copyright credit line are given on the appropriate pages. For permission to make photocopies of any item in this volume for other noncommercial, nonprofit purposes, contact The Geological Society of America. Written permission is required from GSA for all other forms of capture or reproduction of any item in the volume including, but not limited to, all types of electronic or digital scanning or other digital or manual transformation of articles or any portion thereof, such as abstracts, into computer-readable and/or transmittable form for personal or corporate use, either noncommercial or commercial, for-profit or otherwise. Send permission requests to GSA Copyright Permissions, 3300 Penrose Place, P.O. Box 9140, Boulder, Colorado 80301-9140, USA. GSA provides this and other forums for the presentation of diverse opinions and positions by scientists worldwide, regardless of their race, citizenship, gender, religion, sexual orientation, or political viewpoint. Opinions presented in this publication do not reflect official positions of the Society. Copyright is not claimed on any material prepared wholly by government employees within the scope of their employment. Published by The Geological Society of America, Inc. 3300 Penrose Place, P.O. Box 9140, Boulder, Colorado 80301-9140, USA www.geosociety.org Printed in U.S.A. GSA Books Science Editors: Marion E. Bickford and Donald I. Siegel Library of Congress Cataloging-in-Publication Data Geological criteria for evaluating seismicity revisited : forty years of paleoseismic investigations and the natural record of past earthquakes / edited by Franck A. Audemard M., Alessandro Maria Michetti, James P. McCalpin. p. cm. — (Special paper ; 479) Includes bibliographical references. ISBN 978-0-8137-2479-9 (pbk.) 1. Paleoseismology—Methodology. I. Audemard M., Franck A. II. Michetti, Alessandro M. III. McCalpin, James. QE539.2.P34G46 2011 551.22—dc22 2011007062 Cover: 2006 study at the northern strand of the Boconó fault at the Lagunillas pull-apart basin (state of Mérida, Mérida Andes, Venezuela). (Top) location of the trench site (trench under white tent), (bottom) trench view. Photos courtesy of Franck A. Audemard M.
10 9 8 7 6 5 4 3 2 1
Contents
Introduction: Geological criteria for evaluating seismicity revisited: Forty years of paleoseismic investigations and the natural record of past earthquakes . . . . . . . . . . . . . . . . . . . . . . . . . 1 Franck A. Audemard M. and Alessandro Maria Michetti 1. Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA: The Calabacillas fault, Albuquerque, New Mexico . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 J.P. McCalpin, J.B.J. Harrison, G.W. Berger, and H.C. Tobin 2. Late Quaternary earthquakes on the Hubbell Spring fault system, New Mexico, USA: Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift . . . . . . . . . . 47 Susan S. Olig, Martha C. Eppes, Steven L. Forman, David W. Love, and Bruce D. Allen 3. Large-magnitude late Holocene seismic activity in the Pereira-Armenia region, Colombia . . . . 79 Claudia Patricia Lalinde P., Gloria Elena Toro, Andrés Velásquez, and Franck A. Audemard M. 4. Evidence of Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Myriam C. López C. and Franck A. Audemard M. 5. Style and timing of late Quaternary faulting on the Lake Edgar fault, southwest Tasmania, Australia: Implications for hazard assessment in intracratonic areas . . . . . . . . . . . . . . . . . . . . . 109 Dan Clark, Matt Cupper, Mike Sandiford, and Kevin Kiernan 6. Multiple-trench investigations across the newly ruptured segment of the El Pilar fault in northeastern Venezuela after the 1997 Cariaco earthquake . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Franck A. Audemard M. 7. Lake sediments as late Quaternary paleoseismic archives: Examples in the northwestern Alps and clues for earthquake-origin assessment of sedimentary disturbances . . . . . . . . . . . . . 159 Christian Beck 8. Late Pleistocene–early Holocene paleoseismicity deduced from lake sediment deformation and coeval landsliding in the Calchaquíes valleys, NW Argentina . . . . . . . . . . . . . . . . . . . . . . . 181 Reginald L. Hermanns and Samuel Niedermann 9. Rupture length and paleomagnitude estimates from point measurements of displacement— A model-based approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 Glenn Biasi, Ray J. Weldon II, and Kate Scharer
iii
The Geological Society of America Special Paper 479 2011
Geological criteria for evaluating seismicity revisited: Forty years of paleoseismic investigations and the natural record of past earthquakes Franck A. Audemard M.* Fundación Venezolana de Investigaciones Sismológicas (FUNVISIS), Final Prolongación Calle Mara, Quinta Funvisis, El Llanito, Caracas 1073, Venezuela, and School of Geology, Mines & Geophysics, Universidad Central de Venezuela, Ciudad Universitaria, Los Chaguaramos, Caracas 1010, Venezuela Alessandro Maria Michetti* Dipartimento di Scienze Chimiche e Ambientali, Università degli Studi dell’Insubria, Via Valleggio, 11, 22100 Como, Italy
ABSTRACT The identification of individual past earthquakes and their characterization in time and space, as well as in magnitude, can be approached in many different ways with a large variety of methods and techniques, using a wide spectrum of objects and features. We revise the stratigraphic and geomorphic evidence currently used in the study of paleoseismicity, after more than three decades since the work by Allen (1975), which was arguably the first critical overview in the field of earthquake geology. Natural objects or geomarkers suitable for paleoseismic analyses are essentially preserved in the sediments, and in a broader sense, in the geologic record. Therefore, the study of these features requires the involvement of geoscientists, but very frequently it is a multidisciplinary effort. The constructed environment and heritage, which typically are the focus of archaeoseismology and macroseismology, here are left aside. The geomarkers suitable to paleoseismic assessment can be grouped based on their physical relation to the earthquake’s causative fault. If directly associated with the fault surface rupture, these objects are known as direct or on-fault features (primary effects in the Environmental Seismic Intensity [ESI] 2007 scale). Conversely, those indicators not in direct contact with the fault plane are known as indirect or off-fault evidence (secondary effects in the ESI 2007 scale). This second class of evidence can be subdivided into three types or subclasses: type A, which encompasses seismically induced effects, including soft-sediment deformation (soil liquefaction, mud diapirism), mass movements (including slumps), broken (disturbed) speleothems, fallen precarious rocks, shattered basement rocks, and marks of degassing (pockmarks, mud volcanoes); type B, which consists of remobilized and redeposited sediments (turbidites,
*E-mails:
[email protected],
[email protected];
[email protected]. Audemard M., F.A., and Michetti, A.M., 2011, Geological criteria for evaluating seismicity revisited: Forty years of paleoseismic investigations and the natural record of past earthquakes, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, p. 1–21, doi:10.1130/2011.2479(00). For permission to copy, contact
[email protected]. © 2011 The Geological Society of America. All rights reserved.
1
2
Audemard M. and Michetti homogenites, and tsunamites) and transported rock fragments (erratic blocks); and type C, entailing regional markers of uplift or subsidence (such as reef tracts, microatolls, terrace risers, river channels, and in some cases progressive unconformities). The first subclass of objects (type A) is generated by seismic shaking. The second subclass (type B) relates either to water bodies set in motion by the earthquake (for the sediments and erratic blocks) or to earthquake shaking; in a general way, they all relate to wave propagation through different materials. The third subclass (type C) is mostly related to the tectonic deformation itself and can range from local (next to the causative fault) to regional scale. The natural exposure of the paleoseismic objects—which necessarily conditions the paleoseismic approach employed—is largely controlled by the geodynamic setting. For instance, oceanic subduction zones are mostly submarine, while collisional settings tend to occur in continental environments. Divergent and wrenching margins may occur anywhere, in any marine, transitional, or continental environment. Despite the fact that most past subduction earthquakes have to be assessed through indirect evidence, paleoseismic analyses of this category of events have made dramatic progress recently, owing to the increasingly catastrophic impact that they have on human society.
INTRODUCTION Paleoseismology is a currently growing and developing earth sciences discipline. The recognition that paleoseismology has in fact become a branch of learning in itself is a rather recent development (Slemmons, 1957; Allen, 1975; Sieh, 1978; Crone and Omdahl, 1987; Vittori et al., 1991; Yeats et al., 1997; Audemard, 1998; McCalpin, 1996, 2009). This is borne out by the recent publication of dedicated textbooks (Wallace, 1986; McCalpin, 1996; Yeats et al., 1997; Keller and Pinter, 1998; Burbank and Anderson, 2001; McCalpin, 2009) and special issues of scientific journals (e.g., Serva and Slemmons, 1995; Michetti and Hancock, 1997; Pavlides et al., 1999; Pantosti et al., 2003; Michetti et al., 2005a; Reicherter et al., 2009, this volume) or conference proceedings (e.g., Michetti, 1998; Cello and Tondi, 2000; HAN2000–PALEOSIS Project; Okumura et al., 2000, 2005; Toda and Okumura, 2010). Also, special issues or proceedings, and even books, on more specific paleoseismic issues have also come to light (e.g., Maltman [1994] on soft sediment deformation; Stratton Noller et al. [2000] on Quaternary dating techniques; Comerci [2001] on seismically induced mass movements; Gilli and Audra [2001] on speleoseismology; and Tappin [2007] on the sedimentary record of tsunamis and tsunamites). Finally, concerted efforts within the International Union for Quaternary Research (INQUA) community for several years have led to the recent proposal and acceptance (currently under validation) of a new macroseismic scale based exclusively on earthquake environmental effects, named the Environmental Seismic Intensity Scale 2007, known under the acronym ESI 2007 (Michetti et al., 2007). Most researchers agree that the main aim of paleoseismology is the identification and characterization of past earthquakes from geo-archives. In a broader sense, this type of research intends to establish the seismic history of a given fault or given region from deformed sediments or rocks, beyond the limited tempo-
ral resolution of instrumental and historical seismicity. However, this does not exclude major temporal overlapping with these two latter disciplines, as proposed by many authors (e.g., Gürpinar, 1989; Audemard, 1998, 2005; Levret, 2002; Carrillo et al., 2009). In regions of the world where historical and/or instrumental seismicity studies are really scarce, or where earthquake recurrence is too long to allow any detection of large past events, the study of past seismic behavior of faults has to rely exclusively on paleoseismology. Furthermore, paleoseismic research can even reveal very distinct trends of seismic activity over longer periods of time (longer than the current interglacial period), as is the case in large regions of the world subject to glaciation-deglaciation cycles (e.g., Adams, 1996b; Mörner et al., 2003; Mörner, 2005). On the other hand, a common misconception about paleoseismology relates this discipline to the study of prehistoric earthquakes only, as originally defined by Wallace (1981). This surely used to be, and still is, its main focus, but it does not cover all current potential spectra and unnecessarily restrains paleoseismology to a strict time window. In fact, paleoseismology should be understood as the study of the ground effects from past earthquakes preserved in the geologic and geomorphic record (Michetti et al., 2005b), regardless of the time of occurrence. In a more general way, it aims at studying any geologic deformation related to earthquakes (Audemard, 2005). In that sense, the separation among disciplines studying past individual earthquakes, i.e., instrumental seismology, historical seismology, archaeoseismology, and paleoseismology, should not be based on strict time windows of observation, but on the type of source information that is being used and managed by the researchers. In other words, the distinction between these four different disciplines is fundamentally methodological, both in terms of applied tools and research objects. Consequently, we share the proposal of Caputo and Helly (2008), in which each discipline focuses on different objects: (1) analogue or digital instrumental records, in modern
Geological criteria for evaluating seismicity revisited seismology; (2) oral or written witnesses, in historical seismology; (3) artifacts, in archaeoseismology, where artifact is defined as any product of human activity; and (4) natural features, in paleoseismology. Particularly, “natural object” in this paper is a generic name given to all geologic and geomorphologic features, indicators, evidence, effects, and markers pinpointing the occurrence of an earthquake. The methods and techniques employed for paleoseismic characterization of source parameters are diverse and numerous (for a thorough review, refer to reference textbooks such as McCalpin, 1996; Yeats et al., 1997; McCalpin, 2009) but are still evolving. Methodologies are continuously changing and adapting to encompass very broad and multidisciplinary approaches for the characterization of past earthquakes. Fault trenching investigations have become a cornerstone for paleoseismic analysis because they have the potential to provide a direct assessment of the amount and timing of recent coseismic and total (coseismic + aseismic) fault slip. This has led to a common misconception that trenching and paleoseismology are almost synonymous or equivalent (e.g., Caputo and Helly, 2008; Carrillo et al., 2009). As a matter of fact, trenching is only one technique among many other paleoseismic approaches, with its own specific advantages and limitations. For further details on this issue, we refer to McCalpin (1996), Yeats et al. (1997), Audemard (2005), Michetti et al. (2005b), and McCalpin (2009), among other reviews. The study objects that are the focus of paleoseismic investigations have also evolved with time and have particularly increased in number as a response to new demands in the field of seismic hazard assessment (SHA). From the understanding that a tectonic deformation may induce a sedimentological response, which can then be fossilized and preserved in the geologic record, paleoseismologists, and particularly geologists and tectonic geomorphologists, have continuously developed new paleoseismic objects to derive the timing and magnitude of hazardous earthquakes. These developments in paleoseismology have led to the incorporation of an ever-growing number of earth sciences disciplines, such as seismology, Quaternary geology, tectonics, structural geology, sedimentology, (sequence) stratigraphy, pedology, geomorphology, geochronology, remote sensing and geophysical prospecting methods, such as seismic reflection, side-scan sonar, ground-penetrating radar (GPR), and electric tomography, as well as geodetic surveying and monitoring techniques, such as global positioning system (GPS) and electron distance meter (EDM), among several others. In fact, the present contribution aims at answering the following question that any researcher may pose himself or herself anytime they are requested to assess the seismic hazard of a given fault or region: What object(s) can I study in this region in order to know the level of seismic hazard to which the planned construction project will be subjected? MAIN POTENTIAL STUDY OBJECTS The study and understanding of contemporary earthquakes and their associated effects have provided the basis for the search,
3
recognition, and characterization of the types of evidence that paleoseismology needs to unravel from the geologic record. Owing to the physics of earthquakes, two major types of deformations typically occur at ground or sea-bottom surfaces. One set of features relates directly to the earthquake rupture process itself, while other features are the result of the accompanying seismic shaking. Serva (1994), McCalpin (1996), and Michetti et al. (2007) referred to these as primary and secondary evidence, respectively. Depending on their location (or proximity to the causative fault or surface break), McCalpin (1996) subdivided these mechanisms into on-fault (near-field) and off-fault (farfield) evidence. Since the secondary features can occur throughout the entire affected region, including on or across the surface break, Audemard (2005) preferred to subdivide the coseismically induced permanent ground deformations into two groups based on their genesis and degree of association with the earthquake, namely direct and indirect ground deformations. The first set of features is directly related to the fault plane and its kinematics. The second group of features includes both the seismically induced effects (mass instability and soft-sediment deformation, including liquefaction) and all other ground modifications at the local or regional scale, such as uplift and subsidence, warping, buckling, bulging, etc. In order to incorporate the many newly employed paleoseismic study objects or targets, this latter classification needs to be given a wider meaning that does not restrict it to ground deformation only and that provides room to include other indirect types of evidence of earthquakes, such as seismites, which do not imply any ground-surface deformation, but refer to sediment remobilization and redeposition, broken or affected speleothems, truncated bioherms, tree growth changes recorded in dendrochronology samples, tsunamites, and even some types of soft-sediment deformations that include sills and convoluted bedding. McCalpin (1996) added a third level of subdivision based on the timing of the feature with respect to the time of occurrence of the earthquake, namely instantaneous (coseismic) and delayed response (postseismic). The need for this third level of classification does not appear to have any practical application because paleoseismic studies typically can make no distinction between coseismic and postseismic processes. Furthermore, the time taken by delayed processes (e.g., deposition of colluvial wedges, filling of open cracks, remobilization and redeposition of sediments, tree growth recovery or healing, among others) is negligible compared to the recurrence interval between large earthquakes in most faults. We herein propose to subdivide the indirect or off-fault objects into three subgroups or types: Type A encompasses seismically induced effects, including soft-sediment deformation (soil liquefaction, mud diapirism), mass movements (including slumps), broken or disturbed speleothems, fallen precarious rocks, shattered basement rocks, and marks from degassing (mud volcanoes, seeps, pockmarks). Type B consists of remobilization and redeposition of sediments (turbidites, homogenites, and tsunamites) and rocks (e.g., large boulders launched by tsunami waves on rocky coasts, i.e., erratic blocks). Type C includes regional markers of
4
Audemard M. and Michetti
tectonic uplift or subsidence, such as flexural-slip thrust faults, reef tracts, (truncated) bioherms, micro-atolls, terrace risers, river channels, progressive unconformities, and mountain fronts. ON-FAULT PALEOSEISMIC OBJECTS On-fault features, i.e., those related to fossil or fresh surface rupture of earthquakes, attracted the attention of researchers from the early days of paleoseismology. Owing to the scarcity of natural exposures of these features (with very rare exceptions, such as the case studied by Audemard et al. [1999] and also the spectacularly preserved surface ruptures of normal fault earthquakes in Nevada [e.g., Slemmons, 1957; Wallace, 1984]), or the rare availability of anthropogenic cuts across them (e.g., Nicol and Nathan, 2001; Mörner et al., 2003; Kuhn, 2005; López, 2006; López C. and Audemard M., this volume), paleoseismic trenching has become the most widespread technique of viewing these ground or sedimentary disturbances in the late Quaternary geologic record. We could actually state that paleoseismology at its early stages essentially developed around trenching studies. Trenching is typically practiced across active fault traces that are currently well identified by means of their geomorphic expression via fault-related landforms (D in Fig. 1A and I in Fig. 1B). Several hundreds of these studies are reported in the literature (e.g., Sieh, 1978; Machette et al., 1987; Crone and Luza, 1990; Van Dissen et al., 1992; McCalpin et al., 1993; Okumura et al., 1994; Audemard, 1996; Michetti et al., 1996; Townsend, 1998; among many others), but even larger numbers of paleoseismic studies are of a confidential nature or can only be found in unpublished reports, particularly in the United States, as many of these studies have been performed in compliance with regulations or legislation. This volume includes several new cases of these types of studies (Audemard, Clark, Lalinde P. et al., López C. and Audemard M., McCalpin, Olig et al.). Paleoseismic trenching strongly relies on a previous very thorough and detailed neotectonic assessment of the fault or region (Audemard and Singer, 1996, 1997, 1999; Audemard, 2005; Michetti et al., 2005b) to identify the best potential trenching sites. The main aim of this technique is to expose the interplay of fault activity with recent sedimentation in trench walls. Young (late Pleistocene to Holocene) sediments record perturbations that are directly related to the fault interference with the sequence. We emphasize that successful results from trench assessments heavily rely on the recognition of active or capable fault traces and the understanding of the interaction between tectonics and sedimentation at the chosen trench site, prior to excavation. The only way of bypassing this lengthy procedure of trench-site selection is when a trench can be directly excavated across a newly ruptured fault surface break (i.e., the Cariaco 1997 earthquake surface rupture studied by Audemard [1999a, 1999b, 2007, this volume], and the ChiChi 1999 rupture studied by Chen et al. [2004]), in which case an accurate location of the seismogenic fault trace(s) and the tectonic style of the active fault are provided. Finally, the trenching approach is not only applied to on-fault features but also to all
those perturbations indirectly induced by active faulting or by seismic shaking that may be suitably recorded in the sedimentary sequence (Fig. 1). This issue shall be dealt with in more detail later in this paper. In particular settings, geophysical investigations have been of great help in trench studies. The most commonly applied geophysical methods are ground-penetrating radar, which is widely known by its acronym GPR (e.g., Cai et al., 1996; Jol et al., 2000; Reicherter et al., 2003; Ollarves et al., 2004; Maurya et al., 2005), and electrical resistivity tomography (ERT; e.g., Silva et al., 2001; Caputo et al., 2003; Colella et al., 2004; Fazzito et al., 2009). These methods have helped to locate the active fault trace with a much higher precision; these traces often exhibit subdued or very subtle earthquake-related geomorphic features or are situated in land that has been heavily modified by anthropogenic activities (e.g., Cushing et al., 2000; Dost and Evers, 2000; Jongmans et al., 2000; Lehmann et al., 2000; Meghraoui et al., 2000; Verbeeck et al., 2000; Demanet et al., 2001). In the particular case of the surface trace of a gently dipping thrust fault that tends to parallel and mimic the topographic contours, the localization of the fault trace solely on the basis of morphologic expression is very difficult and requires the support of these geophysical surveys prior to trenching (e.g., Chow et al., 2001; Fazzito et al., 2009). In addition, GPR and/or ERT have also been applied to onshore strikeslip faults, where often the fault splays into several branches, and these techniques serve to determine the branches that are the most active (e.g., Gross et al., 2000; Wise et al., 2003; Audemard et al., 2006; Kürçer et al., 2008). The most common on-fault geomorphic features used as paleoseismic indicators, from a geomorphologic viewpoint, are fault scarps or counterscarps, sag or fault ponds, pop-ups or pressure ridges (at smaller scale, these are known as mole tracks), shutter ridges, and open fissures, among others. Regardless of the tectonic style (thrust, normal, or strike-slip faulting), fault scarps are definitely the most widely assessed feature due to their common occurrence (D and G in Fig. 1A and I in Fig. 1B). Furthermore, they constitute a perturbation at the ground surface, which triggers morphodynamic processes such as erosion, redeposition, and surface smoothing that have a sedimentary signature and are prone to fossilization. Depending on the balance between tectonic and sedimentation/erosion rates at the scarp, three different situations are recognized that are recorded distinctively: (1) When tectonic rate is higher than sedimentation rate, the scarp grows through time and sedimentation is bounded to the “lowtopography” block; (2) when tectonic and sedimentation rates are similar, available space is filled by sediments and the fault-related landform is leveled; and (3) when sedimentation rate is higher than the tectonic rate, the landform is buried by sediments. This last condition leads to blind normal or thrust faulting, which in turn induces growing fault-propagation folding. Depending on whether the difference between sedimentation and tectonic rates is small or large, this deformation ought to be assessed either as on-fault evidence or an off-fault feature, respectively. In regions where event dating is difficult to impossible due to lack or scarcity
Geological criteria for evaluating seismicity revisited of datable materials (K in Fig. 1B), techniques based on the diffusion equation (e.g., Begin, 1992; Hanks, 2000) have been applied to date scarps by estimating the degree of scarp degradation. This technique experienced a boom in the western United States in the 1970s and 1980s (e.g., Bucknam and Anderson, 1979; Nash, 1980; Colman and Watson, 1983), but its generalized application on a worldwide scale has become very limited because it is strongly dependent on local climate, which may vary quite rapidly within a basin or along a mountain front. It definitely depends on a local diffusivity (Phillips et al., 2003), which is different from place to place. In recent years, cosmogenic dating (radionuclides 10 Be, 26Al, and 36Cl, and stable nuclides 3He and 21Ne) of fault scarps by exposure to cosmogenic radiation has largely replaced the scarp degradation technique. This approach has been applied to fault scarps both in Quaternary alluvial deposits (e.g., Phillips et al., 2003; Siame et al., 2006) and bedrock (e.g., Zreda and Noller, 1998, 1999; Mitchell et al., 2001; Siame et al., 2006). Most commonly, this technique provides data on tectonic slip rates during the time span of scarp exposure to cosmogenic radiation (e.g., Ritz et al., 1995; Van der Woerd et al., 1998; Palumbo et al., 2004). In a similar manner, it has been used for dating nontectonic scarps, such as sackung uphill-facing scarps (B in Fig. 1A) by Hippolyte et al. (2006), which could also eventually provide pertinent paleoseismic information in the cases where these large mass instabilities were seismically triggered (refer to section “Seismically Induced Effects” later herein). As mentioned earlier, scarps are the features most commonly trenched for paleoseismic purposes. The success of this type of investigation is enhanced if sag- or fault-bounded ponds form against them. However, tectonic scarps are seldom preserved in the case of thrust faults, unless the fault cuts across very competent rocks. Most Quaternary thrust faults cut across soft or unconsolidated deposits, since they tend to crop out at the foot of the mountain front. It is frequently the case that tectonic slip rates of such thrust faults are much less than the sedimentation rate, which can lead to an almost complete masking of the active thrust fault by younger sediments, creating the situation of a blind thrust fault. If no other option is available, paleoseismologists have overcome this situation by trenching sympathetic faults—secondary faults that form in mechanical and kinematical association with the main thrust fault, such as scarps of flexuralslip thrust faults (Fig. 1B; e.g., Yeats, 1986a, 1986b; Costa et al., 1999; Livio et al., 2009). Back-thrust faults are also a potential trenching target. They happen to exhibit a higher dip than the main thrust and very frequently face in an upstream direction, which offers a better scenario for the interaction between tectonics and sedimentation. The major limitation in these situations is the low degree of certainty the researcher has on the actual kinematic activation of the sympathetic and/or antithetic faults in relation to the coseismic slip of the seismogenic thrust fault. In other words, the sympathetic fault does not necessarily move each time the main thrust fault does, which converts it into a poorly reliable chronometer. Moreover, although mechanically connected, there are as yet no published cases of the relation-
5
ship between actual slip on the major fault and the induced slip on the flexural-slip faults. Earthquake size estimates in this case should be based on: (1) intensity data (Serva, 1994), which are then converted into magnitude values, when available, and (2) the application of the concept of seismic landscape (Michetti et al., 2005b). The definition of the relationship between macroseismic intensity and earthquake ground effects was in fact the rationale for the introduction of the ESI 2007 Intensity Scale proposed by INQUA (Michetti et al., 2007; Reicherter et al., 2009). Generally speaking, all the previous geomorphic objects, which favor the continuous fine-grained, thinly bedded sedimentation against the fault plane, constituting the ideal setting for the recording of future ground-surface deformations, have proved to be really useful for paleoseismic studies. After overcoming water problems (through draining/water pumping in tropical regions and digging frozen ground in temperate regions), sag and fault ponds have generally provided the best and most complete paleoseismic records (e.g., Audemard, 1997; Lindvall et al., 2002; Audemard, this volume). Several favorable geologic conditions (for more details, refer to Audemard, 2005) have to come together at these “trenchable” sites in order to increase the chances of successful seismic hazard assessment. The paleoseismic assessment of any of the aforementioned objects can yield information on most or all of the following aspects (Audemard, 2005): confirmation of Holocene fault activity, slip per event and average slip rate of a given fault or fault segment, slip vector decomposition from fault-plane kinematic indicators, recurrence intervals and magnitude of the larger earthquakes (seismic potential) on known faults, coseismic rupture length, fault segmentation, fault interaction as a consequence of stress loading by stick-slip on contiguous faults, time-space distribution of seismic activity along a given tectonic feature, elapsed time since latest event and estimated time to the next event, which determines the likelihood of occurrence of a future earthquake, seismotectonic association of historical earthquakes, and short- and long-term landscape evolution. In cut or trench exposures, the most frequent kinds of geologic evidence of the interaction of sedimentation and faulting (Amit et al., 1995; Audemard, 2005), which are interpreted as the result of individual earthquakes, are: (1) filling of groundsurface open fissures (tension or T cracks, synthetic Riedel or R shears, or their combination); (2) fault-scarp–bounded deposition on a downthrown block; (3) a colluvial wedge derived from scarp degradation/subduing, occasionally interspersed with sedimentation on the downthrown fault compartment; (4) event horizon (faulting sealed by a leveling depositional episode); and (5) normal-fault propagation folding, later buried by overlying beds. Trees have been excellent time recorders—chronometers— of the occurrence of large contemporary earthquakes, in association with, as well as away from, their surface breaks (e.g., Sheppard and Jacoby, 1989). The study of the annual growth rings of certain particular plant species, known as dendrochronology, which is somewhat equivalent to varve chronology in lakes of temperate regions (which has applications in higherlatitude regions of the globe that exhibit a marked seasonality),
Figure 1. Generalized sketches of different potential settings of coexisting natural indicators of seismic activity (see text for explanation). (A–G) Upper cartoon depicts a subduction setting, where the paleoseismic geomarkers in relation to the marine environments are to the left, while those related to the orogen are to the right. The lower cartoon emphasizes the objects related to several subaqueous subenvironments. (H–L) The upper diagram tries to show the potential paleoseismic objects existing in a compressional environment, dominated by either outcropping or blind thrust faults. The lower diagram summarizes the paleoseismic objects to seek in a cratonic environment subject to glaciation-deglaciation processes, such as Scandinavia. LGM—Last Glacial Maximum.
6
Figure 1. (Continued.)
7
8
Audemard M. and Michetti
has revealed that certain growth changes (ring eccentricity, dead and healing zones, among others) that are the result of seismically induced perturbations can be recorded in big, long-lived tree trunks (e.g., Meisling and Sieh, 1980; Yamaguchi et al., 1997). In response to the natural positive phototropism exhibited by most large trees, loss of tree verticality—tilting of a tree—is soon naturally repaired in the years directly after the perturbing event, which is nicely recorded in the growth of annual rings. This has been shown in localities close to or right on top of contemporary earthquake ruptures (e.g., Page, 1970; Sheppard and White, 1995; Jacoby et al., 1997; Lin and Lin, 1998; Carver et al., 2004). In some cases, disruption of tree roots by ground cracks (e.g., Tasdemiroifelgülu, 1971; Carver et al., 2004) is marked by a deceleration of ring growth, which records the partial or total death of the tree (e.g., Toppozada et al., 2002; Doser, 2004). Even more astonishingly, large trunks, especially of sequoia trees, have been split into two by a crosscutting earthquake surface rupture (G in Fig. 1A; e.g., Carver et al., 2004). This illustrates that growth anomalies of tree rings can be considered as direct or indirect evidence of surface faulting, depending on the degree of association with the ground rupture. The limitation of this kind of evidence, as seen from a paleoseismic viewpoint, resides in the low probability of its preservation and the even lower probability of actually finding the preserved evidence. It has certainly proved to be useful as long as the tree lives. In that sense, this technique in very recent times is being used widely for mass movement monitoring and detection (e.g., Pop et al., 2008; Surdeanu et al., 2008), although its applicability has been known for quite a while (e.g., Terasmae, 1975; Hupp et al., 1987). In coastal areas along subduction zones, it has been observed that after the occurrence of large subduction earthquakes, large coastal areas have drowned after the seismic energy was released, as evidence of coseismic subsidence, due to elastic rebound. This has led to the death of extensive stands of trees as a result of the flooding by marine-water incursions of the low-lying areas and is commonly referred to as “drowned or ghost forests” (e.g., Atwater, 1987; Hamilton et al., 2005). This type of event is preserved in the sedimentary record as a couplet consisting of a horizon of organic-rich terrestrial sediments overlain by a marine sediment layer (A in Fig. 1A; e.g., Nelson et al., 1996; Cundy et al., 2000). OFF-FAULT PALEOSEISMIC OBJECTS The term “off-fault evidence” as defined by McCalpin (1996), or “indirect perturbations” as originally defined by Audemard (2005) and enlarged herein, covers all the earthquake-related or earthquake-triggered evidence different from ground-surface rupturing coinciding with the surface trace of the causative fault, although they can occasionally concur with the fault plane as nearfield features. To illustrate this eventual coalescence of processes, a sand dike, which results from earthquake shaking that triggers liquefaction, takes advantage of the recently displaced causative fault plane, as has been observed in trench walls by Rockwell et al. (2001) and Audemard et al. (2008), because it probably has
a transient higher permeability that facilitates the venting process several seconds to a few minutes after the earthquake occurrence. Since the off-fault evidence is commonly not associated with the causative fault plane, its paleoseismic interpretation needs more elaboration than the on-fault objects. For instance, the interpretation of these features in terms of earthquake magnitude is rarely unequivocal, unless their evaluation is carried out spatially (e.g., Ricci Lucchi, 1995). Each object of the same nature or type (e.g., slide, sand blow, or broken speleothem) should be envisaged as a single “seismometer,” and only the integration of them in a “network” relation can provide a reliable estimate of earthquake magnitude. This tacitly implies that all objects of the same type need to be ascribed to the same event. They must be temporally related; otherwise they cannot belong to a network. Significant progress has been achieved by this approach for mass movements and seismically induced liquefaction. Using modern earthquakes as analogues, relationships between earthquake magnitude and size or spatial distribution of slides have been derived (e.g., Youd and Hoose, 1978; Keefer, 1984; Wilson and Keefer, 1985; Keefer and Manson, 1989; Rodríguez et al., 1999; Crozier, 1992; Keefer, 2000) that can be applied, as an inverse method, to fossil landslides in paleoseismic studies to estimate the magnitude and localization of the epicenter of the triggering event, and by association, the causative fault as well. The same procedure has been carried out for liquefaction features (e.g., Kuribayashi and Tatsuoka, 1975; Youd and Perkins, 1987; Ambraseys, 1988; Barlett and Youd, 1992; Papadopoulos and Lefkopoulos, 1993; Munson et al., 1995; Allen, 1986; Galli, 2000; Castilla and Audemard, 2002; Rodríguez et al., 2002; Rodríguez Pascua et al., 2003; Papathanassiou et al., 2005; Rodríguez et al., 2006; Castilla and Audemard, 2007). On the other hand, and this is a very sensitive issue, the size of these objects may have no relation with the earthquake magnitude or the seismic attenuation laws. For instance, the lateral extent and thickness of homogenites, and tsunamites, tend to reflect pre-earthquake local conditions of the area where they occur. This is also the case for sand dikes formed in association with lateral spreading (Fig. 1A). Their final width, contrary to claims made by Obermeier (1996), is strongly conditioned by the size of the nearby available void (both in width and height) that sets in motion the horizontally sliding mass, creating the fractures through which venting will occur. In many cases, these offfault features provide a good time constraint for the occurrence of the inducing earthquake. So, they can be considered as reliable chronometers. The features implying redeposition appear to be best suited in this respect. These include tsunamites, homogenites, and erratic (tsunami-transported) boulders and blocks, but also broken speleothems, affected bioherms, mass movements, particularly those with ground-surface disruption, and even sand blows. In fact, their occurrence gives a good indication of the timing of the earthquake. On the other hand, other objects can merely reveal that an area is seismo-tectonically active, only due to their occurrence within the sedimentary record. Many soft-sediment deformations, even after reliably confirming their seismic origin,
Geological criteria for evaluating seismicity revisited fall into this category. All those soft-sediment deformations that maintain parallelism to the stratification after remobilization, including convolute bedding, contorted layers, basal surface of slumped beds, flame structures, ball and pillars, sand-vented sills, etc., are useless in terms of earthquake chronometers. Also, penetrative liquefaction features, such as sand dikes, like the ones depicted by Rodríguez-Pascua et al. (2000), need careful handling in terms of age determination of the triggering earthquake. If there is no convincing proof that such a dike reached the preearthquake ground surface by being connected to a sand blow at the top, the age of the youngest disrupted bed should be considered as the oldest age of occurrence of the associated earthquake. In other words, the age of the event in such a case is younger than the last affected stratigraphic marker. This is also valid in the cases when the sand-venting dike is later truncated by erosion. Based on their generating mechanism, in combination with the paleoseismic approach that could be applied to their study, we have subdivided the indirect or off-fault objects into three subcategories: Type A includes seismically induced effects; type B consists of remobilization and redeposition of sediments and/or rocks; and type C entails regional markers of vertical deformation (surface level changes). These are the features that have been the subject of significant development in the past years; in some cases, huge progress has been made. Paradoxically, the inaccessibility of subduction fault zones can be primarily held responsible for the recent progress in the study of all these indirect types of earthquake evidence, as anticipated by Adams (1996b). The reality that subduction zones lie underwater and are responsible for the largest known and most destructive contemporary earthquakes worldwide (e.g., 22 May 1960, Valdivia; 27 March 1964, Anchorage; 26 December 2004, Sumatra; and 27 February 2010, Concepción, earthquakes), in combination with their very large extent around the Pacific rim, as well as their position in relation to large population centers, with their costly infrastructure along the Pacific shores, make the assessment of their seismic potential an urgent necessity. This all has led to the realization that it is necessary to study ground deformations and on-land sedimentary disturbances as evidence of earthquakes in the coastal zones (e.g., Atwater et al., 1995), as well as along the submarine trench during earthquakes. Adams (1996b) had already conducted a rather thorough overview of the ways in which the seismic hazard assessment of such a complex active tectonic and geodynamic setting could be tackled, by presenting the case of the West Canadian coast. A multiproxy investigation, using several of the potential indirect objects (Fig. 1), actually appears to be the most appropriate approach in order to derive the seismic history (earthquake chronology) of a given subduction segment or a given region. It includes any or some of the phenomena, such as earthquake-triggered liquefaction, coseismic land-level changes, and associated sedimentary response (interbedding of marine and continental deposits, growth increase or truncation of living [brain] corals, emerged or submerged coseismic marine terraces), and onshore tsunamites, as well as underwater mass wasting, pockmarks from degassing/dewatering, mud and sand
9
volcanoes, offshore turbidites, and homogenites. This concept of a multiproxy approach has been applied to continental regions as well, for instance, in the interior of Canada (Adams, 1996b), in the alpine lakes of Switzerland and Italy (Becker et al., 2005; Fanetti et al., 2008), and in Venezuela (Audemard, 2005). Eventually, by applying some of the existing relationships treated in the literature for mass wasting and/or liquefaction, mentioned earlier, we will be able to make a minimum estimate of the triggering paleo-earthquake size. Seismically Induced Effects Mass movements and ground liquefaction are the two most common and widespread natural phenomena associated with earthquakes worldwide. These effects, induced by earthquake shaking, have been commonly recognized and described for onshore environments in historical documents over the centuries. For instance, description of liquefaction features by Sarconi (1784) and illustration of mass movements by Simón (1627) are classical examples of historical reports for seismically induced geological effects. Mass movements, because of their requirement of an energy gradient, typically occur in mountainous regions, but they can also affect large flat areas in the form of lateral spreading (Fig. 1A). Conversely, liquefaction tends to affect geologically young, low-lying, flat alluvial fill areas of fluvial or coastal environments. However, these two phenomena are not exclusive of onshore environments. On the contrary, they appear to develop underwater, in sea-bottom and lake-bottom sediments, and are more frequent and more widespread, as observed during contemporary earthquakes (e.g., Sims, 1973, 1975; Mosher et al., 1994; Nakajima and Kanai, 2000), as well as known from the past geologic record (e.g., Hempton and Dewey, 1983; El-Isa and Mustafa, 1986; Ringrose, 1989; Roep and Everts, 1992; Maltman, 1994; Hampton et al., 1996; Marco et al., 1996; Mörner, 1996; Rodríguez-Pascua et al., 2000; Mörner et al., 2000, 2003). These induced effects have proved, during many destructive earthquakes, to be more harmful than the earthquake shaking itself. As an illustration of this, there is the large destruction produced by the Prince William Sound, Anchorage-Alaska, 1964, earthquake at Turnagain Heights due to lateral spreading (Hansen, 1965). In the same way, either large mass-wasting bodies (e.g., Ancash, Perú, 1970, earthquake, with 18,000 casualties; Cluff, 1971; Plafker et al., 1971) or widespread mass movements have caused generalized destruction and numerous fatalities in urban areas, such as during the Deixi, Sichuan-China, 1933, earthquake, with 6800 casualties (Li et al., 1986). Associated with slope instabilities, an even more harmful factor has been the breaching of slide-dammed lakes days or months after the earthquake (Fig. 1B; e.g., the Mocotíes 1610 event in the Mérida Andes—Singer, 1998; Ferrer, 1999; the Deixi, Sichuan-China, 1933, earthquake, with 2500 casualties—Li et al., 1986). As for outcropping active fault planes, trenching has been practiced onshore for both earthquake-induced gravitational scarps (B in Fig. 1A; e.g., Wallace, 1984; Crozier, 1992; Nolan
10
Audemard M. and Michetti
and Weber, 1992, 1998; McCalpin, 1999; Onida et al., 2001; McCalpin and Hart, 2002; Tibaldi et al., 2004; Gutiérrez et al., 2005, 2008) and liquefaction features (e.g., Tuttle et al., 1990; Audemard and De Santis, 1991; Tuttle and Seeber, 1991; Clague et al., 1992; Walsh et al., 1995; Tuttle, 2001; Mörner et al., 2003; Cox et al., 2004; Guccione, 2005) in many different tectonic settings. Not only have these features proved to be the only means of assessing the seismic history of subduction zones (A in Fig. 1A), but they have been used elsewhere where faults have not ruptured the ground surface. For instance, trenching for liquefaction features has been carried out in the New Madrid area to recognize not only the effects of the 1811–1812 earthquake sequence and their extent, but also those of its precursors (Tuttle, 2001; Cox et al., 2004; Guccione, 2005). In cratonic areas subject to glaciation and deglaciation processes during the Quaternary, such as northern North America and Scandinavia, where Holocene sediments act as young recorders of recent deformation, and recognized onshore (intraplate) faults with Holocene surface rupture are scarce (e.g., Lagerbäck, 1979; Adams, 1989; Grant, 1990; Adams et al., 1991; Fenton, 1994), the study of indirect paleoseismic objects, as well as other features or approaches (e.g., tsunami deposits, age determination of fault-scarp exposure and scarp-profile degradation studies, dendrochronology, lichenometry; Fig. 1B), has clearly established that these socalled stable regions are seismically active. Adams (1996b) and Mörner (2005) have found that unexpected contrasting seismic patterns at the time of deglaciation, characterized by large and very frequent earthquakes, also took place. In addition, the study of earthquake-induced effects (mass wasting and soft-sediment deformation in a more general way) is becoming a complementary tool to on-fault trench studies. In Venezuela, several periglacial (paleo-)lakes directly offset by the active Boconó fault have been cored and sampled in order to demonstrate the applicability of this approach, with the intention of extending its use to seismically active regions with no outcropping seismogenic faults (Carrillo, 2006; Carrillo et al., 2006a, 2006b, 2009). In Israel, very long earthquake histories have been reconstructed through the study of soft-sediment deformations in late Pleistocene to Holocene laminated lacustrine sequences of the Dead Sea region (Marco et al., 1996; Agnon et al., 2006). In addition to directly exposing the seismic evidence through trenching, these past seismically induced effects in the geologic record have been assessed through indirect methods, either invasive (coring and geotechnical methods; E in Fig. 1A and H in Fig. 1B) or noninvasive (geophysical methods). Fossil liquefaction features (particularly sand blows) in the subsurface have been uncovered by GPR (e.g., Liu and Li, 2001). Since these earthquake-induced effects may also occur underwater, their investigation under such conditions has mostly relied on the combination of geophysical methods and sediment core recovery (Figs. 1A and 1B). The overall geometry of the sedimentary bodies and their stratigraphy are obtained from the geophysical investigation, which can call upon methods such as side-scan sonar to provide highly detailed sea-bottom topography, high-resolution shallow seismic, including multi-
beam, sparker, boomer, and pinger (subbottom profiler), among others, while coring reveals the physical characteristics of the sediments and deformations, and also brackets the time of occurrence of such effects (if dated). These underwater objects have been studied both at sea (Schafer and Smith, 1987; Piper et al., 1999) and in lakes (Beck et al., 1992; Shilts and Clague, 1992; Thomas et al., 1993; Beck et al., 1996; Chapron et al., 1996; Becker et al., 2005; Schnellmann et al., 2005; Carrillo, 2006; Carrillo et al., 2006a, 2006b, 2009; Beck, this volume), employing some of the aforementioned methods. The timing of activation (or reactivation) of earthquakerelated mass movements has been derived from trenching across the scarp or scar at the slide crown of such mass wasting (e.g., Wallace, 1984; Onida et al., 2001; McCalpin and Hart, 2002; Tibaldi et al., 2004; Gutiérrez et al., 2010). In this sense, large sackungen have received particular attention (B in Fig. 1A; e.g., Gutiérrez et al., 2005, 2008). From these studies, it is noteworthy that a significant uncertainty remains as to the completeness of the earthquake chronology derived by these methods. It still remains a difficult task to ascertain that a particular slide of a given size and certain characteristics will always be prone to slide during an earthquake of a particular magnitude or intensity above a certain threshold. It is equally difficult to assign a seismic origin to many sackungen in high mountain regions, especially if their occurrence is concurrent with deglaciation periods (Hippolyte et al., 2006). If the displacement of the mass movement is large enough and the mobilized mass does not completely fall down, the consequent depression generated at the foot of the scarp may create new space for the emplacement of small lakes. In such a case, the earthquake chronology can be assessed by the study of the perturbations recorded in the bottom sediments of these lakes (B in Fig. 1A; soft-sediment deformation, turbidites, [seicheinduced] homogenites, slumps, among others). If these lakes or ponds happen to be small in size, only coring can be performed. If a lake is large and deep enough, geophysical prospecting can be undertaken. Some large mass movements, particularly rock avalanches, but also huge debris and mud flows, may have important runoff and may block the valleys by their downstream deposits. This may result in the damming of water bodies of considerable volume upstream. This process has been observed in several historical and contemporary earthquakes (e.g., Simón, 1627; Li et al., 1986; Clague and Evans, 2000). Consequently, the recognition of the aforementioned markers (soft-sediment deformation, turbidites, seiche-induced homogenites, slumps, among others) in lake-bottom deposits in slide-dammed lakes has also been used as a natural indicator of past earthquakes (Costa and Schuster, 1988; Weidinger, 1998; Hermanns et al., this volume). This same approach can be applied to water bodies that are pounded against a coseismically growing active fold (H in Fig. 1B). An example of this happened at Oued Foda during the 1980 El Asnam–Algeria earthquake, above an active blind thrust fault (Philip and Meghraoui, 1983). The analysis of such lake fills may help to reconstruct the seismic history of these faults
Geological criteria for evaluating seismicity revisited that show no evidence of brittle expression at the surface as direct evidence. Geophysical methods, such as vertical electrical soundings (VES) and ERT, have been used jointly to calibrate multielectrode profiles and to estimate the depths of landslide deposits in the town of Celano (Fucino Basin, central Italy; Rinaldini et al., 2008). Assuming that the quantum leaps forward in multichannel seismic acquisition technology are maintained in the years to come, we could easily envisage a situation in which much higherresolution three-dimensional (3-D) seismics (seismic cube) would allow the regional-scale mapping of most of these seismically induced effects, particularly those paleoseismic objects that rest on a perturbed ground surface or seabed (e.g., sand blows, slides, but also mud volcanoes and pockmarks; Fig. 1). This modern seismic method currently allows the construction of time slices, although not yet of sufficiently high resolution. These constitute a reconstruction of the environment or topography at a given time and thus would fulfill the same objectives as current macroseismic surveys, from which magnitude and epicenter of the causative earthquake could be derived by applying known relationships. In other words, at each identified event horizon, corresponding to a time slice in the seismic cube, the paleoseismic objects would be mapped in as much detail as the resolution of the method would permit. Those phenomena described here are known in the literature as seismically induced effects, but other, less well-known, natural features also form as a result of seismic shaking, among which we shall discuss the following: mud volcanoes, oil seeps and pockmarks, broken speleothems, and toppled precariously balanced rocks. Mud volcanoes/diapirs (F in Fig. 1A), and even salt domes, can be used as analogs for earthquake-triggered sand venting in the form of dikes connected to sand blows. Although it is known that mud diapirs may form and flow to the surface during earthquakes (e.g., Arnold et al., 1960), their use as paleoseismic indicators, to our knowledge, has not yet been reported in the literature. This can be attributed to the fact that they are not exclusively associated with earthquakes. Furthermore, many of these features do not have an episodic motion, but rather they have a steady-state ascending motion, which would make it more difficult to ascribe a seismic origin, although their upward motion and also final eruption are known to accelerate during earthquake shaking. The process of seabed degassing or gas seepage, which produces pockmarks as surface evidence (Hovland and Judd, 1988; Hovland et al., 2002; Pinet et al., 2008) is also known to accelerate during earthquakes (Fig. 1A). Pockmarks are shallow depressions (5–300 m in diameter and 2–20 m deep) formed by the explosive release of gas (Gluyas and Swarbrick, 2004). Seismic shaking can provoke unusually large flows from gas or oil seeps (Levorsen, 1967; Field and Jennings, 1987), similar to what happens with water springs that may resume or stop flowing as a consequence of earthquake action (e.g., González et al., 2004). For instance, following the 1971 San Fernando Valley (California) earthquake, several of the previously inactive oil seeps in the area resumed activity (Mandel, 2001). Unfortunately, there is
11
no way that degassing evidence will be recorded in the geologic record, but pockmarks are perfectly well preserved on the seabed, as revealed by side-scan sonar. These features potentially provide the same paleoseismic information as earthquake-triggered sand blows. Underwater, they can be detected by high-resolution single-channel seismics (sparker or subbottom profiler) or multibeam surveys, whereas age bracketing can be obtained on coring. So far, it has not been applied for specific paleoseismic purposes. Speleothems are also excellent recorders, as well as highly precise chronometers of seismic shaking (C in Fig. 1A). For over 20 yr, speleothems have been used as indicators of seismic activity (Postpischl et al., 1991; Bini et al., 1992; Gilli, 1996, 1999; Delaby, 2001; Pérez López et al., 2009). This has been stimulated by the fact that they are easily datable (by 14C, for instance), they record growth perturbations very neatly, and they also grow by annual precipitation, similar to tree rings. Their principal limitation is that their occurrence is restricted to soluble carbonatic rocks. Anyhow, their steady way of growing permits detection of growth pattern disturbance as a result of catastrophic events with a temporal resolution of only 1 yr, similar to varve chronology and dendrochronology. Because of these characteristics, speleothems have been an object of many paleoseismic studies, as confirmed by the large number of studies in which historic (Gilli, 1999; Lemeille et al., 1999) and prehistoric (Postpischl et al., 1991; Delaby, 2001; Lignier and Desmet, 2002; Lacave et al., 2004; Kagan et al., 2005) earthquakes have been dated or constrained. Good overviews on the use of speleothems as paleoseismic indicators have been published by Forti (2001), Gilli and Delange (2001), and Becker et al. (2006). However, their interpretation is not straightforward, as demonstrated by Cadorin et al. (2001) and Gilli (2004). The latter author indicated that the breakage or failure of speleothems may also be caused by other processes different from seismic shaking, such as glacial advances or retreats. The possibility that sliding of the limestone masses containing the caves occurred as a result of gravitational processes also needs to be ruled out before ascribing a seismic origin, unless sliding can be assumed to have been triggered by a given earthquake. Precarious rocks or precariously balanced rocks indicate that strong earthquake motions have not occurred at a site since their particular precarious situation developed. In that particular negative sense, these rocks or groups of rocks can be effectively used as earthquake strong ground-motion seismoscopes that have been operating on solid rock outcrops for thousands of years (Brune and Whitney, 1992; Brune, 1994, 1996, 1999; Shi et al., 1996; Brune et al., 2003). Thus, estimates of the threshold ground acceleration for the toppling of these rocks can provide constraints on the peak ground accelerations experienced during previous earthquakes. Consequently, the presence of precarious rocks is an indicator for the absence of earthquakes above a certain magnitude. Preservation of these rocks in their precarious position is thus a function of the seismicity of the region under study with a propensity for the larger events, and this reduces its applicability to high-seismicity regions of the world. On the other
12
Audemard M. and Michetti
hand, toppled precarious rocks can be used as chronometers of the latest earthquake, the magnitude of which can be estimated as having exceeded the threshold value needed for toppling. The age of the place on which the precariously balanced rocks used to stand (commonly made of hard rock) could be determined by cosmogenic dating, thus providing time of occurrence of the latest event. Likewise, the age of the precarious rock toppling could be derived from dating the age of the ground surface on which the block fell (similar to J in Fig. 1B). If it fell on soil, the 14C method of dating could be applied, whereas if the block has fallen on hard rock or on sediments, even luminescence (thermoluminescence or optically stimulated luminescence) techniques could be applied. Remobilized and Redeposited Sediments We propose to include in this type of indirect paleoseismic object all those deposits of any grain size that have been transiently mobilized by the earthquake or associated phenomena (e.g., underwater or subaerial mass wasting, tsunami waves) and redeposited shortly after. Usually, this type of deposit, which includes turbidites, homogenites, and tsunamites, is generated in aqueous environments but results from different processes or intervening mechanisms. Seismically induced turbidites and homogenites require that underwater (eventually, subaerial) sediments were set in motion by the earthquake shaking (e.g., Francis, 1971; Siegenthaler et al., 1987; Shiki et al., 2000). Turbidites are generally the result of submarine mass-wasting processes causing hyperpycnal inflow that generates turbidity currents, which transport sediments downslope along the seabed (e.g., Adams, 1990, 1996a; Doig, 1991; Inouchi et al., 1996; Gorsline et al., 2000) in the days or months following an earthquake, as evidenced in the case of several twentieth-century earthquakes (e.g., Heezen and Ewing, 1952; Thunell et al., 1999). Homogenites, instead, would rather require the oscillatory motion (seiche) of a confined water body, although this motion is also fed from underwater mass wasting (e.g., slumps; e.g., Beck, this volume) or nearshore subaerial mass movements (e.g., collapse of the edifice of the Santorini Volcano as proposed by Cita et al., 1996). Conversely, most tsunamites are fed from nearshore sediments that are carried inland by tsunami waves (e.g., Atwater, 1987; Clague and Bobrowsky, 1994a, 1994b; Clague et al., 1994; Chagué-Goff and Goff, 1999; Bourgeois et al., 1999; Sawai et al., 2008; among many others). This does not exclude the situation where tsunamites can also appear as homogenites in deep and/or confined water bodies (e.g., Kastens and Cita, 1981; Beck et al., 2007). Typically, tsunamites are thought to be fine- to medium-grained deposits (e.g., Dawson et al., 1988; Shi et al., 1995; Paris et al., 2007), but in fact, the size of the transported material is a function of the tsunami wave energy and availability of materials in the nearshore environment in the case of onshore tsunamites (for a review, refer to Scheffers and Kelletat, 2003), or in the sea bottom for deep-water tsunamites (homogenites). Tsunami waves have been capable of removing and transporting large boulders
inland (e.g., Bourrouilh-Le Jan and Talandier, 1985; Jones and Hunter, 1992) and huge (several m3 in size) blocks of coral reefs, eolianites, or other nearshore deposits (Scheffers, 2004; Scheffers and Kelletat, 2005; Scheffers and Scheffers, 2007). Onshore tsunamites can be assessed in outcrop or by shallow coring (A in Fig. 1A). The coast of the Cascadia subduction zone of the northwestern United States has been a natural laboratory for over the past 40 yr, during which their study has much evolved (Atwater et al., 1995). In the early days, recognition of the tsunamites in that area relied mostly on sedimentary features recognized with the naked eye. The identification of these marine incursions nowadays relies on the combination of several disciplines: sedimentology, geochemistry, paleontology, including micropaleontology, malacology, palynology, and geochronology, among others, which has enhanced and facilitated their recognition in the geologic record. Since all other paleoseismic objects of this sort are mostly preserved in still-active natural environments, particularly in lake- or sea-bottom sequences, they require an integrated paleoseismic assessment relying on joint subaqueous geophysical prospecting methods and piston coring (e.g., Beck et al., 1996; Cita and Rimoldi, 1997; Doig, 1998; Carrillo et al., 2006b; Beck et al., 2007; Beck, this volume). However, paleoseismic indicators of this subset have also been retrieved from outcropping lake sequences (e.g., Doig, 1991; Carrillo et al., 2006a), which unfortunately have been truncated and as a consequence provide only incomplete paleoseismic records for the recent time window. Paleoseismic Indicators of Vertical Motion Most of the indirect paleoseismic objects discussed here constitute an invaluable help when direct evidence of coseismic faulting is unavailable. In other words, their recognition has been greatly enhanced by their study in subduction-related seismogenic source regions and occasionally also in association with blind thrusts. Earlier, we explained how elastic rebound acts coseismically in subduction zones and can be recognized by the presence of the so-called ghost or drowned forests after the occurrence of large subduction earthquakes in the coastal zone, where it is clearly recorded in the onshore sediments by a couplet of organic soils capped by a horizon of marine sediments (A in Fig. 1A). In addition, it can be perfectly recorded by marine biogenic constructions such as coral reefs and (micro-)atolls on the foreshore side of the coastal bulge (Fig. 1B). In inverse relation to the onshore sediment couplet, which registers emergence during the interseismic period followed by subsidence during the earthquake, the coral growth offshore first indicates steady subsidence during the interseismic period followed by abrupt uplift during the seismic energy release (Taylor et al., 1987; Sieh et al., 1999; Zachariasen et al., 1999, 2000). This sudden uplift, when of sufficient magnitude, can result in aerial exposure and subsequent coral head truncation by wave erosion or natural death of the coral species. Similar to tree rings and varves in temperate regions, growth of corals indicates seasonality in tropical regions, and
Geological criteria for evaluating seismicity revisited coral growth is registered in annual bands (e.g., Knutson et al., 1972; Woodroffe and McLean, 1990), but their occurrence is of course limited to the warm regions of the world. Furthermore, certain species of the coral genera Porites and Goniastrea are sensitive natural recorders of lowest tide levels, which make them ideal natural instruments for measuring emergence or subsidence relative to a tidal datum (Scoffin and Stoddart, 1978; Taylor et al., 1987). These Porites and Goniastrea coral heads grow radially upward and outward until they reach an elevation that allows their highest corallites to be exposed to the atmosphere during lowest tides. Consequently, rather small changes in elevation of the coral substratum or sea level induce very clear changes in the growth patterns of these coral colonies or micro-atolls (Zachariasen et al., 2000; Natawidjaja et al., 2007). Thanks to the excellent temporal and vertical resolution of these micro-atolls, it has been possible to decipher the seismic history of particular segments of subduction zones in tropical regions (e.g., Sieh et al., 2008). In the past 30 yr or so, the development of new, modern, and more precise dating techniques such as uranium series, thermoluminescence (TL), optically stimulated luminescence (OSL), and electron spin resonance (ESR), among others, in combination with a more accurate reconstruction of the history of Quaternary sea levels based on variations in the oxygen isotopic composition of seawater (Chappell, 1974; Chappell et al., 1996a), has stimulated and accelerated the study of both emerged and submerged coastal marine features in tectonically active regions worldwide, as well as improved their temporal resolution (Fig. 1B). This has led to an improved determination of vertical slip rates based on marine terraces, either constructional (biogenic or detrital) or erosional. Many appropriate areas have been studied in that respect, and some regions of the world have actually become natural laboratories, such as Huon Peninsula in Papua New Guinea, where the matching of these marine terraces to sea-level curves (Chappell, 1974; Chappell et al., 1996a) has allowed the recognition of other marine terraces, the origin of which has been ascribed to coseismic vertical tectonic motion (Ota et al., 1993; Chappell et al., 1996b; Ota and Chappell, 1996). Other regions of the world have been assessed likewise, such as Chile by Vita-Finzi (1996), New Zealand by Pillans and Huber (1995), and the Gibraltar Strait zone by Zazo et al. (1999). The Ventura Avenue anticline in California deserves particular mention in this respect, where Stein and Yeats (1989) established that this active fold is repeatedly raised by earthquakes recurring roughly every 600 yr, as registered in a series of coseismically uplifted marine terraces. The equivalent of marine terraces in coastal areas, alluvial terraces of the inland environment have been occasionally assessed for the purpose of determining coseismic vertical or horizontal deformations (e.g., Wellman, 1955; Suggate, 1960; Lensen, 1968; Lensen and Vella, 1971; Van Dissen et al., 1992) in continental interiors. However, the most notable contribution of these flights of staircase-style marine, lake, or alluvial terraces is in permitting us to arrive at reliable estimates of the vertical slip rate of tectonic blocks or folds that have been uplifted or subsided.
13
CONCLUDING REMARKS Over the past 40 yr, there has been a steady increase in the number of natural objects (generic name herein given to geologic and geomorphologic evidence, indices, markers, effects, indicators, etc.), both onshore and offshore on the seabed, that have been identified as being of distinct relevance to paleoseismology. This increase in the number and variability of natural objects has been in response to the need for an improved and wider-ranging characterization of the seismic potential of inaccessible seismogenic sources, such as subduction zones and intracontinental seismic sources (e.g., New Madrid seismic zone). This has led to the identification of numerous indirect indicators and to the understanding of their functioning. Consequently, the appropriate paleoseismic approach is intimately related to the seismic landscape under assessment and is defined as the cumulative geomorphological and stratigraphic effect of the signs left on the environment of an area by its past earthquakes over a geologically recent time interval (Vittori et al., 1991; Michetti et al., 2005b). Likewise, the use of other natural paleoseismic markers has partly decreased, been abandoned, or fallen into disuse because of their limited applicability or limited results in terms of quantifying the seismic potential, for any or several of the following reasons: (1) The occurrence of the object is restricted to a specific region, a climate, or process, like lake varves, tree rings, or coral bands. (2) There is high specificity of the object to local conditions, e.g., scarp degradation to highly variable diffusion equations, lichenometry. (3) There is limited or unlikely preservation of the object in the geologic record, such as dead trees for dendrochronology. (4) There is difficulty in determining the magnitude and location of past earthquakes based on the regional distribution of the natural objects, which results in great uncertainty in assigning the observed seismically induced effects to a single common earthquake. (5) There is a degree of uncertainty in the cause-effect association between the natural indicator and earthquakes, for instance, liquefaction or mass movements. In the ultimate instance, the final selection of the natural indicators available for use is the responsibility of the paleoseismologist in accordance with those objects that each environment offers in terms of natural features. Finally, if available, direct objects are certainly everyone’s preference because they provide straightforward answers to the seismic history and potential for future rupture of a given fault or fault segment. ACKNOWLEDGMENTS We profusely thank Hans Diederix (private consultant, Colombia) and Pablo Silva (Universidad de Salamanca, Spain) for fruitful comments and suggestions on an earlier version of this contribution. Many thanks are also due to Marina Peña for her wonderful china-ink, hand-made drawings. As invited editors of this Geological Society of America Special Paper, we would like to acknowledge the help and work provided by James McCalpin during this book’s earlier stages. This volume is a
14
Audemard M. and Michetti
contribution to the “Neotectonics and Paleoseismology” Focus Area of the Commission on Terrestrial Processes, Deposits and History (TEPRO) of INQUA and the former “Paleoseismicity” Subcommission of INQUA, as well. REFERENCES CITED Adams, J., 1989, Postglacial faulting in eastern Canada: Nature, origin and seismic hazard implications: Tectonophysics, v. 163, p. 323–331, doi:10.1016/ 0040-1951(89)90267-9. Adams, J., 1990, Paleoseismicity of the Cascadia subduction zone: Evidence from turbidites off the Oregon-Washington margin: Tectonics, v. 9, p. 569–583, doi:10.1029/TC009i004p00569. Adams, J., 1996a, Great Earthquakes Recorded by Turbidites off the Washington-Oregon Coast: Assessing Earthquake Hazards and Reducing Risk in the Pacific Northwest: U.S. Geological Survey Professional Paper 1560, v. 1, p. 147–158. Adams, J., 1996b, Paleoseismology in Canada: A dozen years of progress: Journal of Geophysical Research, v. 101, no. B3, p. 6193–6207, doi:10.1029/ 95JB01817. Adams, J., Wetmiller, R.J., Hasegawa, H.S., and Drysdale, J., 1991, The first surface faulting from a historical intraplate earthquake in North America: Nature, v. 352, p. 617–619, doi:10.1038/352617a0. Agnon, A., Migowski, C., and Marco, S., 2006, Intraclast breccia layers in laminated sequences: Recorders of paleo-earthquakes, in Enzel, Y., Agnon, A., and Stein, M., eds., New Frontiers in Dead Sea Paleoenvironmental Research: Geological Society of America Special Paper 401, p. 195–214, doi:10.1130/2006.2401(13). Allen, C.R., 1975, Geologic criteria for evaluating seismicity: Geological Society of America Bulletin, v. 86, p. 1041–1057, doi:10.1130/0016-7606(1975 )86<1041:GCFES>2.0.CO;2. Allen, J.R.L., 1986, Earthquake magnitude-frequency, epicentral distance, and soft-sediment deformation in sedimentary basins: Sedimentary Geology, v. 46, no. 1–2, p. 67–75, doi:10.1016/0037-0738(86)90006-0. Ambraseys, N., 1988, Engineering seismology: International Journal of Earthquake Engineering and Structural Dynamics, v. 17, p. 1–50, doi:10.1002/ eqe.4290170101. Amit, R., Harrison, J.B.J., and Enzel, Y., 1995, Use of soils and colluvial deposits in analyzing tectonic events—The southern Arava valley, Israel: Geomorphology, v. 12, p. 91–107, doi:10.1016/0169-555X(94)00081-2. Arnold, R., Macready, G.A., and Barrington, T.W., 1960, The First Big Oil Hunt, Venezuela, 1911–1916: New York, Vantage Press, 353 p. Atwater, B.F., 1987, Evidence for great Holocene earthquakes along the outer coast of Washington State: Science, v. 236, p. 942–944, doi:10.1126/ science.236.4804.942. Atwater, B.F., Nelson, A.R., Clague, J.J., Carver, G.A., Yamaguchi, D.K., Bobrowsky, P.T., Bourgeois, J., Darienzo, M.E., Grant, W.C., HemphillHaley, E., Kelsey, H.M., Jacoby, G.C., Nishenko, S.P., Palmer, S.P., Peterson, C.D., and Reinhart, M.A., 1995, Summary of coastal geologic evidence for past great earthquakes at the Cascadia subduction zone: Earthquake Spectra, v. 11, p. 1–18, doi:10.1193/1.1585800. Audemard, F.A., 1996, Paleoseismicity studies on the Oca-Ancon fault system, northwestern Venezuela: Tectonophysics, v. 259, p. 67–80, doi:10.1016/ 0040-1951(95)00144-1. Audemard, F.A., 1997, Holocene and historical earthquakes on the Boconó fault system, southern Venezuelan Andes: Trench confirmation: Journal of Geodynamics, v. 24, no. 1–4, p. 155–167. Audemard, F.A., 1998, Contribución de la paleosismología a la sismicidad histórica: Los terremotos de 1610 y de 1894 en los Andes venezolanos meridionales, in Ferrer, C., Laffaille, J., and Rengifo, M., eds., Selección de Artículos de las Primeras Jornadas de Sismicidad Histórica en Venezuela, Núcleo Universitario Rafael Rangel, ULA-Trujillo, Mayo 1997: Revista Geográfica Venezolana, Universidad de Los Andes, Venezuela, v. 39, no. 1–2, p. 87–105. Audemard, F.A., 1999a, El sismo de Cariaco del 09 de julio de 1997, edo. Sucre, Venezuela: Nucleación y progresión de la ruptura a partir de observaciones geológicas, in Proceedings of the VI Congreso Venezolano de Sismología e Ingeniería Sísmica [abs.]: Mérida, Venezuela (CD-ROM). Audemard, F.A., 1999b, Nueva percepción de la sismicidad histórica del segmento en tierra de la falla de El Pilar, Venezuela nororiental, a partir de
primeros resultados paleosísmicos, in Proceedings of the VI Congreso Venezolano de Sismología e Ingeniería Sísmica [abs.]: Mérida, Venezuela (CD-ROM). Audemard, F.A., 2005, Paleoseismology in Venezuela: Objectives, methods, applications, limitations and perspectives: Tectonophysics, v. 408, no. 1–4, p. 29–61, doi:10.1016/j.tecto.2005.05.034. Audemard, F.A., 2007, Revised seismic history of El Pilar fault, northeastern Venezuela, after the Cariaco 1997 earthquake and from recent preliminary paleoseismic results: Journal of Seismology, v. 11, no. 3, p. 311–326, doi:10.1007/s10950-007-9054-2. Audemard M., F.A., 2011, this volume, Multiple trench investigation across the newly ruptured segment of the El Pilar fault in northeastern Venezuela, after the 1997 Cariaco earthquake, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, doi:10.1130/2011.2479(06). Audemard, F.A. and De Santis, F., 1991, Survey of liquefaction structures induced by recent moderate earthquakes: Bulletin IAEG/AIGI (International Association of Engineering Geology/Association Internationale de Géologie de l’Ingenieur), v. 44, p. 5–16. Audemard, F.A., and Singer, A., 1996, Active fault recognition in northwestern Venezuela and its seismogenic characterization: Neotectonic and paleoseismic approach: Geofísica Internacional, v. 35, no. 3, p. 245–255. Audemard, F.A., and Singer, A., 1997, La Ingeniería de Fallas Activas en Venezuela: Historia y Estado del Arte, in Seminario Internacional de Ingeniería Sísmica: Aniversario del Terremoto de Caracas de 1967: Caracas, Venezuela, Universidad Católica Andrés Bello, p. 11–27. Audemard, F.A., and Singer, A., 1999, La Ingeniería de Fallas Activas en Venezuela: Historia y Estado del Arte, in Guevara, L.T., ed., Memorias Ponencias Presentadas en el “Curso Internacional Protección del Patrimonio Construido en Zonas Sísmicas”: Caracas, Venezuela, Facultad de Arquitectura y Urbanismo de la Universidad Central de Venezuela, p. 45–67. Audemard, F.A., Bousquet, J.-C., and Rodríguez, J.A., 1999, Neotectonic and paleoseismicity studies on the Urumaco fault, northern Falcón Basin, northwestern Venezuela: Tectonophysics, v. 308, p. 23–35, doi:10.1016/ S0040-1951(99)00081-5. Audemard, F.A., Ollarves, R., Díaz, G., Bechtold, M., and Cataldi, A., 2006, El geo-radar como herramienta para la definición de fallas activas: Aplicación en el sector central de la falla de Boconó, estado Mérida, Venezuela: Revista de la Facultad de Ingeniería de la Universidad Central de Venezuela, v. 21, no. 4, p. 57–70. Audemard, F.A., Ollarves, R., Betchtold, M., Díaz, G., Beck, C., Carrillo, E., Pantosti, D., and Diederix, H., 2008, Trench investigation on the main strand of the Boconó fault in its central section, at Mesa del Caballo, Mérida Andes, Venezuela: Tectonophysics, v. 459, p. 38–53, doi:10.1016/ j.tecto.2007.08.020. Barlett, S.F., and Youd, T.L., 1992, Empirical analysis of horizontal ground displacement generated by liquefaction-induced lateral spreads: State University of New York at Buffalo, National Center for Earthquake Engineering Research Technical Report NCEER-92-0021. Beck, C., 2011, this volume, Lake sediments as late Quaternary paleoseismic archives: Examples in the northwestern Alps and clues for earthquakeorigin assessment of sedimentary disturbances, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, doi:10.1130/2011.2479(07). Beck, C., Rochette, P., and Tardy, M., 1992, Interprétation en termes de paléoséismicité de niveaux destructurés dans des rythmites lacustres quaternaires des Alpes nord-occidentales: Comptes Rendus de l’Academie des Sciences du Paris, v. 315, no. II, p. 1525–1532. Beck, C., Manalt, F., Chapron, E., Van Rensbergen, P., and De Batist, M., 1996, Enhanced seismicity in the early post-glacial period: Evidence from the post-würm sediments of Lake Annecy, northwestern Alps: Journal of Geodynamics, v. 22, no. 1/2, p. 155–171, doi:10.1016/0264-3707(96)00001-4. Beck, C., Mercier de Lépinay, B., Schneider, J.-L., Cremer, M., Wendenbaum, E., Çagatay, N., Boutareau, S., Ménot, G., Schmidt, S., Weber, O., Eris, K., Armijo, R., Meyer, B., Nicolas, P., Gutscher, M.-A., and MARMACORE Cruise Party, 2007, Late Quaternary coseismic sedimentation in the Sea of Marmara’s deep basins: Sedimentary Geology, v. 199, p. 65–89, doi:10.1016/j.sedgeo.2005.12.031.
Geological criteria for evaluating seismicity revisited Becker, A., Ferry, M., Schnellmann, M., and Giardini, D., 2005, Multiarchive paleoseismic record of late Pleistocene strong earthquakes in Switzerland: Tectonophysics, v. 400, p. 153–177, doi:10.1016/j.tecto.2005.03.001. Becker, A., Davenport, C., Eichenberger, U., Gilli, E., Jeannin, P.-Y., and Lacave, C., 2006, Speleoseismology: A critical perspective: Journal of Seismology, v. 10, p. 371–388, doi:10.1007/s10950-006-9017-z. Begin, Z.B., 1992, Application of quantitative morphological dating to paleoseismicity of the northwestern Negev, Israel: Israel Journal of Earth Sciences, v. 41, p. 95–103. Bini, A., Quinif, Y., Sules, O., and Uggeri, A., 1992, Les mouvements tectoniques récents dans les grottes du Monte Campo Dei Fiori (Lombardie, Italie): Karstologia, v. 19, p. 23–30. Bourgeois, J., Petroff, C., Yeh, H., Titov, V., Synolakis, C., Benson, B., Kuroiwa, J., Lander, J., and Norabuena, E., 1999, Geologic setting, field survey and modeling of the Chimbote, northern Peru, tsunami of 21 February 1996: Pure and Applied Geophysics, v. 154, p. 513–540, doi:10.1007/ s000240050242. Bourrouilh-Le Jan, F.G., and Talandier, J., 1985, Sédimentation et fracturation de haute énergie en milieu récifal: Tsunami, ouragans et cyclones et leurs effets sur la sédimentologie et la gemorphologie d’un atoll: Motu et Hoa, à Rangiroa, Tuamotu, Pacifique SE: Marine Geology, v. 67, p. 263–333, doi:10.1016/0025-3227(85)90095-7. Brune, J.N., 1994, Constraints on ground motion from great earthquakes provided by groups of precarious rocks: Seismological Research Letters, v. 65, no. 1, p. 64. Brune, J.N., 1996, Precariously balanced rocks and ground motion maps for Southern California: Bulletin of the Seismological Society of America, v. 86, p. 43–54. Brune, J.N., 1999, Precarious rocks along the Mojave section of the San Andreas fault, California: Constraints on ground motion from great earthquakes: Seismological Research Letters, v. 70, no. 1, p. 29–33. Brune, J.N., and Whitney, J., 1992, Precariously balanced rocks with rock varnish: Paleoindicators of maximum ground acceleration?: Seismological Research Letters, v. 63, no. 1, p. 21. Brune, J.N., Anooshehpoor, A., Zeng, Y., and Anderson, J.G., 2003, Precarious Rocks Methodology and Preliminary Results, Final Report: Reno, University of Nevada. Bucknam, R., and Anderson, R., 1979, Estimation of fault-scarp-height-slopeangle relationship: Geology, v. 7, p. 11–14, doi:10.1130/0091-7613(1979)7 <11:EOFAFA>2.0.CO;2. Burbank, D., and Anderson, R., 2001, Tectonic Geomorphology: Maiden, Massachusetts, Blackwell Science, 274 p. Cadorin, J.F., Jongmans, D., Plumier, A., Camelbeeck, T., Delaby, S., and Quinif, Y., 2001, Modelling of speleothems failure in the Hotton cave (Belgium): Is the failure earthquake induced?: Netherlands Journal of Geosciences, v. 80, no. 3–4, p. 315–321. Cai, J., McMechan, G., and Fisher, M., 1996, Application of ground penetrating radar to investigation of near-surface fault properties in the San Francisco Bay region: Bulletin of the Seismological Society of America, v. 86, p. 1459–1470. Caputo, R., and Helly, B., 2008, The use of distinct disciplines to investigate past earthquakes: Tectonophysics, v. 453, p. 7–19, doi:10.1016/j.tecto .2007.05.007. Caputo, R., Piscitelli, S., Oliveto, A., Rizzo, E., and Lapenna, V., 2003, The use of electrical resistivity tomography in active tectonics: Examples from the Tyrnavos Basin, Greece: Journal of Geodynamics, v. 36, no. 1–2, p. 19–35, doi:10.1016/S0264-3707(03)00036-X. Carrillo, E., 2006, L’enregistrement Sédimentaire de la Sismicité Récente de la Frontière Sud-Occidentale de la Plaque Caraïbe (Faille de Boconó): Modalités et Chronologie, Contribution à l’Estimation de l’Aléa Sismique Régional [Ph.D. thesis]: Chambery, France, Université de Savoie, 367 p. Carrillo, E., Audemard, F.A., Beck, C., Cousin, M., Jouanne, F., Cano, V., Castilla, R., Melo, L., and Villemin, T., 2006a, A late Pleistocene natural seismograph along the Boconó fault (Mérida Andes, Venezuela): The moraine-dammed Los Zerpa paleo-lake: Bulletin de la Société Géologique de France, v. 177, no. 1, p. 3–17, doi:10.2113/177.1.3. Carrillo, E., Beck, C., Audemard, F.A., Moreno, E., and Ollarves, R., 2006b, Disentangling late Quaternary climatic and seismo-tectonic controls on Lake Mucubají sedimentation (Mérida Andes, Venezuela): Palaeogeography, Palaeoclimatology, Palaeoecology, Special Issue, v. 259, no. 2–3, p. 284–300.
15
Carrillo, E., Beck, C., and Audemard, F.A., 2009, La paleosismología lacustre: Una nueva herramienta en la evaluación del peligro sísmico en Venezuela: Boletín de la Academia de las Ciencias Físicas, Matemáticas y Naturales: Caracas, Venezuela, v. 64, no. 1, p. 9–20. Carver, G., Plafker, G., Metz, M., Cluff, L., Slemmons, B., Johnson, E., Roddick, J., and Sorensen, S., 2004, Surface rupture on the Denali fault interpreted from tree damage during the 1912 Delta River Mw 7.2–7.4 earthquake: Implications for the 2002 Denali fault earthquake slip distribution: Bulletin of the Seismological Society of America, v. 94, no. 6B, p. S58–S71, doi:10.1785/0120040625. Castilla, R., and Audemard, F.A., 2002, Licuación de sedimentos: Relaciones empíricas y su aplicación a la paleosismología, in Proceedings of III Jornadas Venezolanas de Sismología Histórica: Caracas, Venezuela, Serie Técnica FUNVISIS No. 1-2002, 4 p. (CD-ROM). Castilla, R., and Audemard, F.A., 2007, Sand blows as tools for magnitude estimation of pre-instrumental earthquakes: Journal of Seismology, v. 11, no. 4, p. 473–487, doi:10.1007/s10950-007-9065-z. Cello, G., and Tondi, E., eds., 2000, The resolution of geological analysis and models for earthquake faulting studies: Journal of Geodynamics, v. 29, no. 3–5, p. 151–564. Chagué-Goff, C., and Goff, J., 1999, Geochemical and sedimentological signature of catastrophic saltwater inundations (tsunami) New Zealand: Quaternary Australia, v. 17, p. 38–48. Chappell, J., 1974, Geology of coral terraces, Huon Peninsula, New Guinea: A study of Quaternary tectonic movements and sea level changes: Geological Society of America Bulletin, v. 85, p. 553–570, doi:10.1130/0016 -7606(1974)85<553:GOCTHP>2.0.CO;2. Chappell, J., Omura, A., Esat, T., McCulloch, M., Pandolfi, J., Ota, Y., and Pillans, B., 1996a, Reconciliation of late Quaternary sea levels derived from coral terraces at Huon Peninsula with deep sea oxygen isotope records: Earth and Planetary Science Letters, v. 141, p. 227–236, doi:10.1016/0012-821X(96)00062-3. Chappell, J., Ota, Y., and Berryman, K.R., 1996b, Late Quaternary coseismic uplift history of Huon Peninsula, Papua New Guinea: Quaternary Science Reviews, v. 15, p. 7–22, doi:10.1016/0277-3791(95)00062-3. Chapron, E., Van Rensbergen, P., Beck, C., De Batist, M., and Paillet, A., 1996, Lacustrine sedimentary records of brutal events in Lake Le Bourget (northwestern Alps–southern Jura): Quaternaire, v. 7, p. 155–168, doi:10 .3406/quate.1996.2068. Chen, W.-S., Lee, K.-J., Lee, L.-S., Ponti, D., Prentice, C., Chen, Y.-G., Chang, H.-C., and Lee, Y.-H., 2004, Paleoseismology of the Chelungpu fault during the past 1900 years: Quaternary International, v. 115, p. 167–176, doi:10.1016/S1040-6182(03)00105-8. Chow, J., Angelier, J., Hua, J., Lee, J., and Sun, R., 2001, Paleoseismic event and active faulting: From ground penetrating radar and high resolution seismic reflection profiles across the Chihshang fault, eastern Taiwan: Tectonophysics, v. 333, p. 241–259, doi:10.1016/S0040-1951(00)00277-8. Cita, M.B., and Rimoldi, B., 1997, Geological and geophysical evidence for a Holocene tsunami deposit in the eastern Mediterranean deep-sea record: Journal of Geodynamics, v. 24, p. 293–304. Cita, M.B., Camerlenghi, A., and Rimoldi, B., 1996, Deep-sea tsunami deposits in the eastern Mediterranean: New evidence and depositional models: Sedimentary Geology, v. 104, p. 155–173, doi:10.1016/0037-0738 (95)00126-3. Clague, J., and Bobrowsky, P., 1994a, Evidence for a large earthquake and tsunami 100–400 years ago on western Vancouver Island, British Columbia: Quaternary Research, v. 41, p. 176–184, doi:10.1006/qres.1994.1019. Clague, J., and Bobrowsky, P., 1994b, Tsunami deposits beneath tidal marshes on Vancouver Island, British Columbia: Geological Society of America Bulletin, v. 106, p. 1293–1303, doi:10.1130/0016-7606(1994)106<1293 :TDBTMO>2.3.CO;2. Clague, J., and Evans, S., 2000, A review of catastrophic drainage of morainedammed lakes in British Columbia: Quaternary Science Reviews, v. 19, p. 1763–1783, doi:10.1016/S0277-3791(00)00090-1. Clague, J., Naesgaard, E., and Sy, A., 1992, Liquefaction features on the Fraser delta: Evidence for prehistoric earthquakes?: Canadian Journal of Earth Sciences, v. 29, p. 1734–1745. Clague, J., Bobrowsky, P., and Hamilton, T.S., 1994, A sand sheet deposited by the 1964 Alaska tsunami at Port Alberni, British Columbia: Estuarine, Coastal and Shelf Science, v. 38, p. 413–421, doi:10.1006/ecss.1994.1028. Clark, D., Cupper, M., Sandiford, M., and Kiernan, K., 2011, this volume, Style and timing of late Quaternary faulting on the Lake Edgar fault,
16
Audemard M. and Michetti
southwest Tasmania, Australia: Implications for hazard assessment in intracratonic areas, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, doi:10.1130/2011.2479(05). Cluff, L.S., 1971, Peru earthquake of May 31, 1970: Engineering geology observations: Seismological Society of America Bulletin, v. 61, no. 3, p. 511–521. Colella, A., Lapenna, V., and Rizzo, E., 2004, High-resolution imaging of the High Agri Valley basin (southern Italy) with electrical resistivity tomography: Tectonophysics, v. 386, p. 29–40, doi:10.1016/j.tecto.2004.03.017. Colman, S.M., and Watson, K., 1983, Ages estimated from diffusion equation model for scarp degradation: Science, v. 221, p. 263–265, doi:10.1126/ science.221.4607.263. Comerci, V., ed., 2001, Seismically induced ground ruptures and large scale mass movements: APAT-INQUA (Agenzia per la Protezione dell’Ambiente e per i servizi Tecnici–International Union for Quaternary Research) Subcommission on “Paleoseismology”: Working Group on “Mountain Building” Field Excursion and Meeting, 21–27 September 2001. Costa, C., Rockwell, T., Paredes, J., and Gardini, C., 1999, Quaternary deformations and seismic hazard at the Andean orogenic front (31°–33°, Argentina): A paleoseismological perspective, in Proceedings 4th International Symposium on Andean Geodynamics: Paris, p. 187–191. Costa, J.E., and Schuster, R.L., 1988, The formation and failure of natural dams: Geological Society of America Bulletin, v. 100, p. 1054–1068, doi:10 .1130/0016-7606(1988)100<1054:TFAFON>2.3.CO;2. Cox, R.T., Larsen, D., Forman, S., Woods, J., Morat, J., and Galluzzi, J., 2004, Preliminary assessment of sand blows in the southern Mississippi Embayment: Bulletin of the Seismological Society of America, v. 94, no. 3, p. 1125–1142, doi:10.1785/0120030130. Crone, A.J., and Luza, K., 1990, Style and timing of Holocene surface faulting on the Meers fault, southwestern Oklahoma: Geological Society of America Bulletin, v. 102, p. 1–17. Crone, A.J., and Omdahl, E.M., eds., 1987, Proceedings of Conference XXXIX, Directions in Paleoseismology: U.S. Geological Survey Open-File Report 87-673, 456 p. Crozier, M., 1992, Determination of paleoseismicity from landslides, in Bell, D.H., ed., Landslides (Glissements de Terrain): Proceedings of the Sixth International Symposium (Christchurch, New Zealand): Rotterdam, Balkema, v. 2, p. 1173–1180. Cundy, A.B., Kortekaas, S., Dewez, T., Stewart, I.S., Collins, P.E.F., Croudace, I.W., Maroukian, H., Papanastassiou, D., Gaki-Papanastassiou, P., Pavlopoulos, K., and Dawson, A., 2000, Coastal wetlands as recorders of earthquake subsidence in the Aegean: A case study of the 1894 Gulf of Atalanti earthquakes, central Greece: Marine Geology, v. 170, p. 3–26, doi:10.1016/S0025-3227(00)00062-1. Cushing, M., Lemeille, F., Cotton, F., Grellet, B., Audru, J.-C., and Renardy, F., 2000, Paleo-earthquakes investigations in the Upper Rhine graben in the framework of the PALEOSIS project, in Workshop Proceedings, “HAN2000—Evaluation of the Potential for Large Earthquakes in Regions of Present-Day Low Seismic Activity in Europe”: Han-surLesse, Belgium, p. 39–43. Dawson, A.G., Long, D., and Smith, D.E., 1988, The Storegga slides: Evidence from eastern Scotland for a possible tsunami: Marine Geology, v. 82, p. 271–276, doi:10.1016/0025-3227(88)90146-6. Delaby, S., 2001, Palaeoseismic investigations in Belgian caves: Netherlands Journal of Geosciences, v. 80, no. 3–4, p. 323–332. Demanet, D., Renardy, F., Vanneste, K., Jongmas, D., Camelbeek, T., and Megharaoui, M., 2001, The use of geophysical prospecting for imaging active faults in the Roer graben, Belgium: Geophysics, v. 66, no. 1, p. 78–89, doi:10.1190/1.1444925. Doig, R., 1991, Effects of strong seismic shaking in lake sediments, and earthquake recurrence interval, Temiscaming, Québec: Canadian Journal of Earth Sciences, v. 28, p. 1349–1352. Doig, R., 1998, Paleoseismological evidence from lake sediments for recent movement on the Denali and other faults, Yukon Territory, Canada: Tectonophysics, v. 296, no. 3, p. 363–370, doi:10.1016/S0040-1951(98 )00152-8. Doser, D.I., 2004, Seismicity of the Denali-Totschunda fault zone in central Alaska (1912–1988) and its relation to the 2002 Denali fault earthquake sequence: Bulletin of the Seismological Society of America, v. 94, p. S132–S144.
Dost, B., and Evers, L., 2000, Site selection and interpretation of trenching results along the Peel boundary fault, in Workshop Proceedings, “HAN2000—Evaluation of the Potential for Large Earthquakes in Regions of Present-Day Low Seismic Activity in Europe”: Han-surLesse, Belgium, p. 53–56. El-Isa, Z.H., and Mustafa, H., 1986, Earthquake deformations in the Lisan deposits and seismotectonic implications: Royal Astronomical Society Geophysical Journal, v. 86, p. 413–424. Fanetti, D., Anselmetti, F.D., Chapron, E., Sturm, M., and Vezzoli, L., 2008, Megaturbidite deposits in the Holocene basin fill of Lake Como (southern Alps, Italy): Palaeogeography, Palaeoclimatology, Palaeoecology, v. 259, p. 323–340, doi:10.1016/j.palaeo.2007.10.014. Fazzito, S., Rapalini, A., Cortés, J.M., and Terrizzano, C., 2009, Characterization of Quaternary faults by electric resistivity tomography in the Andean Precordillera of western Argentina: Journal of South American Earth Sciences, v. 28, p. 217–228, doi:10.1016/j.jsames.2009.06.001. Fenton, C., 1994, Postglacial Faulting in Eastern Canada: An Annotated Bibliography: Geological Survey of Canada Open-File 2774, 98 p. Ferrer, C., 1999, Represamientos y rotura de embalses naturales (lagunas de obturación) como efectos cosísmicos: Algunos ejemplos en los Andes venezolanos: Revista Geográfica Venezolana, v. 40, p. 119–131. Field, M.E., and Jennings, A.E., 1987, Seafloor gas seeps triggered by a northern California earthquake: Marine Geology, v. 77, p. 39–51, doi:10.1016/ 0025-3227(87)90082-X. Forti, P., 2001, Seismotectonic and paleoseismic studies from speleothems: The state of the art: Netherlands Journal of Geosciences, v. 80, no. 3–4, p. 175–185. Francis, T.J., 1971, Effects of earthquakes on deep-sea sediments: Nature, v. 233, p. 98–102, doi:10.1038/233098a0. Galli, P., 2000, New empirical relationships between magnitude and distance for liquefaction: Tectonophysics, v. 324, p. 169–187, doi:10.1016/S0040 -1951(00)00118-9. Gilli, E., 1996, Effets des séismes dans l’endokarst. Application aux études de sismicité historique: Revue d’Analyse Spatiale Quantitative et Appliquée, v. 38–39, p. 121–132. Gilli, E., 1999, Evidence of palaeoseismicity in a flowstone of the Observatoire cave (Monaco): Geodinamica Acta, v. 12, no. 3–4, p. 159–168, doi:10 .1016/S0985-3111(00)88655-6. Gilli, E., 2004, Glacial causes of damage and difficulties to use speleothems as palaeoseismic indicators: Geodinamica Acta, v. 17, no. 3, p. 229–240, doi:10.3166/ga.17.229-240. Gilli, E., and Audra, P., eds., 2001, Tectonique Active et Géomorphologie: Revue d’Analyse Spatiale Quantitative et Appliquée, Special Publication 2001. Gilli, E., and Delange, P., 2001, Utilisation des spéleothemes comme indicateurs de néotectonique ou de la paléosismicité, in Gilli, E., and Audra, P., eds., Tectonique Active et Géomorphologie: Revue d’Analyse Spatiale Quantitative et Appliquée, Special Publication 2001, p. 79–90. Gluyas, J.G., and Swarbrick, R.E., 2004, Petroleum Geoscience: Malden, Massachusetts, Blackwell Publishing, 359 p. González, J., Schmitz, M., Audemard, F.A., Contreras, R., Mocquet, A., Delgado, J., and De Santis, F., 2004, Site effects of the 1997 Cariaco, Venezuela, earthquake: Engineering Geology, v. 72, p. 143–177, doi:10.1016/j .enggeo.2003.07.002. Gorsline, D.S., De Diego, T., and Nava-Sanchez, E.H., 2000, Seismically triggered turbidites in small margin basins: Alfonso Basin, western Gulf of California and Santa Monica Basin, California Borderland: Sedimentary Geology, v. 135, p. 21–35, doi:10.1016/S0037-0738(00)00060-9. Grant, D., 1990, Late Quaternary movement of Aspy fault, Nova Scotia: Canadian Journal of Earth Sciences, v. 27, p. 984–987. Gross, R., Holliger, K., Green, A., and Begg, J., 2000, 3D ground penetrating radar applied to paleoseismology: Examples from the Willington fault, New Zealand, in Noon, D., Stickley, G., and Longstaff, D., eds., Proceedings of the 8th International Conference on Ground Penetrating Radar (GPR 2000), Gold Coast, Australia, 23–26 May 2000, p. 478–481. Guccione, M.J., 2005, Late Pleistocene and Holocene paleoseismology of an intraplate seismic zone in a large alluvial valley, the new Madrid seismic zone, central USA: Tectonophysics, v. 408, p. 237–264. Gürpinar, A., 1989, Importance of historical earthquakes in research in the siting of nuclear facilities, in Workshop, “Historical Earthquakes in the Ibero-Maghrebian Region. Methodological Approach and Case Studies,” 15–17 November 1989: Lisbon, p. 3–7.
Geological criteria for evaluating seismicity revisited Gutiérrez, F., Acosta, E., Ríos, S., Guerrero, J., and Lucha, P., 2005, Geomorphology and geochronology of sackung features (uphill-facing scarps) in the central Spanish Pyrenees: Geomorphology, v. 69, p. 298–314. Gutiérrez, F., Ortuño, M., Lucha, P., Guerrero, J., Acosta, E., Coratza, P., Piacentini, D., and Soldati, M., 2008, Late Quaternary episodic displacement on a sackung scarp in the central Spanish Pyrenees. Secondary paleoseismic evidence?: Geodinamica Acta, v. 21, no. 4, p. 187–202, doi:10.3166/ ga.21.187-202. Gutiérrez, F., Lucha, P., and Galve, J., 2010, Reconstructing the geochronological evolution of large landslides by means of the trenching technique in the Yesa Reservoir (Spanish Pyrenees): Geomorphology, v. 124, p. 124– 136, doi:10.1016/j.geomorph.2010.04.015. Hamilton, S., Shennan, I., Combellick, R., Mulholland, J., and Noble, C., 2005, Evidence for two great earthquakes at Anchorage, Alaska, and implications for multiple great earthquakes through the Holocene: Quaternary Science Reviews, v. 24, p. 2050–2068, doi:10.1016/j.quascirev .2004.07.027. Hampton, M.A., Lee, H.J., and Locat, J., 1996, Submarine landslides: Reviews of Geophysics, v. 34, p. 33–59, doi:10.1029/95RG03287. HAN2000, 2000, Evaluation of the Potential for Large Earthquakes in Regions of Present-Day Low Seismic Activity in Europe: Han-sur-Lesse, Belgium. Hanks, T., 2000, The age of scarplike landforms from diffusion-equation analysis, in Stratton Noller, J., Sowers, J.M., and Lettis, W.R., eds., Quaternary Geochronology: Methods and Applications: Washington, D.C., American Geophysical Union, p. 313–338. Hansen, W.R., 1965, Effects of the Earthquake of March 27, 1964, at Anchorage, Alaska: U.S. Geological Survey Professional Paper 542-A, 68 p. Heezen, B.C., and Ewing, M., 1952, Turbidity currents and submarine slumps, and the 1929 Grand Banks earthquake: American Journal of Science, v. 250, p. 849–873, doi:10.2475/ajs.250.12.849. Hempton, M.R., and Dewey, J.F., 1983, Earthquake-induced deformational structures in young lacustrine sediments, East Anatolian fault, southwest Turkey: Tectonophysics, v. 98, p. 7–14, doi:10.1016/0040-1951(83 )90294-9. Hermanns, R.L., 2011, this volume, Late Pleistocene–early Holocene paleoseismicity deduced from lake sediment deformation and coeval landsliding in the Calchaquíes valleys, NW Argentina, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, doi:10.1130/2011.2479(08). Hippolyte, J.-C., Brocard, G., Tardy, M., Nicoud, G., Bourlès, D., Braucher, R., Ménard, G., and Souffaché, B., 2006, The recent fault scarps of the Western Alps (France): Tectonic surface ruptures or gravitational sackung scarps? A combined mapping, geomorphic, levelling, and 10Be dating approach: Tectonophysics, v. 418, p. 255–276, doi:10.1016/j.tecto .2006.02.009. Hovland, M., and Judd, A.G., 1988, Seabed Pockmarks and Seepage: London, Graham and Trotman, 293 p. Hovland, M., Gardner, J.V., and Judd, A.G., 2002, The significance of pockmarks to understanding fluid flow processes and geohazards: Geofluids, v. 2, p. 127–136, doi:10.1046/j.1468-8123.2002.00028.x. Hupp, C., Osterkamp, W., and Thornton, J., 1987, Dendrogeomorphic Evidence and Dating of Recent Debris Flows on Mount Shasta, Northern California: U.S. Geological Survey Professional Paper 1396-B, p. 1–39. Inouchi, Y., Kinugasa, Y., Kumon, F., Nakano, S., Yasumatsu, S., and Shiki, T., 1996, Turbidites as records of intense paleoearthquakes in Lake Biwa, Japan: Sedimentary Geology, v. 104, p. 117–125, doi:10.1016/0037-0738 (95)00124-7. Jacoby, G.C., Bunker, D.E., and Benson, B.E., 1997, Tree-ring evidence for an A.D. 1700 Cascadia earthquake in Washington and northern Oregon: Geology, v. 25, no. 11, p. 999–1002, doi:10.1130/0091-7613 (1997)025<0999:TREFAA>2.3.CO;2. Jol, H., Junck, M., and Kaminsky, G., 2000, High resolution ground penetrating radar imaging (225–900 MHz) of geomorphic and geologic settings: Examples from Utah, Washington and Wisconsin, in Noon, D., Stickley, G., and Longstaff, D., eds., Proceedings of the 8th International Conference on Ground Penetrating Radar (GPR 2000), Gold Coast, Australia, 23–26 May 2000, p. 69–74. Jones, B., and Hunter, I.G., 1992, Very large boulders on the coast of Grand Cayman; the effects of giant waves on rocky shorelines: Journal of Coastal Research, v. 8, p. 763–774.
17
Jongmans, D., Teerlynck, H., and Renardy, F., 2000, The use of geophysical prospecting for paleoseismology in the Rhine graben: Potentiality and perspectives: Workshop Proceedings, “HAN2000—Evaluation of the Potential for Large Earthquakes in Regions of Present-Day Low Seismic Activity in Europe”: Han-sur-Lesse, Belgium, p. 83–88. Kagan, E., Agnon, A., Bar-Matthews, M., and Ayalon, A., 2005, Dating large infrequent earthquakes by damaged cave deposits: Geology, v. 33, no. 4, p. 261–264, doi:10.1130/G21193.1. Kastens, K., and Cita, M.B., 1981, Tsunami-induced sediment transport in the abyssal Mediterranean Sea: Geological Society of America Bulletin, v. 92, p. 845–857, doi:10.1130/0016-7606(1981)92<845:TSTITA>2.0.CO;2. Keefer, D.K., 1984, Landslides caused by earthquakes: Geological Society of America Bulletin, v. 95, p. 406–421, doi:10.1130/0016-7606(1984)95 <406:LCBE>2.0.CO;2. Keefer, D.K., 2000, Statistical analysis of an earthquake-induced landslide distribution—The 1989 Loma Prieta, California, event: Engineering Geology, v. 58, p. 231–249, doi:10.1016/S0013-7952(00)00037-5. Keefer, D.K., and Manson, M.W., 1989, Regional distribution and characteristics of landslides generated by the earthquake, in Keefer, D.K., ed., The Loma Prieta, California, Earthquake of October 17, 1989, Landslides: U.S. Geological Survey Professional Paper 1551-C, p. C7–C32. Keller, E., and Pinter, N., 1998, Active Tectonics: Earthquakes, Uplift, and Landscape: Upper Saddle River, New Jersey, Prentice Hall, 337 p. Knutson, D.W., Buddemeier, R.W., and Smith, S.V., 1972, Coral chronometers: Seasonal growth bands in reef corals: Science, v. 177, p. 270–272, doi:10.1126/science.177.4045.270. Kuhn, G., 2005, Paleoseismic features as indicators of earthquake hazards in north coastal San Diego County, California, USA: Engineering Geology, v. 80, p. 115–150, doi:10.1016/j.enggeo.2005.04.006. Kürçer, A., Chatzipetros, A., Tutkun, S., Pavlides, S., Ateş, Ö., and Valkaniotis, S., 2008, The Yenice–Gönen active fault (NW Turkey): Active tectonics and palaeoseismology: Tectonophysics, v. 453, p. 263–275, doi:10.1016/ j.tecto.2007.07.010. Kuribayashi, E., and Tatsuoka, F., 1975, Brief review of liquefaction during earthquakes in Japan: Soil and Foundation, v. 15, p. 81–92. Lacave, C., Koller, M.G., and Egozcue, J.J., 2004, What can be concluded about seismic history from broken and unbroken speleothems?: Journal of Earthquake Engineering, v. 8, no. 3, p. 431–455, doi:10.1142/ S1363246904001493. Lagerbäck, R., 1979, Neotectonic structures in northern Sweden: Geologiska Föreningens i Stockholm Förhandlingar, v. 100, p. 263–269. Lalinde P., C.P., Toro, G.E., Velásquez, A., and Audemard M., F.A., 2011, this volume, Large-magnitude late Holocene seismic activity in the PereiraArmenia region, Colombia, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, doi:10.1130/2011.2479(03). Lehmann, K., Klostermann, J., Pelzing, R., and Hinzen, K.-G., 2000, Paleoseismological investigations at the Rurrand fault, FRG, in Workshop Proceedings, “HAN2000—Evaluation of the Potential for Large Earthquakes in Regions of Present-Day Low Seismic Activity in Europe”: Han-surLesse, Belgium, p. 93–96. Lemeille, F., Cushing, M., Carbon, D., Grellet, B., Bitterli, T., Flehoc, C., and Innocent, C., 1999, Co-seismic ruptures and deformations recorded by speleothems in the epicentral zone of the Basel earthquake: Geodinamica Acta, v. 12, no. 3–4, p. 179–191, doi:10.1016/S0985-3111(00)88657-X. Lensen, G., 1968, Analysis of progressive fault displacement during downcutting at the Branch River terraces, South Island, New Zealand: Geological Society of America Bulletin, v. 79, no. 5, p. 545–556, doi:10.1130/0016 -7606(1968)79[545:AOPFDD]2.0.CO;2. Lensen, G.J., and Vella, P., 1971, The Waiohine River faulted terrace sequence: Bulletin of the Royal Society of New Zealand, v. 9, p. 117–119. Levorsen, A.I., 1967, Geology of Petroleum: San Francisco, W.H. Freeman and Co., 724 p. Levret, A., 2002. Histoire et aléa sismique, in Actes des Vème Rencontres du Groupe APS (Association pour l’identification et l’étude des Pathologies d’origine sismique dans le bâti ancien), 6–7 April 2000: Perpignan, France, p. 107–118. Li, T., Schuster, R.L., and Wu Jishan, 1986, Landslide dams in southwestern China, in Schuster, R.L., ed., Processes, Risk, and Mitigation: American Society of Civil Engineers Special Publication 3, p. 146–162.
18
Audemard M. and Michetti
Lignier, V., and Desmet, M., 2002, Les archives sédimentaires quaternaires de la grotte sous les Sangles (Bas-Bugey, Jura méridional, France): Indices paléo-climatiques et sismotectoniques: Karstologia, v. 39, no. 1, p. 27–46. Lin, A., and Lin, S.J., 1998, Tree damage and surface displacement: The 1931 M 8.0 Fuyun earthquake: The Journal of Geology, v. 106, p. 751–758, doi:10.1086/516058. Lindvall, S.C., Rockwell, T.K., Dawson, T.E., Helms, J.G., and Weaver Bowman, K., 2002, Evidence for two surface ruptures in the past 500 years on the San Andreas fault at Frazier Mountain, California: Bulletin of the Seismological Society of America, v. 92, no. 7, p. 2689–2703, doi:10.1785/0120000610. Liu, L., and Li, Y., 2001, Identification of liquefaction and deformation features using ground penetrating radar in the New Madrid seismic zone, USA: Journal of Applied Geophysics, v. 47, p. 199–215, doi:10.1016/S0926 -9851(01)00065-9. Livio, F., Berlusconi, A., Michetti, A.M., Sileo, G., Zerboni, A., Trombino, L., Cremaschi, M., Mueller, K., Vittori, E., Carcano, C., and Rogledi, S., 2009, Active fault-related folding in the epicentral area of the December 25, 1222 (Io = IX MCS), Brescia earthquake (northern Italy): Seismotectonic implications: Tectonophysics, v. 476, no. 1–2, p. 320–335, doi:10.1016/j.tecto.2009.03.019. López C., M.C., 2006, Análisis de Deformación Tectónica en el Piedemonte de las Cordilleras Central y Occidental, Valle del Cauca, Colombia: Contribuciones Paleosísmicas [Master’s thesis]: Medellín, Colombia, Universidad EAFIT, 131 p. + appendices. López C., M.C., and Audemard M., F.A., 2011, this volume, Evidence of Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, doi:10.1130/2011.2479(04). Machette, M., Personius, S.F., and Nelson, A., 1987, Quaternary geology along the Wasatch fault zone: Segmentation, recent investigations and preliminary conclusions, in Gori, P.L., and Hays, W.W., eds., Evaluation of Urban and Regional Earthquake Hazard and Risk in Utah: U.S. Geological Survey Open-File Report 97-585, p. A1–A72. Maltman, A., ed., 1994, The Geological Deformation of Sediments: London, Chapman and Hall, 362 p. Mandel, R., 2001, When Low Technology Is the Answer: Response to a Remote Oil Seep: Report International Oil Spill Conference 2001-306, 22 p., available at http://www.epa.gov/oem/docs/oil/fss/fss04/mandel_04.pdf. Marco, S., Stein, M., and Agnon, A., 1996, Long-term earthquake clustering: A 50,000-year paleoseismic record in the Dead Sea graben: Journal of Geophysical Research, v. 101, no. B3, p. 6179–6191, doi:10.1029/95JB01587. Maurya, D., Patidar, A., Mulchandani, N., Goyal, B., Thakkar, M., Bhandari, S., Vaid, S., Bhatt N., and Chamyal, L., 2005, Need for initiating ground penetrating radar studies along active faults in India: An example from Kachchh: Current Science, v. 88, no. 2, p. 231–240. McCalpin, J.P., 1999, Episodic earthquake-induced movement on the Stillwater scarp “sacking,” central Nevada: Geological Society of America Abstracts with Programs, v. 31, no. 7, p. 474. McCalpin, J.P., ed., 1996, Paleoseismology: San Diego, Academic Press, 588 p. McCalpin, J.P., ed., 2009, Paleoseismology (2nd ed.): San Diego, Academic Press, 647 p. McCalpin, J.P., and Hart, E., 2002, Ridge-Top Spreading Features and Relationship to Earthquakes, San Gabriel Mountain Region, Southern California: Parts A and B: U.S. Geological Survey National Earthquake Hazard Reduction Program Technical Contract 99HQGR0042, 44 p. McCalpin, J.P., Zuchiewicz, W., and Jones, A., 1993, Sedimentology of scarpderived colluvium from the 1983 Borah Peak rupture, central Idaho: Journal of Sedimentary Petrology, v. 63, no. 1, p. 120–130. McCalpin, J.P., Harrison, J.B.J., Berger, G.W., and Tobin, H.C., 2011, this volume, Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA: The Calabacillas fault, Albuquerque, New Mexico, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, doi:10.1130/2011.2479(01). Meghraoui, M., Camelbeeck, T., Vanneste, K., Borndeel, M., and Jongmans, D., 2000, Active faulting and paleoseismology along the Bree fault, lower Rhine graben, Belgium: Journal of Geophysical Research, v. 105, no. B6, p. 13,809–13,841, doi:10.1029/1999JB900236.
Meisling, K.E., and Sieh, K.E., 1980, Disturbance of trees by the 1857 Fort Tejon earthquake, California: Journal of Geophysical Research, v. 85, p. 3225–3238, doi:10.1029/JB085iB06p03225. Michetti A., ed., 1998, Proceedings of the Resolution of Geological Analysis and Models for Earthquake Faulting Studies, 3–6 June, Rome and Camerino, Italy. Michetti, A.M., and Hancock, P.L., 1997, Paleoseismology: Understanding past earthquakes using Quaternary geology: Journal of Geodynamics, v. 24, no. 1–4, p. 3–10, doi:10.1016/S0264-3707(97)00004-5. Michetti, A.M., Brunamonte, F., Serva, L., and Vittori, E., 1996, Trench investigations of the 1915 Fucino earthquake fault scarps (Abruzzo, central Italy): Geological evidence of large historical events: Journal of Geophysical Research, v. 101, no. B3, p. 5921–5936, doi:10.1029/95JB02852. Michetti, A.M., Audemard, F.A., and Marco, S., eds., 2005a, Paleoseismology: Integrated Study of the Quaternary Geological Record for Earthquake Deformation and Faulting: Tectonophysics, Special Issue, v. 408, no. 1–4, 307 p. Michetti, A., Audemard, F.A., and Marco, S., 2005b, Future trends in paleoseismology: Integrated study of the seismic landscape as a vital tool in seismic hazard analyses, in Michetti, A.M., Audemard, F., and Marco, S., eds., Paleoseismology: Integrated Study of the Quaternary Geological Record for Earthquake Deformation and Faulting: Tectonophysics, Special Issue, v. 408, no. 1–4, p. 3–21. Michetti, A.M., Esposito, E., Guerrieri, L., Porfido, S., Serva, L., Tatevossian, R., Vittori, E., Audemard, F.A., Azuma, T., Clague, J., Comerci, V., Gürpinar, A., McCalpin, J., Mohammadioun, B., Mörner, N.A., Ota, Y., and Roghozin, E., 2007, Environmental Seismic Intensity Scale 2007 “ESI 2007”: Memorie Descrittive della Carta Geologica d’Italia, v. 74, p. 7–54 (http://www.apat.gov.it/site/en-GB/Projects/INQUA_Scale/). Mitchell, S., Matmon, A., Bierman, P., Enzel, Y., Caffee, M., and Rizzo, D., 2001, Displacement history of a limestone normal fault scarp, northern Israel, from cosmogenic 36Cl: Journal of Geophysical Research, v. 106, no. B3, p. 4247–4264, doi:10.1029/2000JB900373. Mörner, N.-A., 1996, Liquefaction and varve deformation as evidence of palæoseismic events and tsunamis: The autumn 10,430 BP case in Sweden: Quaternary Science Reviews, v. 15, p. 939–948, doi:10.1016/S0277-3791 (96)00057-1. Mörner, N.-A., 2005, An interpretation and catalogue of paleoseismicity in Sweden: Tectonophysics, v. 408, no. 1–4, p. 265–307, doi:10.1016/j.tecto .2005.05.039. Mörner, N.-A., Tröften, P.E., Sjöberg, R., Grant, D., Dawson, S., Bronge, C., Kvamsdal, O., and Sidén, A., 2000, Deglacial paleoseismicity in Sweden: The 9663 BP Iggesund event: Quaternary Science Reviews, v. 19, p. 1461–1468, doi:10.1016/S0277-3791(00)00095-0. Mörner, N.-A., Audemard, F.A., Bronge, Ch., Dawson, S., Grant, D., Kvamsdal, O., Nikonov, A., Sidén, A., Sjöberg, R., Strandth, L., Sun, G., Tröften, P.-E., Wigren, H., and Zykov, D., 2003, The Hudisksvall area: The Boda Cave and its surroundings, the 9663 BP paleoseismic event, in Mörner, N-A., ed., Paleoseismicity of Sweden: A Novel Paradigm: Stockholm, Sweden, JOFO Grafiska AB, p. 29–224. Mosher, D.C., Moran, K., and Hiscott, R.N., 1994, Late Quaternary sediment, sediment mass-flow processes and slope stability on the Scotian slope, Canada: Sedimentology, v. 41, p. 1039–1061, doi:10.1111/j.1365-3091.1994 .tb01439.x. Munson, P., Munson, C.A., and Pond, E.C., 1995, Paleoliquefaction evidence for a strong Holocene earthquake in south-central Indiana: Geology, v. 23, no. 4, p. 325–328, doi:10.1130/0091-7613(1995)023<0325:PEFASH>2 .3.CO;2. Nakajima, T., and Kanai, Y., 2000, Sedimentary features of seismoturbidites triggered by the 1983 and older historical earthquakes in the eastern margin of the Japan Sea: Sedimentary Geology, v. 135, p. 1–19, doi:10.1016/ S0037-0738(00)00059-2. Nash, D., 1980, Morphological dating of degraded normal fault scarp: The Journal of Geology, v. 88, p. 353–360, doi:10.1086/628513. Natawidjaja, D.H., Sieh, K., Galetzka, J., Suwargadi, B.W., Cheng, H., Edwards, R.L., and Chlieh, M., 2007, Interseismic deformation above the Sunda megathrust recorded in coral microatolls of the Mentawai Islands, West Sumatra: Journal of Geophysical Research, v. 112, B02404, doi:10 .1029/2006JB004450. Nelson, A.R., Shennan, I., and Long, A.J., 1996, Identifying coseismic subsidence in tidal-wetland stratigraphical sequences at the Cascadia subduction zone of western North America: Journal of Geophysical Research, v. 101, no. B3, p. 6115–6135, doi:10.1029/95JB01051.
Geological criteria for evaluating seismicity revisited Nicol, A., and Nathan, S., 2001, Folding and the formation of bedding-parallel faults on the western limb of Grey Valley syncline near Blackball: New Zealand Journal of Geology and Geophysics, v. 44, no. 1, p. 127–135, doi: 10.1080/00288306.2001.9514929. Nolan, J., and Weber, G., 1992, Evaluation of surface cracking caused by the 1989 Loma Prieta earthquake, Santa Cruz County, California: Case histories, in Sharma, S., ed., Proceedings of the 28th Symposium on Engineering Geology and Geotechnical Engineering: Pocatello, Idaho, University of Idaho, p. 272–286. Nolan, J., and Weber, G., 1998, Evaluation of coseismic ground cracking accompanying the earthquake: Trenching studies and case histories, in Keefer, D.K., ed., The Loma Prieta, California, Earthquake of October 17, 1989, Landslides: U.S. Geological Survey Professional Paper 1551-C, p. 145–163. Obermeier, S.F., 1996, Use of liquefaction-induced features for paleoseismic analysis—An overview of how liquefaction features can be distinguished from other features and how their distribution and properties of source sediment can be used to infer the location and strength of Holocene paleoearthquakes: Engineering Geology, v. 44, p. 1–76, doi:10.1016/S0013-7952 (96)00040-3. Okumura, K., Shimokawa, K., Yamazaki, H., and Tsukuda E., 1994, Recent surface faulting events along the middle section of the Itoigawa-Shizuoka tectonic line—Trenching survey of the Gofukuji fault near Matsumoto, central Japan: Journal of the Seismological Society of Japan (Zisin), 2nd Series, v. 46, p. 425–438. Okumura, K., Takada, K., and Goto, H., eds., 2000, Symposium and School on Active Faulting: The Kobe Earthquake Five Years Later: “The Frontier of Active Fault Research,” 18–26 January 2000: Asa-Kita-Ku, Hiroshima, Japan, Letter Press Co. Ltd., 606 p. Okumura, K., Kondo, H., and Toda, S., 2005, International Symposium on Active Faulting, 17–24 January 2005: Tsukuba, Ibaraki, Japan, Isebu Co. Ltd., 185 p. Olig, S.S., Eppes, M.C., Forman, S.L., Love, D.W., and Allen, B.D., 2011, this volume, Late Quaternary earthquakes on the Hubbell Spring fault system, New Mexico, USA: Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, doi:10.1130/2011.2479(02). Ollarves, R., Audemard, F.A., Sánchez, C., Camacho, P., Cataldi, A., and Zambrano, F., 2004. Uso del GPR como herramienta para la búsqueda de fallas activas: Calibración con un afloramiento cuaternario [abs.], in Proceedings XII Congreso Venezolano de Geofísica: Caracas, Venezuela, 8 p. (CD-ROM). Onida, M., Galadini, F., and Forcella, F., 2001, Application of paleoseismological techniques to the study of late Pleistocene–Holocene deep-seated gravitational movements at the Mortirolo Pass (central Alps, Italy): Netherlands Journal of Geosciences, v. 80, no. 3–4, p. 209–227. Ota, Y., and Chappell, J., 1996, Late Quaternary coseismic uplift events on the Huon Peninsula, Papua New Guinea, deduced from coral terrace data: Journal of Geophysical Research, v. 101, no. B3, p. 6071–6082, doi:10 .1029/95JB02245. Ota, Y., Chappell, J., Kelley, R., Yonekura, N., Matsumoto, E., Nishimura, T., and Head, J., 1993, Holocene coral reef terraces and coseismic uplift of Huon Peninsula, Papua New Guinea: Quaternary Research, v. 40, p. 177– 188, doi:10.1006/qres.1993.1070. Page, R., 1970, Dating episodes of faulting from tree rings: Effects of the 1958 rupture of the Fairweather fault on tree growth: Geological Society of America Bulletin, v. 81, p. 3085–3094, doi:10.1130/0016-7606 (1970)81[3085:DEOFFT]2.0.CO;2. Palumbo, L., Benedetti, L., Bourlès, D., Cinque, A., and Finkel, R., 2004, Slip history of the Magnola fault (Apennines, central Italy) from 36Cl surface exposure dating: Evidence for strong earthquakes over the Holocene: Earth and Planetary Science Letters, v. 225, no. 1–2, p. 163–176, doi:10.1016/j.epsl.2004.06.012. Pantosti, D., Berryman, K., Yeats, R.S., and Kinugasa, Y., eds., 2003, Ten years of paleoseismology in the ILP: Progress and prospects: Annals of Geophysics, v. 46, no. 5, p. 733–1194. Papadopoulos, G.A., and Lefkopoulos, G., 1993, Magnitude-distance relations for liquefaction in soil from earthquakes: Bulletin of the Seismological Society of America, v. 83, p. 925–938.
19
Papathanassiou, G., Pavlides, S., Christaras, B., and Pitilakis, K., 2005, Liquefaction case histories and empirical relations of earthquake magnitude versus distance from the broader Aegean region: Journal of Geodynamics, v. 40, p. 257–278, doi:10.1016/j.jog.2005.07.007. Paris, R., Lavigne, F., Wassmer, P., and Sartohadi, J., 2007, Coastal sedimentation associated with the December 26, 2004, tsunami in Lhok Nga, west Banda Aceh (Sumatra, Indonesia): Marine Geology, v. 238, p. 93–106, doi:10.1016/j.margeo.2006.12.009. Pavlides, S., Pantosti, D., and Peizhen, Z., eds., 1999, Earthquakes, Paleoseismology and Active Tectonics: 29th General Assembly of the Association of Seismology and Physics of the Earth’s Interior (IASPEI), August 1997, Thessaloniki, Greece: Tectonophysics, v. 308, no. 1–2, 308 p. Pérez López, R., Rodríguez-Pascua, M.A., Giner, J.L., Martínez Díaz, J.J., Marcus Nuez, C., Silva, P.G., and Calvo, J.P., 2009, Speleoseismological study of the “Benis Cave” (Murcia, Spain): Coseismic collapses from the earthquake crisis of Mula (mb= 4.8, MKS VII, 1999), in Reicherter, K., Michetti, A., and Silva, P.G., eds., Paleoseismology: Historical and Prehistorical Records of Earthquake Ground Effects for Seismic Hazard Assessment: Geological Society of London Special Publication 316, p. 207–216. Philip, H., and Meghraoui, M., 1983, Structural analysis and interpretation of the surface deformation of the El Asnam earthquake of October 10, 1980: Tectonics, v. 2, p. 17–49, doi:10.1029/TC002i001p00017. Phillips, F.M., Ayarbe, J.P., Harrison, J.B.J., and Elmore, D., 2003, Dating rupture events on alluvial fault scarps using cosmogenic nuclides and scarp morphology: Earth and Planetary Science Letters, v. 215, no. 1–2, p. 203– 218, doi:10.1016/S0012-821X(03)00419-9. Pillans, B., and Huber, P., 1995, Interpreting coseismic deformations using Holocene coastal deposits, Wellington, New Zealand: Quaternary International, v. 26, p. 87–95, doi:10.1016/1040-6182(94)00049-B. Pinet, N., Duchesne, M., Lavoie, D., Bolduc, A., and Long, B., 2008, Surface and subsurface signatures of gas seepage in the St. Lawrence Estuary (Canada): Significance to hydrocarbon exploration: Marine and Petroleum Geology, v. 25, p. 271–288, doi:10.1016/j.marpetgeo.2007.07.011. Piper, D.J., Cochonat, P., and Morrison, M.L., 1999, The sequence of events around the epicentre of the 1929 Grand Banks earthquake: Initiation of debris flows and turbidity current inferred from sidescan sonar: Sedimentology, v. 46, p. 79–97, doi:10.1046/j.1365-3091.1999.00204.x. Plafker, G., Ericksen, G.E., and Fernandez Concha, J., 1971, Geological aspects of the May 31, 1970, Peru earthquake: Seismological Society of America Bulletin, v. 61, no. 3, p. 543–578. Pop, O., Popa, I., and Surdeanu, V., 2008, Dendrogeomorphological analysis of human-induced debris-flows in the Călimani Mountains (România), in International Association of Geomorphologists Regional Conference “IGAR” [abs.]: Brasov, Rumania (CD-ROM). Postpischl, D., Agostini, S., Forti, P., and Quinif, Y., 1991, Palaeoseismicity from karst sediments: The “Grotta del Cervo” cave case study (central Italy): Tectonophysics, v. 193, p. 33–44, doi:10.1016/0040-1951(91)90186-V. Reicherter, K., Jabaloy, A., Galindo-Zaldívar, J., Ruano, P., Becker-Heidmann, P., Morales, J., Reiss, S., and González-Lodeiro, F., 2003, Repeated palaeoseismic activity of the Ventas de Zafarraya fault (S Spain) and its relation with the 1884 Andalusian earthquake: International Journal of Earth Sciences, v. 92, p. 912–922, doi:10.1007/s00531-003-0366-3. Reicherter, K., Michetti, A.M., and Silva, P.G., eds., 2009, Palaeoseismology: Historical and Prehistorical Records of Earthquake Ground Effects for Seismic Hazard Assessment: Geological Society of London Special Publication 316, 320 p. Ricci Lucchi, F., 1995, Sedimentological indicators of paleoseismicity, in Serva, L., and Slemmons, D.B., eds., Perspectives in Paleoseismology: Association of Engineering Geologists Special Publication 6, p. 7–17. Rinaldini, A., Marino, A., and Ciucci, M., 2008, Characterization of a rock avalanche deposit for risk assessment in the town of Celano (Fucino Basin, central Italy): Advances in Geosciences, v. 14, p. 155–158, doi:10.5194/ adgeo-14-155-2008. Ringrose, P.S., 1989, Palaeoseismic (?) liquefaction event in late Quaternary lake sediments at Glen Roy, Scotland: Terra Nova, v. 1, p. 57–62, doi:10.1111/j.1365-3121.1989.tb00326.x. Ritz, J.F., Brown, E.T., Bourlès, D.L., Philip, H., Schlupp, A., Raisbeck, G.M., Yiou, F., and Enkhtuvshin, B., 1995, Slip rates along active faults estimated with cosmic-ray-exposure dates: Application to the Bogd fault, Gobi-Altaï, Mongolia: Geology, v. 23, p. 1019–1022, doi:10.1130/0091 -7613(1995)023<1019:SRAAFE>2.3.CO;2.
20
Audemard M. and Michetti
Rockwell, T., Barka, A., Dawson, T., Akyuz, S., and Thorup, K., 2001, Paleoseismology of the Gazikoy-Saros segment of the North Anatolia fault, northwestern Turkey: Comparison of the historical and paleoseismic records, implications of regional seismic hazard, and models of earthquake recurrence: Journal of Seismology, v. 5, p. 433–448, doi:10.1023/A:1011435927983. Rodríguez, C., Bommer, J., and Chandler, R., 1999, Earthquake-induced landslides: 1980–1997: Soil Dynamics and Earthquake Engineering, v. 18, p. 325–346, doi:10.1016/S0267-7261(99)00012-3. Rodríguez, L.M., Audemard, F.A., and Rodríguez, J.A., 2002, Casos históricos y contemporáneos de licuación de sedimentos inducidos por sismos en Venezuela desde 1530, in Proceedings of III Jornadas Venezolanas de Sismología Histórica: Caracas, Venezuela, Serie Técnica FUNVISIS No. 1-2002, 4 p. (CD-ROM). Rodríguez, L.M., Audemard, F.A., and Rodríguez, J.A., 2006, Casos históricos y contemporáneos de licuación de sedimentos inducidos por sismos en Venezuela desde 1530: Revista de la Facultad de Ingeniería de la Universidad Central de Venezuela, v. 21, no. 3, p. 5–32. Rodríguez-Pascua, M.A., Calvo, J.P., De Vicente, G., and Gómez-Gras, D., 2000, Soft-sediment deformation structures interpreted as seismites in lacustrine sediments of the Prebetic zone, SE Spain, and their potential use as indicators of earthquake magnitudes during the late Miocene: Sedimentary Geology, v. 135, p. 117–135, doi:10.1016/S0037-0738(00)00067-1. Rodríguez-Pascua, M.A., De Vicente, G., Calvo, J.P., and Pérez-López, R., 2003, Similarities between recent seismic activity and paleoseismites during the late Miocene in the external Betic Chain (Spain): Relationship by “b” value and the fractal dimension: Journal of Structural Geology, v. 25, p. 749–763, doi:10.1016/S0191-8141(02)00078-0. Roep, T.B., and Everts, A.J., 1992, Pillow-beds: A new type of seismite? An example from an Oligocene turbidite fan complex, Alicante, Spain: Sedimentology, v. 39, p. 711–724, doi:10.1111/j.1365-3091.1992.tb02148.x. Sarconi, M., 1784, Istoria de’ fenomeni del tremoto avvenuto nelle Calabrie, e nel Valdemone nell’anno 1783 posta in luce dalla Reale Accademia delle Scienze e delle Belle Lettere di Napoli: Atlante Iconografico: Napoli, Giuseppe Campo Impressore, 528 p. Sawai, Y., Fujii, Y., Fujiwara, O., Kamataki, T., Komatsubara, J., Okamura, Y., Satake, K., and Shishikura, M., 2008, Marine incursions of the past 1500 years and evidence of tsunamis at Suijin-numa, a coastal lake facing the Japan Trench: The Holocene, v. 18, no. 4, p. 517–528, doi:10 .1177/0959683608089206. Schafer, C.T., and Smith, J.N., 1987, Hypothesis for a submarine landslide and cohesionless sediment flows resulting from a 17th century earthquake-triggered landslide in Quebec, Canada: Geo-Marine Letters, v. 7, p. 31–37, doi:10.1007/BF02310463. Scheffers, A., 2004, Tsunami imprints on the Leeward Netherlands Antilles (Aruba, Curacao, Bonaire) and their relation to other coastal problems: Quaternary International, doi:10.1016/j.quaint.2004.01.015 Scheffers, A., and Kelletat, D., 2003, Sedimentologic and geomorphologic tsunami imprints worldwide—A review: Earth-Science Reviews, v. 63, p. 83–92, doi:10.1016/S0012-8252(03)00018-7. Scheffers, A., and Kelletat, D., 2005, Tsunami relics on the coastal landscape west of Lisbon, Portugal: Science of Tsunami Hazards, v. 23, no. 1, p. 3–17. Scheffers, A., and Scheffers, S., 2007, Tsunami deposits on the coastline of west Crete (Greece): Earth and Planetary Science Letters, v. 259, p. 613– 624, doi:10.1016/j.epsl.2007.05.041. Schnellmann, M., Anselmetti, F., Giardini, D., and McKenzie, J., 2005, Mass movement–induced fold-and-thrust belt structures in unconsolidated sediments in Lake Lucerne (Switzerland): Sedimentology, v. 52, p. 271–289, doi:10.1111/j.1365-3091.2004.00694.x. Scoffin, T.P., and Stoddart, D.R., 1978, The nature and significance of microatolls: Philosophical Transactions of the Royal Society of London, ser. B, v. 284, p. 99–122. Serva, L., 1994, Ground effects in intensity scales: Terra Nova, v. 6, p. 414–416, doi:10.1111/j.1365-3121.1994.tb00515.x. Serva, L., and Slemmons, D.B., eds., 1995, Perspective in Paleoseismology: Association of Engineering Geologists Special Publication 6, 138 p. Sheppard, P.R., and Jacoby, G.C., 1989, Application of tree-ring analysis to paleoseismology: Two case studies: Geology, v. 17, no. 3, p. 226–229, doi:10.1130/0091-7613(1989)017<0226:AOTRAT>2.3.CO;2. Sheppard, P.R., and White, L.O., 1995, Tree-ring responses to the 1978 earthquake at Stephens Pass, northeastern California: Geology, v. 23, no. 2, p. 109–112, doi:10.1130/0091-7613(1995)023<0109:TRRTTE>2.3.CO;2.
Shi, B., Anooshehpoor, A., Zeng, Y., and Brune, J.N., 1996, Rocking or overturning of precariously balanced rocks by earthquake: Bulletin of the Seismological Society of America, v. 86, no. 5, p. 1364–1371. Shi, S., Dawson, A.G., and Smith, D.E., 1995, Coastal sedimentation associated with the December 12th, 1992, tsunami in Flores, Indonesia, in Satake, K., and Imamura, F., eds., Tsunami 1992–1994: Their Generation, Dynamics and Hazard: Pure and Applied Geophysics, Topical Volume, v. 145, no. 3–4, p. 525–536. Shiki, T., Kumon, F., Inouchi, Y., Kontani, Y., Sakamoto, T., Tateishi, M., Matsubara, H., and Fukuyama, K., 2000, Sedimentary features of the seismoturbidites, Lake Biwa, Japan: Sedimentary Geology, v. 135, no. 1–4, p. 37–50, doi:10.1016/S0037-0738(00)00061-0. Shilts, W., and Clague, J., 1992, Documentation of earthquake-induced disturbance of lake sediments using subbottom acoustic profiling: Canadian Journal of Earth Sciences, v. 29, p. 1018–1042. Siame, L.L., Sébrier, M., Bellier, O., and Bourlès, D., 2006, Can cosmic ray exposure dating reveal the normal faulting activity of the Cordillera Blanca fault, Peru?: Revista de la Asociación Geológica Argentina, v. 61, no. 4, p. 536–544. Siegenthaler, C., Finger, W., Kelts, K., and Wang, S., 1987, Earthquake and seiche deposits in Lake Lucerne, Switzerland: Eclogae Geologicae Helvetiae, v. 80, p. 241–260. Sieh, K.E., 1978, Prehistoric large earthquakes produced by slip on the San Andreas fault at Pallett Creek, California: Journal of Geophysical Research, v. 83, p. 3907–3939, doi:10.1029/JB083iB08p03907. Sieh, K., Ward, S., Natawidjaja, D., and Suwargadi, B., 1999, Crustal deformation at the Sumatran subduction zone revealed by coral rings: Geophysical Research Letters, v. 26, p. 3141–3144, doi:10.1029/1999GL005409. Sieh, K., Natawidjaja, D.H., Meltzner, A.J., Shen, C.-C., Cheng, H., Li, K.-S., Suwargadi, B.W., Galetzka, J., Philibosian, B., and Edwards, R.L., 2008, Earthquake supercycles inferred from sea-level changes recorded in the corals of west Sumatra: Science, v. 322, no. 5908, p. 1674–1678, doi:10.1126/science.1163589. Silva, P.G., González Hernández, F.M., Goy, J.L., Zazo, C., and Carrasco, P., 2001, Paleo- and historical seismicity in Mallorca Island (Spain): A preliminary approach: Acta Geológica Hispánica, v. 36, no. 3–4, p. 245–266. Simón, P., 1627, Noticias Historiales de Venezuela, Volume II. Re-edited in 1963: Caracas, Venezuela, Academia Nacional de la Historia, No. 67. Sims, J., 1973, Earthquake-induced structures in sediments of Van Norman Lake, San Fernando, California: Science, v. 182, p. 161–163, doi:10.1126/ science.182.4108.161. Sims, J., 1975, Determining earthquake recurrence intervals from deformational structures in young lacustrine sediments: Tectonophysics, v. 29, p. 141–152, doi:10.1016/0040-1951(75)90139-0. Singer, A., 1998, Evaluación retrospectiva de los efectos geológicos destructores del terremoto de 1610 en los Andes venezolanos por medio de la confrontación de testimonios del siglo 17 y de observaciones de campo actuales: Revista Geográfica Venezolana, v. 39, p. 289–296. Slemmons, D.B., 1957, Geological effects of the Dixie Valley–Fairview Peak, Nevada, earthquakes of December 16, 1954: Bulletin of the Seismological Society of America, v. 47, no. 4, p. 353–375. Stein, R., and Yeats, R., 1989, Hidden earthquakes: Scientific American, v. 260, p. 48–57, doi:10.1038/scientificamerican0689-48. Stratton Noller, J., Sowers, J.M., and Lettis, W.R., eds., 2000, Quaternary Geochronology: Methods and Applications: Washington, D.C., American Geophysical Union, 582 p. Suggate, R.P., 1960, The interpretation of progressive fault displacement of flights of terraces: New Zealand Journal of Geology and Geophysics, v. 3, no. 3, p. 364–374. Surdeanu, V., Dulgheru, M., Pop, O., Anghel, T., and Todică, S., 2008, Spatial and temporal activity of landslides using dendrogeomorphological methods in the Călimani mining area (România): International Association of Geomorphologists Regional Conference “IGAR” [abs.]: Brasov, Rumania (CD-ROM). Tappin, D., 2007, Sedimentary features of tsunami deposits—Their origin, recognition and discrimination: An introduction: Sedimentary Geology, v. 200, p. 151–154, doi:10.1016/j.sedgeo.2007.01.001. Tasdemiroifelgülu, M., 1971, The 1970 Gediz earthquake in Western Anatolia, Turkey: Bulletin of the Seismological Society of America, v. 61, p. 1507– 1527. Taylor, F.W., Frohlich, C., Lecolle, J., and Strecker, M., 1987, Analysis of partially emerged corals and reef terraces in the central Vanuatu arc:
Geological criteria for evaluating seismicity revisited Comparison of contemporary coseismic and nonseismic with Quaternary vertical movements: Journal of Geophysical Research, v. 92, p. 4905– 4933, doi:10.1029/JB092iB06p04905. Terasmae, J., 1975, Dating of landslides in the Ottawa River valley by dendrochronology: A brief comment, in Yatsu, E., Ward, A., and Adams, F., eds., Mass Wasting: Proceedings of 4th Guelph Symposium on Geomorphology: Norwich, UK, GeoAbstracts Ltd., p. 153–158. Thomas, R.L., Wallach, J.L., McMillan, R.K., Bowlby, J.R., Frape, S., Keyes, D., and Mohajer, A.A., 1993, Recent deformation in the bottom sediments of western and southeastern Lake Ontario and its association with major structures and seismicity: Géographie Physique Quaternaire, v. 47, p. 325–335. Thunell, R., Tappa, E., Varela, R., Llano, M., Astor, Y., Muller-Karger, F., and Bohrer, R., 1999, Increased marine sediment suspension and fluxes following an earthquake: Nature, v. 398, p. 233–236, doi:10.1038/18430. Tibaldi, A., Rovida, A., and Corazzato, C., 2004, A giant deep-seated slope deformation in the Italian Alps studied by paleoseismological and morphometric techniques: Geomorphology, v. 58, p. 27–47, doi:10.1016/ S0169-555X(03)00184-3. Toda, S., and Okumura, K., eds., 2010, International Symposium on Active Faulting: Forecasting Large Earthquakes from Active Faults in Time and Space, 17–21 January 2010: Hiroshima, Japan, 123 p. Toppozada, T.R., Branum, D.M., Reichle, M.S., and Hallstrom, C.L., 2002, San Andreas fault zone, California: M≥5.5 earthquake history: Bulletin of the Seismological Society of America, v. 92, p. 2555–2601, doi:10.1785/0120000614. Townsend, T., 1998, Paleoseismology of the Waverley fault zone and implications for earthquake hazard in South Taranaki, New Zealand: New Zealand Journal of Geology and Geophysics, v. 41, p. 467–474, doi:10.1080/ 00288306.1998.9514823. Tuttle, M., 2001, The use of liquefaction features in paleoseismology: Lessons learned in the New Madrid seismic zone, central United States: Journal of Seismology, v. 5, no. 3, p. 361–380, doi:10.1023/A:1011423525258. Tuttle, M., and Seeber, L., 1991, Historic and prehistoric earthquake-induced liquefaction in Newbury, Massachusetts: Geology, v. 19, p. 594–597, doi: 10.1130/0091-7613(1991)019<0594:HAPEIL>2.3.CO;2. Tuttle, M., Law, T., Seeber, L., and Jacob, K., 1990, Liquefaction and ground failure in Ferland, Quebec, triggered by the 1988 Saguenay earthquake: Canadian Geotechnical Journal, v. 27, p. 580–589, doi:10.1139/t90-073. Van der Woerd, J., Ryerson, F.J., Tapponnier, P., Gaudemer, Y., Finkel, R., Meriaux, A.S., Caffee, M., Guaguang, Z., and Qunlu, H., 1998, Holocene leftslip rate determined by cosmogenic surface dating on the Xidatan segment of the Kunlun fault (Qinghai, China): Geology, v. 26, no. 8, p. 695–698, doi:10.1130/0091-7613(1998)026<0695:HLSRDB>2.3.CO;2. Van Dissen, R.J., Berryman, K., Pettinga, J.R., and Hill, N.L., 1992, Paleoseismicity of the Wellington–Hutt Valley segment of the Wellington fault, North Island, New Zealand: New Zealand Journal of Geology and Geophysics, v. 35, p. 165–176, doi:10.1080/00288306.1992.9514511. Verbeeck, K., Beatse, H., Vanneste, K., Renardy, F., Van Der Meer, H., RoyChowdhury, K., and Camelbeeck, T., 2000, Geomorphic and geophysical reconnaissance of the Reppel and Bocholt faults, NE Belgium, in Workshop Proceedings, “HAN2000—Evaluation of the Potential for Large Earthquakes in Regions of Present-Day Low Seismic Activity in Europe”: Han-sur-Lesse, Belgium, p. 167–170. Vita-Finzi, C., 1996, Paleoseismology in coastal Chile: Journal of Geophysical Research, v. 101, no. B3, p. 6109–6114, doi:10.1029/95JB01586. Vittori, E., Serva, L., and Sylos Labini, S., 1991, Paleoseismicity: Critical review of the state of the art: Tectonophysics, v. 193, p. 9–32, doi:10.1016/0040 -1951(91)90185-U. Wallace, R., 1981, Active faults, paleoseismology, and earthquake hazards in the western United States, in Simpson, W., and Richards, P., eds., Earthquake Prediction: An International Review: Washington, D.C., American Geophysical Union, Maurice Ewing Series 4, p. 209–216.
21
Wallace, R., 1984, Fault Scarps Formed during the Earthquakes of October 2, 1915, Pleasant Valley, Nevada, and Some Tectonic Implications: U.S. Geological Survey Profession Paper 1274-A, 33 p. Wallace, R., 1986, Active Tectonics: Impact on Society: Washington, D.C., National Academy Press, 266 p. Walsh, T., Combellick, R., and Black, G., 1995, Liquefaction Features from a Subduction Zone Earthquake: Preserved Examples from the 1964 Alaska Earthquake: Division of Geology and Earth Resources, Washington State Department of Natural Resources Report of Investigation 32, 80 p. Weidinger, J., 1998, Case history and hazard analysis of two lake-damming landslides in the Himalayas: Journal of Asian Earth Sciences, v. 16, no. 2–3, p. 323–331, doi:10.1016/S0743-9547(98)00026-9. Wellman, H.W., 1955, New Zealand Quaternary tectonics: International Journal of Earth Sciences, v. 43, no. 1, p. 248–257. Wilson, R.C., and Keefer, D.K., 1985, Predicting areal limits of earthquakeinduced landsliding: U.S. Geological Survey Professional Paper 1360, p. 317–345. Wise, D., Cassidy, J., and Locke, C., 2003, Geophysical imaging of the Quaternary Wairoa North fault, New Zealand: A case study: Journal of Applied Geophysics, v. 53, p. 1–16, doi:10.1016/S0926-9851(03)00013-2. Woodroffe, C., and McLean, R., 1990, Microatolls and recent sea level change on coral atolls: Nature, v. 344, p. 531–534, doi:10.1038/344531a0. Yamaguchi, D.K., Atwater, B.F., Bunker, D.E., Benson, B.E., and Reid, M.S., 1997, Tree-ring dating the 1700 Cascadia earthquake: Nature, v. 389, p. 922–923, doi:10.1038/40048. Yeats, R., 1986a, Faults related to folding with examples from New Zealand, in Reilly, W.I., and Hartford, B.E., ed., Recent Crustal Movements of the Pacific Region: Royal Society of New Zealand Bulletin, v. 24, p. 273–292. Yeats, R., 1986b, Active faults related to folding, in Wallace, R.E., ed., Active Tectonics: Washington, D.C., National Academy Press, p. 63–79. Yeats, R., Sieh, K., and Allen, C., 1997, The Geology of Earthquakes: Oxford, UK, Oxford University Press, 576 p. Youd, T.L., and Hoose, S.N., 1978, Historical Ground Failures in Northern California Triggered by Earthquakes: U.S. Geological Survey Professional Paper, 993 p. Youd, T.L., and Perkins, D.M., 1987, Mapping of liquefaction severity index: Journal of Geotechnical Engineering, v. 113, no. 11, p. 1374–1392, doi:10 .1061/(ASCE)0733-9410(1987)113:11(1374). Zachariasen, J., Sieh, K., Edwards, R.L., Taylor, F.W., and Hantoro, W.S., 1999, Pre- and coseismic vertical displacement associated with the giant 1833 Sumatran subduction earthquake: Evidence from coral microatolls: Journal of Geophysical Research, v. 104, p. 895–919, doi:10.1029/ 1998JB900050. Zachariasen, J., Sieh, K., Taylor, F.W., and Hantoro, W.S., 2000, Modern vertical deformation above the Sumatran subduction zone: Paleogeodetic insights from coral microatolls: Bulletin of the Seismological Society of America, v. 90, no. 4, p. 897–913, doi:10.1785/0119980016. Zazo, C., Silva, P.G., Goy, J.L., Hillaire-Marcel, C., Lario, J., Bardají, T., and González, A., 1999, Coastal uplift in continental collision plate boundaries: Data from the Last Interglacial marine terraces of the Gibraltar Strait area (south Spain): Tectonophysics, v. 301, p. 95–109, doi:10.1016/ S0040-1951(98)00217-0. Zreda, M., and Noller, J., 1998, Ages of prehistoric earthquakes revealed by cosmogenic chlorine-36 in a bedrock fault scarp at Hebgen Lake: Science, v. 282, p. 1097–1099, doi:10.1126/science.282.5391.1097. Zreda, M., and Noller, J., 1999, Timing of late Quaternary earthquakes on the Hebgen Lake fault by cosmogenic chlorine-36 dating of bedrock fault scarp: Geological Society of America Abstracts with Programs, v. 31, no. 4, p. A-63.
MANUSCRIPT ACCEPTED BY THE SOCIETY 7 DECEMBER 2010
Printed in the USA
The Geological Society of America Special Paper 479 2011
Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA: The Calabacillas fault, Albuquerque, New Mexico J.P. McCalpin GEO-HAZ Consulting, Box 837, Crestone, Colorado 81131, USA J.B.J. Harrison Department of Earth and Environmental Sciences, New Mexico Institute of Mining and Technology, Socorro, New Mexico 87801, USA G.W. Berger Desert Research Institute, 2215 Raggio Parkway, Reno, Nevada 89506-1095, USA H.C. Tobin Department of Geoscience, University of Wisconsin–Madison, Madison, Wisconsin 53706, USA
ABSTRACT The Calabacillas fault is a 40-km-long, down-to-the-east normal fault that trends N-S on the western edge of the Llano de Albuquerque, in western Albuquerque, New Mexico. It is one of several east-dipping normal faults that define the western margin of the Rio Grande rift at the latitude of Albuquerque. In the past 0.5–1 m.y., since the abandonment of the Llano de Albuquerque surface by the Rio Puerco and Rio Grande, vertical displacement on the Calabacillas fault has created a 27-m-high, eastfacing fault scarp on the western edge of the llano, equating to a long-term slip rate of 0.027–0.054 mm/yr. Our two trenches were located ~1 km from the south end of the fault, where a 1-km-wide graben has formed east of the main fault scarp. Trenching of the graben across the southern Calabacillas fault was 50% successful. The paleoearthquake event history on the 5.3-m-high antithetic scarp could not be reconstructed in detail because a strong carbonate soil profile had overprinted the entire 3-m-thick colluvial wedge deposit. It appears that numerous submeter displacements created this scarp, but the displacement was partitioned across several faults, so no single free face was higher than 10–20 cm. Free faces so small did not create colluvial wedges, and thus faulting did not trigger the pattern of footwall erosion and hanging-wall deposition needed to identify individual faulting events. On the 27-m-high main fault scarp, a 60-m-long trench straddled a minor slope break that overlies the main strand of the Calabacillas fault. The upper four soils exposed in the trench could be correlated across the main fault and indicated displacements of 10 cm, 30 cm, 55 cm, and 20 cm in the latest four paleoearthquakes.
McCalpin, J.P., Harrison, J.B.J., Berger, G.W., and Tobin, H.C., 2011, Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA: The Calabacillas fault, Albuquerque, New Mexico, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, p. 23–46, doi:10.1130/2011.2479(01). For permission to copy, contact
[email protected]. © 2011 The Geological Society of America. All rights reserved.
23
24
McCalpin et al. Six infrared-stimulated luminescence (IRSL) dates on eolian sands range from 14 ka at a depth of 0.5 m to 219 ka at a depth of 5 m. Secondary calcium carbonate has accumulated in soils here at a rate of 0.17–0.35 g/k.y. The latest four faulting events are dated at ca. 14 ka, 23 ka, 35 ka, and 55 ka. Thus, the displacement and recurrence times increase with increasing age, yielding relatively consistent slip rates of 0.011–0.028 mm/yr. There is evidence at this trench for a late Pleistocene (14 ka) small faulting/cracking event, similar in displacement and timing to the youngest warping event interpreted for the County Dump fault, which lies ~5 km to the east. The displacements measured in the main scarp trench are even smaller than those inferred on the County Dump fault, despite the length of the Calabacillas fault (40 km) being similar to that of the County Dump fault (35 km). If our trenches had been located farther north toward the center of the Calabacillas fault, the displacements may have been larger. The ages and recurrence intervals of the four events that occurred subsequent to 55 ka are similar to those seen at the County Dump site. The youngest event on the Calabacillas fault had only 5–10 cm of throw, which is considerably smaller than the 20–55 cm throws of the three previous events. This situation parallels the County Dump chronology, where the youngest warping event was abnormally small compared to earlier events.
INTRODUCTION Purpose and Scope of Study This study is part of a continuing effort by the U.S. Geological Survey’s National Earthquake Hazard Reduction Program (NEHRP) to characterize the mid- to late Quaternary activity of normal faults in the Rio Grande rift near and in Albuquerque, New Mexico. This study continues efforts begun in 1996 (McCalpin et al., 2006) to trench faults that displace the surface of the Llano de Albuquerque (also known locally as West Mesa), an abandoned valley floor of the Rio Grande valley that now stands ~100–150 m above the Rio Grande on the western margin of metropolitan Albuquerque. The fault scarps that traverse the llano surface trend north-south and face east, ranging from 10 to 30 m high, but they have very gentle slope angles (typically less than 5°). During May 1999, we spent 3 weeks excavating, logging, and sampling two trenches on the southern end of the Calabacillas fault on the Llano de Albuquerque (West Mesa) of the Albuquerque Basin. The trenches were oriented east-west and transected north-south–trending fault scarps on either side of a north-south–trending, 1-km-wide graben (Fig. 1). The goal of this trenching investigation was to reconstruct the chronology of surface-faulting events on the Calabacillas fault since the abandonment of the Llano de Albuquerque as a depositional surface, in the past 0.5–1 m.y. The Albuquerque–Santa Fe area contains numerous late Quaternary normal faults, but it has not experienced a surfacerupturing earthquake in the 146 yr since the first recorded New Mexico earthquake (1849). However, there have been 10 historic earthquakes of Modified Mercalli Intensity V or greater in the urban corridor (Von Hake, 1975). The largest of these events
was the 18 May 1918 earthquake (MMI VII–VIII) at Cerillos (near Santa Fe), during which people were thrown off their feet and a break in the ground surface was noted (Von Hake, 1975). These earthquakes, plus the presence of Quaternary fault scarps, indicate the potential for larger, surface-rupturing earthquakes (M >6.5) in the corridor. Machette (1998, p. 89) stated: Because the level of seismicity of the Rio Grande rift is generally low, the populace (in general) believes that earthquakes do not pose a significant threat to them, whereas the presence of abundant young faults tells a different story.… A myopic view of earthquake hazards posed by individual structures (i.e., having recurrence intervals of 10,000 yr or more) can lead to a complacent attitude that strengthens a perception of low seismic potential for the region. Without proper caution, this attitude can be manifested in inappropriate construction styles, building codes, land-use policies, and the siting (or relocation) of important critical facilities.
Machette (1998) touches on three unique geologic situations that have led to the perception of low earthquake risk in Albuquerque: (1) long recurrence intervals on Quaternary faults, (2) the short, segmented nature of faults in the basin, and (3) the large number of faults. Situation 2 and trench studies described later herein suggest that most basin faults rupture in M 6.5–7 earthquakes near the threshold of surface rupture, resulting in 0.1–0.5 m of surface offset, rather than the >1 m observed on longer range-bounding faults. Situation 1 means that these small surface scarps are nearly obliterated by erosion in the long time interval between events. Situations 1 and 2 explain why there are hundreds of meters of cumulative vertical displacement of the Tertiary rift sediments (Santa Fe Group) on many intrabasin faults but fault scarps are unimpressive.
Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA
Figure 1. Map showing faults with known and suspected displacements in Quaternary deposits near metropolitan Albuquerque, New Mexico (light gray). Dark-gray pattern shows pre-Neogene bedrock of the Sandia and Manzano Mountains. Fault numbers correspond to those in Table 1 of Personius et al. (1999). Faults with known displacements in the late Pleistocene (10–130 ka) or Holocene (<10 ka) are shown with heavier line weight. Hollow bars on the East Paradise, County Dump, and Hubbell Spring faults mark locations of previous detailed trench studies. Box labeled “study area” marks the site of the two trenches across the Calabacillas fault described in this report. Figure is from Personius et al. (1999). fz—fault zone.
25
26
McCalpin et al.
However, the aggregate hazard potential arising from 27 Quaternary faults within 40 km of downtown Albuquerque, even if they are low-slip-rate faults, is considerable. This rather unique situation suggests we should try to characterize the paleoearthquake history and surface deformation style of at least the closest faults to the populated areas of Albuquerque, such as the Calabacillas fault. Tectonic Setting The Albuquerque–Belen Basin is one of the largest structural depressions in the Rio Grande rift, measuring 135 km in northsouth dimension and up to 60 km in width (east-west dimension) (Kelley, 1977). This basin, along with others in the rift, began to form in the latest Oligocene–earliest Miocene, based on the age of the oldest basin-fill deposits (Chapin and Cather, 1994). Explosive andesitic volcanism and caldera formation accompanied the early and middle stages of rifting, with the latest major caldera collapse at 1.2 Ma (Valles Caldera; Heiken et al., 1990), whereas Quaternary volcanism has been basaltic (Lipman and Mehnert, 1975). The Albuquerque–Belen Basin is bounded by the steep range fronts of the Sandia and Manzano–Los Pinos Mountains on the east and by the Lucero and Ladrone Mountains on the southwest. In a gross sense, the basin is an asymmetric east-tilted half graben (Fig. 2). In 1994, Russell and Snelson (1994a, 1994b) proposed that the west-dipping master fault in the northern graben, termed the “Rio Grande master fault,” underlay the Holocene floodplain of the Rio Grande River and became listric at a depth of ~10 km. The numerous east-dipping faults found west of the river (County Dump, Zia, Calabacillas, and Sand Hills fault zones; Fig. 2) were interpreted to be antithetic to this master fault. The steep range front of the Sandia Mountains is underlain by west-dipping faults. These faults were interpreted to have defined the original eastern rift boundary early in its history of development but were assumed to be largely abandoned now (Lozinsky et al., 1991; Russell and
Snelson, 1994a, 1994b; Lozinsky, 1994). The inferred basinward younging of normal faults from the Sandia range front to the buried Rio Grande master fault matches a trend documented elsewhere in rift zones (Rosendahl, 1987; Dart et al., 1995). Between the “older” Sandia and Rincon faults and the newer Rio Grande master fault, a secondary footwall block exists, termed the North Albuquerque structural bench (“Albuquerque bench” on Fig. 2). This bench is a pediment surface veneered with piedmont-facies alluvial-fan deposits roughly 300,000 yr old. Most of suburban Albuquerque lies on this bench. The pattern of faulting in the Albuquerque–Belen Basin was reinterpreted after acquisition of detailed gravity and magnetic data in the late 1990s by Grauch (1999) and Grauch et al. (1999, 2002). The subsurface geometry is defined by three subbasins (from north to south, the Santo Domingo, Calabacillas, and Belen subbasins) separated by gravity highs or “shelves” (Fig. 3). As shown by Connell (2001) and Connell et al. (2001), these subbasins have a rhomboid shape and are bounded by northwesttrending lineaments. Figure 1 shows that the eastern margin faults of the rift have been active in the “shelves” north (Rincon fault) and south (Hubbell Spring fault) of the Calabacillas subbasin, in the Holocene–late Pleistocene. The Rincon fault shows evidence of two faulting events in the past ~100 k.y., the latest of which was mid-Holocene in age (Connell, 1995, 2000). The Hubbell Spring fault shows evidence of three late Quaternary events, with average vertical offsets of 1.6 m per event (Personius et al., 2000; Personius and Mahan, 2003). In contrast, there is only weak evidence for late Pleistocene slip on the eastern margin of the Calabacillas subbasin (Sandia fault zone; McCalpin and Harrison, 2006). This lack of mid- to late Quaternary faulting between the Rincon and Hubbell Spring faults defines a gap in recent faulting on the eastern side of the rift, termed the “East Heights gap” (McCalpin, 2001). This gap may be a site of future faulting, if the multiple post–300 ka slip events on the Rincon and Hubbell Spring faults have transferred stress to this fault block.
Figure 2. Cross section of the northernmost Albuquerque Basin based on seismic and well data. The line of section trends E-W just north of Rincon Ridge. Although the rift is shown as a nearly symmetrical graben at this latitude, it becomes more an east-tilted half-graben to the south. Figure is from Russell and Snelson (1994b).
Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA On the western margin of the Calabacillas subbasin, previous trenching studies have documented late Quaternary faulting events. These include three events in the past 120 k.y. on the County Dump fault (at ca. 30, 45, and 80 ka; McCalpin et al., 2006); three events in the past 286 k.y. on the East Paradise fault (Personius and Mahan, 2000); and four events in the past 55 k.y. on the Calabacillas fault (this paper). EAST TRENCH Location and Local Geology The east trench is located in the north-central part of the Llano de Albuquerque in the west-central part of “The Volca-
27
noes” quadrangle (USGS 7.5-minute topographic map series) (Figs. 3 and 4). The Llano de Albuquerque represents an abandoned constructional surface that marks the highest level of basin aggradation in the Santa Fe Group west of the Rio Grande, and it represents a local top of the Miocene–Pliocene basin fill (Santa Fe Group). Most of the Llano surface is mantled with a thin or discontinuous cover of Holocene–late Pleistocene eolian sand. Beneath the sand, a strong carbonate-rich paleosol up to 2 m thick (the “Llano de Albuquerque soil”) is developed on the fluvial sands and gravels of the uppermost Santa Fe Group (Machette, 1978). The study area is located at the southern end of the Calabacillas fault, which terminates against the Loma Colorado transfer zone of Hawley (1996). The trench site is at the northern margin of the City of Albuquerque Soil Amendment
Figure 3. Gravity map of the Albuquerque Basin, showing the three subbasins (white lettering) separated by shelves (from Grauch et al., 1999). Thin red lines show 7.5ʹ topographic quadrangle boundaries; thicker purple line outlines metropolitan Albuquerque. Our study area (labeled) is at the southern end of the Calabacillas fault.
28
McCalpin et al.
Figure 4. Perspective slope map of The Volcanoes and Calabacillas Arroyo quadrangles, showing fault scarps and the 1999 trench sites. Lighter tones indicate steeper slope angles; darker tones indicate gentler slopes. This image was produced from the 10 m digital elevation model of the Albuquerque area provided by the U.S. Geological Survey in Denver. The southern end of the Calabacillas fault is clearly truncated by the Loma Colorado transfer zone of Hawley (1996). South of the Loma Colorado zone, there is an area of deranged topography containing small isolated hills and depressions with ephemeral lakes. At the latitude of our two 1999 trenches, the Calabacillas fault is expressed as an asymmetrical, 1-km-wide graben. The western boundary is formed by the 27-m-high main (east-facing) scarp, which is fronted by a 0.6-km-wide apron of alluvium (coalescing alluvial fans) that descends eastward into the graben axis. The eastern boundary is formed by the 5.3-m-high antithetic scarp. The graben axis is a series of closed depressions up against the eastern scarp.
Facility (SAF). At the SAF, sewage sludge from the city’s sewage treatment plants is tilled into the eolian sand that fills the graben of the Calabacillas fault. Presumably, this site was chosen for the SAF because the graben is a closed depression here, and runoff can be controlled. The 7.5ʹ topographic map of the Volcano Ranch quadrangle (renamed in 1995 to The Volcanoes quadrangle) shows three ephemeral lakes in the axis of the graben. Geomorphology The graben is bounded on the west by the broad 27-m-high, 250-m-wide (from crest to toe) main scarp of the Calabacillas fault, and on the east by a 5.3-m-high, west-facing (antithetic) scarp (Fig. 4). It is this smaller scarp that is transected by our east trench. The fault scarp is relatively linear and sharp compared to other scarps on the llano, but it is still subdued compared to most Quaternary normal fault scarps. The scarp is 5.3 m high but 215 m wide, and it attains a maximum scarp slope angle of only 2.5° (Fig. 5). The profile is relatively symmetrical, and the main fault exposed in the trench underlies the central, steepest part of the scarp profile, as is typical for normal faults.
Stratigraphy and Soils The east trench is 130 m long, up to 8 m wide, and 6 m deep, and it was dug in a benched design (Fig. 6). The eastern and western thirds of the trench are composed of a single bench level above a 1-m-wide inner slot, with each of the two vertical wall levels 1.5 m high, for a total depth of ~3 m. The central third of the trench had to be deeper to expose the fault zone and the Miocene–Pliocene Santa Fe Group beneath the colluvial wedge sequence, so it is composed of two bench levels above the inner slot. The maximum depth reached in this section is 6 m. The most striking feature of this trench is the strong stage IV (Gile et al., 1966) carbonate soil that underlies the ground surface across the entire scarp. On the upthrown block, this soil is 2.5 m thick, but it thickens to engulf the entire 4.5-m-thick colluvial deposit abutted against the fault plane. The strong soil development made distinguishing different parent materials difficult. Nevertheless, we did define five major parent materials units (numbered 1–5, where 1 = oldest), three soils (the surface soil and two buried soils) containing 17 horizons, and krotovinas of five different ages (kr0–kr2, where kr0 = oldest).
Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA
29
Figure 5. Topographic profile of the antithetic fault scarp at the site of the east trench.
Figure 6. Photograph of the east trench, looking east. Excavator is deepening and widening the trench at the main fault zone.
The unconsolidated parent materials exposed in the trench fall into five groups (Fig. 7). These groups are, from oldest to youngest: (1) well-stratified sands and gravels belonging to the Santa Fe Group, which underlies the entire llano and is Pliocene– Pleistocene (?) in age; (2) a transitional unit between the Santa Fe Group and the younger Pleistocene eolian sand blanket that covers the llano (found only on the upthrown block); (3) scarpderived colluvium and correlative distal colluvium/slope wash that lies directly atop the Santa Fe Group on the downthrown block; (4) a thick sand deposit totally engulfed by soil K horizon development on the downthrown block; and (5) a thin (0.5–1 m) blanket of Holocene or latest Pleistocene eolian sand that blankets the scarp. The Santa Fe Group (parent material 1) is divided into four subunits (1a–1d). Unit 1a is composed of well-stratified mediumcoarse sands and small pebble gravels that are unaffected by pedo-
genesis. This stream alluvium represents the last phase of deposition by the ancestral Rio Puerco as it flowed on the Llano de Albuquerque surface, before the beginning of incision in the early to mid-Pleistocene. Unit 1b is a poorly stratified mediumcoarse sand that is transitional with underlying unit 1a. Unit 1c is an unusual unit, composed of poorly stratified sand that contains pebble-sized clasts of pure white carbonate. We interpret the clasts as retransported remains of a stream-bottom carbonate deposit, i.e., a precipitate that formed beneath channels of losing streams due to infiltration of carbonate-rich water. According to Sean Connell (New Mexico Bureau of Mines and Mineral Resources, Albuquerque, 1999, personal commun.), such pure carbonate beds are common in the Santa Fe Group. Unit 1d is a poorly stratified sand similar to unit 1b. Unit 2 occupies all the stratigraphic section on the upthrown block above the Santa Fe Group and below the thin blanket of Holocene eolian sand. This parent material has been so strongly affected by the accumulation of pedogenic carbonate that it is difficult to infer its original sedimentology. The K11 and K12 horizons contain 59% and 51% carbonate by weight, indicating that more than half of the present volume of unit 2 is a precipitate. We infer that parent material 1 (Santa Fe Formation) on the upthrown block grades upward into eolian sand that was deposited on the llano surface over a period of many hundreds of thousands of years. Due to soil formation and physical mixing of the two parent materials (bioturbation), it is no longer possible to map this Santa Fe Group–eolian parent material boundary. Unit 2 and its contained soil profile span the entire time period from the abandonment of the llano (0.5–1 Ma?) to the deposition of the Holocene eolian blanket. Unit 3 exists only on the downthrown block and in the fault zone. It lies directly atop unit 1d, and its stratigraphic position indicates it is either the youngest part of the Santa Fe Group alluvium or the oldest part of the overlying, llano-wide eolian deposit. However, the geometry of unit 3 within 2 m of the main fault zone (at 67 m on the horizontal scale of the trench log, or
Figure 7. Log of the east trench. Footwall strata of the Santa Fe Group unaffected by pedogenesis are shown in dark blue. Pink and violet areas are filled krotovinas or carbonate dissolution-collapse features.
30
Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA 67 m H) shows that this unit dips ~10° west and unconformably overlies the Santa Fe Group on both the upthrown and downthrown blocks. This angular unconformity indicates that unit 3 is a scarp-derived deposit of sand that was deposited on a gently sloping fault free face developed in loose sand. This free face must have exposed unit 1c, because white carbonate clasts are found scattered throughout unit 3. In addition, several small faults on the upthrown block are truncated by the bottom of unit 3, confirming that it was deposited after a faulting event. Therefore, we interpret unit 3 proximal to the fault as retransported sand derived from a degraded free face in sandy Santa Fe Group. The part of parent material unit 3 farther from the fault is a massive sand that could be wash-facies colluvium, alluvium, or eolian sand. Unconformably overlying unit 3 on the downthrown block, there is a 3-m-thick strong carbonate soil (unit 4). K horizons of this soil contain 40%–52% carbonate by weight, the precipitation of which has obscured the original sedimentology of the parent material. However, most of the noncarbonate volume of this deposit is sand, so we infer that most of this parent material is part of the 1–2-m-thick eolian deposit that covers the entire llano. As such, it is partly time-correlative with unit 2 on the upthrown block, but it may also contain components of scarpderived deposits (like unit 3) that have been obscured by K horizon development. The youngest parent material exposed in the trench (unit 5) is a thin (0.5–1 m thick) blanket of eolian sand. This parent material is ~0.5 m thick on the upthrown block and ~1 m thick on the downthrown block. This deposit clearly lies unconformably atop the K horizon developed on unit 4 and carries a much weaker soil (described in the next section). The trench walls expose many areas where the K horizons have been removed and replaced by much younger, weakly calcareous, reddish sand. These pockets of sand are labeled as “krotovinas” on the trench log, but they probably have several origins. For example, most of the krotovinas on the upthrown block (0–60 m H on the log; see Fig. 7) are elongated in the horizontal plane, have irregular margins, and are not associated with faults; these are probably true rodent burrows. In contrast, the upward-flaring krotovinas overlying the main normal fault at 67 m H (Fig. 8) grade downward into the fault plane, have very sharp subvertical sides, and are probably tectonic fissures that were later filled with sand. The krotovinas at 63–65 m H overlie secondary normal faults and are elongated vertically, yet they have irregular shapes. These may represent tectonic fractures or fissures in the K horizons that were enlarged later by burrowing rodents. Another indication of the tectonic inheritance of some krotovinas is the increase in density of krotovinas in the footwall progressively closer to the main normal fault zone. The “krotovinas” fall into three broad age categories. Unit kr0 represents very old krotovinas that predate the formation of the soil horizons in unit 4 (i.e., the soil horizons of unit 4 are also developed through these krotovinas; see Fig. 7). Unit kr1 represents younger filled krotovinas. Unit kr2 represents the youngest krotovinas, some of which are younger than the bottom half of
31
Figure 8. Enlarged log of the main fault in the east trench.
parent material unit 5 (Fig. 7). These krotovinas are filled with very loose sand with no detectable soil development other than a red color staining. Soils The main map units in the east trench are soil horizons. The surface soil is defined as occupying only unit 5, the Holocene– late Pleistocene eolian sand blanket, and is composed of AB, Bk, and Bk2 horizons (Figs. 7 and 8; Table 1). The AB horizon contains 5%–6% carbonate by weight, and the Bk horizons contain 8%–11% carbonate by weight (Table 1). Cumulative secondary carbonate in these horizons is 6.8–6.9 g cm–2 column. The uppermost buried soil (buried soil 1, or b1) is developed on parent material unit 2 on the upthrown block and on parent material unit 4 on the downthrown block. The top of this soil is marked by the abrupt appearance of stage III–IV carbonate. In several places, the uppermost horizon in this soil (typically the K11 horizon) has a rubblized texture. We interpret this rubblization to represent a period of past dissolution of the top of the K horizon, probably due to removal of the overlying soil horizons and subaerial exposure. This rubblization is only observed on the upthrown block, where it is reasonable that a period of postfaulting erosion may have stripped off the loose eolian surface sands to expose the top of a hard K horizon. The uppermost buried soil on the upthrown block extends from 63 cm depth to 4.71 m depth and is developed on parent material unit 2 (horizons 2K11b1, 2K12b1, 2K2b1, 2K3b1) and parent material unit 1 (horizons 1dCk1b1, and other Ck horizons developed on other subunits of unit 1). Cumulative secondary carbonate in buried soil 1 at 11 m H is ~94 g cm–2 column.
#W2
#W1
W1-2K1b5
WMF9
21
15
35
30
27 90
WMF27 W2-12Kb6
WMF28 W2-11Kb7
25
WMF24 W2-16k1b4
10
100
WMF23 W2-17Bkb3
WMF26 W2-16Ckb4
23
WMF22 W2-18Bkb2
WMF25 W2-16k2b4
15 19
WMF21 W2-18Bwb2
16
WMF19 W2-20Ck
WMF20 W2-20Ck2
13 17
WMF17 W2-20(Btk)k1
27
WMF16 W2-20Bt2
WMF18 W2-20k2
8 22
WMF15 W2-20Bt1
284
WMF14 W2-21AB
SUM
17
W1-2Bkb5
WMF8
WMF13 W1-1Cnb5-7
W1-17Bkb3
WMF7
30 22
25
W1-18Bwb2
WMF6
30
W1-20Ck
WMF5
8 24
WMF12 W1-1Ckb5-7
W1-20k
WMF4
WMF11 W1-1K3B5-7
W1-20Bt2
WMF3
7 26
24
W1-20Bt1
WMF2
Thickness (cm)
WMF10 W1-1K2b5-7
W1-21AB
WMF1
Albuquerque West Trench
Soil profile no. Lab no. Sample no.
–4.42
–3.52
–3.25
–3.15
–2.85
–2.60
–1.60
–1.37
–1.18
–1.03
–0.87
–0.70
–0.57
–0.30
–0.08
–17.87
–2.77
–2.60
–2.35
–2.05
–1.81
–1.60
–1.45
–1.10
–0.88
–0.58
–0.34
–0.26
–0.07
Depth (m)
8.80
8.31
4.57
10.67
18.87
3.78
5.09
2.52
1.90
2.72
2.97
8.37
1.57
0.58
1.11
65.39
1.58
1.21
11.95
17.05
10.54
1.97
3.89
3.41
1.69
9.06
2.34
0.39
0.31
% CaCO3
1.59
1.37
3.38
1.74
1.72
1.82
1.55
1.69
1.74
1.78
1.69
1.57
1.40
1.51
1.44
1.71
1.68
1.76
1.70
1.74
1.74
1.77
1.72
1.65
1.68
1.34
1.51
1.61
12.59
3.07
1.54
5.57
8.11
6.88
1.81
0.81
0.50
0.77
0.85
1.71
0.59
0.19
0.13
27.23
0.46
0.51
6.31
6.96
3.85
0.51
2.41
1.29
0.84
3.65
0.25
0.15
0.03
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
Bulk Total Orig. density CaCO3 (g) % CaCO3
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
Orig. bulk density
1.81
0.54
0.20
0.60
0.50
2.01
0.46
0.38
0.30
0.32
0.34
0.26
0.54
0.44
0.16
5.71
0.34
0.50
0.60
0.48
0.42
0.30
0.70
0.44
0.60
0.48
0.16
0.52
0.14
8.88
7.08
6.53
6.33
5.73
5.23
3.22
2.75
2.37
2.07
1.75
1.41
1.15
0.60
0.16
5.71
5.37
4.86
4.26
3.78
3.36
3.06
2.35
1.91
1.31
0.82
0.66
0.14
10.78
2.53
1.34
4.97
7.61
4.87
1.35
0.43
0.19
0.45
0.51
1.45
0.05
–0.25
–0.03
21.52
0.12
0.01
5.71
6.47
3.43
0.21
1.71
0.85
0.23
3.17
0.09
–0.37
–0.11
Orig. Cumulative CaCO3 orig. Secondary CaCO3 (g) CaCO3 (g) (g)
36.49
25.71
23.18
21.83
16.87
9.26
4.39
3.03
2.61
2.41
1.96
1.45
0.00
0.00
0.00
21.90
21.79
21.78
16.07
9.60
6.17
5.96
4.25
3.40
3.17
0.00
0.00
0.00
Cumulative secondary CaCO3 (g)
TABLE 1. SUMMARY OF SOIL ANALYTICAL DATA FROM THE WEST AND EAST TRENCHES OF 1999
216.00
55.60
32.20
14.00
IRSL age (ka)
139.28
131.20
101.35
55.60
46.19
36.26
33.13
31.70
28.38
24.62
14.00
5.49 216.00
4.49
4.15
NA
67.93
32.19
22.26
19.13
17.70
14.38
10.62
14.00
(Continued)
52.41
32.20
26.59
24.82
24.01
18.21
20.01
14.00
EstiEstiEstimated mated mated age age age Age (ka) (ka) (ka) trend Method Method Method (k.y./g*) 3 2 1
32
#W4
#W3
52
WMF33 W3 13Ckb5
98 35 35 18 12 25 50
WMF45 W4 17Bkb3
WMF46 W4 16K1b4
WMF47 W4 16K2b4
WMF48 W4 16K3b4
WMF49 W4 16Ck2b4
WMF50 W4 15Ckb4
WMF51 W4 14Ckb5 426
25
WMF44 W4 18Bkb2
SUM
6
WMF41 W4 20C
15
18
WMF40 W4 20 Ck
WMF43 W4 18Bwb2
24
WMF39 W4 20 K2
WMF42 W4 19Bkb1
8 15
WMF38 W4 20 K1
8 18
WMF37 W4 20BtK
WMF35 W4 20Bt1
WMF36 W4 20Bt2
3 13
WMF34 W4 21AB
110
WMF32 W3 13 bk2b5
SUM
30 28
WMF31 W3 13Bk1b5
520
25
SUM
53
WMF30 W2-11Ck2b7
Thickness (cm)
WMF29 W2-11Ckb7
#W2 (Continued)
Soil profile no. Lab no. Sample no.
–31.40
–4.26
–3.76
–3.51
–3.39
–3.21
–2.86
–2.51
–1.53
–1.28
–1.13
–1.07
–0.89
–0.65
–0.50
–0.42
–0.24
–0.16
–0.03
–1.98
–1.10
–0.58
–0.30
–37.64
–5.20
–4.95
Depth (m)
102.87
6.80
5.76
3.17
5.74
13.13
19.79
3.34
2.31
2.02
2.23
2.15
4.43
6.73
9.52
7.89
3.18
4.53
0.15
20.52
7.01
1.74
11.77
95.16
4.52
8.81
% CaCO3
1.79
1.70
1.72
1.77
1.82
1.76
1.82
1.67
1.62
1.67
1.74
1.57
1.66
1.72
1.53
1.53
1.53
1.48
1.72
1.57
1.78
1.67
1.72
48.01
6.09
2.45
0.65
1.83
8.36
12.19
5.96
0.96
0.49
0.22
0.67
1.67
1.68
1.31
2.17
0.39
0.90
0.01
13.32
6.27
0.76
6.29
55.06
1.89
8.03
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
Bulk Total Orig. density CaCO3 (g) % CaCO3
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
Orig. bulk density
8.56
1.01
0.50
0.24
0.36
0.70
0.70
1.97
0.50
0.30
0.12
0.36
0.48
0.30
0.16
0.36
0.16
0.26
0.06
2.21
1.05
0.56
0.60
10.45
0.50
1.07
8.56
7.56
7.06
6.81
6.45
5.75
5.05
3.08
2.57
2.27
2.15
1.79
1.31
1.01
0.84
0.48
0.32
0.06
2.21
1.17
0.60
10.45
9.95
39.44
5.08
1.95
0.41
1.47
7.66
11.49
3.99
0.46
0.19
0.10
0.31
1.19
1.37
1.15
1.81
0.23
0.64
–0.05
11.11
5.22
0.20
5.68
44.61
1.38
6.97
Orig. Cumulative CaCO3 orig. Secondary CaCO3 (g) CaCO3 (g) (g)
44.31
38.63
33.55
31.60
31.19
29.72
22.06
10.57
6.59
6.13
5.94
5.83
5.52
4.33
2.96
1.81
0.00
0.00
0.00
11.11
5.88
5.68
44.84
43.46
Cumulative secondary CaCO3 (g)
151.00
55.60
32.20
14.00
IRSL age (ka)
3.09
3.93
2.76
NA
(Continued)
EstiEstiEstimated mated mated age age age Age (ka) (ka) (ka) trend Method Method Method (k.y./g*) 3 2 1
TABLE 1. SUMMARY OF SOIL ANALYTICAL DATA FROM THE WEST AND EAST TRENCHES OF 1999 (Continued)
33
38 65 30 25
WMF60 E1 1.5Ck1 E1.09
WMF61 E1 1.5Ck2 E1.10
WMF62 E1 unit2 E1.11
15 37
WMF70 E2 U3 E2.07
WMF71 E2 U4 E2.08 276
23
WMF69 E2 U2 E2.06
SUM
58
WMF66 E2 K12 E2.03
30
54
WMF65 E2 K11 E2.02
WMF68 E2 1Ck E2.05
23
WMF64 E2 Bk E2.01
WMF67 E2 K2 E2.04
7 29
WMF63 E2 AB E2.00
496
–3.38
10
WMF58 E1 1Ck E1.07
WMF59 E1 1.5Bk E1.08
SUM
–3.28
50
WMF57 E1 1K3 E1.06
–1.73
5.01 183.42
–13.26
8.04
10.18
8.33
40.34
52.70
40.76
11.38
6.68
209.63
10.86
8.19
13.53
8.02
11.52
19.40
13.35
51.28
59.35
8.28
5.85
% CaCO3
–2.76
–2.39
–2.24
–2.01
–1.71
–1.13
–0.59
–0.36
–0.07
–31.10
–4.96
–4.71
–4.41
–3.76
–2.78
55
–1.18
105
55
WMF54 E1 K11 E1.03
–0.63
–0.28
WMF55 E1 K12 E1.04
35
WMF53 E1 Bk E1.02
Depth (m)
WMF56 E1 K2 E1.05
28
WMF52 E1 AB E1.01
Thickness (cm)
*Age of sample divided by secondary carbonate.
#E2
# E1
Albuquerque East Trench
Soil profile no. Lab no. Sample no.
1.50
1.57
1.59
1.66
1.49
1.40
1.40
2.10
1.57
1.84
1.71
1.63
2.09
1.96
1.72
1.58
1.84
1.57
1.93
1.50
108.04
2.78
1.89
3.72
4.15
34.86
39.84
13.12
6.93
0.73
182.19
5.00
4.20
14.34
6.37
2.26
16.68
22.15
51.90
51.25
5.59
2.46
1.50
1.50
1.50
4.50
1.50
1.50
1.50
1.50
1.50
4.50
4.50
4.50
4.50
4.50
1.50
1.50
1.50
1.50
1.50
1.50
Bulk Total Orig. density CaCO3 (g) % CaCO3
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
1.34
Orig. bulk density
6.75
0.74
0.30
0.46
1.81
1.17
1.09
0.46
0.58
0.14
16.72
1.51
1.81
3.92
2.29
0.60
1.01
2.11
1.11
1.11
0.70
0.56
6.75
6.01
5.71
5.25
3.44
2.27
1.19
0.72
0.14
16.72
15.22
13.41
9.49
7.20
6.59
5.59
3.48
2.37
1.27
0.56
101.28
2.04
1.59
3.26
2.34
33.70
38.76
12.66
6.35
0.59
165.46
3.49
2.39
10.42
4.08
1.65
15.68
20.04
50.79
50.14
4.89
1.89
Orig. Cumulative CaCO3 orig. Secondary CaCO3 (g) CaCO3 (g) (g)
101.28
99.25
97.66
94.39
92.06
58.36
19.60
6.94
0.59
165.46
161.97
159.58
149.17
145.09
143.43
127.75
107.72
56.93
6.78
1.89
Cumulative secondary CaCO3 (g)
IRSL age (ka)
EstiEstiEstimated mated mated age age age Age (ka) (ka) (ka) trend Method Method Method (k.y./g*) 3 2 1
TABLE 1. SUMMARY OF SOIL ANALYTICAL DATA FROM THE WEST AND EAST TRENCHES OF 1999 (Continued)
34
Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA On the downthrown block, buried soil 1 is up to 3 m thick. This thickening is due to carbonate accumulation within parent material unit 4 (horizons 4K1b1, 4K2b1, 4K3b1, 4Ckb1). Cumulative secondary carbonate in buried soil 1 at 67 m H is 138 g cm–2 column, or ~47% more carbonate than in buried soil 1 on the upthrown block. This greater amount of carbonate resulted from: (1) erosion of carbonate from upthrown block soils and redeposition as “recycled” parent material on the downthrown block, after faulting events, and (2) greater infiltration of carbonate dust on the lower half of the scarp than on the upper half, due to a progressive increase of downslope sheetwash and increased permeability due to brecciation of the soil. This catena effect has been observed on many normal fault scarps (McCalpin and Berry, 1996; Birkeland, 1999). The oldest buried soil (buried soil 2) is developed only on parent material units 3 and 1 on the downthrown block. This relatively weak soil represents soil formation that occurred on proximal and distal colluvium after the initial faulting event on this scarp. Structure The east trench is dominated by the main normal fault zone, which is located at 64–67 m H, exactly beneath the scarp midpoint, and which consists of four west-dipping normal faults. Together with a smaller fault zone at 32 m H, normal faults in the footwall have a cumulative throw of ~3 m. The main normal fault plane at 67 m H is responsible for most of this throw, and it consists of a 10-cm-wide zone of sheared sand and gravel that dips 75°W (Fig. 8). The main fault juxtaposes very loose sandy gravels of the uppermost Santa Fe Group against eolian-derived colluvium heavily impregnated with calcium carbonate. Within 2 m (west) of the main fault, there are four secondary normal faults in the footwall that dip 70°W and have an aggregate throw of 30 cm. The western two of these faults do not extend upward into unit 3, suggesting that they predate that unit. The main normal fault grades upward into a 30-cm-wide filled fissure on the middle 1.5-m-high trench wall (Fig. 8), and that fissure in turns grades upward into a 1-m-wide krotovina in the upper 1.5-m-high trench wall. The planar, subvertical margins of the fissure and krotovinas strongly suggest that they were defined by tectonic fractures. Because these margins extend upward through all of unit 4 (up through horizon 4K1b1), it appears that some fault movement is younger than unit 4 and its soil (Fig. 7). In fact, several prominent fractures associated with the large krotovinas extend upward into unit 5Bk, although they do not offset the lower contact of that unit. It is unclear whether these fractures are tectonic or pedogenic. If they are tectonic, they imply that some type of fracturing event occurred after deposition of at least the lower half of parent material unit 5. In the hanging wall, the Santa Fe Group in the bottom of the trench, there appears to be warp of 1 m up-to-the-west between 90 and 100 m H (Fig. 7). There are no up-to-the-west faults visible in this warped area, so it is unclear whether this warp overlies
35
faults beneath the trench floor (i.e., it is a drape fold), or whether the warp is a depositional feature. However, the warped area is overlain by a huge carbonate-filled void that extends through the entire thickness of parent materials units 3 and 4 (Fig. 7). This void appears to be a large burrowed-out area where animals took advantage of fractures caused by the warping. Finally, unit 3 is anomalously tilted down-to-the-west by at least 1 m between 120 and 123 m H, at the far western end of the trench (Fig. 7). This tilting also affects the 4Bk3-4K1 contact, although not as much. Small groups of east-dipping fractures at 120 m H and 125 m H may overlie better-defined faults that lie beneath the trench floor. If these faults have the same east dip as the fractures, they would have an apparent reverse geometry. Such “reverse” faults were observed in the 1996 County Dump trenches (McCalpin et al., 2006) and in the 1999 west trench described later. Geochronology Because the east trench did not contain a good stratigraphic record of faulting events, we did not collect any radiocarbon or luminescence dating samples from it, preferring to concentrate our limited dating budget on the west trench. However, we did analyze the amount of secondary calcium carbonate in two soil profiles, one in the footwall at 11 m H and one in the hanging wall at 68 m H (Fig. 7). The entire soil profile exposed in the trench contains 101 g cm–2 column in the footwall and 165 g cm–2 column in the hanging wall (Table 1). If we apply the carbonate-accumulation rates of 0.25–0.5 g cm–2 column k.y.–1 cited by Machette (1985) for long time periods (pluvial plus interpluvial), we can calculate a soil age of 202–404 ka for the footwall soil and 330–660 ka for the hanging-wall soil. The footwall soil has probably been periodically eroded after faulting events, so its soil-age estimate should be viewed as a minimum. The hanging-wall soil has been augmented by carbonate eroded from the footwall, so its carbonate content is greater than that of an equivalent-age soil on a stable geomorphic surface, such as the surfaces from which Machette’s (1985) calibration curve was derived. Accordingly, the estimates of 330–660 ka for soil formation should be viewed as maximum age estimates. Interpretation Our preliminary interpretation is that individual faulting events on this 5.3-m-high scarp only averaged ~0.2–0.5 m of throw distributed across several faults. The free faces above each fault were developed in loose eolian sand, which quickly collapsed to form thin, amorphous colluvial deposits; no gravel of the Santa Fe Group was ever exposed in the free face. The thin colluvial wedge was then impregnated by pedogenic carbonate accumulation from the surface. As a result, despite repeated surface-faulting events in the past 500 k.y. on the fault, the resulting carbonate soil was cumulic on the colluvial wedge parent materials and indurated itself into a 4.5-m-thick massive
36
McCalpin et al.
carbonate soil composed of multiple K horizons. Due to this carbonate overprinting effect, we were not able to distinguish discrete colluvial wedges except for unit 3. The main lesson to be learned at this trench is that, in order to distinguish individual colluvial deposits, scarp-derived colluvial deposits must be thick enough to escape cumulic soil formation and soil “welding.” Conversely, if vertical displacement is <<0.5 m, as probably occurred on this antithetic fault, then soil overprinting prevents identification of individual colluvial wedge deposits. This phenomenon of small free fault faces being created on an eolian-dominated landform was also documented by McCalpin et al. (2006) on the Llano de Albuquerque, and by Personius and Mahan (2003) on the Llano de Manzano on the eastern margin of the rift.
lineament and exposed a large normal fault directly beneath the slope break. The scarp face has been incised by widely spaced gullies that end at the slope break and become the loci of the heads of low-gradient alluvial fans. The topographic profile of the scarp (Fig. 9) shows that although the scarp is very gentle (maximum scarp slope angle is 5°), the scarp face is composed of two elements of subequal height. The steeper lower scarp face averages 4°–5°, compared to the gentler upper scarp face, which slopes 3°–4°. The scarp crest is fairly abrupt and gives a good vantage point that looks eastward over the graben. We noticed abundant Paleoindian stone artifacts lying on the surface at the scarp crest, as if ancient hunters had established lookouts there. We did not notice any such artifacts elsewhere on the scarp face.
WEST TRENCH
Stratigraphy
Location and Geomorphology
Parent Materials The west trench is 67 m long, up to 8 m wide, and 5.7 m deep (Fig. 10). The trench exposes Santa Fe Group (parent material unit 1) on the upthrown block and a series of eolian, colluvial, and alluvial deposits (parent materials units 11–19) on the downthrown block (Fig. 11). In addition, the entire surface here is mantled with a 1–1.5-m-thick blanket of late Pleistocene and Holocene eolian sand (parent material unit 20). The oldest deposits in the trench are beds of sandy and gravelly alluvium of the Santa Fe Group (unit 1) exposed in the footwall. We subdivided unit 1 into eight subunits (1a–1h) based on differences in grain size and sorting (see descriptions in Appendix 1). Although six of the eight subunits were
The west trench is located on the western side of the graben across the main (east-facing) fault scarp of the Calabacillas fault. The western margin scarp is composed of a 250-m-wide, low-angle scarp face fronted by a 0.6-km-wide apron of lowgradient, coalescing alluvial fans (Fig. 4). There is a fairly sharp break in slope between the scarp face and the head of the alluvial apron (Fig. 9). The steepest slope segment on the scarp face appears to be at this break in slope, and at the west trench site, a line of anomalously large shrubs trends north-south at this slope break, suggesting a linear subsurface moisture anomaly. We centered our 60-m-long trench on this slope break–vegetation
Figure 9. Topographic profile across the main scarp of the Calabacillas fault at the west trench site. Upper part shows the true-scale profile; lower part has 3× vertical exaggeration.
Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA composed of clast-supported channel facies (sandy gravels and gravelly sands), two thin beds were much finer grained (units 1e and 1f) and appeared to be floodplain overbank or marsh deposits. These extremely hard clay-rich beds apparently constitute ground-penetrating radar (GPR) “basement” on the upthrown block (described at the end of this chapter). The next youngest parent materials are units 2–4, which form westward-tapering wedges in the footwall directly west of the main normal fault (F2, Fig. 11). These three parent materials unconformably overlie unit 1 and appear to be the remains of scarp-derived colluvial wedges similar to unit 3 in the east trench. The only parts of the unit 2–4 wedges that are still preserved are “perched” in the footwall; their equivalents in the hanging wall are assumed to lie far beneath the trench floor, based on the presumed 27 m of throw on the main fault (required to create the 27 m of scarp height observed here). The next youngest parent materials in the trench are found in the hanging wall, in a fault-bounded structural block between the main normal fault (32 m H) and a zone of reverse faults (25–27 m H). We numbered these parent materials as units 11 and 12. This jump in the numbering system indicates our inference that this parent material is younger than the remnants of wedges in the footwall (units 2–4), and that an unknown number of units exists beneath the trench floor between unit 11 and the correlatives of units 2–4 in the hanging wall. We do not mean to imply that there are six of these buried units, because we have no idea how many additional colluvial/alluvial strata exist beneath the trench floor in the footwall. Our jump from single-digit unit numbers (footwall)
Figure 10. Photograph of the west trench, looking east. The logged walls are at right. The spoil pile of the east trench is visible in the distance at upper left. At upper center, three of the Albuquerque volcanoes appear on the horizon. The City of Albuquerque Soil Amendment Facility’s main shed is visible just below the right-hand volcano. This shed lies just east of the antithetic fault on the eastern margin of the graben, so a line between the east trench and the shed approximates the eastern margin of the graben. In the far distance at top are the Sandia Mountains. Photograph was taken by J.P. McCalpin on 1 June 1999.
37
to double-digit unit numbers (hanging wall) simply indicates that there are some unexposed strata between parent materials 4 and 11. Units 11 and 12 are both strongly affected by K horizon development, so parent material properties are obscured. However, the sandy texture suggests that both units represent eolianderived sands that were probably transported by sheetwash to their present location. Units 11 and 12 are overlain by three parent materials (units 13–15) that can be traced from the center to the eastern end of the trench. Unit 13 projects beneath the trench floor at 14 m H and pinches out to the west at 27 m H. Likewise, both units 14 and 15 pinch out to the west, the former at 19 m H and the latter at 12 m H. Unit 13 contains abundant floating pebbles in a massive sandy matrix. The pebbles and lack of stratification suggest that unit 13 is a colluvium derived from a scarp free face that exposed gravels in the Santa Fe Group. To expose Santa Fe gravels today in a free face would require a fault scarp higher than 3.5 m, that being the combined thickness of eolian and colluvial units now covering the Santa Fe Group at the main fault. Clearly, when the free face was created that shed unit 13 colluvium, the Santa Fe Group in the footwall was not deeply buried by an eolian blanket (units 16–20), as it is today. Unit 14 also contains rare floating pebbles, but it is much sandier and softer than unit 13. The difference in hardness results from the weak carbonate soil development in unit 14 (Ck) versus that in unit 13 (K). Unit 15 is a nearly pure medium sand of probable eolian origin. At 10 m H, unit 15 forms a funnel-shaped mass that protrudes downward through units 14 and 13. This feature looks like a sand dike caused by liquefaction. Units 13, 14, and 15 are all erosionally truncated by the basal contact of unit 16, which is marked by a prominent concentration of gravel. The overall geometry suggests that units 13–15 originally extended across the reverse fault (F1, Fig. 11) and possibly across the main normal fault, but they have subsequently been removed by erosion. Unit 16 is the only parent material in the hanging wall that is clearly a stream deposit rather than eolian-colluvial deposit. The lower one quarter to one third of the unit is a clast-supported small-medium pebble gravel, which fines slowly upward into a gravelly sand. The basal gravel is the coarsest and has clearly truncated the underlying, east-dipping strata. Between 8 m H and 25 m H, the unit thickens in the trench wall and has the appearance of a broad, flat-floored channel. Perhaps unit 16 represents an ephemeral stream that flowed parallel to the base of the fault scarp. Units 17–21 are all eolian sands that apparently blanketed the scarp and that do not contain any evidence of a scarp-derived colluvial component. The only effect of faulting on these units is to thicken them on the downthrown block. For example, unit 17 is 40 cm thick in the footwall and 80–100 cm thick in the hanging wall, and unit 18 is 20 cm thick in the footwall and 40–60 cm thick in the hanging wall. In contrast, unit 20 thickens less across the fault, from 80 cm in the footwall to 120 cm in the hanging wall. Unit 19 is a thin lens of eolian sand at the far eastern end of the trench that pinches out at 12 m H.
Figure 11. Log of the center of the west trench showing the main fault zone.
38
Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA
39
Soils
Structure
Soils in the west trench generally conform to the previously described parent materials, especially in the upper parent material units 17–21. In these eolian-blanket deposits, soil horizons parallel parent material contacts, which in turn parallel the modern ground surface. The surface soil is contained within parent material units 20 and 21 and contains eight horizons (AB, Bt1, Bt2, Btk, K1, K2, Ck1, Ck2). Parent material unit 19 contains its own soil, which is numbered “b1” (buried soil 1). Parent material units 17 and 18 each contain their own soil, composed of one Bk horizon each (unit 18Bkb2, unit 17Bkb3). We have labeled these two units as separate buried soils (b2 and b3) because there is an apparent parent material unconformity between units 17 and 18. This unconformity has two manifestations: (1) unit 17 thins in the footwall directly west of fault F2, and unit 18 correspondingly thickens, as if free face erosion had occurred between their deposition, and (2) horizon 17Bkb3 contains an extensive angular network of carbonate-filled fractures that generally do not extend upward into horizon 18Bkb2. We suspect that this network of fractures records a deformation event that affected unit 17 but not unit 18. Finally, the displacement of these two units across fault F2 is significantly different, as described in the next section. Below buried soil b3, buried soils in the footwall and hanging wall are developed in parent materials of different ages. This happens because strata are missing in the footwall due to interfaulting erosion, whereas the stratigraphic sequence is more complete in the hanging wall. For example, the fourth buried soil in the hanging wall is developed in parent material units 14–16 and contains the uppermost strong (stage III–IV) K horizon in the hanging wall (horizon 16K1b4). In contrast, this same strong K horizon in the footwall is developed on parent material unit 4 (horizon 4K1b4), which is a much older deposit. We thus infer that buried soil 4 developed when the ground surface here was underlain by unit 16 in the hanging wall and unit 4 in the footwall. In a similar manner, the fifth buried soil in the hanging wall is developed in parent material unit 13, whereas its footwall correlative is developed in parent material unit 2. However, beneath unit 13 in the hanging wall, there are two additional soils, buried soil 6 developed on unit 12 and buried soil 7 developed on unit 11, both of which contain K horizons. The time span represented by these soils (and additional soils beneath the floor of the trench) must be partly correlative to the time span represented by the soil developed in parent material unit 2 in the footwall. Accordingly, we have numbered the horizons developed on parent material unit 2 as 2K1b5-7 and 2K2b5-7, indicating that the period of soil formation probably overlapped with that of soils b5, b6, and b7 in the hanging wall. When traced west of 35 m H, parent material unit 2 thins, pinches out, and then reappears at 44 m H. In this area, buried soil 5-7 is thicker than the remaining part of unit 2, so the lower part of buried soil b5-7 extends down into parent material unit 1 (Santa Fe Group).
The west trench exposes three major deformation zones, labeled F1–F3 on Figure 11. The main normal fault zone (F2) abuts Santa Fe Group units in the footwall against colluvial units 11 and 12 in the hanging wall. The zone contains two parallel normal faults ~30–40 cm apart, and the block between these faults contains eastward-tilted beds correlative with footwall strata. The western fault has 1.4–1.6 m of throw, whereas the eastern fault has more throw and is the main fault strand. The total throw across both fault strands is probably in the tens of meters, if this is the fault responsible for creating the 27-m-high scarp of the Calabacillas fault. However, there could be additional fault strands beneath the scarp face farther west of our trench (see Fig. 9). The eastern fault plane in this zone is fronted by a 60–80-cm-wide zone of disturbed sand (unit 1d) that contains floating blocks of cohesive footwall units such as 1e and 1f. It is unclear whether this material is debris-facies colluvium, fissure fill, or merely blocks of footwall strata that have been rotated and internally deformed by normal faulting. The previous description applies only to fault zone F2 in the lowest of the three walls of the trench. In the middle wall (between the two benches shown on Fig. 11), the fault zone is completely obliterated by a 40–60-cm-wide krotovina. This krotovina is filled with reddish sand derived from units younger than unit 17. The irregular margins of the krotovina indicate that it was probably shaped by burrowing animals, but the overall vertical shape and location of the krotovinas at the fault suggest that the animals were burrowing in a fractured zone softened by faulting or fissuring. The character of F2 changes again on the uppermost trench wall, where there is a single vertical fault strand with short, westdipping faults splaying off and dying out upward into unit 20. The upper 60 cm of the fault are made up of a krotovina or fissure fill bounded by fairly linear fractures. The displacement of parent material units 17–20 across fault F2 can be measured directly because those units are preserved on both sides of the fault. For older units, we must rely on our correlation of buried soils (such as b4) across the fault. Figure 12 shows fault zone F2 and the differential displacements measured on the bottom contacts of units 20Btk, 18Bwb2, 17Bkb3, and soil b4 (developed on unit 16 in the hanging wall and unit 3 in the footwall). The vertical displacements on these contacts increase downward, from 10 cm to 40 cm, 95 cm, and 115–145 cm, respectively. We use the cumulative throw of the bottom of each soil (b2, b3, b4) as a measure of the cumulative throw of the soil, because in each case, the soil on the upthrown block was thinned by postfaulting erosion, so throws measured on the top of a given soil are minimum values. These differential displacements suggest four successive faulting events with throws of (from youngest to oldest) 10 cm, 30 cm, 55 cm, and 20–50 cm each. The second largest fault zone in the trench is fault zone F1, a zone of five faults between 25 and 27 m H (Fig. 11). These faults
40
McCalpin et al. Geochronology
Figure 12. Slip history diagram based on the four differential offsets shown in Figure 11 and the infrared-stimulated luminescence age estimates.
drop units 11 and 12 from 2 m above the trench floor to beneath the trench floor, indicating a minimum down-to-the-east throw of 2 m. The faults are either vertical or dip steeply west, with the latter having the geometry of reverse faults. However, we believe that the westward dips of these faults result from eastward toppling of step-fault blocks into the hanging wall, with the blocks farthest east having subsided the most. This structural style has been observed on other Quaternary normal faults in the United States (County Dump fault—McCalpin et al., 2006; Pajarito fault—McCalpin, 2005; Wasatch fault zone—McCalpin, 2002), in Greece (Mercier et al., 1979), and in Italy (Fucino fault—Blumetti et al., 1988). Movement on fault zone F1 has created an intermediate structural block between faults F1 and F2, which is relatively upthrown compared to the rest of the hanging wall. Faulting here must have uplifted units 13–15, after which the fluvial channel of unit 16 beveled off the top of the fault block, removing units 13–15 from most of the block. There is no evidence for movement on these faults subsequent to deposition of unit 16. The third fault zone (F3) is located in the footwall at 35–39 m H. This zone only affects the lower part of the Santa Fe Group that is exposed (units 1a–1g; Appendix 1), indicating that movement ceased before the end of Santa Fe deposition here (0.5–1 Ma?). The down-to-the-east normal faults in this zone parallel those in fault zone F2, but the total throw across the zone (based on unit 1c) is less than 50 cm.
Two samples were dated by both multi-aliquot, polymineral fine-silt thermoluminescence (TL) and infrared-stimulated luminescence (IRSL), and five samples were dated by only IRSL, at Glenn Berger’s laboratory at the Desert Research Institute, Reno, Nevada. Laboratory methods are described by Berger and Péwé (2001) and Berger and Doran (2001). The samples were collected at night to prevent sunlight exposure, to which IRSL samples are very sensitive. The samples yield a sequence of ages from 14 ka to ca. 200 ka, in correct stratigraphic order (Fig. 12; Table 2). The five IRSL ages were then used to calculate the rate of secondary carbonate accumulation with time (Table 1, three columns at far right). The advantages of constructing a graph of soil carbonate versus age are twofold: (1) It permits us to estimate the ages of all soil horizons, even those that are not amenable to luminescence dating, and (2) by comparing the proportion of carbonate in various soils, we can proportionally assign each soil a fraction of the total age of the exposure, as was done in the pioneering study of Machette (1978) on the County Dump fault. We used three methods (Table 1) to calibrate the secondary carbonate values against IRSL ages. Method 1 assumes that carbonate stopped accumulating at 14 ka, because beds younger than 14 ka (units 20Bt2 and higher) contain no carbonate. Method 2 assumes that the carbonate between unit 18Bkb2 and the ground surface accumulated at a constant rate between 32 ka and the present. Method 3 assumes that carbonate accumulated up to the present, but at variable rates between the IRSL ages of ca. 14 ka, 32.2 ka, and 55.6 ka. Because we have no reason to prefer one of the models over the other, their age range adds a “provenance error” to the analytical uncertainties in the ages of the soil horizons. In addition, the soil ages cited in the last three columns on the right of Table 1 are the ages at which soil formation began in each horizon. In other words, it is the age of the parent material at the bottom of each horizon. Figure 13 shows the increase of secondary (pedogenic) carbonate with IRSL age for soil profiles at 11 m H (soil profile 4) and at 27 m H (soil profile 2). The latter profile is missing units 13–15 (eroded off the intermediate fault block), so their carbonate amounts are missing from the total. Figure 13 shows that the average accumulation rate for secondary carbonate in the (relatively) uneroded soil profile 4 was 0.35 g/k.y. between ca. 151 and 55 ka. Between 55 and 32 ka, the accumulation rate was lower, 0.17 g/k.y. Both of these computed accumulation rates assume that no carbonate has been eroded from the hanging wall at 11 m H, and this is certainly not true for unit 15, the upper two thirds of which were eroded by unit 16 here. These accumulation rates can be compared with rates calculated at the County Dump fault (McCalpin et al., 2006) and cited by Machette (1985) for numerous locations in the western United States. At the County Dump fault, TL ages define a carbonate accumulation rate of 0.26 g/k.y. over the time period 4–82 ka, and of 0.54 g/k.y. between 82 and 293 ka. These rates can be compared to those computed
Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA
41
TABLE 2. LUMINESCENCE AGE ESTIMATES FROM THE WEST TRENCH Sample no. AWT99-1 AWT99-2
Unit
Dose rate (Gy/yr)
13K1b5 12Kb6
2.50 ± 0.11 2.75 ± 0.11
Age estimate (ka, ±1σ) and method 103 ± 15 (TL and IRSL) 196 ± 18 (TL and IRSL)
AWT99-3
12Kb6
2.54 ± 0.12
AWT99-4
13K1b5
3.00 ± 0.37
219 ± 64 (IRSL) 151 ± 35 (IRSL)
AWT99-5
16K1b4
3.123 ± 0.0095
55.6 ± 2.9 (IRSL)
AWT99-6
18Bkb2
3.474 ± 0.088
32.2 ± 1.5 (IRSL)
AWT99-7
20Bt2
3.574 ± 0.094
13.99 ± 0.67 (IRSL)
Note: TL—thermoluminescence; IRSL—infrared-stimulated luminescence.
Figure 13. Secondary CaCO3 as a function of infrared-stimulated luminescence (IRSL) age, west trench.
elsewhere in the southwestern United States (Machette, 1985), and they fall closest to cited rates from Las Cruces, New Mexico of 0.25 g/k.y. for pluvial climates in the Pleistocene, and 0.5 g/k.y. for interpluvial climates. Interpretation The four differential displacements across fault zone F1 (Fig. 12) suggest four small-displacement events with earthquake horizons at the top of soil horizons 20Btk, 18Bwb2, 17Bkb3, and buried soil 4. Using the luminescence age estimates and average accumulation rates of secondary carbonate, we can estimate the ages of these four earthquake horizons (Table 1, right-hand columns), as follows:
Event Z displaces the top of horizon 20Btk and is overlain by 20Bt2. Horizon 20Bt2 is undated; for horizon 20Btk, the top is 14.0 ka according to IRSL. Therefore, event Z has a one-sided, close maximum age constraint of ca. 14 ka. Event Y displaces soil 18 by 38 cm and is overlain by horizon 20Ck2. A close minimum age constraint on event Y is the age of the bottom of horizon 20Ck2 (19–25 ka based on carbonate content; Table 1); a rather loose maximum age constraint is the carbonate-based age of the base of horizon 18Bwb2 at 22–27 ka (Table 1). Therefore, our preferred age estimate for event Y is the average of the overlap of the minimum and maximum age ranges, i.e., ca. 23.5 ka. Event X displaces soil 17 by 95 cm and is overlain by horizon 18Bkb2. A close minimum age constraint is provided by the
42
McCalpin et al.
age of the bottom of horizon 18Bkb2, which is 32.2 ± 1.5 ka by IRSL (Table 2); a very loose maximum age constraint is provided by the age of the bottom of horizon 17Bkb3, which is 55.6 ± 2.9 ka by IRSL (Table 2). Due to the asymmetrical dating constraints, we rather arbitrarily place this event at ca. 35 ka, with large uncertainties. Event W displaces the bottom of soil “b4” by 115–145 cm (depending on whether the small footwall unit 3Bkb4? is considered to be part of soil b4) and is overlain by horizon 17Bkb3. A close minimum age constraint is provided by the bottom of horizon 17Bkb3, which is ca. 52–68 ka by carbonate content (Table 1); a close maximum age is that of the top of horizon 16K1b4, dated at 55.6 ± 2.9 ka by IRSL (Table 2). The overlap of these age ranges is ca. 52–56 ka, so our preferred age for event W is ca. 55 ka. In summary, our preferred ages for the earthquake horizons for events Z through W (i.e., the tops of horizon 20Btk, and buried soils b2, b3, and b4) are 14 ka, 23.5 ka, 35 ka, and 55 ka, respectively. These ages form the basis for the slip history diagram shown in Figure 12 (see also Table 3), in which each closed seismic cycle is named after the event that ends it. Over the latest three seismic cycles, the slip rate at this site on the Calabacillas fault has been 0.011 mm/yr (Z), 0.026 mm/yr (Y), and 0.028 mm/yr (X). These rates are comparable to the overall post-llano slip-rate estimate for the southern Calabacillas fault of 0.027–0.054 mm/yr, based on the 27-m-high scarp on a 0.5–1 Ma surface. The slightly lower slip rates of the latest three events on fault zone F1, compared to the long-term average for the southern Calabacillas fault, may have several causes. First, fault zone F1 may not have been the only fault to experience movement in these latest three events; there may be other, unanalyzed, fault strands beneath the 250-m-wide scarp face farther west of our trench. However, if such strands exist, they have not produced any visible surface relief, which suggests that their cumulative throw would be quite small. Second, event W (and presumably earlier events) produced movement on fault zone F2, so earlier events may have ruptured both faults F1 and F2, with a corresponding increase in net throw per event. Third, if our trench successfully captured the total slip in the latest three events, then it appears that slip has been progressively decreasing in the latest
three events. It may also be true that slip has been progressively decreasing on the entire Calabacillas fault from 0.5 to 1 Ma to present, such as has been observed for the Pajarito fault farther north (McCalpin, 2005). The Ground-Penetrating Radar Experiment After the trench was back filled, we measured a GPR profile adjacent to the south wall of the west trench. The purpose of this experiment was to determine whether GPR could detect faults in unconsolidated Quaternary deposits. Similar studies on strike-slip faults have been successful in detecting fault planes in unconsolidated deposits and vertical offset of reflectors (e.g., Anderson et al., 2003; Green et al., 2003; Slater and Niemi, 2003). In this experiment, we already knew the location of all faults, fissures, and stratigraphic contacts because the trench had previously been logged. Two different antennas were used, a 50 MHz antenna, which gave deeper penetration but poorer resolution, and a 100 MHz antenna. Only the 100 MHz results are described herein. The GPR profile (Fig. 14) imaged a series of reflectors down to an estimated depth of ~3.6 m, below which there were no returns. This GPR “basement” evidently represents two different strata in the footwall and hanging wall. In the hanging wall, the basement coincides with a thin bed of silty clay (unit 1e) in the upper part of the Santa Fe Group. In the footwall, GPR basement coincides with the top of a series of old (150–220 ka) buried soils (units 12, 13) overlain by weaker, younger (50 ka) soils. Thus, it appears that even thin deposits with a high percentage of clay, or thicker, dense soils with abundant calcium carbonate, can absorb GPR energy. The GPR profile also shows a prominent subvertical discontinuity in its center, correlative with the main normal fault exposed in the west trench (Fig. 14). This discontinuity shows significant vertical offset in the lower part of the imaged stratigraphic section and little or no vertical offset in the upper layers. This geometry is the same as that exposed in the trench. Without reference to the trench log, H.C. Tobin made some preliminary correlations of reflectors across the discontinuity, merely based on their similarity on the GPR profile. His upper (yellow) correlation line probably represents the top of unit 17Bkb3, which is
TABLE 3. SUMMARY OF THE FOUR LATEST SEISMIC CYCLES INTERPRETED FOR THE SOUTHERNMOST PART OF THE CALABACILLAS FAULT Seismic cycle*
Throw Approximate start of cycle Approximate end of cycle Mean length of cycle (cm) (ka) (ka) (k.y.) Z 10 23.5 14 9.5 Y 30 35 2 3. 5 1 1. 5 X 55 55 35 20 W 20–50 <<151 55 U n k n o wn *Named for the paleoearthquake that ends the cycle. † This slip rate only represents the southernmost end of the fault and may not apply to the entire fault. § NA—not applicable.
Estimated slip rate (mm/yr) 0.011 0.026 0.028 § NA
†
Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA offset 40 cm both on the GPR profile and in the trench. His lower correlation line (blue) represents very different beds in the hanging wall versus the footwall. Based on this limited experiment, it appears that GPR can image buried faults in weak late Quaternary paleosols developed on eolian sand, at least to a depth of 3.6 m. This ability is suffi-
43
cient for GPR to be a useful tool for locating trenching targets, so that trenches do not extend hundreds of feet merely “searching” for the main displacement zones. The fact that there is ambiguity about the correlation of reflectors in the subsurface is not critical, because these strata will be exposed in the trench anyway and their correlation can be determined directly.
Figure 14. Comparison of ground-penetrating radar (GPR) profile with trench log on the main scarp of the Calabacillas fault. (A) 100 MHz GPR profile. The yellow and blue lines indicate preliminary correlations made by H.C. Tobin without reference to the trench log. (B) Trench log, mirrored and stretched to match the scale of the GPR profile. Black dashed lines are benches. The yellow line from part A represents the top of the orange unit on the trench log and thus is a confirmed correlation. The blue line represents different beds on the upthrown and downthrown blocks and so represents an example of a miscorrelation of a GPR reflector. Further work is necessary to determine why such miscorrelations occur.
44
McCalpin et al.
CONCLUSIONS Our trenching of the graben across the southern Calabacillas fault was 50% successful. The paleoearthquake event history on the 5.3-m-high eastern boundary scarp (east trench) could not be reconstructed in detail, because a strong caliche soil profile had overprinted the entire 3-m-thick colluvial wedge deposit. Our interpretation is that numerous decimeter-size displacements had created this scarp, but the displacement was partitioned across several faults, so no single free face was higher than 10–20 cm. Free faces this small did not create colluvial wedges, and thus faulting did not trigger the pattern of footwall erosion and hanging-wall deposition that is required to identify individual faulting events. On the 27-m-high western main fault scarp our 60-m-long trench straddled a slope break that turned out to overlie a major normal fault; we believe that this is the main strand of the Calabacillas fault, although conceivably other strands could exist farther west. The upper four soils exposed in the trench could be correlated across the main fault and indicated per-event displacements of 10 cm, 30 cm, 55 cm, and 20–50 cm in the latest four events. We did not use these displacements to estimate the magnitude of the paleoearthquakes (e.g., Wells and Coppersmith, 1994), for two reasons: (1) We could not identify or measure individual event displacements on the antithetic fault, so we could not compute the net vertical displacement across the Calabacillas fault graben in the latest four paleoearthquakes, and (2) all displacements were measured at the extreme southern end of the fault, so they are probably less than the average displacement per event, and much less than the maximum displacement per event, that occurred during each paleoearthquake. Seven IRSL dates on eolian sands came out in correct stratigraphic order and range from 14 ± 0.7 ka at a depth of 0.5 m to 219 ± 64 ka at a depth of 5 m. Secondary calcium carbonate has accumulated in soils here at a rate of 0.17–0.35 g/k.y.. The latest four faulting events are dated at ca. 14 ka, 23.5 ka, 35 ka, and 55 ka. Thus, the displacement and recurrence times increase with increasing age, yielding relatively consistent slip rates of 0.011–0.028 m/k.y. This pattern would also result, however, if we missed evidence for some earlier events and our earlier seismic cycles were in fact composed of two events. However, that would not change the slip rate estimates. Event Z (14 ka) was a small faulting/cracking event, different than the earlier four events. This event is similar in displacement and timing to the youngest warping event interpreted for the County Dump fault (McCalpin et al., 2006). The displacements for all four paleoearthquakes measured in the west trench are smaller than those inferred on the County Dump fault, despite the fact that the length of the Calabacillas fault (40 km) is similar to that of the County Dump fault (35 km). Perhaps the displacements are smaller because the 1999 trench was located only 1 km from the southern end of the Calabacillas fault, and displacements on normal faults are typically less than average that close to the fault’s end (McCalpin, 2009; McCalpin and Slemmons,
1998). If our trenches had been located farther north toward the center of the Calabacillas fault, the displacements may have been larger; however, our target graben feature only existed at the southern end. This study documents four faulting events in the past ~55 k.y. on the Calabacillas fault, which forms the western margin of the Calabacillas subbasin of the Albuquerque–Belen Basin. In contrast, the latest Quaternary faulting on the eastern margin of the Calabacillas subbasin has been dated at 53–67 ka (Sandia fault zone; McCalpin and Harrison, 2006). Thus, in the past 55 k.y., the Calabacillas subbasin has experienced most of its surface faulting on its western margin, implying an overall tilt to the west due to faulting. In contrast, in the same time period, the rift subelements to the north (the shelf containing the Rincon fault) and to the south (shelf containing the northern Hubbell Springs fault) have experienced most of their surface faulting on the eastern margin, implying an eastward tilt. These changes in tilt direction of rift subelements over space and time in the Albuquerque–Belen Basin may represent an ongoing example of “seesaw” subsidence, as described by Smith et al. (2001) for the Santo Domingo subbasin. In that basin, Neogene basin-fill strata and the location of the axial Rio Grande channel indicate a west tilt (Miocene to early Pliocene), then an east tilt (late Pliocene and early Pleistocene), and then a west tilt (middle Pleistocene to present). This geometry suggests that the fault along which most basin subsidence was accommodated (the active basin margin) has shifted from one side of the basin to the other, with a periodicity of several million years. We hypothesize that the Calabacillas subbasin may also undergo seesaw subsidence. The rift cross section (Fig. 2) and gravity data (Fig. 3) indicate that the Calabacillas subbasin is roughly symmetrical, yet almost all the post–55 ka displacement has been on the western margin faults, and the eastern margin fault has been scarcely active in that time (McCalpin and Harrison, 2006). The most reasonable explanation is that eastern margin faults were once active and will become active again at the end of this episode of westward tilt. From a seismic hazard perspective, we would like to know how long the current episode of westward tilt might last, because during that episode, seismic hazard will be high for the western margin faults and low for the eastern margin faults, which are much closer to urban Albuquerque. In the Santo Domingo basin, there have been only three tilt episodes since ca. 7 Ma, so each one lasted 2–3 m.y. Considering that recurrence times between paleoearthquakes on the Calabacillas fault average 10–20 k.y., a 2–3 m.y. tilt episode would encompass 100–300 characteristic earthquakes. In other words, a tilt episode includes so many seismic cycles that, for estimating the probable character of the present seismic cycle, we should assume that we are in the same tilt episode as portrayed by the past few characteristic earthquakes. For the Calabacillas subbasin, this assumption indicates that seismic hazard for surface faulting in the near-term will be higher on its western margin and lower on its eastern margin.
Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA APPENDIX 1. UNIT DESCRIPTIONS, WEST TRENCH (UNIT 1 SUBUNITS ONLY)
Unit 1a—Buff gray, alternating lenses of gravelly sand and sandy gravel; for sandy gravel, clasts are 50% of volume, 10 cm maximum, 1 cm average, subangular to subround; moderately sorted; matrix is medium-coarse sand, slightly cemented; beds average 15 cm thick; crossbeds; slight CaCO3 coatings on bottoms of some clasts; for gravelly sand, clasts 10% by volume, maximum 1 cm, average 3 mm, subangular to subround; well sorted; matrix mostly medium sand; weakly cemented; crossbeds; alluvium of the Santa Fe Group (Pliocene–Pleistocene). Unit 1b—Buff to pinkish gray, gravelly sand; 5% clasts by volume, maximum 1.5 cm, average 3 mm, subangular to subround; poorly sorted; matrix is fine-medium sand; friable; no apparent stratification except for thin clay beds 1 cm thick; occasional mud balls; slight oxidation; alluvium of the Santa Fe Group (Pliocene–Pleistocene). Unit 1c—7.5YR7/3 (top) to 7.5YR7/2 (bottom) gravelly sand; clasts 5% by volume, maximum 2 cm, average 3 mm, subangular to subround; poorly sorted; matrix is fine-medium sand, slightly hard, contains Bw horizon over Bk horizon, the latter with stage I CaCO3; contains conjugate fractures; alluvium of the Santa Fe Group (Pliocene–Pleistocene). Unit 1d—10YR7/3 (salt-and-pepper) gravelly sand; clasts 5% by volume, maximum 2 cm, average 2 mm, subangular to subround; moderately sorted; matrix is fine-medium sand, slightly hard; some CaCO3 cement; contains conjugate fractures; alluvium of the Santa Fe Group (Pliocene–Pleistocene). Unit 1e—7.5YR7/4 silty clay; clasts <1% by volume, maximum 2 cm, subangular; well-sorted clay, very hard; small root casts ~1 mm thick; manganese nodules and some CaCO3 along fractures; overbank floodplain or marsh deposit of the Santa Fe Group (Pliocene–Pleistocene). Unit 1f—7.5YR8/1 silt; no clasts; very hard to extremely hard; coarsens upward to fine sand in places; sparse manganese coatings on fractures; overbank floodplain or marsh deposit of the Santa Fe Group (Pliocene–Pleistocene). Unit 1g—Buff gray gravelly sand; clasts 2%–10% by volume, maximum 10 cm, average 2 cm, subangular to subround; moderately to well sorted; matrix is fine-medium sand, slightly hard; beds 1.5–3 cm thick, large crossbeds; CaCO3 induration; contains conjugate fractures; alluvium of the Santa Fe Group (Pliocene–Pleistocene). Unit 1h—7.5YR7/3 gravelly sand with lenses of sandy gravel; for sandy gravel, clasts are 50% of volume, 10 cm maximum, 4 cm average, subangular to subround; moderately sorted; matrix is fine-coarse sand, slightly cemented; no induration but slight CaCO3 coast on bottoms of some clasts; for gravelly sand, clasts 15% by volume, maximum 3 cm, average 3 mm, subangular to subround; poorly sorted; matrix fine-medium sand; alluvium of the Santa Fe Group (Pliocene– Pleistocene).
45
ACKNOWLEDGMENTS This trenching study would not have been possible without the landowner permission granted by Matt Schmader (Open Space Division, Parks & General Services Department, City of Albuquerque) and Chuck Bowman and Steve Glass (Wastewater Division, City of Albuquerque). Trenches were logged by the senior author and Martha Eppes, Paul Wisniewski, Willow Foster, and Mike Kulikowski (all Albuquerque, New Mexico). Ian Madin of the Oregon Department of Geology & Mineral Industries (Portland, Oregon) assisted in logging the east trench. Nelia Dunbar (New Mexico Tech) examined the mysterious white carbonate clasts from the east trench. Sean Connell (New Mexico Bureau of Mines & Geology) put up with our bothersome presence at the bureau’s Albuquerque office. The study was funded by the U.S. Geological Survey under National Earthquake Hazard Reduction Program grant 99HQGR0056. We thank reviewers A.M. Michetti, E. Vittori, and L. Amoroso for their constructive comments on the manuscript. REFERENCES CITED Anderson, K.B., Spotila, J.A., and Hole, J.A., 2003, Application of geomorphic analysis and ground-penetrating radar to characterization of paleoseismic sites in dynamic alluvial environments: An example from southern California: Tectonophysics, v. 368, no. 1–4, p. 25–32, doi:10.1016/S0040-1951 (03)00148-3. Berger, G.W., and Doran, P.T., 2001, Luminescence-dating zeroing tests in Lake Hoare, Taylor Valley, Antarctica: Journal of Paleolimnology, v. 25, no. 4, p. 519–529, doi:10.1023/A:1011144502713. Berger, G.W., and Péwé, T.L., 2001, Last-Interglacial age of the Eva Forest Bed, central Alaska, from thermoluminescence dating of bracketing loess: Quaternary Science Reviews, v. 20, p. 485–498, doi:10.1016/S0277-3791 (00)00103-7. Birkeland, P.W., 1999, Soils and Geomorphology (3rd ed.): Oxford, UK, Oxford University Press, 448 p. Blumetti, A.M., Michetti, A.M., and Serva, L., 1988, The ground effects of the Fucino earthquake of January 13, 1915: An attempt for the understanding of recent geological evolution of some tectonic structure, in Margottini, C., and Serva, L., eds., Historical Seismicity of Central Eastern Mediterranean Region: Rome, Nuove Tecnologie, l’Energie e l’Ambiente, p. 297–319. Chapin, C.E., and Cather, S.M., 1994, Tectonic setting of the axial basins of the northern and central Rio Grande rift, in Keller, G.R., and Cather, S.M., eds., Basins of the Rio Grande rift—Structure, Stratigraphy, and Tectonic Setting: Geological Society of America Special Paper 291, p. 5–21. Connell, S.D., 1995, Quaternary Geology and Geomorphology of the Sandia Mountains Piedmont, Bernalillo and Sandoval Counties, Central New Mexico [M.S. thesis]: Riverside, California, University of California, 414 p. (also New Mexico Bureau of Mines & Mineral Resources Open-File Report 425, 1996). Connell, S.D., 2000, Geologic Map of the Alameda Quadrangle, Bernalillo and Sandoval Counties, New Mexico: New Mexico Bureau of Mines and Mineral Resources Open-File Report OF-DM-10, scale 1:24,000. Connell, S.D., 2001, Stratigraphy of the Albuquerque Basin, Rio Grande rift, central New Mexico; a progress report, in Connell, S.D., et al., leaders, Stratigraphy and Tectonic Development of the Albuquerque Basin, Central Rio Grande Rift: Field Trip Guidebook for the Geological Society of America, Rocky Mountain–South-Central Section Meeting, Albuquerque, New Mexico: New Mexico Bureau of Mines & Mineral Resources Open-File Report 454B, p. A-1–A-26. Connell, S.D., et al., leaders, 2001, Stratigraphy and Tectonic Development of the Albuquerque Basin, Central Rio Grande Rift: Field Trip Guidebook
46
McCalpin et al.
for the Geological Society of America, Rocky Mountain–South-Central Section Meeting, Albuquerque, New Mexico: New Mexico Bureau of Mines & Mineral Resources Open-File Report 454A, p. 1–56. Dart, C., Cohen, H.A., Akyuz, H.S., and Barka, A., 1995, Basinward migration of rift-border faults; implications for facies distributions and preservation potential: Geology, v. 23, p. 69–72, doi:10.1130/0091-7613 (1995)023<0069:BMORBF>2.3.CO;2. Gile, L.H., Peterson, F.F., and Grossman, R.B., 1966, Morphological and genetic sequences of carbonate accumulation in desert soils: Soil Science, v. 101, no. 5, p. 347–360, doi:10.1097/00010694-196605000-00001. Grauch, V.J.S., 1999, Principal features of high resolution aeromagnetic data collected near Albuquerque, New Mexico, in Pazzaglia, F.J., and Lucas, S.G., eds., Albuquerque Geology: New Mexico Geological Society Guidebook 50, p. 115–118. Grauch, V.J.S., Keller, G.R., and Gillespie, C.L., 1999, Discussion of new gravity maps for the Albuquerque Basin area, in Pazzaglia, F.J., and Lucas, S.G., eds., Albuquerque Geology: New Mexico Geological Society Guidebook 50, p. 119–124. Grauch, V.J.S., Rodriguez, B.D., and Deszcz-Pan, M., 2002, How geophysical methods have been used to understand the subsurface, in Bartolino, J.R., and Cole, J.C., eds., Ground-Water Resources of the Middle Rio Grande Basin, New Mexico: U.S. Geological Survey Circular 1222, p. 36–37. Green, A., Gross, R., Holliger, K., Horstmeyer, H., and Baldwin, J., 2003, Results of 3-D georadar surveying and trenching the San Andreas fault near its northern landward limit: Tectonophysics, v. 368, no. 1–4, p. 7–23, doi:10.1016/S0040-1951(03)00147-1. Hawley, J.W., 1996, Hydrogeologic framework of potential recharge areas in the Albuquerque Basin, central New Mexico, in Hawley, J.W., and Whitworth, T.M., compilers, Hydrogeology of Potential Recharge Areas and Hydrogeochemical Modeling of Proposed Aquifer Recharge Methods in Basin- and Valley-Fill Aquifer Systems, Albuquerque Basin, New Mexico: New Mexico Bureau of Mines & Mineral Resources and U.S. Geological Survey Open-File Report 402-D, p. 1–71. Heiken, G., Goff, F., Gardner, J.N., Baldridge, W.S., Hulen, J.B., Nielson, D.L., and Vaniman, D., 1990, The Valles/Toledo caldera complex, Jemez volcanic field, New Mexico: Annual Review of Earth and Planetary Sciences, v. 18, p. 27–53. Kelley, V.C., 1977, Geology of the Albuquerque Basin, New Mexico: New Mexico Bureau of Mines & Mineral Resources Memoir 33, 60 p. Lipman, P.W., and Mehnert, H.H., 1975, Late Cenozoic basaltic volcanism and development of the Rio Grande depression in the southern Rocky Mountains, in Curtis, B.F., ed., Cenozoic History of the Rocky Mountains: Geological Society of America Memoir 144, p. 119–154. Lozinsky, R.P., 1994, Cenozoic stratigraphy, sandstone petrology, and depositional history of the Albuquerque Basin, central New Mexico, in Keller, G.R., and Cather, S.M., eds., Basins of the Rio Grande Rift: Structure, Stratigraphy, and Tectonic Setting: Geological Society of America Special Paper 291, p. 73–81. Lozinsky, R.P., Hawley, J.W., and Love, D.W., 1991, Geologic overview and Pliocene-Quaternary history of the Albuquerque Basin, central New Mexico, in Field Guide to Geologic Excursions in New Mexico and Adjacent Areas of Texas and Colorado: New Mexico Bureau of Mines and Mineral Resources Bulletin 137, p. 157–163. Machette, M.N., 1978, Dating Quaternary faults in the southwestern United States by using buried calcareous paleosols: U.S. Geological Survey Journal of Research, v. 6, p. 369–381. Machette, M.N., 1985, Calcic soils of the southwestern United States, in Weide, D.L., ed., Soils and Quaternary Geology of the Southwestern United States: Geological Society of America Special Paper 203, p. 1–21. Machette, M.N., 1998, Contrasts between short- and long-term records of seismicity in the Rio Grande rift—Important implications for seismic-hazard assessments in areas of slow extension, in Lund, W.R., ed., Western States Seismic Policy Council, Proceedings Volume, Basin and Range Province Seismic Hazards Summit: Utah Geological Survey Miscellaneous Publication 98-2, p. 84–95. McCalpin, J.P., ed., 2009, Paleoseismology (2nd ed.): New York, Academic Press, 629 p. McCalpin, J.P., 2001, Late Quaternary faulting surrounds Albuquerque, defines the East Heights seismic gap: Geological Society of America Abstracts with Programs, v. 33, no. 5, p. A-49.
McCalpin, J.P., 2002, Post-Bonneville Paleoearthquake Chronology of the Salt Lake City Segment, Wasatch Fault Zone, from the 1999 Megatrench Site: Utah Geological Survey Miscellaneous Publication 02-7, 37 p. McCalpin, J.P., 2005, Late Quaternary activity of the Pajarito fault, Rio Grande rift of northern New Mexico, USA: Tectonophysics, v. 408, no. 1–4, p. 213–236, doi:10.1016/j.tecto.2005.05.038. McCalpin, J.P., and Berry, M.E., 1996, Soil catenas to estimate ages of movements on normal fault scarps, with an example from the Wasatch fault zone, Utah, USA: Catena, v. 27, p. 265–286, doi:10.1016/0341-8162 (96)00020-3. McCalpin, J.P., and Harrison, J.B.J., 2006, Paleoseismicity of the Sandia Fault, Albuquerque, New Mexico: Final Technical Report submitted to U.S. Geological Survey by GEO-HAZ Consulting, Inc., Contract 02HQGR0068, 14 March 2006, 38 p. McCalpin, J.P., and Slemmons, D.B., 1998, Statistics of Paleoseismic Data: Final Technical Report submitted to U.S. Geological Survey by GEOHAZ Consulting, Inc., Contract 1434-HQ-96-GR-02752, 20 March 1998, 62 p. McCalpin, J.P., Olig, S.S., Harrison, J.B.J., and Berger, G.W., 2006, Quaternary Faulting and Soil Formation on the County Dump Fault, Albuquerque, New Mexico: New Mexico Bureau of Mines & Mineral Resources Circular 212, 36 p. Mercier, J.L., Mouyaris, N., Simeakis, C., Roundoyannis, T., and Angelidhis, C., 1979, Intra-plate deformation; a quantitative study of the faults activated by the 1978 Thessaloniki earthquakes: Nature, v. 278, p. 45–48, doi:10.1038/278045a0. Personius, S.F., and Mahan, S.A., 2000, Paleoearthquake recurrence on the East Paradise fault zone, metropolitan Albuquerque, New Mexico: Bulletin of the Seismological Society of America, v. 90, p. 357–369, doi:10 .1785/0119990089. Personius, S.F., and Mahan, S.A., 2003, Paleoearthquakes and eolian-dominated fault sedimentation along the Hubbell Spring fault zone near Albuquerque, New Mexico: Bulletin of the Seismological Society of America, v. 93, p. 1355–1369, doi:10.1785/0120020031. Personius, S.F., Machette, M.N., and Kelson, K.I., 1999, Quaternary faults in the Albuquerque area—An update, in Pazzaglia, F.J., and Lucas, S.G., eds., Albuquerque Geology: Albuquerque, New Mexico Geological Society, 50th Field Conference, p. 189–200. Personius, S.F., Eppes, M.C., Mahan, S.A., Love, D.W., Mitchell, D.K., and Murphy, A., 2000, Log and data from a trench across the Hubbell Spring fault zone, Bernalillo County, New Mexico: U.S. Geological Survey Miscellaneous Field Studies Map MF-2348, version 1.1 (available at http://greenwood.cr.usgs.gov/pub/mf-maps/mf-2348/). Rosendahl, B.R., 1987, Architecture of continental rifts with special reference to East Africa: Annual Review of Earth and Planetary Sciences, v. 15, p. 445–503, doi:10.1146/annurev.ea.15.050187.002305. Russell, L.R., and Snelson, S., 1994a, Structural style and tectonic evolution of the Albuquerque Basin segment of the Rio Grande rift, New Mexico, USA, in Landon, S.M., ed., Interior Rift Basins: American Association of Petroleum Geologists Memoir 59, p. 205–258. Russell, L.R., and Snelson, S., 1994b, Structure and tectonics of the Albuquerque Basin segment of the Rio Grande rift; insights from reflection seismic data, in Keller, G.R., and Cather, S.M., eds., Basins of the Rio Grande Rift; Structure, Stratigraphy, and Tectonic Setting: Geological Society of America Special Paper 291, p. 83–112. Slater, L., and Niemi, T.M., 2003, Ground-penetrating radar investigation of active faults along the Dead Sea transform and implications for seismic hazards within the city of Aqaba, Jordan: Tectonophysics, v. 368, no. 1–4, p. 33–50, doi:10.1016/S0040-1951(03)00149-5. Smith, G.A., McIntosh, W., and Kuhle, A.J., 2001, Sedimentologic and geomorphic evidence for seesaw subsidence of the Santa Domingo accommodation-zone basin, Rio Grande rift, New Mexico: Geological Society of America Bulletin, v. 113, no. 5, p. 561–574, doi:10.1130/0016-7606 (2001)113<0561:SAGEFS>2.0.CO;2. Von Hake, C.A., 1975, Earthquake History of New Mexico: U.S. Geological Survey Earthquake Information Bulletin 7, p. 23–26. Wells, D.L., and Coppersmith, K.J., 1994, Empirical relationships among magnitude, rupture length, rupture area, and surface displacement: Bulletin of the Seismological Society of America, v. 84, p. 974–1002. MANUSCRIPT ACCEPTED BY THE SOCIETY 7 DECEMBER 2010
Printed in the USA
The Geological Society of America Special Paper 479 2011
Late Quaternary earthquakes on the Hubbell Spring fault system, New Mexico, USA: Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift Susan S. Olig* Seismic Hazards Group, URS Corporation, 1333 Broadway, Suite 800, Oakland, California 94612, USA Martha C. Eppes† Department of Earth & Planetary Science, University of New Mexico, Albuquerque, New Mexico 87131, USA Steven L. Forman Department of Geological Sciences, University of Illinois at Chicago, Chicago, Illinois 60607, USA David W. Love New Mexico Bureau of Geology and Mineral Resources, New Mexico Institute of Mining and Technology, 801 Leroy Place, Socorro, New Mexico 87801, USA Bruce D. Allen New Mexico Bureau of Geology and Mineral Resources, New Mexico Institute of Mining and Technology, 2808 Central SE, Albuquerque, New Mexico 87106, USA
ABSTRACT The Hubbell Spring fault system lies near the eastern margin of the Albuquerque– Belen Basin in the central Rio Grande rift, and it is one of the most active normal faults in the region. Recent mapping and geophysical studies indicate that fault geometry is more complex and longer than previously thought, with several significant, subparallel, anastomosing, west-dipping splays that form a broad zone as wide as 18 km and ~74 km long. We conducted a paleoseismic investigation of the previously untrenched central Hubbell Spring fault splay (splay L) at the Carrizo Spring site. Our study included mapping, trenching, drilling, and luminescence analyses. We found structural, stratigraphic, and pedologic evidence for the occurrence of at least four, possibly five, large earthquakes that occurred since deposition of piedmont deposits on the Llano de Manzano surface ca. 83.6 ± 6.0 ka. All of these events included warping
*
[email protected]. † Current address: Department of Geography and Earth Sciences, University of North Carolina at Charlotte, 9201 University City Blvd., Charlotte, North Carolina 28223, USA. Olig, S.S., Eppes, M.C., Forman, S.L., Love, D.W., and Allen, B.D., 2011, Late Quaternary earthquakes on the Hubbell Spring fault system, New Mexico, USA: Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, p. 47–77, doi:10.1130/2011.2479(02). For permission to copy, contact
[email protected]. © 2011 The Geological Society of America. All rights reserved.
47
48
Olig et al. across a broad deformation zone, whereas the three largest events also included discrete slip across four fault zones. Behavior appears noncharacteristic (i.e., highly variable slip per event), with preferred vertical displacements ranging from 0.4 to 3.7 m. The total down-to-the-west throw of piedmont deposits is 7.3 ± 1.0 m. Luminescence ages indicate that the timing of the four largest surface-deforming events on fault splay L overlaps with the timing of the four youngest faulting events from previous studies of the western Hubbell Spring fault splay (or splay J), suggesting simultaneous rupture of fault splays J and L. Displacement data and correlation of buried soils on event horizons between sites also support simultaneous rupture; however, timing constraints are on the order of thousands of years, and so triggering of events between splays cannot be precluded. The smallest warping event on fault splay L, event Y(?), does not appear to correlate to any events on splay J, suggesting that independent rupture of fault splay L also occasionally occurs. Assuming simultaneous rupture of splay J and L, the average recurrence interval over the past three complete seismic cycles is 19 (+5/−4) k.y., consistent with recurrence intervals estimated for individual cycles, which are 17 k.y., 27 k.y., and 14 k.y. We estimate an average vertical slip rate for the past four complete seismic cycles on splays J and L of ~0.2 mm/yr. In comparison, recent average late Quaternary slip rate estimates for the entire Hubbell Spring fault system range between 0.2 and 1.0 mm/yr, based on topographic profiles transecting all the splays. Slip rates for individual complete seismic cycles for splays L and J vary through time by an order of magnitude, ranging from 0.044 mm/yr to 0.46 mm/yr. This is not due to temporal clustering of earthquakes but instead is primarily due to large variations in slip per event, a finding that may have significant implications for seismic hazards elsewhere in the Rio Grande rift. Additional investigations are needed to determine the paleoseismic behavior of the many other splays of the Hubbell Spring fault system and better characterize this complex fault system for seismic hazard evaluations in the Albuquerque region.
INTRODUCTION Formerly referred to as the Hubbell Spring fault (e.g., Machette et al., 1998; Personius and Mahan, 2003), here we use the name Hubbell Spring fault system to refer to the broad zone of multiple, branching, and anastomosing late Quaternary fault strands south of Albuquerque (Fig. 1). The Hubbell Spring fault system is one of the most active fault systems in the Albuquerque–Belen Basin in central New Mexico (Personius et al., 1999; Olig and Zachariasen, 2010). It is the most significant seismic source for the southern Albuquerque area (Wong et al., 2004), particularly for the rapidly growing communities of Los Lunas and Belen. Recent mapping, geophysical, and paleoseismic studies have shed some new light on this complex fault system, but these studies have also raised significant questions about its late Quaternary behavior and earthquake potential, which are discussed further herein. This study answers some aspects of these questions through a paleoseismic trench investigation of the previously untrenched central splay (or splay L) of the Hubbell Spring fault system. Geologic Setting The Hubbell Spring fault system is a north-striking, westdipping, intrabasin zone of normal faults that lie near the east-
ern margin of the Albuquerque Basin in the central Rio Grande rift (Fig. 1). The rift is a physiographic and structural depression that consists of a series of north-trending, en echelon structural basins that are flanked by mountain ranges or uplifted plateaus (Chapin, 1971). This continental rift zone extends for ~1000 km from central Colorado, through central New Mexico, and into west Texas and Mexico (Keller and Cather, 1994). The namesake river, the Rio Grande, follows this seismically, tectonically, and volcanically active depression, which is actually part of the Basin and Range Province (Hawley, 1986). The Rio Grande rift is characterized by: (1) late Cenozoic extension accommodated by faulting and volcanism that are as young as Holocene in age; (2) shallow (≤13 km) diffuse background seismicity that generally is not associated with specific structures except for some zones that may be correlated with magmatic activity; (3) focal mechanisms that indicate a mix of normal and strike-slip faulting, and a horizontal least principal stress direction of WNWESE; (4) high heat flow; (5) deep asymmetric half grabens and grabens that tend to show opposing symmetries (tilting to the west vs. tilting to the east); and (6) large negative gravity anomalies (Chapin and Cather, 1994; Keller and Cather, 1994; Morgan et al., 1986; Sanford et al., 1991). The Albuquerque Basin is nearly 120 km long and 40–60 km wide, and it is the largest and deepest rift basin in New Mexico (Hawley et al., 1995). Clastic deposits (alluvial, colluvial, eolian,
Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift 106°37'30"W
106°30'0"W
2135
2128
20 21
FA ULT
Ri
Faults modified from Maldonado et al. (2007) after Love et al. (1996) and Love (1998)
ult
g g
2121
g g
g
NM
g
34°37'30"N
34°30'0"N
g g g
Å
ño
s
fa
ul t
US 60 Black Butte
Abo Scholle
Lo sP iñ os
Other features: Trench sites Boundaries of Tribal Lands of the Isleta Pueblo and Sevilleta Wildlife Refuge
Km
0 2 4 8 NAD 27 UTM 13N 34°15'0"N
2110
2117
I-25
34°15'0"N
Faults from the USGS Quaternary Fault and Fold Database, including: 2033 Tijeras-Canoncito fault system 2037 Sandia fault 2118 Los Pinos fault 2119 Manzano fault 2121 Intrabasin faults on the Llano de Albuquerque
34°22'30"N
2111
211
8
Sevilleta National Wildlife Refuge
Lo s
2121
Pi
34°22'30"N
Mo un ta in s
NM 304
Ä
Å
Faults from the USGS Quaternary Fault and Fold Database including McCormick Ranch (#2135) and Hubbell Spring faults (#2120) of Machette et al. (1998) and modifications to the Palace-Pipeline fault of Maldonado et al. (2007)
Other Quaternary faults:
47
Q yo
g
Ä
Man zan o
no za an M
Å
Å
Å
Contreras Cemetery fault of McCraw et al. (2006), dashed where uncertain; dotted where concealed; ticks on downthrown side
2119
Ä g
Ä
P A bo Ar ro
Suspected Quaternary fault scarp (possibly nontectonic); dashed where uncertain or inferred; dotted where concealed; ticks on downthrown side
21 19
un tain s
Mo
34°45'0"N
D E
HUBBELL
Ä Ä
Ä Ä
Ä
d
G r an
O
o
L
L L A N O
g
Ä
From Olig and Zachariasen (2010): Quaternary fault scarp, dashed where uncertain or inferred; queried where highly uncertain; dotted where concealed; ticks on downthrown side; arrows show direction of back tilt; letters (A through Q) refer to specific splays
fa
S PR ING
Ä
I G H
C
e
Bosque Peak
Ä
Ä Ä
BDE F
I-25
M A N Z A N O
Ä Ä
Ä Ä
Ä
Ä
D E
Ä
K?
J
g
Lineament (no scarp)
g
2121
L
M
34°52'30"N
Capilla Peak
Ä
g
Veguita
N
Ä
Ä
Ä Ä
Ä
Ä
MANZANO
EXPRES
Ä
Ä
F
Meadow Lake
Ä
L L A N O
Tribal Lands of Isleta Pueblo
g
g
2115
o n Wa sh
Carrizo J Spring I trench site H
g
g
Pu e r c o R io
SWAY
El Cerro Tome
A
2116
34°30'0"N
Hubbell Spring trench site
SYST E M
Ä
Ä
J
Belen
2116
34°37'30"N
40
I-25
2135 47 NM
I-25
2121
NM 31
4
2121 34°45'0"N
3 Ä
1 212 2121
He
Ca n y
l ls
Los Lunas
2121
2122
6
rr s A oy o
3 20
35°0'0"N
Tribal Lands of Isleta Pueblo
NM
Sandia fault
2121
A L B U Q U E R Q U E
2122
er a
Hubbell Spring fault system:
337
2049
2122
34°52'30"N
Ti j
Explanation
NM 333
NM
2121
35°0'0"N
I 40
Albuquerque
I4 0I
0
I4 I 40
106°22'30"W NM 14
106°45'0"W
20 37
106°52'30"W
49
Roads White lines are municipal boundaries
106°52'30"W
106°45'0"W
106°37'30"W
106°30'0"W
106°22'30"W
Figure 1. Map of the Hubbell Spring fault system (HSFS) and nearby Quaternary faults in the southern Albuquerque area, New Mexico, USA (after Olig and Zachariasen, 2010).
50
Olig et al.
lacustrine, and volcaniclastic sediments) and volcanic rocks comprise the Santa Fe Group, the Miocene–Pleistocene synrift sedimentary fill of Rio Grande rift basins (e.g., Hawley et al., 1969). These basin-fill deposits are as thick as 4570 m (~15,000 ft) in the Albuquerque Basin (Hawley et al., 1995). Although extension in the region initiated at 27–32 Ma, rift basins were not integrated by the through-going ancestral drainage of the Rio Grande until much later. The axial fluvial and tributary deposits of the ancestral Rio Grande in the Albuquerque Basin are part of the Sierra Ladrones Formation of Machette (1978), and they were deposited from 7 Ma to sometime after 1.2 Ma (Connell et al., 2001). The top of these deposits formed a Pleistocene basin floor that is now 100–200 m above the present Rio Grande, indicating substantial subsequent incision that left extensive alluvial surfaces abandoned (Machette and McGimsey, 1983; Machette, 1985). Based on recent mapping and stratigraphic studies, Connell et al. (2001) estimated that the Rio Grande initiated this incision sometime between 0.7 and 1.2 Ma. The Albuquerque Basin is flanked on the east by the easttilted, fault-block uplifts of the Sandia, Manzanita, Manzano, and Los Pinos Mountains (Kelley, 1977). These ranges expose Precambrian plutonic and metamorphic rocks that are unconformably overlain by Paleozoic limestones, sandstones, and shales. The resulting structural relief is as much as 8500 m (Woodward, 1982). The basin is flanked to the west by the lower-relief uplifts of the Colorado Plateau. Based on seismic lines, drill holes, and gravity data, Lozinsky (1994) and Russell and Snelson (1994) separated the Albuquerque Basin into two subbasins: one north of Tijeras Arroyo and one south of Los Lunas. Based on gravity data, Grauch (1999) further separated the Albuquerque Basin into three subbasins: the Santo Domingo, Calabacillas, and Belen. The Hubbell Spring fault system transects both the Belen and Calabacillas subbasins. It lies well within these subbasins, located ~3–10 km west of the range-bounding Manzano and Los Pinos faults (Fig. 1). The Hubbell Spring fault system cuts the TijerasCañoncito fault system near its northern end and is along strike of the less-active Sandia fault to the north (Fig. 1). Even though the Hubbell Spring fault system is an intrabasin structure, its prominent geomorphic expression, structural relief, and relation to adjacent faults suggest that it forms the active rift margin (Machette and McGimsey, 1983; Olig and Zachariasen, 2010). Permian, Triassic, and Tertiary rocks are exposed in some portions of the footwall of the Hubbell Spring fault system, indicating unusually substantial structural relief across this intrabasin fault zone (Reiche, 1949; Kelley, 1977; Love et al., 1996). Recent mapping indicates that the Hubbell Spring fault system is characterized by prominent late Quaternary fault scarps that extend for 74 km along the Llano de Manzano (Olig and Zachariasen, 2010). The Llano de Manzano is an early to late Pleistocene alluvial surface that extends for over 90 km south of Albuquerque between the Rio Grande to the west and the Manzano Mountains to the east (Fig. 1). This gently west-sloping surface was considered by Machette (1985) to be graded to an alluvial terrace that lies
92–113 m above the modern Rio Grande. Based on soil studies, and geomorphic and stratigraphic relations, he estimated an age on the order of ca. 300 ka. Based on more recent detailed mapping, and stratigraphic studies, Maldonado et al. (1999) broke out two additional older surfaces, the Sunport and Cañada Colorado, north of Hells Canyon Wash. They estimated a Pliocene age for the Cañada Colorado and early Pleistocene ages for the Sunport and Llano de Manzano surfaces. As an extension of these studies, Connell et al. (2001) considered the Llano de Manzano to be a complex surface of a basin-fill succession that includes middle Pleistocene piedmont deposits shed off the Manzano Mountains, overlying and truncating axial Rio Grande deposits. They provisionally assigned these deposits to the Sierra Ladrones Formation of Machette (1978). Eolian cover sands blanket much of the Llano de Manzano and other surfaces in the region. These eolian sands are particularly significant to fault studies because they tend to dominate over colluvial sedimentation along faults (e.g., Personius and Mahan, 2003), they complicate age estimates of the many alluvial surfaces in the region, they mute the geomorphic expression of faults, and they are excellent candidates for luminescence dating. Despite the eolian cover sands, scarps of the Hubbell Spring fault system are as high as 40 m on the Pliocene Cañada Colorado (Machette and McGimsey, 1983), and as high as 31 m on the Pleistocene Llano de Manzano (Olig and Zachariasen, 2010). Previous Work and Unresolved Issues Read et al. (1944) first mapped the Hubbell Spring fault system and named it the Ojuelos fault for Los Ojuelos Springs on the central Hubbell Spring fault splay ~4 km south of our trench site. However, several subsequent researchers (e.g., Kelley, 1977; Machette and McGimsey, 1983; Machette et al., 1998; Maldonado et al., 1999; Personius and Mahan, 2003) used the name Hubbell Spring fault, and we retain that nomenclature here but add the term “system” to emphasize the existence of multiple, subparallel, significantly long fault splays that comprise this complex structure. Machette and McGimsey (1983, p. 6) originally mapped, profiled, and analyzed fault scarps of the Hubbell Spring fault system, describing them as “perhaps the most spectacular fault scarps in the central Rio Grande rift.” They mapped a complex, anastomosing series of three dominant splays that converged to the north and had a total length of 43 km, with only the central trace extending south of Los Lunas. They measured scarp heights of 2–40 m, with larger offsets on older surfaces clearly indicating recurrent Quaternary movements. Based on differences in scarp morphologies, they concluded movement was younger to the south, as originally suggested by Reiche (1949), and they broke the fault into northern and southern segments. Based on comparison to 5 ka and 15 ka scarps studied elsewhere in the Basin and Range Province, Machette and McGimsey (1983) concluded that youngest faulting on the Hubbell Spring fault system was late Pleistocene, but probably considerably older than 15 ka.
Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift Subsequent detailed mapping of late Cenozoic sediments along the northern Hubbell Spring fault system in the Pueblo of Isleta Tribal Lands (Love et al., 1996; Love, 1998; Maldonado et al., 1999, 2007) indicated that the fault geometry is more complex and some traces are much longer than previously mapped by Machette and McGimsey (1983). In particular, Maldonado et al. (1999) mapped multiple, anastomosing fault traces that form three dominant north-south–trending fault splays, which they named western, central, and eastern. Machette and McGimsey (1983) showed the western and eastern traces dying out northeast of Los Lunas, whereas Maldonado et al. (1999) extended these traces at least another 15 km to the south, in part based on their prominent aeromagnetic signature (Fig. 2). Subsequently, Maldonado et al. (2007) reinterpreted the southern portion of the eastern splay (south of Hells Canyon Wash) as a separate buried fault that they named the Meadow Lake fault. The Meadow Lake fault has no geomorphic expression (Olig and Zachariasen, 2010), shows down-to-the-east offset in the subsurface, and has no associated gravity anomaly. It may actually be a relict fault from a pre-Quaternary period of extension, and it is not included on Figure 1 as part of the Hubbell Spring fault system. Most recently, mapping and profiling of Quaternary fault scarps on the Llano de Manzano south of the Isleta Pueblo indicated that the Hubbell Spring fault system is even more extensive and complex than previously believed, with prominent scarps extending to south of Abo Arroyo and Black Butte (Olig and Zachariasen, 2010). Several subparallel, anastomosing and branching fault splays form a broad deformation zone as wide as 18 km and as long as 74 km, which shows overall offset of late Quaternary sediments down to the west (Fig. 1). As mapped by Olig and Zachariasen (2010), the Hubbell Spring fault system includes the Palace-Pipeline fault and McCormick Ranch faults of Maldonado et al. (2007), the unnamed faults on the Llano de Manzano of Machette and McGimsey (1983), and the Contreras Cemetery fault of McCraw et al. (2006). Due to the complexity, later researchers have used letters A through Q to designate fault splays, so that the central splay of Maldonado et al. (2007) is splay L, the western splay of Maldonado et al. (2007) is splay J, and the Palace-Pipeline fault of Maldonado et al. (2007) is splay H. We follow this nomenclature here. From topographic transects across the entire system, Olig and Zachariasen (2010) estimated cumulative vertical surface displacements of 28–83 m and average late Quaternary slip rates of 0.2–1.0 mm/yr. The large uncertainties result from inclusion (in maximum estimates) or exclusion (in minimum estimates) of the unusual back tilting observed in the footwall of many of the Hubbell Spring fault splays. In addition to being longer, a previous paleoseismic study of the western Hubbell Spring fault splay (splay J) indicates that it is younger than previously recognized. A trench investigation of a 7-m-high scarp on splay J near Hubbell Spring at its northern end (Fig. 1) revealed evidence for four surface-faulting events that occurred since deposition of fan deposits on the Llano de Manzano (Personius et al., 2001), probably around 92 ± 7 ka (Personius and Mahan, 2003). These events resulted in 5–8 m of throw
51
and average displacements per event of 1–2 m (Personius et al., 2001). The luminescence ages of colluvial/eolian deposits associated with faulting indicate that the most recent, penultimate, and antepenultimate events occurred around 12 ± 1 ka, 29 ± 3 ka, and 56 ± 6 ka, respectively (Personius and Mahan, 2003). The age of the oldest event is poorly constrained, but it occurred prior to the antepenultimate event and sometime after 92 ± 7 ka. The relatively large per event displacements support either longer rupture lengths for splay J, as mapped by Olig and Zachariasen (2010), or simultaneous rupture with the other splays of the Hubbell Spring fault system, or possibly both. These recent studies raise several questions about the earthquake potential and rupture behavior of the Hubbell Spring fault system, including: (1) which is the most dominant and active fault splay; (2) do splays rupture together or independently; and (3) is the southern portion more recently active than the northern portion, as suggested by Reiche (1949) and Machette and McGimsey (1983)? Prior to the mapping of Olig and Zachariasen (2010), previous studies suggested that the central Hubbell Spring fault (splay L) forms the active margin along this portion of the Rio Grande rift (Machette et al., 1998). Overall, splay L does have significant structural relief and the most prominent geomorphic expression on the Llano de Manzano, forming the Hubbell bench, a higher portion of the Llano de Manzano surface on the upthrown side of the fault. However, this may be partly due to the fact that splay L is the only trace with pre-Tertiary bedrock exposed in the footwall along much of its length, and thus its prominent geomorphic expression may not necessarily be indicative of the greatest late Quaternary rate of activity. Splay L is also most closely associated with a strong north-south–trending gravity gradient along the Llano de Manzano, although gravity and aeromagnetic modeling suggest that the major basement offset appears to be just over 2 km west of splay L, but still east of splay J (Grauch and Hudson, 2002). Despite its prominence, very little is known about the paleoseismic behavior of fault splay L of the Hubbell Spring fault system. Purpose and Scope The purpose of this study was to develop a better understanding of the late Quaternary paleoseismicity of the central Hubbell Spring fault splay (splay L) through a detailed trench investigation. We then wanted to compare results from our study with those of the previous investigation of the western Hubbell Spring fault splay (splay J) by Personius et al. (2001) and Personius and Mahan (2003) to better understand the earthquake potential of the Hubbell Spring fault system. Our study included: (1) interpretation of black-and-white stereo aerial photographs of the trench site at different scales (~1:41,000 scale 1996, and ~1:52,000 scale 1953); (2) detailed mapping of the surficial geology at the trench site; (3) topographic profiling of fault scarps; (4) excavation, interpretation, and logging of trench and soil pit exposures; (5) description of soil profiles and lithologic units; (6) luminescence analyses of samples to determine numerical
52
Olig et al.
w
E
N
W
C
L
−
Figure 2. Shaded-relief aeromagnetic image of a portion of the Hubbell Spring fault system, illuminated from the west (after Grauch, 2001).
Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift ages; and (7) drilling, logging, and interpretation of three shallow boreholes at the trench site. INVESTIGATION OF THE CARRIZO SPRING TRENCH SITE ON THE CENTRAL HUBBELL SPRING FAULT (SPLAY L OF THE HUBBELL SPRING FAULT SYSTEM) Surficial Geology The Carrizo Spring trench site is located near the alongstrike midpoint of fault splay L, ~16 km south-southeast of Los Lunas (Fig. 1), and 1 km north of Carrizo Spring (Fig. 3). The site is ~11 km due west of Bosque Peak, which, at 2929 m, is the third highest peak in the Manzano Mountains. In contrast, the site is at ~1646 m elevation along the Llano de Manzano surface, which is ~175 m above the Rio Grande. At this latitude, Olig and Zachariasen (2010) mapped six additional fault splays: splays F, H, I, and J, west of splay L, and splays M and N, east of splay L (Fig. 1). The closest splay, splay M, is roughly 350 m from splay L but shows little geomorphic expression and is characterized as a tonal lineament at the latitude of the trench site. Near the trench site, splay L is marked by an alignment of springs along a well-defined but broad, single, simple, westfacing scarp on the Llano de Manzano (Fig. 3). There is no evidence for any antithetic faults or back tilting on the surface along this section of the fault. Scarp heights range from 3 to 15 m, generally decreasing to the north near the town of Meadow Lake. To the south, splay L continues as a single scarp, increasing in height and becoming more dissected as it transects the Tome Land Grant, and eventually (roughly 5 km south) exposes Triassic and Permian sedimentary rocks in the footwall (Rawling and McCraw, 2004), becoming a bedrock-alluvium fault contact. Directly to the north and south of the trench site, ephemeral drainages have incised 1–11 m into the Llano de Manzano, generally showing greater incision on the upthrown side of fault splay L. For example, the drainage along Maes Spring is the largest locally (Fig. 3) and is incised 7.6–10.7 m into the footwall, versus 4.6–6.1 m into the hanging wall. However, all of the local drainages are relatively small and are not incised extensively. Thus, they appear to be graded locally to the Llano de Manzano and not to the Rio Grande (Connell et al., 2001). Some very small drainages have formed small Holocene fans at the base of the central Hubbell Spring fault scarp, such as the small drainage and fan directly south of the trench site (Fig. 3). Overall, the Llano de Manzano in the area is underlain by late Quaternary piedmont alluvium shed off the Manzano Mountains, and it is blanketed by a thin cover of eolian sand, creating a remarkably uniform surface (except for drainages and fault scarps) that slopes gently (2° to 4°) westward. The dominant wind direction is from the southwest, and eolian deposits have built up to form small local dunes (unit He on Fig. 3), particularly where deposits have banked up against fault scarps such as at the trench site. Based on age analyses from the trench (discussed in the following sec-
53
tions), these loose eolian sands likely span a range of ages, from mid-Holocene to modern. Deposits not shown on Figure 3, but which turned out to be important to this study, are small, localized spring deposits along the fault. These are visible as lightcolored patches of concentrated carbonate on the surface, and they are most prominent around Carrizo and Maes Springs, but notably small patches are visible in a gully just north of the trench site and also exposed in the drainage to the south. The Llano de Manzano provides a good datum on which to measure long-term late Quaternary offsets from topographic profiles. A long topographic profile (P1 on Fig. 3) measured across fault splays L and M at the trench site yielded a net vertical tectonic displacement of 7.5 ± 1.0 m down-to-the-west (Figs. 4 and 5D) and a maximum scarp angle of 12°. At this location, there is no evidence for antithetic fault scarps or back tilting of the hanging-wall surface toward the fault. The scarp crest is very broad and is located ~35 m east of the maximum scarp angle, which forms the only bevel on the scarp face. No net offset was apparent across either the lineament associated with splay M or across two small swales occupied by ephemeral drainages, located ~115 and 220 m east of the scarp crest of splay L (Fig. 5D). Another scarp profile measured at our alternate trench site about ¾ km to the south (P2 on Fig. 3) yielded a net vertical tectonic displacement of 7.0 ± 1 m down-to-the-west. Here, eolian dune sand is banked up over the scarp crest. The scarp crest is still broad, but the profile shows a double bevel and a maximum scarp angle of only 8°. Subsurface Investigations We excavated one trench across fault splay L (Figs. 3, 4, and 5A–5D) and one soil pit located ~43 m west of the trench (Figs. 5D and 6). We also augered three shallow boreholes in the hanging wall of the fault (B1, B2, and B3 on Fig. 5D). The trench was over 60 m long and 4½ m deep. The soil pit was nearly 3 m deep and ~5 m long. Both the trench and soil pit were excavated with a rubber-tire backhoe using a 3-ft-wide (0.9 m) bucket. Walls were scraped and cleaned to remove bucket smear. The trench was originally logged at a scale of 1 inch = 1 m (~1:40 scale) on a planimetric grid (Figs. 5A–5C), whereas only a profile was logged for the soil pit (Fig. 6). In the trench, we strung level lines and marked stations at 1 m intervals to provide reference lines. Locations of samples, faults, and stratigraphic and pedologic contacts were marked with nails and/or spray paint and were measured relative to a level line to the nearest centimeter. Total errors of measured points on the logs are estimated to be ≤5 cm. Original trench logs were then simplified and reduced during drafting for publication (Figs. 5A–5C). Boreholes were between 5.7 (B3) and 10.4 m (B1) deep. Boreholes were augered with a SIMCO 2800 HS drill rig provided and operated by the New Mexico Bureau of Geology and Mineral Resources. Samples were continuously collected on the 4-inch-diameter (10.2 cm) auger stems. Except for some sloughing in limited zones, this method worked relatively well for holes B1 and B2. Unfortunately, challenges with keeping
54
Olig et al.
Figure 3. Surficial geology of the Carrizo Spring trench site, including fault splays L, M, and N, and associated lineaments of the Hubbell Spring fault system.
M L
the hole vertical for B3 resulted in significant sample disturbance for much of the hole, making stratigraphic interpretations from B3 less reliable. We collected 11 samples from the trench for luminescence analyses to provide numerical age constraints for faulting events (Figs. 5A and 5B; Table 1). Application of thermoluminescence dating of sediments in paleoseismic studies of normal faults began in the 1980s (e.g., Forman et al., 1988). Subsequent devel-
N
opments in the past 20 yr have broadened its use and generally reduced errors, but a good understanding of the stratigraphic context of samples (such as depositional setting, pedogenesis, and bioturbation) is still critical to successful application. Thermoluminescence is the release of light when mineral grains are heated above 150 °C, and sediments acquire thermoluminescence from background radiation. Thermoluminescence in sediments increases steadily with time, and age estimates are made
Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift
Figure 4. Photograph looking southeast at the fault scarp of Splay L and the Carrizo Spring trench site, with the Manzano Mountains in the background. The fault scarp is nearly 8 m high.
by determining the ratio of the equivalent dose (proportional to the luminescence signal accumulated since burial) to the dose rate (or background radiation at the sample site) (for further discussion, see Forman et al., 1999). Thus, analyses provide the time since deposition and burial. Thermoluminescence is released during exposure to sunlight. Therefore, during transport, the “thermoluminescence clock” is reset if enough sunlight reaches individual grains, such as for eolian deposits and finegrained slope wash. These types of deposits are good candidates for luminescence dating. Additionally, the development of infrared- or optically stimulated luminescence (IRSL or OSL), which measures luminescence of the infrared or another portion of the light spectrum, has some significant advantages for paleoseismic applications (Spooner et al., 1990; Forman, 1999). Measuring this portion of the spectrum generally provides smaller errors and broader applications to a greater variety of depositional environments because this portion of the spectrum is zeroed or reset more quickly. For example, in a comparison of traditional thermoluminescence analyses and IRSL analyses of a modern dune near the Hubbell Spring trench, Personius and Mahan (2003) found that the IRSL analyses had smaller errors and an order-of-magnitude smaller residual or inherited signal (~300 vs. 2000 yr) than the thermoluminescence analyses. Samples CHSF02-9 and CHSF02-11 were too coarse for IRSL, and so we used green OSL analyses on these samples. All our other analyses were IRSL (Table 1). Our methods of collection and analyses are described in Olig et al. (2004). Errors of 1σ are reported for all ages. Trench Exposure The trench across splay L of the Hubbell Spring fault system exposed piedmont alluvium, slope-wash colluvium, sag pond deposits, and eolian sands that included several buried soils throughout the section (Figs. 5A–5D). A broad deformation zone consisting of fractures, fault zones (FZ1, FZ2, FZ3, FZ4a, and
55
FZ4b on Fig. 5B), and warping offsets these deposits down to the west. Several of the stratigraphic units and buried soils that were exposed in the footwall could be traced across the deformation zone and into the hanging wall. However, units and, particularly, soil horizons were generally thicker in the hanging wall. Also, some units and soils were partially or entirely eroded away in the area of maximum deformation, and some soil catenas showed dramatic differences in properties downslope. In addition, the trench and soil pit were not deep enough to expose the oldest buried soil (S1) on the downthrown side of the fault. Therefore, we drilled three shallow borings (B1, B2, and B3) in the hanging wall to measure the throw on this oldest buried soil. Despite all of the complexities, the trench, soil pit, and boreholes revealed stratigraphic, pedologic, and structural evidence for the occurrence of at least four, probably five, large earthquakes on splay L of the Hubbell Spring fault system. All of these deformation events included warping down to the west; however, only the three largest events showed definitive discrete slip across faults. The evidence for all events, along with their timing, is described in detail in the following sections. We refer to these events with reversed alphabetical labels, where event Z is the youngest and event V is the oldest. Trench Stratigraphy We logged 14 stratigraphic units in the trench exposure, with unit 1 being the oldest and unit 14 the youngest. Abbreviated unit descriptions are shown on Figure 5C, and detailed descriptions are given in Appendix A of Olig et al. (2004). The stratigraphic units included seven buried soils (S1 through S7), and the tops of these catenas are shown on Figures 5A–5D. Detailed soil profile descriptions for several locations (Figs. 5A–5C) are tabulated in the Appendix (Table A1). The following descriptions summarize characteristics relevant to deciphering the faulting history. Unit 1 consists of older piedmont deposits of the Llano de Manzano and is the oldest unit exposed in the trench. It is clearly warped and faulted and was only exposed east of station (st.) 41 m in the trench (Figs. 5A and 5B). It includes three subunits (1a, 1b, and 1c) as well as a stage II buried soil (S1). Unit 1a consists of fluvial channel gravels interbedded with unit 1b, and it was only exposed locally midtrench near the deformation zone. Clasts are dominantly pebble sized, but they can range as large as boulders. They are subangular to rounded and dominantly composed of granitic, gneissic, greenstone, and other metamorphic rocks, with rare limestone and sandstone clasts. Smaller granitic clasts are generally grussified (crumbling to the touch), and some of the metamorphic clasts are also extremely weathered to clays, so that we could slice through them with a soils knife. Clasts are stratified and weakly imbricated to the northeast. Based on its stratigraphic characteristics, we interpret unit 1a to have been deposited by streams draining the Manzano Mountains east of the trench site, where dominantly Precambrian granitic and metamorphic rocks are exposed at the base of the range front, with Paleozoic limestones exposed near the tops of peaks (e.g., Bosque Peak; Karlstrom et al., 2001).
-5
-4
-3
-2
-1
0
S 3a -Btk horizon
S5-Stage g III
5
-6
-5
-4
-3
-2
-1
K
Modern S Mod soilil
1
d
16
2
17
3
4
5
6
5
18
4
20
7
5
DISTANCE T (m)
21
22
CHSF02-6 C HS
K CHSF02-7 Increasing MnO increasi
S1—Stage II carbonate nodular soil
19
K
Mesquite on crest of scarp
4
23
1c
8
Roots
S4—B Btjtj horizo on
10
11
Bedding: 281° 3°W
?
14
25
26
27
S1
4
K
28
f ine sand F
15
29
1a
30
? with no CO3 h fine Greenish sand no CO3
3a
8
1c
S2
Contact: 190° 10°W W
10
14
A
S3b
S5—Stage Stag Stag ge III 8
13
Coarse sand bed
12
Contact: 186° 2°W
3b
Contact: 323° 4°W
6
*S1 and S2 carbonate appears dissolved west of Carbonate-filled st. 29 and st. 28 m, fractures respectively
decreasing Decreasing nodules*
Decreasing nodules*
24
7
9
S7
Buried soils on units 1 and 3 welded to together to east
S3a inferred to correlate with S3b
Soil Profile CS5 88–280 cm depth
6
10
S2—Stage II carbonate nodular soil
Base of Btt b
S7-Stage II nodular d a soil s
Contact: 56° 7°W
horizon orizon orizon S3b-Bt ho
0
K
Soil Profile CS2 (continued) 170–280 cm depth Stage g II
Logged by S. Olig, M. Eppes, N. Bailey, and A. Tillery 5-22-02 through 06-04-02, 10-21-02
14
?
DIS T ANCE (m)
Figure 5. Logs of (A) stations 0–30 m, (B) stations 30–45 m, and (C) stations 45–60 m of the Carrizo Spring trench on splay L of the Hubbell Spring fault system. The top of colored buried soil horizons (S1—blue, S2—green, S3—pink, S5—orange, and S7—brown) correspond to event horizons for paleoearthquakes V, W, X, Y(?), and Z, respectively. See part C for explanation of units and symbols. See Appendix for detailed soil descriptions.
DIS T ANCE (m)
Soil Profile CS2 0–170 cm depth
Matchline to Figure 5B
Azimuth: 277º
56
DISTANCE (m)
-9
-8
-7
-6
-5
-4
-3
-2
East
6
S3bb
7
10
1a
K
1b
5
3 3a
31
9
1b
Top of T
77 cm dtw
32
3a
33
34
1 10
3a
n
Network of carbonate-filled vertical fractures and gp bedding-parallel fractures
4
5
S6
Stag e II
horizo
S7 –
il
S5 –K
Modern so
Lose Unit 6 into base of K horizon
1a
3a
35
Top of T
Top of T
4
5
DISTANCE (m)
37
2
?
38
39
3a
FZ4a: 192° 87°E no apparent offset f on top of 5 top of 2 offset f 6 cm dtw
1b b
2
?
FZ3: 167° 70–85°W 19 cm dtw dip slip on top of (on north wall is 25 cm dtw)
1b b
9
10
12
S3c becomes ?indistinct from S 5
?
?
?
?
?
K K
?
blocks of 10
Open fractures
8 S3c
K
K
40
41
42
9 S5
S7
CHSF02-8 8
10
cooa C a rs seer ssa annd 13
14
B
K
43
4 and 5 ?
K
44
2
K?
4 or sandier
S3c
?
Friable punky zone in carbonate
S6
11 CHSF02-9 K
CHSF02-11
CHSF02-10
Large mesquite roots
FZ4b: 12 to 15 cm dtw dip slip on base of 10 warping dtw in underlying units, but no apparent discrete slip
CHSF02-1 4 CHSF02-2
Soil Profile CS4 0 to 397cm depth Surface disturbed by backhoe
Figure 5. (Continued.)
36
42 cm dtw
15–20 cm dtw
FZ2: 351° to 11° (splays into 3 strands on north wall) FZ2 To T tal dip slip:
1a
FZ1 To T tal dip slip: Top of 3a 45–50 cm dtw (down-to-west) T Top of 3a 43–56 cm dtw T
FZ1 main fault: 150° ± 5 ° 48°–52°W FZ1 splay fault: 145° 65°–73°W
CHSF02-4 SF02-4 0
4
CHSF02-5 C 0
Coarser
6
Contact: t 195° 12°W
1b 1a 1a cCoarser sand 1b
1b
C CH CHSF02-3
Carbonatet filled 3 3a fracturess
4
14
30
S 4— Btj horizon
Matchline to Figure 5A
Top of S3b T is stripped to west and bottom becomes indistinct
M tchline Ma neFigure to Figure e 5c Matchline to 5C 2
57
Matchline to Figure 5B
14
46
Coarse sand
Punky zone in carbonate
K?
Buried soil (S1 is oldest, S7 is youngest)
Luminescence sample location
Fine laminations
4?
K
9
3b?
47
Pink sand pin d
S6
Modern soil
4
and
8
48
49
5
50
51
S3c
53
Figure 5. (Continued.)
54
S2? Coarser sand
55
56
57
5 Red to p pink med to fine sand. T Top marked by buried soil, S3
4 Red coarser sand—slope-wash and eolian sediments
Younger g piedmont deposits of Llano de 3 Y Manzano 3a—Pink fine sand with no carbonate in matrix 3b—Includes buried soil, S2, with a stage II nodular horizon on unit 3a
2 Playa deposit of dark-brown sandy clay
58
59
60
14
Youngest g eolian sand with 14 Y modern soil
C -9
-8
-7
-6
-5
-4
g y coarse sand, slopewash, 13 Brown-gray in west end only
g silty y sand, slope-wash 12 Light-brown and eolian deposit, p onlyy locallyy present near deformation zone
11 Scarp colluvium and eolian sand
10 Siltyy sand that includes yyoungest buried soil, S7, has stage g II morphology, p gy extends most of trench but thickens to west downslope of deformation zone
y y sand with buried Btk 9 Siltyy clayey horizon, S6, present in west end only
8 Includes buried soil, S5, characterized byy K horizon that is eroded near scarp p crest and thickens west of deformation zone, including g a friable zone of punky carbonate nodules
7 Pink eolian cover sand with weak Btj horizon, S4
f eolian cover sand - well-sorted, 6 Buff silty sand with no carbonate, reduced ?
Stratigraphic units 1 Older piedmont deposits of Llano de Manzano 1a—Channels of stream g gravel in unit 1b. Only locally present near midtrench. 1b—Greenish and pink p sand with lenses of fine g gravels and discontinuous silty, clayey y y beds. Interbedded fluvial, slope-wash p and eolian deposi p ts with the latter increasing g upward. p Only y locallyy distinguishable g from unit 1c near midtrench. 1c—Includes oldest buried soil, S1, developed p on Unit 1b in eastern half of trench
DISTANCE T (m)
52
4 4? 3b?
( (Indistinguishable g b due to soil development of S3c) Coarser ppebbly y sand nd may correlate to 4
13 10
Bedding-parallel, gp carbonate-filled fractures, no apparent offset f
High-angle g g fracture, dashed where approximately pp located or inferred, no ap pp parent offset, f usually carbonate-filled
Fault, dashed where approximately pp y located, or inferred, arrows show direction of movement
Top of buried Bt horizon T
Base of buried carbonate soil
Top T p of buried carbonate soil, dashed where inferred or diffuse f
Soil Profile CS1 0 to 370 cm depth
S2?—Mottled Btk horizon with carbonate overprinting p g redox features
S5
S7
Soil Profile CS1 ((ctd.)) 370 to 430 cm depth
Logged by S. Olig, M. Eppes and A. Tillery, 5-29-02 through 06-04-02, 10-21-02
S1
CHSF02-1
?
Carbonate-cemented krotovina
K
Stratigraphic g contact, dashed where gradational g or approximate, pp queried where inferred
Filled burrow (krotovina)
K
Symbols
Explanation
58
DISTANCE T (m)
0
S7
0m
S3
East 275°
S2
50
Lineament of fault splay M
400
S1
Mesquite on crest of scarp
100
Jog to south
Trench
250
Swale
300
450
B1
S3 S7 S2 S1
350
Carbonate nodules weathering out on surface
DISTANCE ALONG SCARP PROFILE (m)
FZ1 through FZ4
150 200 DISTANCE (m)
Swale
SCARP PROFILE P1
450
400
B3
B1
Trench
0 B3
Average slope of projected Llano de Net vertical Manzano tectonic displacement: surface: 2°W 7.5 ± 1.0 m
500
B2
Soil pit
B2
Fence
West
S2 S1
D
Figure 5. (D) Topographic profile and insert schematically showing trench and borehole locations with buried soil horizons S1, S2, S3, and S7, which respectively correspond to event horizons for paleoearthquakes V, W, X, and Z. Buried soil S3 (the event horizon for the possible paleoearthquake Y[?]) is not shown at this scale for clarity. Bt—soil horizon with accumulation of clay; Btj—juvenile Bt horizon; Btk—soil horizon with accumulation of clay and carbonates.
No vertical exaggeration
Fault splay L
59
60
Olig et al.
C
Btjk Horizon
C
[ ]
Figure 6. Log of the south wall of the soil pit at the Carrizo Spring site. Abbreviations with “e” describe sediment reaction to 10% hydrochloric acid solution (scale is from e– [effervesces weakly] to ev++ [effervesces violently]). Bk—soil horizon with accumulation of carbonates; Btjk—soil horizon with accumulation of carbonates and minor clay.
A
G
Unit 1b is a greenish to pinkish sand with discontinuous pebbly, silty, or clayey lenses. It is weakly stratified, overall fining upward, and is interpreted to be interbedded overbank, slope-wash, and eolian deposits. This subunit is stratigraphically continuous with unit 1c, but it is distinguished by the lack of carbonate nodules characteristic of the buried soil S1, developed in unit 1c east of st. 29 m. Indeed the carbonate in S1 appeared to be dissolved from unit 1b, as evidenced by “ghost” nodules (nodular-shaped zones of silt that did not react to HCl) between st. 28 and 30 m, and the intense amount of manganese oxide and limonitic staining between st. 28 and st. 40 m. Additionally, although fractures and warping are evident where S1 dies out near
st. 29 m, no stratigraphic offsets or significant shearing are evident in units 1a or 1b, so S1 is not cut out by faulting. However, some of the soil carbonate does appear to have been remobilized and precipitated into fracture networks within the deformation zone. These fracture networks are characterized by numerous vertical and bedding-parallel fractures similar to those observed by Personius et al. (2001) at the Hubbell Spring trench site. Two samples from unit 1b (CHSF02-3 and CHSF02-4 on Fig. 5B) yielded an average IRSL age of 83.6 ± 6.0 ka (Table 1). These ages indicate that either the Llano de Manzano surface is much younger than previously thought, or, more likely, it is overlain by younger late Pleistocene alluvium in some locations.
Unit 11 Scarp-derived slope-wash colluvium
Unit 2 Sag pond deposit Unit 1b Eolian sand
UIC1056
UIC1358
UIC1054
UIC1089
UIC1055
UIC1091
UIC1359
UIC1090
CHSF02-8
CHSF02-9
CHSF02-5
CHSF02-7
CHSF02-6
CHSF02-1
CHSF02-2
CHSF02-3
Th (ppm) 8.80 ± 1.24 8.08 ± 1.16 6.42 ± 0.75
8.41 ± 1.23
9.64 ± 1.31 11.29 ± 2.06 7.78 ± 1.19 11.29 ± 2.09 12.55 ± 1.56 5.00 ± 0.91
U (ppm) 2.51 ± 0.45 2.55 ± 0.41 3.53 ± 0.40
3.82 ± 0.46
6.14 ± 0.59 5.76 ± 0.76 3.09 ± 0.43 5.58 ± 0.75 4.19 ± 0.58 3.10 ± 0.36
2.02 ± 0.02
2.10 ± 0.02
2.46 ± 0.02
1.93 ± 0.02
2.05 ± 0.02
1.98 ± 0.02
2.13 ± 0.02
2.13 ± 0.02
2.16 ± 0.02
2.05 ± 0.02
K2O (%)
300.34 ± 2.97
267.21 ± 2.01
362.43 ± 1.12
122.12 ± 0.48
158.90 ± 1.12
124.84 ± 0.82
19.22 ± 0.04
22.66 ± 0.24
24.30 ± 0.18
16.88 ± 0.06
Polymineral IRSL 4–11 µm † De (Grays)
10 ± 3
10 ± 3
10 ± 3
5±2
5±2
5±2
5±2
5±2
5±2
5±2
Moisture content (%)
3.56 ± 0.16
4.52 ± 0.20
5.56 ± 0.23
3.60 ± 0.16
5.44 ± 0.22
4.47 ± 0.20
3.56 ± 0.17
3.66 ± 0.17
4.02 ± 0.19
3.20 ± 0.15
Dose rate § (Grays)
84.6 ± 6.0
>59.1 ± 4.5
65.2 ± 5.6
30.2 ± 2.2
28.7 ± 2.4
24.9 ± 1.7
5.4 ± 0.4
5.5 ± 0.4
6.0 ± 0.4
5.3 ± 0.4
IRSL or OSL age (ka)**
Predates event V and buried soil, S1, on Llano de Manzano
Postdates event V
Postdates event V and predates event W
Postdates event W
Postdates event X and predates event Y(?)
Postdates event X and predates event Y(?)
Postdates event Z
Postdates event Z
Postdates event Z
Postdates event Z
Comments
CHSF02-4
UIC1053
Predates event V and S1 on Unit 1b 2.86 ± 0.38 7.21 ± 1.09 2.14 ± 0.02 315.00 ± 1.16 10 ± 3 3.60 ± 0.17 82.5 ± 6.0 Slope wash Llano de Manzano Note: Samples CHSF02-9 and CHSF02-11 were green optically stimulated luminescence (OSL) analyses, all others were infrared-stimulated luminescence (IRSL) analyses with measurements made using a multiple aliquot additive dose method, measuring blue emissions on the 4 to 11 µm polymineral fraction. *Sample locations are shown on Figures 5A and 5B. † De—equivalent dose. § All errors are at 1σ and were calculated by averaging the errors across the temperature range. **Ages are rounded to the nearest 100 yr, and all errors are at 1σ.
Unit 2 Sag pond deposit
Unit 4 Eolian sand
Unit 6 Loess
Unit 6 Loess
Unit 11 Scarp-derived slope-wash colluvium
Unit 13 Eolian sand
UIC1088
CHSF02-10
Unit 14 Eolian sand
UIC1357
Stratigraphic unit
TABLE 1. LUMINESCENCE DATA FOR THE CARRIZO SPRING TRENCH ACROSS SPLAY L OF THE HUBBELL SPRING FAULT SYSTEM
Lab sample no.
CHSF02-11
Field no.*
61
62
Olig et al.
Unit 2 is only locally present on the downthrown side of FZ3 (Fig 5B). It pinches out to the east at st. 37 m and is faulted and warped down at its west end below the base of the trench west of st. 43.5 m. This dark-brown, sandy silty clay overlies unit 1b and is overlain by units 3a and 4. It is weakly bedded with fine sand partings, mottled with MnO staining, and generally lacks carbonate. Based on its location, limited extent, and stratigraphic characteristics, we interpret unit 2 to be a sag pond deposit at the base of a normal fault scarp created during event V by warping and offset on faults (FZ1, possibly FZ2, and FZ3 on Fig. 5B). Due to the subsequent dissolution of carbonate in unit 1b, stratigraphic relations are unclear as to whether unit 2 was deposited before or after development of S1. One IRSL sample from unit 2 (CHSF02-1 on Fig. 5B) yielded an age of 65.2 ± 5.6 ka, whereas another sample (CHSF02-2 on Fig. 5B) yielded a minimum age of >59.1 ± 4.5 ka due to saturation (Table 1). These ages are consistent with the IRSL ages from the underlying unit 1 and the degree of intervening soil development of S1, suggesting that S1 developed before event V occurred and before deposition of unit 2. Unit 3 is dominantly a pink fine sand that is faulted, warped down to the west, and appears to have been eroded away west of st. 38 m. We interpret unit 3 to be younger piedmont deposits of the Llano de Manzano, and similar to unit 1, unit 3 includes two subunits, 3a and 3b, where the latter is distinguished by a buried soil, S2. Subunits 3a and 3b are similar to and unconformably overlie units 1b and 1c, respectively. Unit 3a is a pinkish silty sand that contains some small gravel stringers and intraclasts. This unit is similar to the upper portion of unit 1b, and we interpret it be a mix of slope-wash and eolian deposits. Unit 3a is stratigraphically continuous with unit 3b, but it lacks the carbonate of the stage II nodular buried soil (S2) included in unit 3b. Similar to S1, the carbonate in S2 appears to have been dissolved from unit 3b west of st. 28 in the deformation zone. East of st. 18 m, the carbonate in S2 increasingly overprints the underlying S1 buried soil, and they became indistinguishable east of st. 11 m. Unit 4 unconformably overlies the buried S2 soil and units 3 and 2. Although it is warped and cut by fractures in the deformation zone, unit 4 did not appear offset by faults. This reddish, coarse to fine sand contains discontinuous coarser beds. It thickens on the downthrown side of faults, where sediments are generally coarser and include some intraclasts. Based on these characteristics, we interpret unit 4 to be colluvium and eolian sediments deposited after a faulting event, event W. However, similar to observations made by Personius et al. (2001) in the Hubbell Spring trench, this postfaulting unit is dominated by eolian deposition, and thus does not show many of the typical characteristics of fault-scarp–derived colluvial wedge deposits along normal faults, such as distinct debris and slope-wash facies forming wedged-shaped deposits (Nelson, 1992). An IRSL sample from unit 4 (CHSF02-6 on Fig. 5A) yielded on age of 30.2 ± 2.2 ka. This sample was collected from an eolian-dominated portion in the footwall because sediments in the deformation zone showed undesirable characteristics for luminescence dating (intraclasts, coarser sand, and extensive FeO and MnO staining).
Thus, although the eolian sediments that were sampled clearly postdate faulting, they may have been deposited well after faulting, and their age may not provide a close minimum limiting age for event W. Unit 5 is a well-sorted, reddish, fine to medium eolian sand that contains faint planar laminations, little carbonate, and extensive FeO and MnO staining. It conformably overlies unit 4 and includes a buried soil, S3, which apparently extends nearly the full length of the trench. However, the S3 catena varies considerably across the trench exposure (cf. soil profiles CS2, CS5, and CS1 in the Appendix), probably due to the different soil-forming conditions across the slope of the fault scarp. S3a extends from st. 0 to st. 8 m and is characterized by a mottled Btk horizon with a stage II to III carbonate morphology. Carbonate coatings on peds suggest overprinting of the original Bt horizon, likely similar to S3b, by carbonate related to development of overlying buried soils, S5 and possibly S6. S3 is apparently eroded away between st. 8 and st. 9 m. S3b extends from about st. 9 m to about st. 35 m and is characterized by a Bt horizon that is as much as 0.5 m thick and well cemented with sesquioxides. S3 is stripped away between st. 35 and st. 37. S3c extends from about st. 37 to the west end of the trench and is characterized by a mottled Btk horizon similar to S3a but more disseminated. S3c entirely overprints unit 5 and the top of unit 4, obscuring the contact between these units in the hanging wall. Although unit 5 is cut by fractures and warped down to the west with a total apparent throw of 2.8 ± 0.6 m down to the west (as measured in the trench and boreholes), we observed no discernible fault offsets of this unit. However, unit 5 thinned dramatically between st. 30 and 38 m. The top, including S3, is clearly stripped at the crest of the zone of maximum warping. The angular discordance between the base of unit 5 and the overlying stratigraphic and pedologic horizons suggests that this erosion occurred after a warping event, event X, and before unit 6 was deposited. Unit 6 unconformably overlies unit 5. Unit 6 is a buffcolored, silty, very fine eolian cover sand that is very well sorted and relatively homogeneous, lacks carbonate, and has some faint planar laminations. It appears to have been eroded away east of st. 9 m. West of st. 35.5 m, it is overprinted by carbonate from the overlying soil horizon, S5, and becomes indistinguishable from underlying and overlying units. Two IRSL samples from unit 6 (CHSF02-5 and CHSF02-7 on Figs. 5B and 5A, respectively) yielded an average age of 26.8 ± 2.4 ka (Table 1). Unit 7 is a pinkish tan, clayey silty fine sand that conformably overlies unit 6, extending from about st. 17 m to st. 32 m. This very well-sorted eolian cover sand is similar to unit 6, but it includes a weakly developed buried soil horizon, S4, that varies between a Bw and Btj horizon. Both units 6 and 7 are cut by fractures and warped, but we observed no discrete fault offsets in these units. In the deformation zone, both units 6 and 7 are overprinted, in angular discordance, by carbonate from the overlying buried soil horizon, S5, developed in unit 8. We believe this angular discordance was created by a small warping event, event Y(?),
Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift because S5 appears to have formed on a slope that was steeper and higher than the slope where units 6 and 7 were deposited. Regardless, it is likely that unit 8 in the hanging wall includes units 6 and 7 near its base, but stratigraphic characteristics and contacts are obscured by soil formation in S5. Indeed, unit 8 is entirely characterized by a buried K horizon, S5, which obscures most of the original depositional characteristics of this pinkish-white, silty sand with clay (likely primarily an eolian deposit). It is present in the footwall and hanging wall but locally missing below the crest of the scarp, from about st. 17 m to 28 m, probably due to stripping. Although unit 8 is cut by fractures, warped down to the west, with a total apparent throw of 2.1 ± 0.5 m, and the top appears back tilted in the hanging wall and locally deformed adjacent to FZ4 (Fig. 5B), no faults with significant discrete slip were discernible in this unit. This is somewhat surprising because the abrupt upper contact with the overlying unit 9 makes a relatively distinct marker that is deformed but does not appear offset, even though unit 10 above is clearly offset by the upper portion of FZ4 as it steps westward. It appears that deformation in this portion of FZ4 is very distributed, and discrete slip is discontinuous. Soil carbonate accumulation may also mask small offsets. Unit 8 varies from a stage II to III+ carbonate morphology, but it is dominantly a stage III. It is also more than double in thickness in the hanging wall (over 1 m), where it appears more disseminated than in the footwall and contains large zones that are punky and friable in texture. It also contains irregular-shaped carbonate nodules, some as large as cobbles, that weather out of the trench wall easily in places due to apparent partial dissolution of carbonate in the surrounding matrix. We believe unit 8 has been extensively affected by groundwater and/or spring water upwelling along the faults and fractures in the deformation zone. This would not only explain the unusual textures, but also the apparently incongruous large amount of carbonate accumulation in a horizon that is younger than 30 ka, as indicated by luminescence ages from unit 6. Unit 9 is a silty, clayey sand that is locally present only on the downthrown side of the deformation zone between st. 32 and 56 m. It is characterized by a buried soil, S6, which consists of a mottled, friable Btk horizon with a stage II to III carbonate morphology. However, similar to unit 8, we believe the carbonate in unit 9 may not all be pedogenic, and thus the morphology is not a reliable indicator of age. Unit 9 contains reworked carbonate nodules, apparently from unit 8, which generally decrease downslope. Based on these characteristics and its limited extent on the downthrown side of the deformation zone, we interpret unit 9 to have been scarp-derived colluvium and eolian material deposited after the warping event Y(?). Unfortunately, due to the extensive carbonate accumulation and soil development, unit 9 was not a good candidate for luminescence age analyses. Unit 9 was also cut by faults of FZ4b during the most recent event, event Z. Additionally, unit 9 shows weak planar laminations in places, and these are tilted and warped on the downthrown side of FZ4. Unit 10 is a silty sand with clay that is characterized by the youngest buried soil in the trench, S7. This soil is a stage II nodu-
63
lar carbonate horizon that extends the full length of the trench, except where it was broken up by faults of FZ4b during event Z and disturbed by krotovina between st. 39.5 and 41.5 m. Additionally, unit 10 appears slightly back tilted toward the east on the downthrown side of FZ4b. Similar to unit 9, the carbonate accumulation and soil development in unit 10 prevented luminescence age analyses. Unit 11 is a light-brown, silty sand that contains blocks of unit 10 and reworked carbonate nodules. It is a wedge-shaped deposit that unconformably overlies the buried soil on unit 10. It is not offset by any faults, and it is present only on the downthrown side of FZ4b. Based on these characteristics, we interpret unit 11 to be primarily fault-scarp–derived colluvium with minor eolian material deposited after the youngest faulting event, event Z. Two luminescence samples (CHSF02-8 and CHSF02-9 on Fig. 5B) yielded an average age of 5.5 ± 0.4 ka for the distal and eolian-dominated portion of unit 11. Units 12 and 13 are, respectively, silty and coarse sand deposits that are dominantly eolian sediments with some slope wash. These lenticular-shaped deposits are cut by open fractures, but they did not appear faulted or warped. They conformably overlie unit 11 and appear to be filling the topographic depression that was likely created during faulting event Z, although it is possible that deposition of units 12 and 13 was related to another younger, minor warping event that could have occurred subsequent to event Z and deposition of unit 11. However, given the lack of any other evidence for this hypothetical event, we believe it is much more likely that the open fractures were related to nontectonic processes (such as settlement, bioturbation, freezethaw, etc.) and that the deposition of distinct eolian packages was related to nontectonic causes, such as climate change. An IRSL sample (CHSF 02-10 on Fig. 5B) yielded an age of 6.0 ± 0.4 ka, which is stratigraphically consistent (within 1σ error) with the age for the underlying unit 11 and the overall lack of soil development in units 11 through 13. Finally, unit 14 is a loose, tan, fine sand that drapes the scarp. This eolian sand includes the modern soil, which is characterized by roots in a weakly developed BW horizon. A luminescence sample (CHSF02-11 on Fig. 5B) collected from below the modern soil in unit 14 yielded a luminescence age of 5.3 ± 0.4 ka (Table 1). Trench Structure The broad deformation zone exposed in the trench is characterized by: (1) a zone of warping down to the west, which is coincident with the scarp face; (2) a narrower zone of near-vertical and bedding-parallel fractures, between st. 28 m and st. 41 m, which does not show discernible discrete slip (shown in black on Figs. 5A and 5B); and (3) a series of four west-dipping to subvertical fault zones (FZ1–FZ4, shown in red on Fig. 5B) that offset strata down to the west and are roughly coincident with the zone of fractures, the maximum zone of warping, and the maximum slope angle on the scarp. Although warping appears to have deformed units 1 through 10 in a broad zone (i.e., from the crest of the scarp westward),
64
Olig et al.
there is a more concentrated zone of warping between st. 28 and 42 m that is coincident with faulting and the maximum scarp angle. Within this zone, older units are generally warped more than younger units, but differential warping and distinct events are much more difficult to distinguish than differential offsets and events on faults. However, a progression of warping (and faulting) events is still evident from the overall thickening of many units and soils in the hanging wall, the erosion of several units and soils in the zone of maximum deformation, and the angular discordances between some units and overlying units and soils, as discussed in the previous section. Based on these relations and total differential offsets on buried soils (S1, S2, S3, S5, and S7) (discussed in “Deformation Event Summary and Chronology” section), it is evident that warping occurred during all of the deformation events on splay L of the Hubbell Spring fault system, albeit to various degrees. Fault terminations at the top of buried soils and differential offsets on specific faults provide evidence for three of the faulting events, events V, W, and Z. Fault-zone orientations and dip-slip measurements are shown on the trench log (Fig. 5B). Faults FZ1 and FZ2 both cut units 1 and 3, terminating at the top of unit 3a, the stratigraphic equivalent to the top of the buried soil S2 and the event horizon for event W. Differential offset between units 1 and 3 suggest that FZ1 and FZ2 were both active during events V and W. However, given the uncertainties in measurement, larger offsets of unit 1 may be partially due to the listric geometry of FZ1. Similar to FZ1 and FZ2, FZ3 cut units 1 and 3, terminating at the top of unit 3, although fractures (without offset) extended into unit 4. Small differential offsets between units 1 and 3 were apparent but not definitive on FZ3, given the warping and complex fault geometry of this zone. FZ4 is even more complex. It is characterized by a wide distributed shear zone located between st. 39 and 41 m (Fig. 5B), with discontinuous fault strands that step to the west stratigraphically upward and include lower (FZ4a) and upper (FZ4b) portions. The lower portion, FZ4a, cuts units 1 and 2 and the base of unit 4 and has many associated fractures that extend upward into unit 5. Although the upper portion, FZ4b, offsets units 9 and 10 with discrete slip on faults, definitive discrete slip offset of underlying units 8 and 5 is not apparent, although these units are clearly warped and deformed down to the west between FZ4a and FZ4b. It would appear that shearing is so distributed in this part of FZ4 as to be visible only as warping, although relations are also somewhat complicated by carbonate accumulation and overprinting of S3 and S5 soil development, which may obscure some discrete slip. The upward termination of FZ4b is at the top of the buried soil S7 in unit 10, the event horizon for event Z, with open fractures extending into overlying units 11 and 12. Differential offsets of unit 10 versus unit 3b indicate that FZ4 was active during events W and Z. Soil Pit Exposure The soil pit exposed an extensively bioturbated package of dominantly eolian sand with minor alluvium that included three buried soils underlying the modern soil on loose eolian sand
(Fig. 6). The lowermost soil is a disseminated K horizon developed on a sandy silt to silty sand with gravel. It contains the most carbonate (stage II to III) and lacks the clay, mottling, and peds that are characteristic of S3 in the trench. Therefore, we think this soil most likely correlates to S5 in the trench. Overlying the K horizon, there is a mottled, silty very fine sand with a buried disseminated Bk horizon. This soil may correlate to S6 in unit 9 or S7 in unit 10 in the trench, although carbonate accumulation appears much more disseminated in the Bk horizon. Overlying the Bk horizon, there is sandy alluvium with a thin Btjk horizon that does not appear to directly correlate to any soils in the trench, although it may be time-equivalent to S7. The sandy alluvium appears to have been deposited by a small drainage that incises the fault scarp north of the trench. Due to the extensive bioturbation and more disseminated character of the buried soils exposed in the soil pit, correlations to soils in the trench were slightly ambiguous. However, it is clear that neither of the buried soils, S2 or S1, nor the channel gravels of unit 1 were exposed in the soil pit, and presumably these soils and deposits are at a greater depth below the pit exposure. Indeed, as discussed in the next section, S1 and S2 were observed at greater depths in the boreholes. Boreholes The locations of the three drill holes, projected onto the scarp profile, along with the interpreted depths of buried soils S1 and S2, are shown on Figure 5D. Logs of borings are included in Appendix C of Olig et al. (2004). Boring B1 was located just south of st. 60 m at the west end of the trench. It was the deepest hole (total depth 10.39 m) and provided relatively good correlations to stratigraphy exposed in the trench, except for some uncertainties due to sloughing at depths around 1.5 and 2.5 m. At a depth of 5.6 m, boring B1 encountered a buried carbonate soil horizon that was mottled and appeared nodular. It was developed on a gravelly sand that coarsened downward to a clean sandy gravel with clasts of gneiss, quartzite, and other dark metamorphic rocks. Based on these characteristics, we correlate these sandy gravels to the channel deposits of unit 1a in the trench, and the soil to S1 in the trench. At a shallower depth of 3.85 m, boring B1 encountered another mottled carbonate soil horizon that appeared nodular. Our preferred interpretation is that this soil correlates to S2 in the trench. Alternatively, a thin, more clayey mottled carbonate horizon encountered at 4.5 m depth possibly correlates to S2. Boring B2 was located 46 m southwest of B1 and was 7.22 m deep. B2 encountered a mottled carbonate soil horizon at 2.53 m depth, which we believe correlates to S2 in the trench. Underlying that was another mottled carbonate horizon at 4.76 m depth, which we correlate to S1 in the trench. Underlying S1 in both B1 and B2 was a thick (>1 m), pinkish, well-sorted, silty fine sand. Borehole B3 was located between B1 and B2. Samples from the upper 2 m of B3 were disturbed due to drilling problems. However, a carbonate soil horizon developed on a sandy gravel with metamorphic clasts was encountered at a depth of
Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift 4.95 m, and we correlate this soil and gravel to S1 and unit 1 in the trench. At a shallower depth of 3.7 m, another carbonate soil horizon was encountered in B3, which we tentatively correlate to S2 in the trench. Unfortunately, at the bottom of B3, we could not punch through unit 1 and presumably into the underlying pinkish silty fine sand that was encountered in boreholes B1 and B2. Deformation Event Summary and Chronology This section summarizes the evidence for, and timing constraints on, the faulting and warping events at the Carrizo Spring trench site. Differential offsets provide important evidence for identifying events and information about event size. Due to the extensive warping at the Carrizo Spring site, we knew that we needed to look beyond just differential offsets on individual faults and compare the total down-to-the-west throw (including fault slip and warping) between event horizons. Using observations from both the trench and boreholes, we projected key buried soil horizons across the entire deformation zone to estimate the apparent throw on these key horizons. In particular, we were able to measure the apparent total throw on key marker horizons, buried soils S1, S2, S3, S5, and S7. Independent stratigraphic and structural evidence, which is summarized in the following sections, indicates or suggests that the tops of these buried soils correspond to the event horizons for paleoearthquakes V, W, X, Y(?), and Z, respectively. We refer to these total throw measurements as “apparent” because our correlations of units across the deformation zone rely heavily on pedogenic characteristics, and soils can form on preexisting slopes such that all of the vertical relief we measured may not be related to offset that occurred after the soil formed. However, the generally subrounded to rounded gravels of unit 1a, with clasts as large as boulders, were deposited by relatively high-energy streams flowing west to southwest from the Manzano Mountains, out across the Llano de Manzano surface, and across any scarp associated with splay L. Given the nature of the channel gravels comprising unit 1a, we believe it is likely that these streams beveled off any preexisting scarp before soil S1 formed. Therefore, in our analysis, we assume that S1 formed on a relatively uniform and gentle slope without a preexisting scarp. Assuming there was relatively little to no initial relief on unit 1, and because the vertical relief measured on each event horizon is progressively larger for each older event, we make the key inference that all this post–unit 1 vertical relief was tectonically created in a series of successive faulting/deformation events. This seems reasonable because we see no other likely cause for repeated creation of new relief on a north-trending scarp at this location, especially given the nature of all the deposits overlying unit 1, which are all generally finegrained eolian and slope-wash sediments. Table 2 summarizes the total apparent throw measured on the tops of the buried soils (S1, S2, S3, S5, and S7) respectively forming the event horizons for events V, W, X, Y(?), and Z. These can be used to calculate differential displacements and estimate the displacements per event as follows. Assuming S1 formed on a
65
gentle slope with no scarp as previously discussed, we measured 7.3 ± 1.0 m of total cumulative throw on this buried soil by projecting it across the deformation zone from the trench exposures and boreholes (Fig. 5D; note that we actually measured displacements on a 1:40 scale version of Fig. 5D that was significantly reduced here for publication). Similarly, from the trench exposure and boreholes, we measured 6.5 ± 0.3 m of total apparent throw on soil S2. We calculated a differential offset between S1and S2 of 0.8 m to estimate the net vertical tectonic displacement for event V (Table 2). The net vertical tectonic displacement is the measure of vertical slip at the fault that accounts for the back tilting and antithetic faulting that is common along normal faults (Swan et al., 1980). Although some minor back tilting was apparent for a couple of events, more importantly, at the Carrizo Spring site, this measure also includes the effects of warping. We similarly used differential offsets between each successive event horizon (S2, S3, S5, and S7) to estimate net vertical tectonic displacement for each subsequent deformation event. Resulting preferred estimates of net vertical tectonic displacement per event range from 0.4 to 3.7 m (Table 2), indicating highly variable and noncharacteristic behavior (discussed further in “Rupture History and Behavior” section). The displacements are smallest for the two warping events, X and Y(?). Indeed, due to the small preferred offset for event Y(?), and because independent stratigraphic and pedologic evidence was not as strong for this event, we recognize the uncertainty in the occurrence of this probable warping event by explicitly using a query in referring to it. Except for the youngest event, Z, uncertainties for displacements per event are large due to the difficulties in projecting horizons across the broad deformation zone, and the cumulative effect of calculating differential offsets. This results in a minimum estimate of 0 m for three of these events (V, X, and Y[?]). However, with the possible exception of event Y(?), we believe that the independent evidence for the occurrence of events V and X argues for some minimum offset larger than 0 m for these events. It is also worth pointing out here that the cumulative throw on S1 is 7.3 ± 1.0 m, which is similar to the 7.5 ± 1.0 m of net vertical tectonic displacement measured across the Llano de Manzano surface on the scarp profile. At first glance, this seems somewhat surprising because offsets measured from scarp profiles on normal faults might be considered minimums due to expected postfaulting erosion of the footwall and deposition in the hanging wall. However, the stratigraphic record preserved in the Carrizo Spring trench shows little evidence for preferential erosion of the footwall aside from at the crest of the deformation zone. Additionally, although fault scarps provide excellent local sediment traps for eolian sands in the region (e.g., Personius and Mahan, 2003; McCalpin et al., 2006), deposits can also blanket scarps, so that sediment is not only accumulating in the hanging wall but also in the footwall (e.g., units 4, 5, 6, and 7). These types of cover sands and intervening buried soils are likely present elsewhere on the Llano de Manzano surface and would provide excellent stratigraphic markers for paleoseismic studies
66
Olig et al. TABLE 2. ESTIMATES OF DISPLACEMENTS ON SPLAY L OF THE HUBBELL SPRING FAULT SYSTEM
Deformation event
Event horizon
Total apparent throw on event horizon* (m)
Estimated net vertical tectonic displacement per event (m) 0.8 † (>0 , 1.6)
Basis for estimate
Event V
Top of buried soil S1 (on unit 1)
7.3 ± 0.3
Event W
Top of buried soil S2 (on unit 3)
6.5 ± 0.5
3.7 (2.6, 4.8)
Differential offset between total apparent throw on top of S2 and top of S3 (6.5–2.8 m)
Event X
Top of buried soil S3 (on unit 5)
2.8 ± 0.6
0.7 † (>0 , 1.8)
Event Y(?)
Top of buried soil S5 (on unit 8)
2.1 ± 0.5
0.4 (0, 1.2)
Event Z
Top of buried soil S7 (on unit 10)
1.7 ± 0.3
1.7 (1.4, 2.0)
Differential offset between total apparent throw on top of S3 and top of S5 (2.8–2.1 m) Differential offset between total apparent throw on top of S5 and top of S7 (2.1–1.7 m) Total apparent throw on top of S7
Differential offset between apparent throw on top of S1 and top of S2 (7.3–6.5 m)
Style of deformation
Normal slip on FZ1, FZ2, and possibly FZ3 (and buried faults west of st. 43 m?) Warping Fracturing Normal slip on FZ1 through FZ4 (and buried faults west of st. 43 m?) Warping Fracturing Warping Fracturing
Warping Fracturing Back tilting (?) Normal slip on FZ4 Warping Fracturing (?) Back tilting (in FZ4 hanging wall)
Note: All displacements are down to the west. *Measured from trench exposures (and in boreholes for S1 and S2) by projecting horizons across the zone of deformation. Assumes channel gravels of unit 1 leveled off any preexisting scarp, and all subsequent vertical relief was created by tectonic deformation (see text for discussion). † See text for discussion of minimum values.
of other faults scarps. The drawback of these deposits is that they mute the geomorphic expression and can even bury faults, making fault scarps appear much more discontinuous and difficult to map (Olig and Zachariasen, 2010). In retrospect, the geomorphic expression of fault splay L of the Hubbell Spring fault system at the Carrizo Spring trench site is consistent with the paleoseismic record exposed in the trench, with a very broad scarp resulting from a broad deformation zone of faulting and warping. Additionally, the maximum scarp angle was located at the zone of most concentrated faulting and warping. Scarp heights appear subdued due to little back tilting and a “bulge” of eolian deposits at the scarp base. The low maximum scarp angle and single bevel on the scarp profile may at first seem inconsistent with the four to five late Pleistocene events interpreted in the trench, but the single bevel is actually consistent with the dominant deformation style of warping, and the small maximum slope angle of 12° is likely more related to the angle of repose for deposition of eolian sand on the scarp rather than degradation of an older fault scarp. Thus, the use of morphometric comparisons of scarp height and maximum slope angle (e.g., Bucknam and Anderson, 1979) does not appear to be a reliable indicator of fault-scarp age in this environment, which explains why Machette and McGimsey (1983) erroneously concluded that
the youngest faulting was much older than 15 ka based on their analysis of scarps of the Hubbell Spring fault system. However, our observations also suggest that scarp profiles can still provide useful slip estimates despite the extensive eolian deposition. A summary discussion of the structural, stratigraphic, and pedologic evidence for each surface-faulting/deformation event and the timing of events are presented next. Event Z Structural, stratigraphic, and pedologic evidence for this youngest surface-faulting event includes: (1) offset (including discrete slip on fault FZ4b and warping) of the top of unit 10, and the buried soil S7, 1.7 ± 0.3 m down to the west; (2) fault terminations at the top of S7; (3) an overlying unfaulted, colluvial-wedge deposit (unit 11) adjacent to the maximum zone of faulting, which contained reworked carbonated nodules from unit 10, blocks of unit 10, as well as eolian sediment banked against the scarp created by event Z; and (4) slight back tilting of unit 10 to the east toward fault FZ4b. Event Z occurred sometime before deposition of the distal portion of unit 11, ca. 5–6 ka, and well after unit 6 was deposited at 24–29 ka, because not only were units 7 through 10 deposited subsequently, but three intervening buried soils (S4, S5, and S6) formed. However, as previously discussed, some of
Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift the carbonate accumulation observed in S5 and S6 may be related to nonpedogenic processes, and so unfortunately the carbonate morphology of these soils is likely not reliable a indicator of their age. Regardless, the general lack of soil development in units 11–14 also suggests that event Z occurred closer to 6 ka than 24 ka and was likely younger than 15 ka. Event Y(?) The evidence for the penultimate event, event Y(?), is suggestive, but not conclusive because the apparent deformation associated with this event is limited to minor warping of unit 8 and the buried soil S5 ~0.4 m down to the west. No discrete slip on any faults was observed. However, the angular relation between the buried soil S5 and the underlying units 6 and 7 suggests that this soil formed on a slope that was steeper than the one units 6 and 7 were originally deposited on. This, and the apparent increased differential offset between S7 and S5 suggest that this deformation event occurred after S5 formed but before unit 9 was deposited. Indeed, unit 9 appears to be a mix of scarp-derived colluvium (including reworked carbonate nodules) and eolian sediment banked up against the scarp after event Y(?) occurred. The timing for event Y(?) is very poorly constrained. The Btk horizon of S6 formed on unit 9 suggests that event Y(?) was somewhat older than Holocene, and of course it is younger than deposition of unit 6 at ca. 24–29 ka. Event X Although event X did not result in any apparent discrete slip on faults, fracturing and warping of units 1–5 did result in 0.7 m of overall differential down-to-the-west net vertical tectonic displacement between the tops of buried soils S3 and S5 (see Appendix). Additionally, stripping of S3 and the top of unit 5 locally at the crest of the zone of maximum warping, along with the angular discordance between the base of unit 5 and the overlying stratigraphic and pedologic horizons, also indicates that erosion occurred after a warping event of unit 5 and S3 but before deposition of unit 6. Based on luminescence ages for units 4 and 6 (Table 1), event X occurred between 26.8 ± 2.4 ka and 30.2 ± 2.4 ka. Note that if event Y(?) did not occur, it would imply that larger displacements occurred during event X, ~1.1 m instead of only 0.7 m. Event W Compelling stratigraphic and structural evidence for event W includes: (1) discrete offset of units 1–3 along faults FZ1–FZ4, with multiple fault terminations at the base of unit 4 (the stratigraphic equivalent to the top of the buried soil S2); (2) warping and faulting resulting in large differential down-to-the-west offset of 3.7 m between the top of buried soils S2 and S3 (Appendix); and (3) thickening and coarsening of unit 4 on the downthrown side of faults FZ1, FZ2, and FZ3. Despite the large net vertical tectonic displacement indicated for event W, displacements on individual faults are small and distributed across multiple splays, forming wide zones. Additionally, similar to other events, warp-
67
ing is also evident across a wide zone. One somewhat anomalous stratigraphic relation is the apparent thinning of unit 4 west of st. 38 m, on the downthrown side of FZ4. This may be due to the presence of buried faults west of st. 43 that were not exposed in the trench. Alternatively, it may be due to stratigraphic uncertainties, because the contact between units 4 and 5 becomes more diffuse and affected by overprinting of soil carbonate west of st. 37 m. Regardless, our estimated value of 3.7 m of net vertical tectonic displacement for event W based on drill-hole and trench data includes any offset on those potential buried faults. The timing of event W is not tightly constrained, but it occurred sometime before deposition of unit 4 (ca. 30.2 ± 2.2 ka) but well after deposition of unit 2 (65 ± 6 ka), and after subsequent deposition of unit 3 and formation of the stage II carbonate horizon of S2 on unit 3. Event V Stratigraphic, structural, and pedologic evidence from the trench and boreholes indicates that the oldest event, event V, resulted in ~0.8 m of net vertical tectonic displacement of the top of S1, the buried soil developed on unit 1. Deformation is characterized by warping, fracturing, and discrete normal slip on FZ1, FZ2, and possibly FZ3. In addition to differential offsets, a slight thickening of unit 3a on the downthrown side of faults FZ1 and FZ2 also supports minor slip on these faults during event V. However, much of the deformation associated with event V appears to have been accommodated by warping of unit 1 and S1, creating a depression at the base of the scarp, where a sag pond deposited unit 2. The timing of event V is constrained to be well after unit 1 was deposited (ca. 84 ± 6 ka) and the S1 soil subsequently formed but likely shortly before deposition of unit 2 ca. 65.2 ± 5.6 ka. ANALYSIS OF PALEOSEISMIC PARAMETERS AND DISCUSSION OF FAULT BEHAVIOR Rupture History and Behavior The paleoseismic record of surface faulting and warping events that we deciphered for fault splay L of the Hubbell Spring fault system at the Carrizo Spring site is summarized in Figure 7, including timing constraints and preferred net vertical tectonic displacements per event. We found stratigraphic, structural, and pedologic evidence that indicates at least four, possibly five, surface-deforming earthquakes (events V through Z) occurred since a stage II buried soil carbonate horizon formed on sediments that were deposited on the Llano de Manzano ca. 84 ± 6 ka. Also summarized on Figure 7 is the record of surface-faulting events deciphered by Personius et al. (2001) and Personius and Mahan (2003) for the western Hubbell Spring fault (splay J of the Hubbell Spring fault system) at the Hubbell Spring site. They found evidence for four surface-faulting events that occurred after carbonate rinds formed on fan gravels ca. 92 ± 7 ka. The available paleoseismic data for the Hubbell Spring fault system suggests complex rupture behavior that includes both
68
Olig et al.
Splay L (Central) - HSFS
Splay J (Western) - HSFS s
k.y.
k.y.
k.y.
s
slope-wash
ka slope-wash
Figure 7. Comparison of the paleoseismic records of fault splay L (this study) and fault splay J (Personius et al., 2001; Personius and Mahan, 2003) of the Hubbell Spring fault system (HSFS).
Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift independent rupture of splay L of the Hubbell Spring fault system and simultaneous rupture of splays L and J of the Hubbell Spring fault system. Although the timing constraints are poor for the first event on splay J and for events Y(?) and W on splay L, comparison of the paleoseismic records (Fig. 7) indicates that the timing of the four largest events on splay L (events Z, X, W, and V) overlaps with the timing of the past four events on splay J (fourth through first events, respectively), suggesting simultaneous rupture of splays L and J together during larger events. The relatively large displacements for these events (1–2 m on fault splay J and 0.7–3.7 m on fault splay L), and the similar amounts of total throw since the oldest event (5–8 m on splay J vs. 7.3 ± 0.5 m on splay L) also support simultaneous rupture of the two splays. Additionally, the buried soils that formed before each surface-deforming event appear to correlate between sites (e.g., our soil S1 correlates to the buried soil in unit 2 at the Hubbell Spring site, etc.; Fig. 7), which also supports simultaneous rupture. However, given the resolution of ages, even for the events with better age constraints (e.g., event X on fault splay L and the third event on fault splay J), we acknowledge that we cannot preclude the possibility that the events on each splay may have occurred separately. Regardless, the smallest event on fault splay L, event Y(?), does not appear to correlate to any events on fault splay J of the Hubbell Spring fault system. Assuming that this event did indeed occur and that the record deciphered for fault splay J is complete, this indicates that fault splay L also does occasionally rupture independently from fault splay J of the Hubbell Spring fault system in smaller events. Further comparison of the two sites also reveals additional similarities and differences in the paleoseismic records that warrant discussion. Some of the more significant stratigraphic and structural similarities between the sites are (1) the domination of eolian sedimentation along the scarps; (2) the wide distributed deformation zone of many faults across a single scarp; and (3) the box-like network of slope-parallel and subvertical carbonate-filled veins within the deformation zones. In regard to the first similarity, this study adds to a growing body of evidence (e.g., McCalpin et al., 2006; Personius and Mahan, 2003) for a stratigraphic signature along normal-slip intrabasin faults in the Rio Grande rift that is different than typical range-bounding normal faults. That is, deposits along intrabasin rift faults generally: (1) lack the debris-facies colluvial deposits that are typically proximal to scarps of normal, range-bounding faults; (2) are dominated by eolian sediments banking up against the scarp rather than colluvial sediments derived from the scarp; and (3) are more lenticular-shaped and can extend completely across the scarp as compared to the classic triangular, colluvial wedge deposit that is primarily limited to the downthrown side of the fault. The success of luminescence dating of the eolian and colluvial sediments at both Hubbell Spring fault system trench sites is promising for applications elsewhere in the Albuquerque Basin, not only for fault studies, but for all types of Quaternary studies, particularly where additional absolute age constraints are sorely needed to understand complex diachronous surfaces.
69
There are also important stratigraphic, pedologic, and structural differences between the two Hubbell Spring fault system trench sites. One important difference between the sites is the probable occurrence of an additional event (event Y?) on fault splay L, and the more complex stratigraphic and pedologic sequence that predates and postdates this small warping event at the Carrizo Spring site. This highlights many questions about the seismogenic relation among all the fault splays within the Hubbell Spring fault system, and the need for additional paleoseismic studies to understand the behavior of the many other unstudied splays. Given the anastomosing geometry and close proximity between some splays of the Hubbell Spring fault system, it is likely that some splays merge together in the subsurface. However, the geometry of this complex system (its long length, great width, and the fact that nearly all the splays are west-dipping) also indicates that all of the fault splays of the Hubbell Spring fault system cannot merge together. Given the complex geometry and the available paleoseismic data, it seems unlikely that the entire Hubbell Spring fault system ruptures simultaneously, especially in every event. Still, preliminary observations of alongstrike displacement variations and the anastomosing geometry of faults do suggest complex patterns of slip transfer between faults, with northwest- and northeast-striking sections serving as relay ramps between principal north-striking faults (Olig and Zachariasen, 2010). Obviously, we need additional coordinated studies to better understand rupture patterns for the entire fault system and to more accurately model rupture behavior in hazard evaluations, which presently use very simplistic rupture models for the Hubbell Spring fault system due to the lack of data (e.g., Wong et al., 2004; Olig et al., 2007). Another structural difference between the trench sites is the greater degree of warping and variability in the displacements per event at the Carrizo Spring site. Displacements per event on fault splay L showed large variability, ranging from 0.4 to 3.7 m (Fig. 7). In contrast, estimated displacements per event for fault splay J only ranged between 1 and 2 m (Personius and Mahan, 2003); however, displacements for this splay are not as well constrained as for splay L. Despite the apparently more uniform displacements per event on splay J, cumulative displacements per event (assuming events were simultaneous in both trench sites) still show large variability. For the youngest (event Z) to the oldest (event V) events, we estimate cumulative displacements of 3.7, 0.4, 1.7, 4.7, and 2.8 m. Calculations of cumulative displacements are complicated by the fact that the two trench sites are over 18 km apart along strike, and displacements on a fault can vary significantly along strike. Here, we assumed that displacements measured at each site are representative averages for each fault splay, and so we simply added displacements per event from each site to estimate cumulative values for splays L and J. Although we believe that the complex rupture behavior interactions between different splays of this intrabasin fault are a significant contributing factor to the large variability in displacements per event, it is worth pointing out that displacement variability is still large between simultaneous rupture events, ranging from 1.7 to 4.7 m.
70
Olig et al.
The large variability in displacements per event indicates noncharacteristic behavior for splay L of the Hubbell Spring fault system, and likely for the Hubbell Spring fault system overall, which has important implications for recurrence models used in probabilistic hazard analyses (Wong et al., 2004). As originally defined by Schwartz and Coppersmith (1984), the characteristic recurrence model was based on paleoseismic observations along the San Andreas and Wasatch faults of similar-sized displacements per event at a particular point along a fault. The characteristic model predicts fewer moderate-size events and generally results in lower hazard estimates than the traditional Gutenberg-Richter exponential frequency-magnitude relationship (Youngs and Coppersmith, 1985). Thus, the large variability in displacements for the Hubbell Spring fault system implies noncharacteristic behavior and higher associated hazard. However, we also recognize that it is possible that along-strike variations in displacement on fault splay J are such that a trench on that splay at the same latitude as the Carrizo Spring site would reveal complementary displacements per event that would result in total displacements per event for both splays that were more similar in size. However, observed variations between displacements per event are so large that noncharacteristic behavior is still strongly suggested, regardless of possible along-strike displacement variations. For example, 2.6 m is an absolute minimum estimate for event W (see Table 2, and assuming the second event shows little or no displacement on fault splay J at the latitude of the Carrizo Spring site). In contrast, assuming event Y(?) did not occur on fault splay
J, 1.2 m is a maximum estimate for this event (Table 2), which still implies noncharacteristic behavior for the fault zone overall. Paleomagnitudes Paleomagnitude estimations for prehistoric deformation events on the Hubbell Spring fault system are complicated by the geometry of its many overlapping fault splays. Although many empirical relations have been developed to estimate paleomagnitudes from various fault parameters, such as length and displacement per event, none of these has been developed for faults with multiple, subparallel overlapping splays like the Hubbell Spring fault system. In particular, for simultaneous rupture of splays, it is not clear whether summation of displacements per event for each splay is appropriate or not, although we did add them here because it seems the most logical approach. Table 3 shows paleomagnitude estimates for the Hubbell Spring fault system using various relations based on surface-rupture length (L), average displacement (AD), and maximum displacement (MD). Paleomagnitude estimates vary from Mw 6.6 to 7.5 depending on the assumed input parameters and rupture behavior of the Hubbell Spring fault system. All of the estimates based on displacement could arguably be considered minimums because unknown possible displacements on other splays (other than J and L) have not been included. Additional investigation of other fault splays and a more complete paleoseismic record are obviously needed for the Hubbell Spring fault system to better estimate paleomagnitudes.
TABLE 3. PALEOMAGNITUDE ESTIMATES FOR SURFACE-FAULTING EARTHQUAKES ON THE HUBBELL SPRING FAULT SYSTEM Fault parameter L = 74 km*
Expected moment magnitude (Mw)* 7.25 ± 0.28
MD = 4.7 m
†
7.19 ± 0.39
AD = 1.7 m
§
7.12 ± 0.39
AD = 4.7 m
#
7.48 ± 0.39
AD = 0.4 m** ††
6.60 ± 0.39
L = 42 km 6.96 ± 0.28 Note: We used empirical relations from Wells and Coppersmith (1994) for all types of slip: Mw = 5.08 + 1.16(log L), σ = 0.28; Mw = 6.93 + 0.82(log AD), σ = 0.39; Mw = 6.69 + 0.74(log MD), σ = 0.04. All surface rupture lengths (L) were measured straight line, end to end in kilometers. MD is maximum displacement along strike, and AD is average displacement along strike per event measured in meters. *For the entire Hubbell Spring fault system and based on mapping by Olig and Zachariasen (2008). † This is the maximum observed displacement for simultaneous rupture of fault splays J and L, which occurred during event W; here assumed to be representative of the maximum along-strike displacement. § This is the minimum observed displacement of simultaneous rupture of fault splays J and L, which occurred during event X; here assumed to provide a lower bound for the average displacement for these splays. # Maximum observed displacement for simultaneous rupture of splays L and J, but here assumed to provide an upper bound of AD for these splays. **Preferred displacement for event Y(?); applies to independent rupture of fault splay L only. †† For the rupture of splay L only and based on mapping by Olig and Zachariasen (2008).
Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift Recurrence and Slip Rates
71
⎛ 84 k.y. − 11k.y. ⎞ = 24.3 k.y. ⎟. ⎜ 3 intervals ⎝ ⎠
and J did rupture separately, this would nearly double the number of events that occurred during the past ~100 k.y. It would also imply strong temporal clustering of events and even suggest that an earthquake rupture on one splay may have triggered rupture on another splay. For comparison, and because seismic hazard analyses often do not consider complex rupture behavior alternatives (e.g., Frankel et al., 2002), we also estimated the average recurrence interval between all surface-deforming events on both splays L and J of the Hubbell Spring fault system, regardless of the type of behavior. For this estimate, we simply included event Y(?), which added another interval. This yielded a preferred average recurrence interval between all surface-deforming events of 15 k.y., with a range of 12 k.y. to 18 k.y. These recurrence interval estimates are not significantly shorter than those between simultaneous rupture events, especially given the uncertainties. Next, we used our previously calculated cumulative displacements per event (see “Rupture History and Behavior” section) to estimate cumulative slip rates for splays J and L of the Hubbell Spring fault system. Obviously, these are minimums for the entire Hubbell Spring fault system, because slip rates for other splays of the Hubbell Spring fault system are not included. However, they are still useful, particularly when looking at variations of rates through time. First, we estimated the average vertical slip rate over the past four complete seismic cycles for splays L and J. Again, assuming the oldest event (event V) occurred 70 ka, we estimated a preferred average rate of 0.18 mm/yr for splays J and L ⎛ 4.7 m + 1.7 m + 0.4 m + 3.7 m ⎞ ⎜ ⎟. 70 k.y. − 12 k.y. ⎝ ⎠
For individual recurrence intervals, we cannot improve on the original estimates of 17 and 27 k.y. by Personius and Mahan (2003) for the two youngest intervals (Fig. 7). For the interval between the first (= event V) and second (= event W) events, we estimate a preferred recurrence interval of 14 k.y., assuming event V occurred at 70 ka and event W occurred ca. 56 ka (Fig. 7). Although there is evidence for temporal clustering of surfacefaulting events on many faults in the Rio Grande rift (e.g., Foley et al., 1988; Machette, 1998; Gardner et al., 2008), we see no evidence for clustering of rupture events along splays L and J of the Hubbell Spring fault system, assuming that splay J did indeed rupture simultaneously with splay L. Individual recurrence interval estimates of 14–27 k.y. are similar to average intervals of 19 (+5/−4) k.y., and soils developed between events are consistent with these intervals. We point out, however, that although we believe that the luminescence ages, stratigraphic and pedologic correlations between trench sites, and displacement data all favor simultaneous rupture of fault splays J and L of the Hubbell Spring fault system during events Z (or fourth), X (or third), W (or second), and V (or first), we cannot preclude the possibility of independent rupture of the two splays given the large uncertainties for event ages. If splays L
Of the 10.5 m of vertical slip associated with these complete seismic cycles, over 96% (10.1 m) occurred as simultaneous rupture of splays J and L, yielding an average slip rate of 0.21 mm/yr for the past three complete seismic cycles of simultaneous rupture. In comparison, Olig and Zachariasen (2010) estimated average late Quaternary vertical slip rates of 0.2–1.0 mm/yr for the Hubbell Spring fault system based on topographic profiles across the entire system. In particular, for the topographic profile along Monterey Boulevard, just south of the Carrizo Spring site, they measured vertical displacements across splays L and J of 7.2 m and 4.0–16.0 m, respectively, with large uncertainties for splay J due to back tilting in the footwall. Next, we calculated slip rates for individual seismic cycles to see if rates have varied through time. These rates for the past four complete seismic cycles are shown in the slip rate diagram in Figure 8. Note that we consider a complete seismic cycle to include the time interval of strain buildup preceding an event and the strain release (or displacement) for that event. This is consistent with Reid’s model of elastic rebound theory for earthquake occurrence on faults. Note that since we do not know the time interval of strain buildup associated with the 2.8 m of displacement in the oldest event (0.8 m on splay L and 2 m on splay J), this is not a complete seismic cycle, and we did
Since we believe that fault splay L has ruptured both simultaneously with, and independently from, splay J of the Hubbell Spring fault system, ideally we should calculate rates of activity for both types of behavior. However, given only one observation for an independent event on splay L of the Hubbell Spring fault system, we cannot calculate recurrence intervals for this type of behavior, particularly since there are many splays of the Hubbell Spring fault system for which the paleoseismic behavior is unknown, and one of these may have actually ruptured with splay L during event Y(?). We can calculate recurrence intervals between simultaneous rupture events of splays J and L. Since timing constraints are best from the Hubbell Spring site for the youngest three events and from the Carrizo Spring site for the oldest event, we used this combination of data to determine the following intervals. Assuming a preferred age of 70 ka for event V at the Carrizo Spring site, we calculated a preferred average recurrence interval between the past four simultaneous ruptures of ~19 k.y. ⎛ 70 k.y. − 12 k.y. ⎞ = 19.3 k.y. ⎟, ⎜ 3 intervals ⎝ ⎠ with a range of 15 k.y. ⎛ 59 k.y. − 13 k.y. ⎞ = 15.3 k.y. ⎟ ⎜ ⎝ 3 intervals ⎠ to 24 k.y.
Olig et al.
not calculate an associated slip rate. Thus, the first (and oldest) complete seismic cycle is the 14 k.y. interval of strain buildup associated with 4.7 m of strain release (Fig. 8), the combined displacements for event W on fault splay L and the second event on fault splay J. The preferred slip rate for this seismic cycle is 0.34 mm/yr ⎛ ⎞ 4.7 m ⎜ ⎟. ⎝ 70 k.y. − 56 k.y. ⎠ This is nearly double the average rate, but it is still not as high as the preferred rate calculated for the youngest (fourth) complete seismic cycle, which is 0.46 mm/yr ⎛ ⎞ 3.7 m ⎜ ⎟. 20 k.y. − 12 k.y. ⎝ ⎠ In contrast, preferred rates for the second and third complete seismic cycles are as much as an order of magnitude lower, at 0.063 ⎛ ⎞ 1.7 m ⎜ ⎟ 56 k.y. − 29 k.y. ⎝ ⎠ and 0.044 ⎛ ⎞ 0.4 m ⎜ ⎟ 29 k.y. − 20 k.y. ⎝ ⎠
mm/yr, respectively (Fig. 8). We note that low slip rates are associated with both types of rupture behavior (independent and simultaneous) and result regardless of the timing uncertainties for event Y(?). Thus, although recurrence intervals have remained relatively consistent through time, cumulative slip rates for the Hubbell Spring fault appear to have varied significantly due to large variations in displacements per event, or noncharacteristic behavior. This may have important implications for hazard evaluations elsewhere in the rift. Interestingly, order-of-magnitude variations in slip rates through time have been observed on many faults throughout the Rio Grande rift (McCalpin, 1995; Machette, 1998). McCalpin (1995) observed that short-term slip rates are generally much higher than long-term rates for dozens of faults. These variations have been possibly attributed to variations in the frequency of occurrence of earthquake events (i.e., temporal clustering) or even variation in the quality or type of data being used (e.g., Wong and Olig, 1998; Machette, 1998). Results from this study indicate that noncharacteristic fault behavior can also cause order-of-magnitude variations in slip rates through time, even though recurrence intervals have remained relatively consistent. Perhaps other faults in the rift also show large slip-rate variations through time due to noncharacteristic behavior, particularly intrabasin faults with potentially complex rupture patterns like the Hubbell Spring fault system.
3.7 m
0.4 m 1.7 m
THROW (m)
72
4.7 m
Date (ka) Figure 8. Plot showing the variation through time of vertical slip rates for fault splays J and L of the Hubbell Spring fault system. The first complete seismic cycle includes the time interval of strain buildup between events V (= first event) and W (= second event) and the strain (or throw) subsequently released during event W. Similarly, the second seismic cycle includes the time interval between events W and X and the cumulative throw for event X, and so on.
CS2
CS1
Locality
0–5
5–15
15–30
30–50
50–64
64–85
85–110
110–127
127–137 137–143 143–165
165–195 195–237 237–268
268–295 295–323 323–370
370–382
382–402
402–420 420–430 0–5
5–16
Bw
Ck
1Bkb
2Bkb
2Ck
3Btkb
3Bkb2
3Bkb3 4Btk 5Btk2
6K 6K2 6K3
6Btkmb 6Bkmb 7kmb2
8Btkmott
8Btkmott2
Bk Bk2 A/C
Bk
Depth (cm)
A/C
Horizon
7.5YR 7/4
7.5YR 6/5 10YR 6/4 7.5YR 5/4
7.5YR 5/6
7.5YR 5/6
7.5YR 5/4 7.5YR 5/4 7.5YR 6/4
7.5YR 8/3 7.5YR 7/3 7.5YR 6/4
7.5YR 5/4 7.5YR 7/4 7.5YR 7/4
7.5YR 5/4
7.5YR 5/4
7.5YR 4/4
7.5YR 5/4
7.5YR 5/4
7.5YR 5/4
7.5YR 5/4
Moist 7.5YR 4/3
10YR 6/3
7.5YR 7/5 7.5YR 7/6 7.5YR 6/4
7.5YR 6/5
7.5YR 6/5
7.5YR 6/4 7.5YR 7/3 7.5YR 7/3
7.5YR 8/1 7.5YR 8/2 7.5YR 7/2
7.5YR 6/4 7.5YR 7/3 7.5YR 5/3
7.5YR 6/4
7.5YR 6/4
7.5YR 5.5/4
7.5YR 5/4
7.5YR 6/4
7.5YR 5/4
10YR 5/4
25
<1 <1 20
<1
<1
<1 <5 <5
<1 <1 <5
<1 <5 <5
<1
<1
<5
<5
<5
<5
<5
Gravel (%) <5
L
SL L SL
SL/L
SL
SL SL SL
CL L CL
L/SiL SiCL L
L
SiL
LS
SL
SL
SL
SL
SL
Texture
2 f/m sbk
m m g/1 vf/f sbk
2 fm abk
m/2 f/m pl
m m m
2 m sbk 2 f sbk 2 m sbk
2 m abk 2 f sbk 3 f sbk
2/3 m abk
3 m abk
m
2 m sbk
3 mc sbk
2 m sbk
2 m sbk
1 vf sbk
Structure
–
– – –
2 d pf/po
v1 f pf/po
v1 f pf/po – –
– – v1 f pf
– 1/2 fd pfpo v1/1 f pfpo
–
v1 f pf
–
–
–
–
–
–
Clay films
so/sh
vh vh so
h/vh
vh
vh vh vh
sh/h h/vh vf
h vh vh vh
vh
vh
sh
h
h
so sh
so sh
ss/s ps
ss/s ps sp ss ps
ss/s ps/p
ss/s ps
ss ps ss ps ss/s ps
s/vs p ss/s ps vs/vp
sp vs vp ss/s ps
sp
sp
so po
ss ps
ss/s ps
ss ps
ss ps
Consistence Dry Wet so ss ps
aw
as – as
cs
cs
gs gs a w/i
cs cs cs
aw cw aw
gs
cs
as
cs
as
as
aw
as
Boundary
TABLE A1. SOIL PROFILE DESCRIPTIONS FOR THE CARRIZO SPRING TRENCH* Dry 10YR 5/4
Color Notes
Gravels are 90% carbonate nodules <5 mm in diameter 1c 1m 2f 3vf Gravels are 90% carbonate nodules <5 mm in diameter 2m 1f 2vf Gravels ~2 mm diameter, quartz and carbonate grains 1m 1f 2vf Gravels ~2 mm diameter, quartz and carbonate grains 1f 1vf Gravels ~2 mm diameter, quartz and carbonate grains 1f 1vf Coarse sand <2 mm diameter, quartz and carbonate grains 1f 1vf Clay films not obvious; may be overprinted with carbonate 1f Crystalline nodules concentrated are top of horizon 1f 1vf Carbonate coatings on clay films – Crystalline carbonate; contains lens of clay in “sage green” deposit; stage II+ to III 1vf Stage III – Stage III – Stage II+ to III– ; 30% of horizon has 3–5-cm-diameter carbonate cemented, irregularly shaped nodules eh sbk; crystalline carbonate growing on nodule faces – Stage I to II – Stage II – Stage II+ grading into stage III at bottom of horizon – Gravels somewhat coarser 2–5 mm; K overprints redox features; Mn coatings; no modern roots, but traces of roots are visible – Gravels somewhat coarser 2–5 mm; K overprints redox features; Mn coatings; no modern roots, but traces of roots are visible – – 1c 2m “Gravel” is reworked K nodules, real gravel is <1% 1c 2m 2f 3vf “Gravel” is reworked K nodules, real gravel is <1% (Continued)
Roots 1c 2m
73
Horizon
CS4
CS3
16–39
39–47 47–64
64–106
106–127 127–170 170–212 212–236 236–280 0–5
5–17 17–45 45–61 61–70 70–87 87–95 95–117
117–130
165–195 0–? ?–27
27–43
43–61 61–70
70–93
93–110
110–122
K K2
Kmb
Bkmb Bkmb2 2Coxk 3Coxk2 Kb A/C
Bw Bwb Bk 2Bk 3Bkb Bkb2 Bwb
Btkb
K A/C Bw
Bwb
Bkb Bkb2
Bkb3
Kb
Kb2
Depth (cm)
Bkb
CS2 (Continued)
Locality
7.5YR 6/4
7.5YR 5/4
7.5YR 6/4
7.5YR 5/4 7.5YR 6/4
7.5YR 5/4
2.5? 2.5/? 7.5YR 5/4 7.5YR 5/4
7.5YR 5/4
7.5YR 5/4 7.5YR 5/4 7.5YR 5/4 7.5YR 5/4 7.5YR 5/4 7.5YR 5/4 7.5YR 5/4
7.5YR 5/4 7.5YR 5/5 7.5YR 5/5 7.5YR 5/6 7.5YR 6/4 7.5YR 4/4
7.5YR 6/4
7.5YR 7/3 7.5YR 6/4
7.5YR 7/4
Moist
Dry
7.5YR 8/4
7.5YR 8/4
7.5YR 7.5/4
7.5YR 5/4 7.5YR 7/4
7.5YR 6/4
? 7.5YR 6/4 7.5YR 6/4
7.5YR 6/4
7.5YR 6/4 7.5YR 6/4 7.5YR 6/5 7.5YR 5/4 7.5YR 6/4 7.5YR 6/4 7.5YR 6/4
7.5YR 6/6 7.5YR 6/5 7.5YR 6/6 7.5YR 6/4 7.5YR 7/4 7.5YR 6/4
7.5YR 7/3
7.5YR 8/2 7.5YR 7/3
<1
<1
<1
<1 <1
<1
<1 <5 <1
<1
<5 <5 <5 <10 <1 <1 <5
<1 <1 <5 <1 <1 <10
<1
<5 <1
<1
Gravel (%)
CL
CL
L/CL
SL L
SL
SiCL SL SL
L
SL SL SL SL SL SL SL
SL SL SL LS/SL L SL
SL
L L/SiL
L
Texture
2/3 f sbk
2 f abk/sbk
3 m/c abk
2 m sbk 3 m sbk
2 f/m sbk
3 c abk 1 f sbk 2 mc sbk
3 m sbk
2/3 m/c sbk 2 f/m sbk 3 mc sbk 2 f/m sbk 2 m sbk 2 f/m sbk 2 m sbk
3 m abk m m m 3 m abk 1 f sbk
m
m sbk 2/3 m sbk
2 f/m sbk
Structure
–
–
v1 f pf/po
– –
–
– – –
vf/1 f pf/po
– – – – – – –
– – – – – –
–
– –
–
Clay films
sh/h
sh
vh
sh vh
sh
vh so so/sh
h/vh
sh sh sh/h h sh so/sh sh
vh sh/h vh vh vh so
vh
vh vh
vh
–
sp
sp
so/ss ps s ps/p
ss ps
sp so ps so/ss ps
sp
ss ps ss ps ss po/ps ss ps ss ps ss ps ss ps
ss po/ps ss po ss ps so/ss po ss/s ps ss po
ss ps
ss/s ps sp
sp
Consistence Dry Wet
a w/i
cs
aw
as cs
cs
cw – aw
cs
aw cs cs cs cs a/c s a s/w
cs c/g s cs aw – –
gs
cs aw
aw
Boundary
Notes
Stage II+ abundant carbonate cemented krotovina, likely reworking of underlying K 1f Stage III 1f 2vf Stage III– carbonate-cemented krotovina 0.5 f Stage III grading into stage II at boundary 1f Mottles 1f 1m Many roots are decayed – Stronger mottles – Faint mottling – Sparse mottles, nodular stage II+ 1c 2m Scraped away gravels; area a mix of rock and reworked K horizon 1c 2m 2f 3vf Top is scraped away; unit 14 – Unit 13 1c 1m 2f 2vf Unit 13 1f 1vf Coarse unit above “green” unit 13 1m 2f 1f 2vf 1m 2f 2vf “Gravels” are reworked carbonate nodules 2–5 mm in diameter, likely eroded from “black” soil; fine sand – Nodular stage II; found a 1-cmdiameter quartz clast; carbonate on clay films – Stage III– to stage II+; slightly salty – Upper boundary scraped 1c 2m 2f 3vf “Baby laminations” in bottom of horizon; “gravels” are reworked carbonate nodules 1m 2f 2vf Occasional carbonate-cemented boroughs; “gravels” are carbonate nodules and quartzite 3f 3vf Occasional carbonate nodules <2 mm 2f 2vf Stage II nodules and carbonatecemented krotovina; difficult to tell if nodules found in situ or if they are reworked from “black” unit 1m 2f 2vf Stage II+; nodules and carbonate cemented krotovina 1vf Stage II+; both “green” horizons have friable/granular appearance in trench wall 1vf Stage III (Continued)
1m 1f 1vf
Roots
TABLE A1. SOIL PROFILE DESCRIPTIONS FOR THE CARRIZO SPRING TRENCH* (Continued )
7.5YR 8/3
Color
74
Horizon
20–36 36–52 52–75 75–97 97–112 112–130
Bkm Ck 2Ck 2Coxk Kb Kb2
370–397
5Coxk
0–20
225–289 289–312 312–370
2Ck 3Coxk 4Btk
Btk
160–191 191–225
Kb4 2Kmb
7.5YR 6/6 7.5YR 6/6 7.5YR 5/4 7.5YR 5/6 7.5YR 5/5 7.5YR 5/6
7.5YR 6/6
10YR 5/6
7.5YR 5/5 7.5YR 5/6 7.5YR 5/6
7.5YR 7/4 7.5YR 6/6
7.5YR 7/4
Moist
Dry
7.5YR 7/6 7.5YR 6/6 7.5YR 7/6 7.5YR 6/6 7.5YR 7/4 7.5YR 7/6
7.5YR 7/6
10YR 6/6
7.5YR 6/6 7.5YR 6/6 7.5YR 6/5
7.5YR 8/4 7.5YR 7/4
<1 <1 <5 <5 <1 <1
<1
<1
<1 <1 0
<1 <1
<1
Gravel (%)
SL SL SL SL L L
SL
L
SL SL SiCL/SiC
CL/SiCL SL
CL/SiCL
Texture
m m m
m m m m 2/3 m abk 2/3 f/m abk
2/3 c sbk
m
m m 3 c abk
3 f sbk m
3 m/c sbk
Structure
– – –
– – – – v1 f pf/po v1 f pf/po
v1 f pf/po
–
– – 2/3 d/p pf/po
– –
–
Clay films
vh vh h
vh vh vh h vh h
vh
vh
vh vh vh
h/vh vh
vh/eh
s ps s ps s ps
ss ps so ps so po so po s ps ss/s ps
ss ps
s ps
so/ss po/ps ss po/ps s vp
sp ss ps
sp
Consistence Dry Wet
cs cs c
cs c/g s cs aw cs cw
cs
–
gs aw a s/w
a w/i aw
cs
Boundary
– – –
– – – – – –
1m 1f
–
– – –
– –
1vf
Roots
TABLE A1. SOIL PROFILE DESCRIPTIONS FOR THE CARRIZO SPRING TRENCH* (Continued )
7.5YR 8/3
Color
Bkb 130–149 7.5YR 6/6 7.5YR 7/6 <1 L Kb3 149–167 7.5YR 6/4 7.5YR 7/4 <1 L Kb4 162–192 7.5YR 6/4 7.5YR 8/4 <1 L *Soil nomenclature and abbreviations used here follow that of Birkeland et al. (1991).
CS5
122–160
Depth (cm)
Kb3
CS4 (Continued)
Locality
Mottles Mottled, nodular, stage II+ More mottled than above; nodular stage II Overprinted in areas from above K Nodular stage II Nodular stage II
Occasional mottling and Mn coatings; effervescent salts “sweating” out of trench wall; + clays are prominent because of abundence in p.m.; no horizonation likely because of bioturbation; veins and isolated nodules of carbonate, otherwise no effervescence; no modern roots but abundant traces w/ oxidation Orange mottles; carbonate stringer; mottling is greatest at top of horizon Orange mottles; carbonate nodules and linings in fractures Boundary marked by end of mottling
Stage III+ & stage II w/ spring; nodules >5 cm diameter; irregular morphology Friable/granular texture in wall 1–5 cm spring nodules, common nodules are eh; a 2–5-cm-thick zone of stage III carbonate at bottom of horizon, otherwise stage II+
Notes
75
76
Olig et al.
ACKNOWLEDGMENTS This study was funded by the U.S. Geological Survey (USGS) through their National Earthquake Hazard Reduction Program (award 99HQGR0089), in cooperation with the New Mexico Bureau of Geology and Mineral Resources (NMBGMR), and with support for report preparation from the URS Corporation Professional Development Fund. The views and conclusions contained in this document are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied, of the U.S. government. We thank the Cordova family, and especially Pete Cordova, for granting access and permission to conduct this study on their ranch. We also thank Ann Tillery and Nicole Bailey (formerly University of New Mexico) for assistance with logging trenches. Mike Machette (USGS) loaned us aerial photographs. Sean Connell (NMBGMR), John Hawley (Hawley Geomatters), and Steve Personius (USGS) provided helpful discussions. Jim McCalpin (GeoHaz Consulting) provided helpful review comments that improved the manuscript. Judy Zachariasen (URS) prepared the geographic information system map for Figure 1, and Fumiko Goss, Mathew Smith, and Melinda Lee (URS) assisted with manuscript preparation. REFERENCES CITED Birkeland, P.W., Machette, M.N., and Haller, K.M., 1991, Soils as a tool for applied Quaternary geology: Utah Geological and Mineral Survey Miscellaneous Publication 91-3, 63 p. Bucknam, R.C., and Anderson, R.E., 1979, Estimation of fault-scarp ages from a scarp-height–slope-angle relationship: Geology, v. 7, p. 11–14, doi:10 .1130/0091-7613(1979)7<11:EOFAFA>2.0.CO;2. Chapin, C.E., 1971, The Rio Grande rift: Part I. Modifications and additions: New Mexico Geological Society Guidebook, v. 22, p. 191–201. Chapin, C.E., and Cather, S.M., 1994, Tectonic setting of the axial basins of the northern and central Rio Grande rift, in Keller, G.R., and Cather, S.M., eds., 1994, Basins of the Rio Grande Rift: Structure, Stratigraphy, and Tectonic Setting: Geological Society of America Special Paper 291, p. 5–25. Connell, S.D., Love, D.W., Sorrell, J.D., and Harrison, J.B.J., 2001, Plio-Pleistocene Stratigraphy and Geomorphology of the Central Part of the Albuquerque Basin; 45th Field Conference of the Rocky Mountain Cell of the Friends of the Pleistocene: New Mexico Bureau of Geology and Mineral Resources Open-File Report 454C and D, variously paginated. Foley, L.L., LaForge, R.C., and Piety, L.A., 1988, Seismotectonic study for Elephant Butte and Caballo Dams, Rio Grande Project, New Mexico: U.S. Bureau of Reclamation Seismotectonic Report 88-9, 60 p., 1 plate, scale 1:24,000. Forman, S.L., 1999, Infrared and red-stimulated luminescence dating of Quaternary sediments from Spitsbergen, Svalbard: Arctic, Antarctic, and Alpine Research, v. 31, p. 34–49, doi:10.2307/1552621. Forman, S.L., Jackson, M.E., McCalpin, J., and Maat, P., 1988, The potential of using thermoluminescence to date buried soils developed on colluvial and fluvial sediments from Utah and Colorado, U.S.A.: Preliminary results: Quaternary Science Reviews, v. 7, p. 287–293, doi:10.1016/0277 -3791(88)90017-0. Forman, S.L., Pierson, J., and Lepper, L., 1999, Luminescence geochronology, in Noller, J.S., Sowers, J.M., and Lettis, W.R., eds., Quaternary Geochronology: Methods and Applications: Washington, D.C., American Geophysical Union, p. 157–176. Frankel, A., Petersen, M., Mueller, C., Haller, K., Wheeler, R., Leyendecker, E., Wesson, R., Harmsen, S., Cramer, C., Perkins, D., and Rukstales, K., 2002, Documentation for the 2002 Update of the National Seismic Hazard Maps: U.S. Geological Survey Open-File Report 02-420, 33 p.
Gardner, J.N., Lewis, C.J., Reneau, S.L., Shultz-Fellenz, E.S., Lavine, A., and Olig, S., 2008, Paleoseismology and geology of the Pajarito fault system, Rio Grande rift, New Mexico: Seismological Research Letters, v. 79, no. 2, p. 275–276. Grauch, V.J.S., 1999, Discussion of new gravity maps for the Albuquerque Basin area, in Pazzaglia, F.J., and Lucas, S.G., eds., Albuquerque Geology: New Mexico Geological Society Guidebook, v. 50, p. 119–124. Grauch, V.J.S., 2001, High-resolution aeromagnetic data, a new tool for mapping intrabasinal faults: Example from the Albuquerque Basin, New Mexico: Geology, v. 29, p. 367–370, doi:10.1130/0091-7613(2001)029 <0367:HRADAN>2.0.CO;2. Grauch, V.J.S., and Hudson, M.R., 2002, Implications of significant distances between major faults and large vertical displacements based on geophysical evidence, central Rio Grande rift, New Mexico: Geological Society of America Abstracts with Programs, v. 34, no. 6, p. 452. Hawley, J.W., 1986, Physiographic provinces (and) landforms of New Mexico, in Williams, J.L., ed., New Mexico in Maps: Albuquerque, University of New Mexico Press, p. 28–31. Hawley, J.W., Kottlowski, F.E., Seager, W.R., King, W.E., Strain, W.S., and LeMone, D.V., 1969, The Santa Fe Group in the South-Central New Mexico Border Region: New Mexico Bureau of Mines and Mineral Resources Circular 104, p. 235–274. Hawley, J.W., Haase, C.S., and Lozinsky, R.P., 1995, An underground view of the Albuquerque Basin, in Ortega-Klett, C.T., ed., The Water Future of Albuquerque and the Middle Rio Grande Basin: Las Cruces, New Mexico Water Resources Research Institute, p. 27–55. Karlstrom, K.E., Brown, C., Armour, J., Lewis, J., and Connell, S., 2001, Geology of the Bosque Peak 7.5-Minute Quadrangle, Valencia and Torrance Counties, New Mexico: New Mexico Bureau of Mines and Mineral Resources Open-File Digital Geologic Map DM 24, scale 1:12,000. Keller, G.R., and Cather, S.M., eds., 1994, Basins of the Rio Grande Rift: Structure, Stratigraphy, and Tectonic Setting: Geological Society of America Special Paper 291, 304 p. Kelley, V.C., 1977, Geology of Albuquerque Basin, New Mexico: New Mexico Bureau of Mines and Mineral Resources Memoir 33, 59 p. Love, D.W., 1998, Geology of the Isleta 7.5-Minute Quadrangle, Bernalillo and Valencia Counties, New Mexico: New Mexico Bureau of Mines and Mineral Resources Open-File Digital Map DM 13, scale 1:24,000. Love, D.W., Hitchcock, C., Thomas, E., Kelson, K., Van Hart, D., Cather, S., Chamberlin, R., Anderson, O., Hawley, J., Gillentine, J., White, W., Noler, J., Sawyer, T., Nyman, M., and Harrison, B., 1996, Geology of the Hubbell Spring 7.5-Min Quadrangle, Bernalillo and Valencia Counties, New Mexico: New Mexico Bureau of Mines and Mineral Resources Open-File Digital Geologic Map 5, scale 1:12,000. Lozinsky, R.P., 1994, Cenozoic stratigraphy, sandstone petrology, and depositional history of the Albuquerque Basin, central New Mexico, in Keller, G.R., and Cather, S.M., eds., 1994, Basins of the Rio Grande Rift: Structure, Stratigraphy, and Tectonic Setting: Geological Society of America Special Paper 291, p. 73–81. Machette, M.N., 1978, Geologic Map of the San Acacia Quadrangle, Socorro County, New Mexico: U.S. Geological Survey Geologic Quadrangle Map GQ-1415, scale 1:24,000. Machette, M.N., 1985, Calcic soils of the southwestern United States, in Weide, D.L., ed., Quaternary Soils and Geomorphology of the American Southwest: Geological Society of America Special Paper 203, p. 1–21. Machette, M.N., 1998, Contrasts between short- and long-term records of seismicity in the Rio Grande rift—Important implications for seismic hazard assessments in areas of slow extension, in Lund, W.R., ed., Western States Seismic Policy Council Proceedings Volume, Basin and Range Province Seismic Hazards Summit: Utah Geological Survey Miscellaneous Publication 98-2, p. 84–95. Machette, M.N., and McGimsey, R.G., 1983, Map of Quaternary and Pliocene Faults in the Socorro and Western Part of the Fort Sumner 1° × 2° Quadrangles, Central New Mexico: U.S. Geological Survey Miscellaneous Field Studies Map MF-1465-A, scale 1:250,000. Machette, M.N., Personius, S.F., Kelson, K.I., Haller, K.M., and Dart, R.L., 1998, Map and Data for Quaternary Faults and Folds in New Mexico: U.S. Geological Survey Open-File Report 98-521, 443 p. Maldonado, F., Connell, S.D., Love, D.W., Grouch, V.J.S., Slate, J.L., McIntosh, W.C., Jackson, P.B., and Byers, F.M., Jr., 1999, Neogene geology of the Isleta Reservation and vicinity, Albuquerque Basin, New Mexico, in
Evidence for noncharacteristic ruptures of intrabasin faults in the Rio Grande rift Pazzaglia, F.J., and Lucas, S.G., eds., Albuquerque Geology: New Mexico Geological Society Guidebook, v. 50, p. 175–188. Maldonado, F., Slate, J.L., Love, D.W., Connell, S.D., Cole, J.C., and Karlstrom, K.E., 2007, Geologic Map of the Pueblo Isleta Tribal Lands and Vicinity, Bernalillo, Torrance, and Valencia Counties, Central New Mexico: U.S. Geological Survey Scientific Investigations Map 2913, scale 1:100,000. McCalpin, J.P., 1995, Frequency distribution of geologically determined slip rates for normal faults in the western U.S.: Bulletin of the Seismological Society of America, v. 85, p. 1867–1872. McCalpin, J.P., Olig, S.S., Harrison, J.B.J., and Berger, G.W., 2006, Quaternary Faulting and Soil Formation on the County Dump Fault, Albuquerque, New Mexico: New Mexico Bureau of Mines and Geology Circular 212, 54 p. McCraw, D.J., Love, D.W., and Connell, S.D., 2006, Preliminary Geologic Map of the Abeytas Quadrangle, Socorro County, New Mexico: New Mexico Bureau of Geology and Mineral Resources Open-File Digital Geologic Map OF-GM 121, scale 1:24,000. Morgan, P., Seager, W.R., and Golombek, M.P., 1986, Cenozoic thermal, mechanical and tectonic evolution of the Rio Grande rift: Journal of Geophysical Research, v. 91, p. 6263–6276, doi:10.1029/JB091iB06p06263. Nelson, A.R., 1992, Lithofacies analysis of colluvial sediments—An aid in interpreting the recent history of Quaternary normal faults in the Basin and Range Province, western United States: Journal of Sedimentary Petrology, v. 63, p. 607–621. Olig, S., and Zachariasen, J., 2008, Additional Analyses and Mapping of the Hubbell Spring and Other Intrabasin Faults South of Albuquerque, New Mexico: New Mexico Bureau of Geology and Mineral Resources OpenFile Report 527, 49 p. and 1 plate (scale 1:100,000). Olig, S.S., Eppes, M.C., Forman, S.L., Love, D.W., and Allen, B.D., 2004, Paleoseismic Investigation of the Central Hubbell Spring Fault, Central New Mexico: Oakland, California, URS Corporation Final Technical Report to the U.S. Geological Survey, NEHRP award no. 99HQGR0089, URS job no. 26813901, variously paginated. Olig, S., Zachariasen, J., Wong, I.G., and Dober, M.C., 2007, Paleoseismic evidence for longer and more complex rupture patterns on the Hubbell Spring fault system, Rio Grande rift, New Mexico: Implications for recurrence models and their use in hazard analysis [abs.]: Seismological Research Letters, v. 78, no. 2, p. 315. Personius, S.F., and Mahan, S.A., 2003, Paleoearthquakes and eolian-dominated fault sedimentation along the HSFS near Albuquerque, New Mexico: Bulletin of the Seismological Society of America, v. 93, p. 1355–1369, doi:10.1785/0120020031. Personius, S.F., Machette, M.N., and Kelson, K.I., 1999, Quaternary faults in the Albuquerque area—An update, in Pazzaglia, F.J., and Lucas, S.G., eds., Albuquerque Geology: New Mexico Geological Society Guidebook, v. 50, p. 189–200. Personius, S.F., Eppes, M.C., Mahan, S.A., Love, D.W., Mitchell, D.K., and Murphy, A., 2001, Log and data from a trench across the HSFS, Bernalillo County, New Mexico: U.S. Geological Survey Miscellaneous Field Studies Map MF-2348, version 1.1. Rawling, G., and McCraw, D.J., 2004, Preliminary Geologic Map of the Tome NE 7.5-Minute Quadrangle: New Mexico Bureau of Geology and Min-
77
eral Resources Draft Open-File Digital Geologic Map OF-GM 00, scale 1:24,000. Read, C.B., Wilpolt, R.H., Andrews, D.A., Summerson, C.H., and Wood, G.H., 1944, Geologic Map and Stratigraphic Sections of Permian and Pennsylvanian Rocks of Parts of San Miguel, Santa Fe, Sandoval, Bernalillo, Torrance, and Valencia Counties, North-Central New Mexico: U.S. Geological Survey Oil and Gas Investigations Preliminary Map 21, scale 1:190,080, 1 sheet. Reiche, P., 1949, Geology of the Manzanita and North Manzano Mountains, New Mexico: Geological Society of America Bulletin, v. 60, p. 1183– 1212, doi:10.1130/0016-7606(1949)60[1183:GOTMAN]2.0.CO;2. Russell, L.R., and Snelson, S., 1994, Structure and tectonics of the Albuquerque Basin segment of the Rio Grande rift: Insights from reflection seismic data, in Keller, G.R., and Cather, S.M., eds., 1994, Basins of the Rio Grande Rift: Structure, Stratigraphy, and Tectonic Setting: Geological Society of America Special Paper 291, p. 82–112. Sanford, A.R., Jaksha, L.H., and Cash, D.J., 1991, Seismicity of the Rio Grande rift in New Mexico, in Slemmons, D.B., Engdahl, E.R., Zoback, M.D., and Blackwell, D.D., eds., Neotectonics of North America: Boulder, Colorado, Geological Society of America, Decade of North American Geology Map, v. 1, p. 229–244. Schwartz, D.P., and Coppersmith, K.J., 1984, Fault behavior and characteristic earthquakes—Examples from the Wasatch and San Andreas fault zones: Journal of Geophysical Research, v. 89, p. 5681–5698, doi:10.1029/ JB089iB07p05681. Spooner, N.A., Aitken, M.J., Smith, B.W., Franks, M., and McElroy, C., 1990, Archaeological dating by infrared-stimulated luminescence using a diode array: Radiation Protection Dosimetry, v. 34, p. 82–86. Swan, F.H., III, Schwartz, D.P., and Cluff, L.S., 1980, Recurrence of moderate to large magnitude earthquakes produced by surface faulting on the Wasatch fault zone, Utah: Bulletin of the Seismological Society of America, v. 70, p. 1431–1462. Wells, D.L., and Coppersmith, K.J., 1994, Analysis of empirical relationships among magnitude, rupture length, rupture area, and surface displacement: Bulletin of the Seismological Society of America, v. 84, p. 974–1002. Wong, I., and Olig, S.S., 1998, Seismic hazards in the Basin and Range Province: Perspectives from probabilistic analyses, in Lund, W.R., ed., Western States Seismic Policy Council Proceedings Volume, Basin and Range Province Seismic Hazards Summit: Utah Geological Survey Miscellaneous Publication 98-2, p. 110–127. Wong, I., Olig, S., Dober, M., Silva, W., Wright, D., Thomas, P., Gregor, N., Sanford, A., Lin, K.W., and Love, D., 2004, Earthquake ground-shaking hazard maps for the Albuquerque–Belen–Santa Fe, New Mexico, corridor, New Mexico Geology: New Mexico Bureau of Geology and Mineral Resources, v. 26, p. 3–33. Woodward, L.A., 1982, Tectonic framework of Albuquerque County: New Mexico Geological Society Guidebook, v. 33, p. 141–146. Youngs, R.R., and Coppersmith, K.J., 1985, Implications of fault slip rates and earthquake recurrence models to probabilistic seismic hazard estimates: Bulletin of the Seismological Society of America, v. 75, p. 939–964. MANUSCRIPT ACCEPTED BY THE SOCIETY 7 DECEMBER 2010
Printed in the USA
The Geological Society of America Special Paper 479 2011
Large-magnitude late Holocene seismic activity in the Pereira-Armenia region, Colombia Claudia Patricia Lalinde P.* Environmental Geology Group, Eafit University, Medellín, A.A. 3300, Colombia Gloria Elena Toro* Geology Department, Eafit University, Medellín, A.A. 3300, Colombia Andrés Velásquez* Observatorio Sismológico del Suroccidente–OSSO / Corporación OSSO, Carrera 101 No. 14-154, Cali, Colombia Franck A. Audemard M.* Fundación Venezolana de Investigaciones Sismológicas (FUNVISIS), El Llanito, Caracas 1073, Venezuela
ABSTRACT The Pereira-Armenia region, located west of the Colombian Central Cordillera, is crosscut by the Romeral fault system, which consists of an active north-south– trending, left-lateral, strike-slip fault system with a secondary thrust component in the Eje Cafetero zone (4°°N–5°°N). The terrain where the Liceo Taller San Miguel high school sits—9 km south of Pereira—is draped with an ~2-m-thick layer of volcanic ash younger than 30 k.y. in age. This locality has been affected by both N40°°E- and E-W–trending faults that correspond to thrust faults or folds and normal rightlateral, strike-slip faults, respectively, in the tectonic model for the zone. Two kinds of strong field evidence for the E-W faults were found at a site named Canchas: (1) the 50°°N tilt of the late Quaternary interbedded sequence of volcanic ash and three paleosols, and (2) a vertical fault throw of ~1.70 m affecting the sequence (layers). A normal vertical throw of ~0.65 m at Parqueadero stands as a proof of the activity of the N40°E-trending faults. This latter faulting does not correspond with the stress tensor proposed for this region, and thus this deformation could be interpreted as being a consequence of flexural slip induced by a NE-SW–striking blind thrust, where reverse faulting along bedding at depth is seen as normal faulting at the surface. Measured offsets could have generated seismic events of at least Mw 6.6 for the NE-trending fault that affected the paleosols and volcanic ash sequence at 13,150 ± 310 14C yr B.P., and a seismic event of Mw 6.9 for the E-W–trending fault that affected the paleosols and volcanic ash sequence at 19,710 ± 830 14C yr B.P. These two recently identified
*E-mails:
[email protected];
[email protected];
[email protected];
[email protected]. Lalinde P., C.P., Toro, G.E., Velásquez, A., and Audemard M., F.A., 2011, Large-magnitude late Holocene seismic activity in the Pereira-Armenia region, Colombia, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, p. 79–89, doi:10.1130/2011.2479(03). For permission to copy, contact
[email protected]. © 2011 The Geological Society of America. All rights reserved.
79
80
Lalinde P. et al. faults are now named the Tribunas (NE-SW) and the Cestillal (E-W) faults. Up to now, the fault and its seismogenic potential determinations in this region have been based solely on morphologic evidence. The maximum seismic magnitude estimated for this region ranged from Mw 6.2 to Mw 6.6 for seismic sources 35 km away from the site. Seismic magnitudes like the one calculated in this work (Mw 6.9) were previously estimated only for source-site distances greater than 50 km. This work provides field evidence that leads to a better understanding of the seismic activity of this region in the last 30 k.y. and confirms the occurrence of local Mw >6.5 seismic events in this region. Although volcanic ash drapes and eventually hides the geomorphic evidence of active deformation, it turns out to be a perfect chronometer of a fault’s activity whenever the deformation is revealed, as in this case. After the Armenia event of 1999, it is imperative to examine the seismic hazard assessments of this region in terms of local crustal seismicity.
INTRODUCTION Northern South America lies at the junction of the Nazca, Caribbean, and South American plates (Fig. 1). The accumulated stress of the plates is released over the Andean block, which includes the Eastern, Central, and Western Cordilleras (Cline et al., 1981; Ego et al., 1996). Holocene intracontinental deformation in the Andean block can reactivate older fault systems like the Romeral, which is one of the most important fault systems extending across Colombian territory (Page, 1986). The Pereira-Armenia region is located west of the Central Cordillera of Colombia, and it is characterized as a fluvial fan system (Thouret, 1983). The Romeral fault system, which crosses the Pereira-Armenia fan in a north-south direction, is a left-lateral, strike-slip fault system with a secondary thrust component. All faults that cross the Pereira-Armenia fan are part of the Romeral fault system. North of 5°N, at present, a maximum NW-SE–oriented horizontal stress (σ1) is assumed (Guzmán et al., 1998). According to the deformation model (Silvestre and Smith, 1976 in Yeats et al., 1997; Keller, 1986), the PereiraArmenia region should display normal faults trending NW, thrust faults trending NE, and right-lateral normal faults trending E-W. The Liceo Taller San Miguel high school is located 9 km south of the city of Pereira, over the upper part of the Pereira-Armenia fan near its apex. In this area, the volcanic ash cover is ~2 m thick and younger than 30 ka (19,710 ± 830 14C yr B.P. and 13,150 ± 310 14 C yr B.P.). Paleoseismic studies in Colombia are needed to complement instrumental records that only cover less than 100 yr and historical records that extend back to the beginning of the sixteenth century. Therefore, the only way of enlarging the seismic record back in time is through the study of paleoseismicity. Our approach is to estimate the maximum seismic event per fault, maximum rupture length, time to the next event, and recurrence interval between equivalent events. The Pereira-Armenia region became the focus of paleoseismic interest in January 1999, when a Mw 6.1 shallow crustal earthquake occurred near the city of Armenia. At that time, the
scientific community became aware of the need for a study of intraplate faults and their seismic potential in the Pereira-Armenia region, thus moving beyond the evaluation of the regional seismic hazards generated by seismicity of Nazca plate subduction. For this study, we collected and assessed seismic information from previous studies in the Pereira-Armenia region. We also studied the volcanic ash sequence that covers the region, going back 50 k.y. Based on the geomorphologic study, fieldwork, aerial photographic interpretation, and the assessment of previous works, we selected different places to study the paleoseismicity of the region. One of these places was the Liceo Taller San Miguel, where we mapped two artificial exposures made during the building of a school. Besides the scientific interest of the Pereira-Armenia region, important future human settlements and economic developments urgently require a rigorous knowledge of past seismic events in order to reduce uncertainties in seismic hazard assessment. STUDY AREA The Pereira-Armenia fan is a torrential volcanic fan, the origin of which is related to the activity of the Ruiz Tolima massif during the past 4.5 m.y., the uprising of the Central Cordillera of Colombia, and movements of the faults that cross the region (Thouret, 1983) (Fig. 1). On the Pereira-Armenia fan surface numerous faults were identified (Fig. 2); anomalous geomorphologic and topographic features are common, such as folds in its distal part, drainage anomalies, benches, and other anomalies that are difficult to explain in relation to the normal evolution of a torrential volcanic fan (e.g., Thouret, 1983; Page, 1986; Guzmán et al., 1998). A more detailed description of the Pereira-Armenia region shows a very complex fan where 14 geomorphologic units are recognized (Lalinde, 2004). Despite these efforts, the sequence of events that formed the Pereira-Armenia fan and their precise chronology are not well known. More detailed investigation is needed to identify the different volcano-sedimentary events, gain more precise knowledge of their chronology, and interpret their
Large-magnitude late Holocene seismic activity in the Pereira-Armenia region, Colombia
81°W 12°N
79°W
N
73°W
75°W Caribbean plate
77°W
81
1.7 cm/yr Oca fault
Santa Marta M Sta arta cara
ve
na Riv er sys tem
Thrust fault
ata
s fa
Ibag
. ue F
Armenia em
A
fa ira m ault a t f Al ras ci lge
Llanos Orientales
Villavicencio
Study zone Pereira-Armenia fan (see Fig. 2)
South American plate
Ro
me
ral
fau
6.4 cm/yr
yst
Buenaventura Cali
lt
Tunja
Bogotá t ul
River Ocean
C.E. fau
Pereira
lt s
p rra Ga
ult
Normal fault Strike-slip fault
al
C.C.
Volcanic Massif Ruiz Tolima
Pacific Ocean
4°
ale
Ma
C.W.
Nazca plate
gd
Medellín
Coiba microplate
C.W. Western Cordillera C.C. Central Cordillera C.E. Eastern Cordillera
Andean block
ult
6°
r
fault
a fa
Ca
Ri
nas
mit
8°
a uc
Maracaibo Basin
ault
Ura
á
Sali
m
0.8 cm/yr
ga f
na
man
Pa
– Bu
Caribbean Sea
Yo p
10°
Figure 1. The study zone, Pereira-Armenia fan, located in the western flank of the Central Cordillera (C.C.), and a part of the Andean block. Figure is modified from Taboada et al. (2000). The movement vectors are from Freymueller et al. (1993) and Kellogg and Vega (1995).
paleoseismic information. We interpret the main scarps in the fan as fault movements. Though there is also an erosional component, based on the available information, it is not possible to resolve this component. By knowing the fan terrace surfaces and their ages, these two components can eventually be separated in the future. Lalinde (2004) described the Pereira-Armenia fan based on topographic profiles (scale 1:25,000) as follows: (1) The Pereira-Armenia fan is highest in the northern and middle parts, and its height gradually decreases to the south from the Roble River. This is evident in the N-S profiles B–Bʹ, C–Cʹ, and F–Fʹ (Figs. 3 and 4). In the distal part, the fan shows a flat surface (Figs. 3 and 4, profiles N–Nʹ, H–Hʹ), with some uplifted blocks that could be the result of erosional and tectonic processes. These rivers are more incised on the eastern part of the fan than to the west. Depth changes in the river profiles suggest active tectonism and vertical movement associated with E-W faults. West of the fan, there is a series of steps that corresponds to N30°–60°E faults.
(2) The NW profiles show the same evidence in the northern and southern parts of the fan and suggest a division at Cestillal drainage or Barbas River. This change corresponds to the trace of Cestillal fault (Figs. 3 and 4). (3) The NE profiles confirm the differences between the north and south parts of the fan (Figs. 3 and 4). As mentioned already, the Pereira-Armenia fan is crossed by numerous faults that belong to the Romeral fault system (Fig. 2) (Thouret, 1983; Page, 1986; Guzmán et al., 1998). Among them, here we report two important faults with evidence of activity at Liceo Taller San Miguel that were not previously defined (Figs. 2, 5, and 6). These faults are the Cestillal fault and Tribunas fault. Cestillal Fault The Cestillal fault is an EW/50°S normal fault that crosses the Pereira-Armenia fan. In the topographic profiles (Figs. 3 and 4), the northern block appears to be 50 m high; part of this displacement could be due to the erosional process. This fault is not
7°
F de Uramita
u F. M rind
N
N
ó
Medellín 6° Quibdó Ju
an
Alcalá F.
n ró e u oq B F. s a un ir b T
R.S.
*
Cuba F. Cerritos
4°
Cali
78°
77°
Dextral strikeslip fault Sinistral strikeslip fault Reverse fault Normal fault
76°
75°
3°N 74°W
Compression force NW-SE Distensive force NE-SW
Ot
ú
Monte
e Pu
D
El Roble River
e ag
F.
ain
Dr
es
lo
bl
ch
Ro
CPL25 F. a i en *
Ar
m
Po
rta
negro
F.
F. Mate ca
ña L.
Alcalá L ver
rto
Sa
ria
a
m
Ri
aC
agu
nR F. Structural model ive Co r C *LTSM on nsot aR s Cestillal D rainage CPL10 ota iver Barba F. s Rive r eta * n u * Lacteos Amaltea Lag
UD
La Vieja
Planes del
UD tigre U Cestillal fault D Kvo
.
Ce
rri
to
sF .
Pereira
5°
a Armenia alim
s F. D
F.
Kvo
ué
g Ibagué F. Iba
pata
arra
F. G
ue
uca Ri v er Ca
F. P ale stin a
Manizales Pereira
F. Honda
an
oS
Rí F.
Salento F.
Explanation Kvo
U
Pereira-Armenia fan
iv
er
El Laurel F.D CPL27 U
U D
R ío
L U D R.S. LTSM
me
Mo
nia
nte
F.
ne
Q
gr
ui
nd
oF .
Armenia
Espejo R
Ar
iver
Diabasic Group Faults from Romeral fault system Lineament Upthrown side Downthrown side Relleno Sanitario site Liceo Taller San Miguel site
* CPL25 Label of field station work River Sources: Ingeominas (1999) Cardona and Ortíz (1994) Present work Graph scale (km)
0
10
20
Figure 2. Schematic map of rivers and faults on the Pereira-Armenia fan. The Liceo Taller San Miguel high school (LTSM) is located 9 km south of the city of Pereira. Figure shows the best exposed fault segments in the region. The Pereira-Armenia fan is located between 4°–5°N and 75°–76°W. F—fault.
Cau ca Riv er
H
N
N F Otún River Consota River
Pereira
Cuba
C B
Cestillal Dra
inage
Barbas River Ulloa Filandia
ch
ue
La
lo
Dr
ai
na
ge
Vi eja Ri ve r
Alcalá
Po
rta
Quimbaya
Circacia
Bʹ
Qu
Armenia
ind
Montenegro
ío
Ri ve r
Roble River
Espejo River
Cʹ
Calarcá
Fʹ
Pereira-Armenia fan Hʹ
Nʹ
0
5
10 km
Figure 3. Location of the N-S topographic profiles. The Pereira-Armenia fan is located between 4°–5°N and 75°–76°W.
Height (m)
Height (m)
Height (m)
850
950
1050
1150
1250
1350
1450
1200
1300
1400
1500
1600
1700
1500
1600
1700
1800
1900
2000
2100
N
0
N
N
0
0
Dosquebradas Drainage Otún River Consota River
Cauca River Tributary of Otún River
Otún River
10
El Oso Drainage
Otún River 5
Consota River
Consota River 10
20
Cestillal fault
Cestillal Drainage Barbas River
Barbas River
30
20
40
H–H´ Profile
15
F–F´ Profile
Distance (km)
10
Distance (km)
Lacha Drainage El Chorro Drainage
Roble River
50
B–B´ Profile
Roble River
Distance (km)
Cestillal Drainage Barbas River
Los Angeles Drainage San Felipe Drainage Campoalegre Drainage Buenavista Drainage Roble River
30
Quindio River
20
Hojasanchas Drainage
Espejo River 60
S
S
S 25
Height (m) 0
Figure 4. N-S topographic profiles. There are a lot of topographic anomalies in the Pereira-Armenia fan. In the H–Hʹ profile, note the topographic anomaly that corresponds to the Cestillal fault. In the other profiles, this fault is not clear; this could be because of the major cover of volcanic ash fall and paleosols or some Quaternary deposits that are not yet identified in the region.
1400
1500
1600
1700
1800
1900
2000
N
Height (m) 1000
1100
1200
1300
1400
1500
1600
Otún River
0
N
5
San Pablo Drainage
Otún River
Consota River
Consota River
Barbas River
10
Cestillal Drainage
Los Angeles Drainage 20
Distance (km)
N–N´ Profile
Aguaserena Drainage
15
Distance (km)
San Felipe Drainage Quimbaya Town Buenavista Drainage
Roble River Montenegro Town
Membrillal Drainage Roble River
40
25
C–C´ Profile
20
Espejo River
Quindío River
S
60
S
30
84
Large-magnitude late Holocene seismic activity in the Pereira-Armenia region, Colombia
Figure 5. Liceo Taller San Miguel high school. View to south. Photograph shows the location of the Canchas and Parqueadero sites, the two sites studied in detail. The Canchas site (left center) is located behind the building where the playground of the high school is located. The Parqueadero site (far right) is in front of the building where the parking place is located. The Canchas site is located 200 m south of the Parqueadero site, and the top of the Canchas site exposure is 2 m below the floor of the Parqueadero site exposure.
Canchas site Parqueadero site
Cestillal fault
N25°W
6
P2 1m
S
1
P2
P1b
5
8
4 3
6
P2
P1c
P0b
8 P1 9 10
19.710 ± 830 yr B.P.
P0
3m
1m
7 8
P0a
11 2m
6
4
0m
P1a
6
7
5
2
5
7
12
13
14
4m
N
12
A
N25°W N
S
Cestillal fault
19.710 ± 830 yr B.P.
85
-Paleosol 1 (P1): light brown to reddish. Hard. -Horizon 6 (6): gray clay (5Y/5/2), thixotropic. -Horizon 5 (5): yellow sand (lapilli) (10YR/5/8), thixotropic. -Horizon 4 (4): gray clay (2.5Y/5/2), thixotropic. -Paleosol 2 (P2): dark, less hard than paleosol 1. Its age is 19.710 ± 830 yr B.P. -Horizon 3 (3): gray clay (5Y/5/2), thixotropic. -Horizon 2 (2): white gray clay (2.5Y/7/1), thixotropic. -Horizon 1 (1): gray clay (5Y/5/1), thixotropic.
B Figure 6. Liceo Taller San Miguel, Canchas site. (A) Sketch of the east wall exposure. Grid lines are 1 m apart. (B) Photograph of the east wall. The Cestillal fault displaces the volcanic ash-fall and paleosol sequence. The sequence is tilted 50°N, and the total displacement along the three fault segments is 1.70 m. Unit 6 is shaded for ease of visualization. The north structure displaces the sequence of volcanic ash falls and paleosols 0.60 m; the middle structure displaces it 0.10 m; and the south structure displaces the sequence 1.00 m. See Figure 7 for description of layers 7–11.
86
Lalinde P. et al.
clear in aerial photographs, which could be related to volcanic ash draping. The Cestillal fault has not been previously mapped or reported, so there is no information to confirm that this fault is 35 km long. Evidence of this segment is a change in course of the La Vieja River (Fig. 2), where it changes its direction from N-S to E-W near the mouths of the Barbas and Consota Rivers, and the change in the course of the Cauca River near Cartago city (Fig. 2). Tribunas Fault This is a N40°E normal fault that could be a secondary fault related to a blind thrust fault. The Tribunas fault is not very clear in the aerial photographs, but some drainage anomalies exist in the trace of the fault, such as a displacement of the Consota River. This fault occurs in the upper part of the Pereira-Armenia fan. DESCRIPTION OF LICEO TALLER SAN MIGUEL The landscape at Liceo Taller San Miguel is rather smooth, exhibiting small hills and undulations. Stream incision is superficial, and in some sites, it is common to find topographic depressions near the streams, which in general are oriented N40°–50°E. This land is used for recreational purposes, cattle grazing, and cultivation. Liceo Taller San Miguel is a high school situated on the Pereira-Armenia fan, and it lies 9 km south of the city of Pereira, near the road that links the cities of Pereira and Armenia (Fig. 2). The geographic location of Liceo Taller San Miguel high school is 4°44ʹ56.17ʺN, 75°40ʹ28.56ʺW (WGS84 international spheroid). During the building of the high school, some artificial outcrops were created, which happened to reveal very useful neotectonic information. We selected two specific places that showed evidence of recent tectonic activity. These places are the so-called Canchas site and the Parqueadero site. The Canchas site is located ~200 m south of the Parqueadero site, and the roof of the Canchas site is ~2 m below the floor of the Parqueadero site. Volcanic ashes cap these sites with a thickness of over 4 m. Canchas Site Geomorphology Near the Canchas site, the landscape is rather smooth, exhibiting small hills and a few undulations. Stream incision is superficial, and in some sites, it is common to find topographic depressions near the streams. This site has human affectation and does not have the complete stratigraphic sequence. The most recent affectation was the artificial cut surface created for a basketball court. Stratigraphy and Soils At the Canchas site, the Pereira-Armenia fan is covered by a sequence of volcanic ashes and paleosols (Fig. 6). We identified two paleosols that represent periods of extended nondeposition of
volcanic ash, or if there were some deposition of volcanic ashes, it was not enough to interrupt soil formation. Between these two pedogenic periods, there are three altered volcanic ash horizons. We consider that the stratigraphic sequence is incomplete due to pre-Columbian human activity, at least before 2500 yr B.P. (Cano, 2004). Because of the artificial cut, we do not know what the natural surface looked like. Structure The sequence of paleosols and volcanic ashes at Canchas site is tilted 50°N and displaced or cut by three faults (Fig. 6). The throws of these three faults, from north to south, are 0.60 m, 0.10 m, and 1.00 m, respectively. The minimum cumulative offset of the volcanic ash layer and paleosol in this fault zone sequence is thus 1.70 m. So, the tilt of the sequence could be the result of a seismic event previous to the one that cut it off. However, this normal faulting does not trend as expected from the regional deformation model, which would be produced by a NW-SE–striking maximum horizontal stress, as proposed by Guzmán et al. (1998). We assume that the displacement of the sequence took place in one seismic event, because if it were the result of two or three different events, we would have found some evidence of that in the layers that were displaced in the first event. The problem is that the sequence is incomplete, so we currently interpret this evidence as the result of only one event that displaced the sequence of volcanic ashes and paleosols at the same time. We are assuming this because it would be the worst-case scenario in a region where we do not know the seismic potential of faults that cross it. The Armenia earthquake (1999) suggests that the seismic potential of the region is higher than we anticipated. Using the displacement identified at the Canchas site, we have applied the equations for all fault types that relate magnitude and maximum surface displacement, i.e., the equation based on moment magnitude (M; Wells and Coppersmith, 1994; McCalpin et al., 1996): M = 6.9 + 0.74(log MD) and log (SRL) = –3.22 + 0.69M, (1) where M is the moment magnitude, MD is maximum displacement in m, and SRL is the surface rupture length in km. We found that the magnitude of the earthquake that affected the sequence was M 6.9, and the length of the fault segment was 35 km. Geochronology The upper (youngest preserved) paleosol was radiocarbon dated at 19,710 ± 830 yr B.P. (Fig. 6). Paleoseismic Interpretation We propose coeval faulting and folding, where normal faulting was produced by moment bending during sequence folding. As mentioned previously, the Pereira-Armenia fan is 25 km wide and is covered by paleosols and volcanic ashes. This means that the Cestillal fault is covered by the same kind of soils along 25 km. An evidence of this fault can be observed in the H–Hʹ
Large-magnitude late Holocene seismic activity in the Pereira-Armenia region, Colombia profile (Figs. 3 and 4), where the northern block is upthrown by at least 50 m. At the Canchas site, the north block is also displaced upward (the source of the profiles is 1:25,000 scale maps). In the B–Bʹ, C–Cʹ, F–Fʹ, and N–Nʹ profiles (Figs. 3, 4, and 6), it is not clear whether this fault is affected by the major cover of volcanic ash and paleosols or by some other Quaternary deposits that have not yet been identified in the region; this hypothesis has to be evaluated in future studies. A seismic hazard assessment for the region estimated that the maximum seismic magnitude would be between 6.2 and 6.6 for a seismic source located 35 km away from the city of Pereira (Guzmán et al., 1998). Magnitudes as large as M 6.9 were estimated for a seismic source located at distances greater than 50 km away.
Weight in section (m) 2.15
Artificially cut surface 5Y/5/1 Clay
1 H
1.80 1.70 1.60
1.40 1.30
2
A
H
1.10
H
Stratigraphy and Soils The description of this slope was restricted to the most important layers for paleoseismic investigation due to logistical problems. So, if the nomenclature of a layer or horizon does not appear, that means we did not collect information for this site, and missing segments could represent one or more layers of volcanic ash (Fig. 7). Structure The stratigraphic succession of the Parqueadero site is affected by a N40°E fault showing the north block displaced upward. In the southern part of the slope, at the base, there is a paleosol that was cut and displaced by tectonic structures. At the bottom of horizon 1 (Fig. 8), there are some sand layers that were liquefied. Geochronology In the southern upper part of the slope, a paleosol (paleosol Z in Fig. 7; not shown in Fig. 6) is dated at 13,150 ± 310 14C yr B.P. Paleoseismic Interpretation The Tribunas fault is a N40°E normal fault. According to the deformation model for the region, NE faults have to be thrust faults. We explain this by saying that the Tribunas fault is a secondary fault associated with a blind thrust fault. In this work, we characterize the secondary fault only. At this stage, we can only affirm that the thrust fault has to generate at least the same earthquake that can be produced by the secondary fault.
Paleosol 1
0.70 0.55 0.45 0.40 0.30
10YR/5/8 Lapilli 5Y/5/2
6
0.60
Clay
Paleosol 2 19,710 ± 830 yr B.P. Clay 4 H 2.5Y/5/2
0.80
Geomorphology Near the Parqueadero site, the geomorphology is rather smooth, exhibiting small hills and a few undulations. This site was artificially cut to build a parking lot, so the original geomorphology of the site is not known. An incomplete sequence of volcanic ashes and paleosols is exposed in a cut slope that contractors made in order to excavate the road that serves as access to the school building.
Clay
2.5Y/7/1
3 H 5Y/5/2
5
Parqueadero Site
87
7 8 9 10 11
A: Amphibole layer H: Hypersthene layer
Paleosol 0 0.00
Liceo Taller San Miguel Canchas site Field station CPL55 Figure 7. Stratigraphic sequence at field station CPL55, Liceo Taller San Miguel, Canchas site. This site is located 9 km south of Pereira. The sequence is tilted 50° to the north.
Based on the fault displacement identified at the Parqueadero site and application of Equation 1, the magnitude of the earthquake that affected the sequence was M 6.6, and the length of the fault segment was 22 km. The Cestillal fault is located south of this structure, implying that it corresponds to the southern boundary of the Tribunas fault. The evidence at the Parqueadero site is in the southern part of the fault segment, in the Pereira-Armenia fan; the northern part of the Tribunas fault is outside the PereiraArmenia fan. The continuity of the Tribunas fault segment is supported by drainage and topographic anomalies. The liquefaction evidence at the base of layer 1 suggests a seismic event of at least Mw 5.5. This evidence could be associated with the M 6.6 seismic event or a previous event. CONCLUSIONS Prior to this study, regional faults and their seismogenic potential had been assessed purely on morphologic evidence. The
88
Lalinde P. et al.
A
Liceo Taller San Miguel - Parqueadero site 2
4
6
5
7
9
8
5m
1m
3
I
10
1
11
12
10 m
1
S
Tribunas fault
2m
13
B
0
1
D C E
4 II
9 P
0m
6
7
1m IIII
4
H
N
1 2
3
5
8
P
G
4
14
A
0
1
N25°W
F
P 10
10
10
N25°W
B Artificial cut surface was not cleaned up specifically for paleoseismic study
S
Tribunas fault Artificial cut surface D
1 4
1
Artificial road cut Artificial road cut surface was not cleaned up specifically for paleoseismic study
Parqueadero and Canchas sites show evidence for seismic events of at least M 6.9 for the Cestillal fault and M 6.6 for the Tribunas fault. The maximum seismic magnitude estimated for this region ranges between 6.2 and 6.6 for a seismic source being 35 km away from the site. Our field evidence confirms that local M 6.6 and 6.9 seismic events have taken place, with a possible seismic source located 9 km south of the city of Pereira. This leads us to a better understanding of the regional seismic activity for the last 20,000 yr. After the Armenia earthquake, it is important to review the seismic hazard assessment for this region and to incorporate new field data into its seismic interpretation. It is necessary to continue paleoseismic studies looking for information that may reduce the uncertainty of fault characterizations and seismic hazard evaluations. Paleoseismic studies are necessary in countries like Colombia because our historical seismic record covers less than 500 yr and instrumental records cover less than 100 yr. The location of the Colombian region in the junction of the Nazca, Caribbean, and South American plates requires detailed paleoseismic studies in order to reconstruct its seismic history and reduce the uncertainties in seismic hazard assessments.
N
Figure 8. Liceo Taller San Miguel, Parqueadero site. (A) Sketch of the west wall exposure. Grid lines are 1 m apart. Horizon 0—lapilli with some rusty iron material; Horizon 1—hard sand; Horizon 2—base is a light gray lapilli; Horizon 3—lapilli with a higher content of rusty iron material; Horizon 4—yellowish white layer with some rusty iron material; Horizon 5—middle gray sandy layer with organic matter; Horizon 6—white, clayish layer; Horizon 7—light gray lapilli; Horizon 8—lapilli with some rusty iron material in the higher part of the layer; Horizon 9—white yellow sandy (lapilli) layer with some rusty iron material; Horizon 10— yellowish white clay layer; Paleosol (P)—affected by fractures G and H (this layer has gradational contacts). (B) Photograph of the west wall. The Tribunas fault is a N40°E normal fault that could be a secondary fault related to a blind thrust fault.
ACKNOWLEDGMENTS We gratefully acknowledge the following persons and institutions for their support: geologist Michael Tistl, who informed us about this site in order to make paleoseismic studies; Teresa Tisnés, director of the Liceo Taller San Miguel high school, who allowed us to investigate this site; Andrés Velásquez, Myriam López, Fabián González, and Carolina Franco, for their help with the fieldwork; and Colciencias, Universidad EAFIT, and “Fundación para el Avance de la Ciencia y la Tecnología del Banco de la Republica–Colombia” for financial support of this investigation. REFERENCES CITED Cano, M., 2004, Los primeros habitantes de las cuencas medias de los ríos Otún y Consota, in Cambios Ambientales en Perspectiva Histórica, Ecorregión del Eje Cafetero, v. 1: Pereira, Colombia, Universidad Tecnológica de Pereira, Facultad de Ciencias Ambientales, Centro de Investigaciones y Extensión, Programa Ambiental GTZ Investigación Aplicada, p. 68–91. Cardona, F.J., and Ortíz, M., 1994, Aspectos Estratigráficos de las Unidades del Intervalo Plioceno Holoceno entre Pereira and Cartago: Propuesta de Definición para la Formación Pereira [thesis]: Manizales, Colombia,
Large-magnitude late Holocene seismic activity in the Pereira-Armenia region, Colombia Caldas University–Corporación Autónoma Regional de Risaralda, 124 p., plus annex. Cline, K.M., Hutchings, L., Page, W.D., and Jaramillo, J.M., 1981, Quaternary tectonics of north west Colombia: Bogotá: Revista Centro de Investigación y Desarrollo en Información Geográfica del Instituto Geográfico Agustín Codazzi (CIAF), v. 6, p. 1–3. Ego, F., Sebrier, M., Lavenu, A., Yepes, H., and Egues, A., 1996, Quaternary state of stress in the northern Andes and the restraining bend model for the Ecuadorian Andes: Physics and Evolution of the Earth’s Interior, v. 1, p. 101–116. Freymueller, J., Kellogg, J., and Vega, V., 1993, Plate motions in the North Andean region: Journal of Geophysical Research, v. 98, p. 21,853– 21,863, doi:10.1029/93JB00520. Guzmán, J., Franco, G., and Ochoa, M., 1998, Informe Final Evaluación Neotectónica. Proyecto para la Mitigación del Riesgo Sísmico de Pereira, Dosquebradas and Santa Rosa de Cabal: Pereira, Carder, 148 p. Ingeominas, 1999, Informe Técnico Científico del Terremoto del Quindío (Morales, C., López, E., Mora, H., Monsalve, H., Nieto, A., Gil, F., Osorio, J., Escallón, J., Bermúdez, M., Martínez, S., Bermúdez, A., Esquivel, J., Vásquez, L., Espinosa, A., Jiménez, E., and Vergara, H., eds.), v. I: Bogotá, Colombia, 34 p. Keller, E., 1986, Investigation of active tectonics: Use of surficial earth processes, in Wallace, R., ed., Active Tectonics: Studies in Geophysics: Washington, D.C., National Academy Press, p. 136–147. Kellogg, J., and Vega, V., 1995, Tectonic development of Panama, Costa Rica and the Colombian Andes, in Mann, P., ed., Geologic and Tectonic Development of the Caribbean Plate Boundary in South Central America: Geological Society of America Special Paper 295, p. 75–89.
89
Lalinde, C.P., 2004, Evidencias Paleosismicas en la Región Pereira-Armenia, Colombia [Master’s thesis]: Medellín, Colombia, EAFIT University, 149 p., plus annex. McCalpin, J.P., Nelson, A.R., Hackett, W.R., Jackson, S.M., Smith, R.P., Carver, G., Weldon, R.J., II, Rockwell, T.K., Obermeier, S.F., and Jibson, R.W., 1996, Paleoseismology: San Diego, California, Academic Press, 588 p. Page, W., 1986, Geología Sísmica del Noroccidente Colombiano: Medellín, Colombia, Report from Woodward Clyde Consultants to ISA and Integral, 156 p., plus appendix and figures. Taboada A., Rivera, L.A., Fuenzalida, A., Cisternas, A., Philip, H., Bijwaard, H., Olaya, J., and Rivera, C., 2000, Geodynamics of the Northern Andes: Subductions and Intracontinental Deformation (Colombia): Bogotá, Special Publication Asociación de Ingeniería Sísmica (AIS), 28 p., plus figures. Thouret, J.C., 1983, Presentación geológica y geomorfoestructural, in Van der Hammen, T., Pérez, A., and Pinto, P., eds., Estudios de Ecosistemas Tropandinos, La Cordillera Central Colombiana, Transecto Parque Los Nevados (Introducción y Datos Iniciales), v. I: Berlin, Gebruder Borntraeger, Proyecto Ecoandes y Ecodinámico, p. 48–55. Wells, D.L., and Coppersmith, K.J., 1994, Empirical relationship among magnitude, rupture length, rupture area, and surface displacement: Bulletin of the Seismological Society of America, v. 84, p. 974–1002. Yeats, R., Sieh, K., and Allen, C., 1997, The Geology of Earthquakes: New York, Oxford University Press, 568 p.
MANUSCRIPT ACCEPTED BY THE SOCIETY 7 DECEMBER 2010
Printed in the USA
The Geological Society of America Special Paper 479 2011
Evidence of Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia Myriam C. López C.* Av. Alberto Mendoza Hoyos No. 89-02, Conjunto Arboletes Casa 38, Manizales, Colombia Franck A. Audemard M.* Fundación Venezolana de Investigaciones Sismológicas (FUNVISIS), El Llanito, Caracas 1073, Venezuela
ABSTRACT Morphotectonic and paleoseismic studies carried out in the surrounds of Tuluá (4°°N, 76°°W) provide strong supporting evidence for ongoing E-W compression in the Cauca Valley, Colombia, during the late Pleistocene and Holocene. This local tectonic regime is kinematically and mechanically connected with the ENE-striking, rightlateral, strike-slip Ibagué fault system, which crosscuts and offsets the Central Cordillera of Colombia. Morphologic, stratigraphic, kinematic, and chronologic evidence obtained on flexural scarps, which are currently shaping the foothills of the Central Cordillera, attests to the recent activity of a compressional fault system. This includes both hinterland-propagating back-thrust faults and foreland-verging thrust faults that reutilize a fold-and-thrust belt, previously considered to be active only during Tertiary times. Kinematic measurements on the back-thrust faults further support an ongoing E-W–oriented maximum horizontal stress at the latitude of Tuluá. In terms of seismic hazard for this region, these investigations provide evidence for Ms ≥7 earthquakes with recurrence in the order of 6 k.y. on this frontal thrust system. In addition, should the A.D. 1766 earthquake have not taken place on these active thrust faults, the probability of occurrence of a forthcoming event with such characteristics would be high.
INTRODUCTION
morphic evidence and geologic evidence for this compression are herein presented (Fig. 1). For certain active geologic structures, paleoseismic evidence is described and discussed. Finally, this compressive style is discussed in relation to the present-day tectonic activity, as well as style, of the right-lateral strike-slip Ibagué fault, which crosscuts in an ENE-WSW direction the almost N-S–trending Central Cordillera.
The purpose of this work is to document active Holocene compression taking place in the town of Tuluá and its surrounds, located on the outer edge of the western foothills of the Central Cordillera of Colombia at about the latitude of 4°N. Both geo-
*E-mails:
[email protected];
[email protected]. López C., M.C., and Audemard M., F.A., 2011, Evidence of Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, p. 91–107, doi:10.1130/2011.2479(04). For permission to copy, contact
[email protected]. © 2011 The Geological Society of America. All rights reserved.
91
Vie ja r ive r Cartago Bár bar a ra
s
ata
Bolivar
Sevilla
est Pal
Cordoba Pijao
g Iba
ué
fau
Ibagué
Cajamarca
IF lt Rovira
Silv
Ka
lt au
ua-
Pzc
ia-P
Bugalagrande
of
La Marina
er
Sa
Ortega
uana
fault 80°W
70°W
CARIBBEAN PLATE Chaparral 1.3 cm/yr
Pzc
Q Amaime
Guamo
PzJ
nJ
fault
uer lmag Cuc
ón
im s-P Ginebra
Tv
Kca
La Albania
aba
Loboguerrero Vijes
ca-A
era
Ko
rad
Sonso
Cau
fau
Kc Buga
?
Yotoco
lt
Kn
se bal Em ima Cal
Gu
Darién Dari
Dagua
Alvarado
lt fau
La Paila
Tuluá Tulu
Buenaventura Buenaventu ra
Pacific Ocean
Machín volcano
Circasia
Armenia
Trujillo
Cali
QF
Dag ?
4º00´N
t
Kq
ijao
Kub
ul
Tnp
Falla
fault ma
Kv
fa
Kca
Verde fault Zarzal Río
Tuluá
rrap
Ga
lt Fau
a
ina
Cau ca R iver
La Union
4º30´N
ot
Sta
Kc
Ke
ns
Pereira
nge
Kc
Co
Ka
La
Magdalena valley
Ansermanuevo
?
López C. and Audemard M.
fault
92
Caribbean Sea
10°N
10°N
PANAMA Natagaima
Yumbo Palmira
VENEZUELA
3º30´N
tía fa
ult
Bolo Azul
0
ific
BRAZIL
SOLIMOES BASIN
Huila volcano
Toribio
Neiva 76º00´N
Kc Ke Kub
Late Cretaceous Dagua structural complex Bolivar ultramafic complex
75º30´N
Fault
Normal fault
Q
Quaternary sediments
Paleogene Vijes Formation
IF
Late Cretaceous Volcanic Formation
Ibagué Fan
Neogene La Paila Formation
Kn Kv
Drainage
Central Cordillera
Quindío Fan
Tv
70°W
80°W
Tnp
Paleogene sediments
400 KM
PERU
Western Cordillera
Tp
0°
0
Kq
Strike-slip fault
QF
SOUTH AMERICAN PLATE ECUADOR
Santander de Quilichao
76º30´N
GUIANA SHIELD
6.6 cm/yr
Alg
Planadas 0°
Ka
Thrust fault
COLOMBIA
ult s fa ira ec
Pac Puerto Tejada
Tp
50 km
tas apa Garr fault ult fa gué Iba
Florida
Jamundi
Timba
6.3 cm/yr
Pradera
Cauca valley
Cauc
Pance
Cali
a-Pa
Ke
NAZCA PLATE
Oce an
Kv
Late Cretaceous Nogales Formation
Ko
Basic rocks Ginebra Ophiolitic Massif
Kc
Early Cretaceous Buga intrusive
Ka
Early Cretaceous Amaime Formation
Kq
Cretaceous Quebradagrande Complex
Kca Early Cretaceous
Arquía Complex
Pzc
Paleozoic Cajamarca Complex
PzJ
Basement
Figure 1. Location of the region and regional geology (adapted from Ingeominas, 1988). Slip senses of Quaternary faults are taken from Woodward-Clyde Consultants (1983), modified from Paris et al. (2000), and data of López (2006). Inset in upper left corner shows regional geopolitical location of Colombia (shaded); inset in lower right corner shows geotectonic map of Colombia (white represents mobile blocks with respect to stable South America) and geodetic offset measurements sensu Trenkamp et al. (2002); central inset is area of Figure 2 where detailed paleoseismic features were documented during this study. Maximum constriction of the Cauca Valley happens at the latitude of Buga. The western foothills of the Central Cordillera develop west of the Cauca-Almaguer fault.
Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia The present study was carried out in two independent but complementary phases. The first phase consisted of the identification and characterization, from aerial photographs, of landforms that are indicators of active tectonics along strike-slip faults, and that are typical features of fold-and-thrust belts. In the latter sense, this analysis relied on evidence such as: (1) the geomorphology of folds and piggyback basins as a sensitive indicator of fold and fault growth, as shown by Burbank et al. (1996); (2) evidence presented by Audemard (1999) and Holbrook and Schumm (1999), who maintain that rivers tend to be deflected when the stream power is less than the tectonic uplift rate; and (3) classical concepts of fluvial geomorphology and tectonic landforms developed by many authors over the years (Howard, 1967; Wesson et al., 1975; Burnett and Schumm, 1983; Philip, 1983; Ouchi, 1985; Philip et al., 1992; Audemard M. and Robertson, 1996; Audemard M., 1999). These geomorphic data were later verified during fieldwork. The second phase of this assessment focused on the study of deformed outcrops in the surrounds of Tuluá, where paleoseismic techniques were applied. No trench was actually dug. We studied a set of new road cuts created for the Tuluá highway north interchange, and also had the opportunity to check a main outcrop thanks to excavation work for a sewer at the foot of the El Ahorcado scarp. Finally, in order to integrate all these data into a regional model and to understand the interplay of the studied structures with more regional structures, we applied known structural patterns (e.g., Medwedeff and Suppe, 1997; Biddle and ChristieBlick, 1985) to a mechanical deformation model (e.g., Wilcox et al., 1973; Harding and Lowel, 1979; Crowell, 1982; Lowell, 1985), considering that some of this evidence could be related to old structural styles (tectonic inheritance), and looking for the features that reflect the active stress regime instead. TECTONIC SETTING The Holocene compression zone documented in this study is part of the more regional transpressional system acting on the NW corner of South America, which has been supported by geodetic measurements (Trenkamp et al., 2002), among many other lines of evidence. It is located in a region where many authors have identified a transition zone indicated by a change in a stress regime (James, 1985; Toussaint and Restrepo, 1987; Ego and Sébrier, 1995; Meyer and Mejía, 1995; MacDonald et al., 1996; Taboada et al., 2000; Audemard M., 2002; Corredor, 2003; Montes et al., 2003). Oblique convergence at the oceanic subduction zone between the Nazca and South America plates is partitioned in Colombia, as postulated by Audemard M. (2003), into a roughly E-W–oriented shortening, accommodated by both type B subduction and orogen growth of three mountain chains, and strike-slip shear trending roughly subparallel to (e.g., Romeral fault system [Ingeominas, 1988] or San Jerónimo, Silvia-Pijao, Cauca-Almaguer fault systems [sensu Maya and González, 1995]) or highly oblique to the orogens (Garrapatas and Ibagué
93
faults). However, this partitioning does not necessarily act uniformly all over the deformation belt, because some major tectonic features may combine both thrust and along-strike slip, such as the eastern boundary fault system of the Eastern Cordillera (Butler and Schamel, 1988). As for the Central Cordillera of Colombia, it extends roughly N-S at latitude 4°N, between the Cauca and Magdalena Rivers to the west and east, respectively. Significant elongated basins and plains are associated with these two very large rivers, which both flow north. This chapter focuses on a small sector of the western foothills of the Central Cordillera, where the Cauca valley is almost strangled by a salient of the foothill unit, which has its maximum constriction at the latitude of Buga and is named the Buga Salient (López, 2006; Fig. 1). The western foothills of the Central Cordillera have been interpreted as a thick-skinned fold-and-thrust belt built during the last phase of the Andean orogeny (Alfonso et al., 1994), and they are believed by many authors to be inactive at present. However, this paper provides evidence for ongoing tectonic activity. Also, according to many authors, only the Valle Group and the underlying Mesozoic basement are involved in the across–Central Cordillera shortening. The Valle Group, as defined by Nivia (2001), consists of the La Paila, Cartago, and Buga Formations, as originally defined by Schwinn (1969), and the Cinta de Piedra Formation sensu McCourt (1984). This sedimentary unit is made of clastic and volcaniclastic sequences of Oligocene–Miocene age (Van der Hammen, 1958). From a morphotectonic viewpoint, the western foothills of the Central Cordillera display a widespread rather flat surface resulting from peneplanation of the Tertiary (Valle Group) units (uplifted flat surface in Fig. 2). This surface rises some 100 m above the Cauca valley floor. Two other lower surfaces are also recognizable, but these are mainly the product of sediment accumulation. The lowermost surface corresponds to the present-day valley floor of the alluvial plain of the Cauca River. As a general trend, this plain dips and drains very gently toward the north. The second surface, intermediate in elevation between the two others, lies ~10 m above the valley floor and essentially corresponds to the top of the Tuluá alluvial fan (lower hills in Fig. 2). This paper mostly focuses on this morphologic and geological unit, but we also look at its interplay with the more regional geologic setting. Other minor morphologic units can also be mapped in more detail, such as staircased alluvial terraces and colluvial aprons (Figs. 3 and 4). PREVIOUS ACTIVE TECTONIC STUDIES AROUND TULUÁ Very little neotectonic evidence had been previously reported in the study area and its surrounds, with the exception of the data published in the Quaternary Tectonic Map of Colombia compiled by Paris et al. (2000), which displays the major active tectonic structures in the study area. In more detail, Marín and Romero (1988) and Paris et al. (1989) referred to some recent scarps along
94
López C. and Audemard M.
Bu
ga
lag
ra
r ive
nd
eR
ive
aR
r
uc
Ca
á
bal
Sa
La Morelia
ʹ
eta
Mo rale sR iver
Main frontal range
r
ive
sR
á
Tu l
uá
Ri ve r
ʹ
á
Elev. (m)
Figure 2. Map (above) and cross section (below) of the main geomorphological features that constitute the western foothills of the Central Cordillera (CC) in the region of study: the uplifted flat surface overhanging 100 m over the Cauca valley floor (stippled), lower hills at 10 m over the valley floor (shaded and stippled), and lower alluvial plain at the same level of the alluvial valley of the Cauca River (shaded), 900 m above sea level. A–Aʹ schematic cross section shows the inferred surficial structures. Insets correspond to Figures 3 and 4.
Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia
95
á Tulu r Rive Mor
ales
Rive
r
Tu l
uá
Ri
ve r
Figure 3. Morphological features and structural patterns associated with active faulting in the surrounds of Tuluá. Location of trenches: (A) Oreja Norte road cut, (B) Cara Norte Oreja, (C) Cara Sur Oreja, (D) Variante Tuluá S, (E) Variante Tuluá N, (F) El Ahorcado, and (G) Sur Río Tuluá.
96
López C. and Audemard M.
the W-verging, NNE-trending Guabas-Pradera thrust fault at the Bugalagrande and Tuluá sites. In addition, some relationships between the different trending faults with suspected recent activity were shown by Nivia et al. (1997) and Nivia (2001), where the NW-striking structures were proposed as the most recent ones. On the other hand, Woodward-Clyde Consultants (1983), based on paleoseismic investigations near Amaime (some 40 km south of Tuluá), offered strong supporting evidence for Holocene activity of a more southerly segment of this western bounding fault of the Central Cordillera. López et al. (2002, 2003a, 2003b, 2004, 2005) and López (2006) have documented geomorphic, stratigraphic, and paleoseismic evidence of back-thrust faults that support the present activity of the main fold-and-thrust belt, which is developed at the western termination of the ENE-trending, rightlateral, strike-slip Ibagué fault. In addition, López (2006) stated that the common extensional faults affecting the Quaternary alluvial sediments in the Cauca valley are directly linked to liquefaction structures. This has been proved by Suter et al. (2008). López et al. (2009) found a direct relationship between the liquefaction features and the Holocene activity of the thrust faults. Evidence of the eastern termination of the Ibague fault was documented by Montes et al. (2003) at the eastern foothills of the Central Cordillera where the Piedras-Girardot fold belt is developed at a left step of the Ibagué fault. These authors calculated a minimum horizontal component of 30 km, according with the displacement of the Ibagué Batholith, and 32 km according with the contraction of the folded belt. Neotectonic activity of the Ibagué fault has been documented in the eastern foothills of the Central Cordillera at the Ibagué fan (Diederix et al., 1987; Vergara, 1988; Montes et al., 2005a, 2005b). Paleoseismic studies done by Montes et al. (2005b) showed that the Ibagué fault behaves as a Riedel system. Analysis of the trench done at a pull-apart basin in the Ibagué fan let Montes et al. (2005a) calculate a maximum magnitude Ms 7 ± 0.1 for a return period of 1300 yr and a mean slip-rate velocity of 0.77 mm/yr. Based on analysis of the Harvard Centroid Moment Tensor (CMT) catalogue, Corredor (2003) stated that the Ibagué fault acts as a transfer zone and plays an important role in the way strain is distributed in these blocks. Mejía and Meyer (2004) maintained that the highest surficial seismicity in southwestern Colombia coincides with the intersection between N-S and ENEWNW lineaments. REGIONAL ACTIVE TECTONICS In the present study, geomorphic evidence of both active strike-slip faulting and fold-and-thrust faulting at the regional scale were identified and mapped in the Central Cordillera, and particularly in the western foothills of the Central Cordillera. Geomorphic Evidence of Strike-Slip Faults These features are more suitably identified at the regional scale. Faults interpreted in this work as having lateral movements
correspond to the ENE-trending system known as the Cucuana type (after McCourt et al., 1984). Five of those major right-lateral, strike-slip fault systems, detectable in Landsat images, obliquely cross the Central Cordillera from the Consota River in the Pereira region in the north to the Timba region in the south. Several of these faults show a left-stepping structural pattern. In addition, the width of the Cauca valley is commonly reduced where these faults intercept or approach the valley, being the most noticeable in the area under study (Fig. 1). It is also common to find large active alluvial fans in association with these ENE-trending faults. Examples of this are found at: El Quindío fan, bounded by both the Garrapatas-Otún faults to the north and the Río Verde–Ibagué faults to the south, the Ibagué fan in association with the Ibagué fault on the eastern flank of the Central Cordillera, and the Tuluá fan at the western tip of the Ibagué fault system. Active Compressional Landforms Active compressional landforms can be observed at the regional scale, as well as at the local scale. As mentioned already, the drainage is very sensitive to active compressional tectonics, since it adjusts to deformation quickly and responds in different manners to uplift or subsidence depending on its erosive energy. Consequently, drainage anomalies are the most frequent indicators of neotectonic movements on both folds and thrust faults (Audemard M., 1999). Some examples of this at the regional scale are (1) the hairpin of the La Vieja River (known as El Codo de la Vieja, Quindío fan) in association with the N-S–trending Santa Barbara range (Fig. 1), and (2) the nonaxial position of the Cauca River inside its valley, as well as the meander belts (López, 2006). However, these are not the only diagnostic evidence: Flexural scarps, warping, buckling, and tilting, among others, reveal the activity, in this region, of underlying or partly buried folds and thrust faults. These features will be discussed in detail later herein, focusing particularly on the Tuluá surrounds. LOCAL EVIDENCE OF ACTIVE TECTONICS The town of Tuluá sits on the western edge of the western foothills of the Central Cordillera, between the foothill unit and the Cauca plain. To the east and as far as the foot of the Central Cordillera, the western foothills extend for some 10 km across as a series of uplifted flattened surfaces. Most of these surface tops correspond to either Pleistocene–Holocene volcanic ash falls from the Machín volcano complex (R. Méndez, Vulcanological Observatory of Manizales, and G. Toro, Eafit University, 2002, personal commun.) or a thin Quaternary alluvial cover that unconformably overlies the folded Neogene sedimentary sequence of the Valle Group. Ductile deformation of this Quaternary cover at a more regional scale (a few kilometers in size, at the most), in combination with the local brittle deformation, attests to the recent activity (or reactivation) of this late Neogene foldand-thrust belt, as will be shown in the following sections. The identified landforms of recent tectonic activity, to be considered
Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia as a consistent and coherent set of supporting evidence, are discussed later. Active Growth of Preexisting Folds in the Western Foothills The stretch of the western foothills of the Central Cordillera, extending between Bugalagrande to the north and Sonso to the south (Fig. 1), displays anticlines and synclines with N-S– trending axes, plunging to the north and south, respectively (Figs. 2, 3, and 4). Close to the Central Cordillera frontal range, the axis of folding tends to trend E-W and plunge E. Here, a rather conspicuous but discontinuous counterscarp can also be followed for over 3 km SW of Galicia (Fig. 2), and it is also indicated by several anomalies in the drainage pattern. This locality is likely to be trenched in the future. Although this is not the subject of this work, we can observe that the frontal range along this stretch preserves elevated flat remnants at different elevations, indicating that the Central Cordillera is still uplifting, provided those remnants can be genetically related to the peneplain lying at lower altitude in the western foothills of the Central Cordillera. In addition, it would also indicate that the peneplanation has developed through different phases (polygenic peneplain). The active growth of several N-S–trending, N-plunging, few-kilometer-long anticlines made of Neogene Valle Group rocks (Figs. 2, 3, and 4), as well as some synclines, is illustrated by: (1) the gentle flexing of the overlying Quaternary deposits above structure flanks (e.g., Cg in Figs. 5 and 6); and (2) the formation of small Quaternary piggyback basins in association with the hinterland flank or back limb of some of these anticlines (e.g., La Llanada surface in Fig. 4). These basins most commonly sit above large, open synclines. (3) In addition, those rivers that have enough stream power to cut through these anticline axes locally exhibit flights of erosional terraces preserved mainly at the anticline cores (Fig. 2). (4) Downstream of some anticlines, the drainage incises progressively steeper, leaving elevated terraces, along with increasing dip (the older the terrace, the steeper the dip), that pinch out toward the anticline located upstream (Qt6 to Qt1 in Fig. 4). This implies that the anticline is growing up and functions as a hinge zone for the drainage network. It is very common for a piggyback basin to be preserved in the upstream side of such anticline; and/or, (5) in some rare cases (e.g., 4 in Fig. 4), the moment-bending normal faulting on anticline crests may even display open cracks filled with Holocene organic-rich soils. In other words, the western foothills of the Central Cordillera east of Tuluá as a whole seem to resemble an active (or recently reactivated) Jura-type foldand-thrust belt. These anticlines frequently show a positive but not prominent relief. They can either display a deeply dissected topography where beds are easily traceable (Ts in Fig. 4), or a very smooth warped surface (Qs in Fig. 4). The farther the anticlines are from the Central Cordillera range front, the deeper is the dissection. In both cases, the second-order drainage is perturbed, whereas
97
the main rivers only form deep gorges across their cores, when not abandoning flights of elevated terraces alongside. In the case of cutting around a periclinal closure, unpaired elevated terraces form (Fig. 2). In the case of anticlines, the number of terraces is always larger toward the inside (concave side of the bend, closer to where the anticline is more elevated). All the drainage anomalies discussed by Audemard M. (1999) on the second-order, or lower, drainages are present. Synclines, if not completely absent in the landscape because they are blanketed by Quaternary sediments, are partly excavated by a centripetal drainage system (C in Fig. 4). Geomorphic Evidence of Activity along the Western Foothills Frontal Thrust System The divide between the high peneplain and the Cauca valley is usually very sharp and corresponds to a morphological scarp of several tens of meters in height, at least between the towns of Bugalagrande and Sonso, to the north and south, respectively (Fig. 1). This feature develops on Quaternary deposits of both volcanic and alluvial origin overlying the Valle Group unit. It is commonly highly dissected, exhibiting a convex profile and a slightly wavy trace. This scarp shows clear evidence of being a flexural scarp, like those described by Audemard M. (1999), just north (at the motorway north interchange) and south (left or south margin of the Tuluá River) of Tuluá (A–G in Fig. 3). There are also as many as four or five other minor W-facing, N-S–trending flexural scarps in the Tuluá region (for instance, the El Ahorcado scarp, which lies east of Tuluá, is clearly distinguishable in landscape) (Fig. 3). In contrast, other flexural scarps are very subtle from the morphologic viewpoint, such as the one lying just west of Tuluá (see the border between lower hills and alluvial plain in Fig. 2). This could be interpreted in terms of the youth and westward (basinward) in-sequence progression of the western foothills of the Central Cordillera thrust sheets. Hanging drainages (or wind gaps) are a common feature on both the main and minor flexural scarps (Hd in Fig. 3), while drainage captures and inversions occur behind them (Cd and Id in Fig. 3), as well as broom-shaped drainage patterns (Bs in Fig. 3). Tectonic gutters and river inversions can occur either at the front or at the back of these flexural scarps, thus implying tilting by loading or rotational back tilting of ground surface, respectively. Some of these flexural scarps occur in the foreland side of pressure ridges (as defined by Philip et al., 1992), such as northeast of Tuluá at the La Oreja pressure ridge, or south of the Tuluá River, just south of Tuluá (Figs. 3 and 6–8). These pressure ridges grow vertically between a W-verging thrust and a back-thrust fault, both trending roughly N-S (G in Fig. 3; Fig. 8). Behind the positive relief located south of the Tuluá River, the drainage, including the Tuluá River itself, forms a broom-shaped pattern to overcome the vertical growth of that ridge (Bs in Fig. 3). The Tuluá River is not completely diverted (or pushed northward) around the northern termination of such a ridge, because a small
98
López C. and Audemard M.
R
R
Figure 4. (Top) Digital elevation model showing the ENE structures truncating Tertiary units (Ts) and affecting Quaternary sediments (Qs) and colluvial aprons (Ca). 1, 2, 3, and 4 are fault plane measures. C—centripetal drainage. (Bottom) The trapezoid area is represented with an interpretative drawing showing the main geomorphological features related to active folding (view from the north). Ts—Tertiary bedding plane surfaces; Qs—Quaternary deposits dipping toward the east; Qt1 to Qt6—remnant terrace surfaces of Qs dipping toward the west. The amplitude of the drawing is 5 km.
Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia
2
99
3
1
3
2
1
A B
14
C 17,800 ± 660 yr B.P.
C Figure 5. (A) Photograph of the morphological aspect of the pressure ridge at Oreja Tuluá with location of outcrops 1, 2, and 3. (B) Left: interpretative sketch of outcrop 1, where the fault overthrust a paleosol 1.6 m, which was sampled at location P. The age of this paleosol is 14C 17,800 ± 660 yr B.P. Right: interpretative sketch of outcrop 3, where fault 7 offsets the basal conglomerate (Cg) of the Quaternary sequence by 3 m. Csp—paleosol derived from volcanic ashes. (C) Photograph (top) of the morphological aspect of the pressure ridge at outcrop 2 and its interpretative sketch (bottom). East-vergent imbricated N-S thrust faults outcrop through the bedding planes of La Paila Formation (Tp), affecting the unconformably overlying Quaternary sequence. The conglomerate horizon (Cg) is offset 3 m by one of the easternmost faults (fault 7); reworked volcanic ashes (Cc) seem to be offset too. Paleosol (P) in the footwall was dated at 14 C 12,840 ± 40 yr B.P.
,
. .
100
López C. and Audemard M.
A
B
14
C 7.930 ± 60 yr B.P.
Figure 6. (A) Photograph (left) of the outcrop at Variante Tuluá N and its interpretative sketch (right). A W-vergent thrust fault offsets the basal conglomerate (Cg) of the Quaternary sequence. (B) Photograph (left) of the outcrop at Variante Tuluá S and its interpretative sketch (right). A paleosol (P) is seen along one of the fault bedding planes (2) that is emerging through the Tertiary units (Tp); this paleosol is sealed by the conglomeratic basal unit (Cg) of the unconformably overlying sequence, which is offset by a more recent fault (4). Csp—sandstones with inclined stratification; Csm—massive sandstones; Cc—reworked ashes.
portion of the ridge is left north of the river crossing. In addition, a tectonic gutter parallels the back-thrust fault to the east. Similar to the Tuluá River, the Morales River also requires the gathering of several drainages (broom-shaped pattern) to cross the El Ahorcado ridge. Tectonic and stratigraphic evidence of the intermediate-dip back-thrust faults at La Oreja and Variante sites (both north of Tuluá), supporting these present-day landforms, is imaged in recent road cuts for the La Oreja motorway exchange and the eastern Tuluá bypass, respectively (discussed in detail in TectonoStratigraphic Evidence of Quaternary Compression section). A particular and very local setting can be seen at the Tuluá fan, where at least five terrace levels can be mapped (t1–t5, in Fig. 3) between the valley floor and the peneplain top. These staircased flat surfaces are N-S elongated and occasionally bounded to the west by flexural scarps. These terraces, built mainly onto the Tuluá fan, only extend in a N-S direction between two ENEWSW–trending lineaments. Without any Quaternary cover, these surfaces would be called cuestas, and the other side would be called contracuestas, according to the dip sense of the Neogene units, and the drainage developed on either side would be called resequent and obsequent. Additionally, they show most of the drainage anomalies proposed by Audemard M. (1999) for the identification and characterization of blind-thrust faulting. Geomorphic Evidence of Activity of Other Fault Sets In the Tuluá region, it is also possible to identify other fault trends. For instance, E-W straight lineaments sharply separate mor-
phologic units with substantial textural differences (e.g., El Zanjón del Sastre, south of Tuluá), which also control part of the drainage network. These lineaments locally show N-facing trapezoidal facets and faceted spurs. Right-lateral deflected divides in one of the youngest uplifted surfaces have also been mapped in association with the western prolongation of the ENE-WSW– trending faults (discussed earlier), on the left margin, south of the Tuluá River (Fig. 3). TECTONO-STRATIGRAPHIC EVIDENCE OF QUATERNARY COMPRESSION A set of six new anthropogenic outcrops, as well as a natural riverbank cut, exposes a set of tectono-stratigraphic elements (labeled A–G in Fig. 3) that strongly suggest ongoing Quaternary compression across the western foothills of the Central Cordillera. In addition, these geologic data are coincident with and confirm the geomorphic evidence of the previously described N-S–trending thrust faulting. Five outcrops lie NE of the town of Tuluá, at the north Tuluá motorway interchange, named herein as Carreteable Oreja Norte, Cara Norte Oreja, Cara Sur Oreja, Variante Tuluá N, and Variante Tuluá S. They expose the same tectonic feature at five different sites (A–E in Fig. 3). From A to D, they all consistently show an E-verging, intermediate-dip, N-S–striking, thrust fault ranging from 30° to 70° dip, cutting through the Neogene Valle Group sedimentary rocks, as well as through the overlying unconformable Quaternary deposits (Figs. 5 and 6); the last outcrop (E in Fig. 3) exposes a W-verging thrust fault (Fig. 6). Separate paleoseismic investigations were
Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia
East
West
Trench
A
B West
East P6
1m
As
Ac
P5
Agu
P4 Ac
P3
Ga
P2
P1
Agr
0
P2
Ac
Gc
As: black clay (present soil) Ac: gray clay Agu: clay with 50% sharp pebbles Agr: brown clay with 40% sharp pebbles Ga: sharp gravels involved in a 15% brown matrix Ac: dark-brown paleosol Gc: clast-supported and subrounded gravels with maximum diameters of 40 cm
4.8 m 14
14
P6: C 5770 ± 130 yr B.P.; P5: C 22,000 ± 160 yr B.P.; P4: 14C 21,500 ± 140 yr B.P.; P3: 14C 7460 ± 330 yr B.P.; P2: 14C 17,900 ± 130 yr B.P.; P1: 14C 13,070 ± 80 yr B.P.
C
Figure 7. (A) Photograph of the morphological aspect of the El Ahorcado pressure ridge and location of the trench at the foot of the scarp. (B) Photograph of the trench (a sewer ditch) with the interpretation of the structures on the east face of the trench. The vertical scale measures 1.5 m. (C) Interpretative sketch of the trench. Two colluvial wedges were revealed: P1—in situ paleosol dates the top of the terrace, P2—reworked paleosol at the bottom of the first colluvium of earthquake 1, P3—in situ paleosol developed at the top of the main body of the first colluvium, P4—reworked paleosol deposited at the second basal colluvium of earthquake 2, P5—reworked paleosol at the main body of the second colluviums, P6—in situ paleosol defining the top of the second colluvial wedge and the beginning of the development of the present soil.
101
102
López C. and Audemard M.
B
A
East
2
West
1
Tectonic gutter
C
2
1
Figure 8. (A) Photograph of the morphological aspect of the pressure ridge at the stream cut on the south side of Tuluá River. Maximum height of the scarp is 30 m. (B) Interpretative sketch of the pressure ridge scarp. The W-vergent thrust fault was seen in the outcrop, and the E-vergent back thrust is inferred. (C) Close-up of the main thrust fault: (1) detail of dragged folds on gravel horizon; and (2) detail of a pop-up structure developed near the top of the scarp. Close-ups are 2 m in amplitude.
performed on these cuts based on fault striation data, radiocarbon datable materials, and measurable offsets. From the geomorphic viewpoint, the W-verging fault matches with an E-facing flexural scarp that dams and diverts smaller streams. A sixth outcrop, located farther south and east of Tuluá (F in Fig. 3), was also studied, corresponding to a sewer system ditch for housing, which happened to be under construction in the same period as our fieldwork. This 1–1.5-m-deep, over a 100-m-long, E-W– trending ditch revealed the intercalation of two colluvial wedges among organic-rich paleosols and the present soil at the foot of the El Ahorcado flexural scarp (Fig. 7). This was complemented with a seventh, natural outcrop (G in Fig. 3), found farther south (due south of Tuluá), cut by the Tuluá River (Fig. 8), which coincided with a ridge bounded by opposing flexural scarps on the west and east. The E-facing scarp is highlighted by a tectonic gutter and a broom-shaped river pattern. In general, the Quaternary sequence is composed of a basal conglomerate (Cg) that is less than 3 m thick and includes clast-supported subrounded gravels. This horizon (Cg) is overlain by coarse to pebbly sand with cross-lamination (Csp); the thickness of this horizon laterally varies between 1 and 2 m, having a columnar structure and gray-brown color, which makes it distinguishable from afar; the contact between Cg and Csp horizon is wavy. Toward the top, there is a massive coarse-sandy horizon (Csm), orange-brown in
color and less than 1 m thick. This horizon (Csm) has a diffuse contact with the overlying horizon (Ccg), which is composed of subangular sandy gravel with few depositional structures. These two horizons (Csm and Ccg) are related, with debris flows coming from the frontal range. At the top of the Quaternary sequence, there is a gray horizon of fine to coarse sands with high volcanic content (Cc) that is covered by the recent soil (s). More details are provided next for each locality. Oreja Norte Road Cut The La Oreja pressure ridge is cut on the northern side by a minor road serving the village of La Iberia (A in Fig. 3). A large N-facing road cut along this road, and next to this secondary road intersection with the motorway, exposes (Figs. 5A and 5B) an E-verging, N-S–striking thrust-fault plane (N5°E, 33°W, striation contained by N75°E vertical plane) cutting across the entire W-dipping sedimentary sequence (intercalation of fanglomerates, silts, and sands). Near the ground surface, the fault cuts through the Quaternary sequence, overthrusting a paleosol that overlies the Csp horizon. Along this paleosol, the slip measures 1.6 m (Fig. 5B). This paleosol, which yields a radiocarbon age of 17,800 ± 660 yr B.P., was sampled at site P (Fig. 5B). The shape of the paleosol horizon resembles a fault wedge that results from
Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia overthrusting. At this place, the fault and the paleosol are buried by the Csm horizon. Cara Norte Oreja and Cara Sur Oreja Slightly southeast of the previous outcrop and on the motorway interchange itself, the road cut, which makes a loop, has two independent exposures that face each other. In the north exposure or Cara Norte Oreja outcrop (B in Fig. 3), an imbricated E-verging, N-S–striking thrust fault system is observed (B and C in Figs. 3 and 5A) that re-utilizes bedding planes within the Neogene Valle Group (Tp in Figs. 5B and 5C). Most of these faults (1–6 in Fig. 5C) are sealed by the unconformably overlying clastic sequence of inferred Quaternary age. One of the planes farthest to the east (7 in Fig. 5C) cuts the basal conglomerate (Cg in Fig. 5C). The offset of Cg measures 3 m (Figs. 5B and 5C). Following toward the east, the next fault (8 in Figs. 5B and 5C) overthrusts onto an organic-rich colluvial wedge 14C dated at 12,840 ± 40 yr B.P. (P in Fig. 5C). In the south face of this motorway interchange, the Cara Sur Oreja outcrop (C in Fig. 3) exposes the same faults (7 and 8) as the previous outcrop. In this outcrop (Fig. 5B), the offset of the Csp horizon is more visible, and it seems that after the deposition of the Cc horizon, a more recent event occurred. Variante Road Cut South of the previous sites, the E-W–trending road exhibits two 200-m-long, N- and S-facing exposures (D and E in Fig. 3). The north outcrop shows the subtle folding of the few-meterthick, Quaternary alluvial sedimentary cover as a result of a W-verging fault plane, as well as the thickening of the Csp horizon in front of the thrust fault and behind the little piggyback basin of the fold (Fig. 6A). On the south exposure (D in Fig. 3), the Quaternary horizons are exposed (Cg and Csp in Fig. 6B), and they truncate the underlying Valle Group clastic sequence (Tp in Fig. 6B). These overlying horizons seem to be gently folded because of the lateral variations in thickness (they thin up toward the anticline crest and thicken away from it), although some complications are introduced by the presence of conjugate fault sets in the Neogene rocks due to moment-bend normal faulting. The western end of this long outcrop (D in Figs. 3 and 6B) exhibits a package of the Valle Group sandstones roughly dipping 30°–45°W. Bedding planes within this unit display fault striation (N-S 25°W, striation trend N80°W; N-S 54°W, striation pitch 73°S; and N17°E, 52°W, striation pitch 60°S). Consequently, these measured fault planes are thrust faults with a very minor horizontal component, slightly sinistral and dextral, respectively. In particular, the first one disrupts the basal gravels of the Quaternary unit by ~3 m (fault 4 in Fig. 6B). The fault plane close to the top of the hanging-wall block exhibits an organic-rich fault wedge that contains some clasts from the basal gravel, which was radiocarbon dated at 7930 ± 60 yr B.P. (fault 2 in Fig. 6B).
103
El Ahorcado Scarp At the western foot of the El Ahorcado flexural scarp (F in Figs. 3 and Fig. 7A), a sewer ditch exposed two colluvial wedges derived from the scarp top by erosion (Figs. 7B and 7C). These colluvial deposits are intercalated between paleosols. The radiocarbon ages of this late Pleistocene–Holocene stratigraphic sequence do not follow a conventional younging-to-top chronologic order (Fig. 7C). However, the paleosols and soils, which are formed in situ (P1, P3, and P6 in Fig. 7C), do young upward in sequence, whereas the colluvial wedges, which accumulate at the scarp foot from erosion products, including reworked paleosols (P2, P4, and P5 in Fig. 7C), progressively get older from bottom to top. We have interpreted this as evidence of exhumation and progressive downcutting of the top of El Ahorcado scarp during incremental flexure (vertical growth of the hanging-wall compartment linked to stick-slip coseismic deformation along the thrust fault). In contrast, the paleosols and soils were generated during the interseismic period of the earthquake cycle, while the fault was loading elastic strain. Consequently, the two latest events occurred just prior to the burial of the older paleosols by colluvial wedges. Paleosols were dated at 13,070 ± 80 yr B.P. and 7460 ± 330 yr B.P., implying that earthquakes have happened at around 13–12 and 7–6 k.y. B.P. This return period is at least in agreement with the time of formation of the present-day soil, which started at ca. 5770 ± 130 yr B.P. This would imply that either the forthcoming equivalent earthquake is approaching or it just happened in the recent past. This second option is also likely because the Amaime region, 40 km to the south, was struck by a rather harmful historic earthquake in A.D. 1766 (Arboleda, 1956; Ramírez, 1975; Meyer, 1983; Espinosa, 1996), and the causative fault has not yet been identified. Espinosa (1996, p. 249) called this event “El Terremoto de Buga del 9 de Julio de 1766.” Severe damages were reported at Buga and Cali; from the original accounts in Ramírez (1975), it would seem that the epicenter was closer to Buga than to Cali, although the lack of geological effects in the historical record makes this difficult to prove. River Cut at Tuluá River The Tuluá River, nearing the Cauca valley, cuts the northern tip of an active N-S–trending compressive ridge that exhibits opposing flexural scarps on both longer sides (G in Figs. 3, 8A, and 8B). A several-tens-of-meters-high erosional cliff has formed against the southern margin of the Tuluá River. This outcrop, although quite vegetated, allows the identification of a W-verging, N-S–striking, intermediate-dip thrust fault crosscutting the entire cliff wall. Drag folding is conspicuous along the fault plane (1 in Fig. 8C). Very locally, this thrust fault exhibits a small antithetic fault (a back thrust) bounding a mesoscale popup structure (2 in Fig. 8C). Its shape and the compressional ridge morphology seem to be much alike. Its geometry also supports the interpreted structure for this ridge, where both flexural scarps would result from the activity of the main W-verging thrust and
104
López C. and Audemard M.
the E-verging back thrust. Because of poor outcropping conditions, it is hard to tell the sedimentary unit that is being crosscut. However, the absence of greenish colors in the alluvial sequence, which characterize the Tertiary La Paila Formation, suggests that this is a younger alluvial unit, most likely Quaternary in age. DISCUSSION The geomorphic and geologic elements near Tuluá and surroundings discussed in the previous sections strongly attest to the Quaternary, and even Holocene, tectonic activity of a set of N-S– trending thrust faults in the western foothills of the Central Cordillera, the main vergence of which is to the west. Locally, these faults show antithetic back thrusts. Except for the paleoseismic work by Woodward-Clyde Consultants (1983) at the Venecia and Piedechinche sites, very little evidence of ongoing compressional deformation in the western foothills of the Central Cordillera is known or has been published so far. Most of the reported Quaternary activity in the region is related to the roughly N-S–striking strike-slip Romeral fault system (Woodward-Clyde Consultants, 1983; Paris et al., 2000), including any of its known traces or segments like Silvia-Pijao, Cauca-Almaguer, or Guabas-Pradera, with a secondary thrust component that lies essentially along the western front of the Central Cordillera. On the other hand, a secondary normal component had been assigned to the Romeral fault system at the Silvia-Pijao and Cauca-Almaguer N-S–striking, left-lateral, strike-slip faults on the Montenegro and Armenia segments (Paris et al., 2000). The types of evidence herein discussed, together with the previous paleoseismic investigations by Woodward-Clyde Consultants (1983), unequivocally demonstrate that a Holocene E-W– directed compression, other than inducing simple active transpression along the different active traces of the Romeral fault system, is shortening the entire western foothills of the Central Cordillera unit. This unit, which mainly sits west of the Romeral fault system and specifically west of the Cauca-Almaguer fault, is still being uplifted and overthrusted over the Cauca valley; the westernmost structural and stratigraphic evidence of this recent activity has been additionally documented by López and MorenoSánchez (2005) and López (2006) at the latitude of Sonso, in the El Vínculo Quarry, near the maximum strangle of the Cauca valley by the Buga Salient. This set of evidence suggests the Holocene shortening of the western foothills of the Central Cordillera in this region, in clear contradiction with what had been previously stated by other authors, such as Alfonso et al. (1994) and Nivia (2001). The latter authors claim that the foothills are currently inactive. From a paleoseismic viewpoint, there is no doubt that several major earthquakes have taken place on thrust faults in Holocene times near Tuluá, thus bringing supporting evidence to the results that Woodward-Clyde Consultants (1983) gathered in the Amaime region. At the Venecia trench site, the WoodwardClyde Consultants interpreted the occurrence of two Holocene thrust events: (1) the latest one, younger than the present-day organic-rich soil formation, having an age estimated at 2 ka (based
on unknown criteria), and (2) a previous earthquake bracketed between the estimated age of the soil (2 ka) and a buried paleosol radiocarbon dated at 6320 yr. We feel that those two event ages could be better bracketed. We interpret that the penultimate event is just slightly younger than the last buried paleosol, when soil formation was abruptly stopped by scarp erosion or degradation, i.e., some 6 k.y. ago. The latest event is definitely very young and could well correlate to the Buga A.D. 1766 earthquake that struck the western foothills of the Central Cordillera region. We can confidently state that the return period is by no means longer than 6 k.y. As to the event magnitudes, they can be estimated from the vertical throws measured in trench walls. To illustrate this, the latest event corresponds to a throw of 1 m (WoodwardClyde Consultants, 1983). No fault plane was actually observed in the trench. So, assuming that the thrust fault typically dips 30°, the coseismic slip is twice as much as the measured throw. Two meters of coseismic slip on thrust faults may well produce earthquakes with magnitude greater than 7, in the order of 7.3–7.4 (e.g., Slemmons, 1977; Wells and Coppersmith, 1994). Instead, the 6-k.y.-old event may be slightly smaller because the measured throw is 0.75 m. It is worth mentioning that these two paleoseismically evaluated localities (Venecia and Piedechinche in the region) are located at the southern tip of the major structure herein described at Tuluá. This major feature extends for over 50 km between Bugalagrande in the north and Amaime in the south, and it shows a concave-to-the-E arcuate shape. Based on the shape and structural style observed along this structure, we interpret it as a large thrust sheet, where Tuluá would be at its frontal ramp and Amaime would sit on its southern lateral ramp (see figures 53–56 in López, 2006). The orientation of the main stress axes deduced at the latitude of Sonso by López (2006) is representative of the main influence of the transcurrent regime. Accordingly, using a deformation model (Wilcox et al., 1973; Crowell, 1984; Harding and Lowel, 1979), this compressive system is developed at a left step of the western termination of the right-lateral, strike-slip Ibagué fault. The prehistoric seismic activity revealed at both Tuluá (in this paper) and Amaime (Woodward-Clyde Consultants, 1983) seems to support our structural interpretation. An event at ca. 6 ka is common to the two regions (Venecia trench in Amaime and Variante outcrop and El Ahorcado trench in the surrounds of Tuluá). Furthermore, these two regions (Amaime and Tuluá) are some 55 km apart. A total surface rupture of such a length along thrust faults is consistent with both a magnitude in the order of 7.3–7.4 and a coseismic slip of ~2 m, after relationships by Slemmons (1977), Shimazaki (1986), and Wells and Coppersmith (1994). In the Tuluá region, two other older events have been identified. Overthrusted organic-rich soils at La Oreja Norte road cut and Cara Norte Oreja outcrops dated at 17,800 ± 660 and 12,820 ± 40 yr B.P., respectively, attest to two late Pleistocene earthquakes. The events should be slightly younger than the buried soils, as is shown by faults 8 and 9 in the Cara Norte Oreja and Cara Sur Oreja outcrops (Figs. 5B and 5C). In addition to these outcrops, a younger earthquake was also dated at the toe
Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia of the El Ahorcado flexural scarp. Here, the fault activity has been revealed by the formation of two colluvial wedges derived from the erosion of the scarp crest that buried the organic-rich soils. The two soils have been dated at 13,070 ± 80 yr B.P. and 7460 ± 330 yr B.P., implying that these two earthquakes happened at around 13–12 and 7–6 kyr B.P. This 7–6 kyr B.P. event reported at the El Ahorcado trench could well correspond to the most recent event identified at the Variante outcrop, which should be somewhat younger than 7930 ± 60 yr B.P. These two sites are less than 5 km apart. As mentioned earlier, this earthquake was also recorded at the Venecia trench in Amaime, located some 55 km southward. No reliable magnitude estimates can be provided from the Tuluá outcrops alone for diverse reasons: (1) The throw may correspond to more than one event (e.g., La Oreja Norte, La Oreja Sur, Variante, etc.); (2) throws are measured on back-thrust fault planes (e.g., La Oreja Norte, La Oreja Sur, Variante, etc.) but not on the main feature; or (3) the fault plane is not visible, and events are interpreted from stratigraphic indicators (e.g., El Ahorcado). However, since these deformations affect the ground surface directly or indirectly, we confidently presume that all these events had magnitudes in the range of 7 or larger. This is supported by the 7–6 kyr B.P. event that displays evidence as far away as 55 km. With regard to the earthquake recurrence on these thrust faults bounding the western foothills of the Central Cordillera, the four events interpreted for the entire Tuluá-Amaime region seem to repeat rather characteristically every 6 k.y. The most recent earthquake could well be the “Buga 9 July of 1766 earthquake” (Arboleda, 1956; Ramírez, 1975; Espinosa, 1996), the epicenter of which must have been close to Buga, because not only were the aftershocks felt for a longer time span there, but also because of the shape and distribution of the isoseismals determined by Espinosa (1996). According to this author, these isoseismal lines show an increase toward the east (toward Ibagué) and little effects north of the latitude of Ibagué, such as in Cartago (Valle). The other three prior earthquakes roughly occurred around 7–6, 13–12, and 18–17 k.y. ago. If the Buga 1766 event happened to be of smaller magnitude, a forthcoming earthquake of this magnitude should then take place in the near future.
105
of staircased terraces, piggyback basins, and drainage migration, inversion, and diversion, among several other features. In some cases, they are the direct expression of either the up-dip propagation of fault planes or regional uplift. (4) Fault-plane kinematic indicators on E-verging thrusts attest to an ongoing regional E-W–trending maximum horizontal stress. (5) Hanging flat remnants at different elevations on the Central Cordillera western flank support regional chain uplift. In addition, if those remnants could be genetically correlated to the lower hills, this would also indicate that the foothills’ peneplanation occurred through different phases (polygenic peneplain). (6) Prehistoric earthquakes spanning the late Pleistocene and Holocene have taken place on thrusts associated with conspicuous flexural scarps at the front of the foothills. These Ms 7.0 events recur every 6–7 k.y. As a final remark, the analysis of drainage patterns and anomalies in this case study proved to be a helpful tool in the identification of not only the active blind-thrust faults and their associated folding but also of the active regional tectonic setting as a whole. Obviously, this evidence guided us to find the main trenching sites, where it was possible to establish time constraints for the earthquakes and determine their magnitude from the measured displacements or offsets. In this particular study, outcrops and existing trenches happened to be very useful for paleoseismic purposes. ACKNOWLEDGMENTS These results are from a larger study “Hacia un modelo de la Sismicidad del Sur-Occidente Colombiano: Investigaciones Paleosísmicas en la Región del Valle del Cauca,” which was partly funded by COLCIENCIAS (project 1106-05-11117-CT: 60-2000), for which we are very thankful. These investigations were also carried out in the frame of the first author’s M.Sc. thesis at EAFIT University. Thanks are also given to the Seismological Observatory of the Southwest-OSSO of the Valle University, as well as Corporación Observatorio Sismológico del Suroccidente, for funding and logistics. Special thanks go to Andrés Velásquez for field assistance, and to FUNVISIS, which allowed the second author to collaborate with his Colombian counterparts as project advisor. Reviews from Federico A. Pasquarè and an anonymous reviewer are greatly appreciated.
CONCLUSIONS REFERENCES CITED The recent tectonic activity of a compressional zone in the western foothills of the Central Cordillera of Colombia at the latitude of Tuluá and adjacent areas is documented by: (1) flexural scarps on thrust fault planes that re-utilize the Neogene fold-andthrust belt, which had previously been postulated to be inactive in recent times. These scarps are the western foothills boundary between the Amaime and Bugalagrande latitudes, and they affect the alluvial plain of the Cauca river. (2) Pressure ridges, which are shaped by both main W-verging thrust faults and their antithetic faults, all involve the Holocene cover. (3) Internal deformations in the foothills, due to active folding, induce the development
Alfonso, C.A., Sacks, P.E., Secor, D.T., Rine, J., and Pérez, V., 1994, A Tertiary fold and thrust belt in the Valle del Cauca Basin, Colombian Andes: Journal of South American Earth Sciences, v. 7, no. 3–4, p. 387–402, doi: 10.1016/0895-9811(94)90023-X. Arboleda, G., 1956, Historia de Cali. Desde los Orígenes hasta la Expiración del Período Colonial: Cali, Colombia, Carvajal y Compañía, 420 p. Audemard M., F.A., 1999, Morpho-structural expression of active thrust fault systems in the humid tropical foothills of Colombia and Venezuela: Zeitschrift für Geomorphologie N. F., v. 118, p. 227–244. Audemard M., F.A., 2002, Ruptura de los grandes sismos históricos venezolanos de los siglos XIX y XX; revelados por la sismicidad instrumental contemporánea, in XI Congreso Venezolano de Geofísica, 17–20 November: Caracas, Venezuela, Sociedad Venezolana de Ingenieros Geofísicos, 8 p. (extended abstract on CD).
106
López C. and Audemard M.
Audemard M., F.A., 2003, Geomorphic and geologic evidence of ongoing uplift and deformation in the Merida Andes, Venezuela: Quaternary International, v. 101–102, p. 43–65, doi:10.1016/S1040-6182(02)00128-3. Audemard M., F.A., and Robertson, K., 1996, Quaternary tectonic activity of the Llanos foothills thrust system, Eastern Cordillera of Colombia: Geomorphological and geological evidences from La Florida anticline, between the Upía and Cusiana Rivers, in Third International Symposium on Andean Geodynamics: St. Malo, France, p. 137–140. Biddle, K., and Christie-Blick, N., 1985, Strike-slip deformation, basin formation, and sedimentation, in Biddle, K., and Christie-Blick, N., eds., Strike-Slip Deformation, Basin Formation, and Sedimentation: Society of Economic Paleontologists and Mineralogists Special Publication 37, p. 375–385. Burbank, D.W., Meigs, A., and Brozovic, N., 1996, Interactions of growing folds and coeval depositional systems: Basin Research, v. 8, p. 199–223, doi:10.1046/j.1365-2117.1996.00181.x. Burnett, A.W., and Schumm, S.A., 1983, Alluvial river response to neotectonics deformation in Louisiana and Mississippi: Science, v. 222, p. 49–50, doi:10.1126/science.222.4619.49. Butler, K., and Schamel, S., 1988, Structure along the eastern margin of the Central Cordillera, Upper Magdalena Valley, Colombia: Journal of South American Earth Sciences, v. 1, p. 109–120, doi:10.1016/0895-9811(88 )90019-3. Corredor, F., 2003, Seismic strain rates and distributed continental deformation in the northern Andes and three-dimensional seismotectonics of northwestern South America: Tectonophysics, v. 372, p. 147–166, doi:10.1016/ S0040-1951(03)00276-2. Crowell, J.C., 1982, The Violin Breccia, Ridge Basin, southern California, in Crowell, J.C., and Link, M.H., eds., Geologic History of the Ridge Basin, Southern California: Los Angeles, Pacific Section, SEPM (Society of Economic Paleontologists and Mineralogists), p. 89–98. Diederix, H., Gómez, H., Knobzi, J., and Singer, A., 1987, Indicios neotectónicos de la Falla de Ibagué en el sector Ibagué-Piedras: Departamento del Tolima, Colombia, Revista Centro Interameriano de Fotointerpretación, Bogotá, v. 11, p. 242–252. Ego, F., and Sébrier, M., 1995, Is the Cauca–Patía and Romeral fault system left or right lateral?: Geophysical Research Letters, v. 22, p. 33–36. Espinosa, A., 1996, El Terremoto de Buga del 9 de Julio de 1766. Análisis Histórico y Geotectónico: Revista de la Academia Colombiana de Ciencias Exactas, Físicas y Naturales, v. 77, p. 247–258. Harding, T.P., and Lowel, J.D., 1979, Structural styles, their plate-tectonic habitats, and hydrocarbon traps in petroleum provinces: American Association of Petroleum Geologists Bulletin, v. 63, p. 1016–1058. Holbrook, J., and Schumm, S.A., 1999, Geomorphic and sedimentary response of rivers to tectonic deformation: A brief review and critique of a tool for recognizing subtle epeirogenic deformation in modern and ancient settings: Tectonophysics, v. 305, p. 287–306, doi:10.1016/S0040-1951(99 )00011-6. Howard, A.D., 1967, Drainage analysis in geologic interpretation: A summation: American Association of Petroleum Geologists Bulletin, v. 51, p. 2246–2259. Ingeominas, 1988, Mapa Geológico de Colombia: Bogotá, Ingeominas, scale 1:1,500,000. James, M.E., 1985, Evidencia de colisión entre la miniplaca bloque Andino y la placa Norteamericana desde el Mioceno Medio, in Wagner, J.V., ed., VI Congreso Latinoamericano de Geología, Volume Tomo I: Bogotá, Colombia, Ingeominas, p. 58–75. López C., M.C., 2006, Análisis de Deformación Tectónica en los Piedemontes de las Cordilleras Central y Occidental, Valle del Cauca, Colombia— Contribuciones Paleosísmicas [M.Sc. thesis]: Medellín, Colombia, Universidad Eafit, 113 p. López C., M.C., and Moreno-Sánchez, M., 2005, Tectónica y sedimentación en el piedemonte occidental de la Cordillera Central de Colombia, un ejemplo en la Cantera El Vínculo, in X Congreso Colombiano de Geología: Bogotá, Colombia, Ingeominas, Universidad Nacional de Colombia, Sociedad Colombiana de Geología, abstract 113, p. 1–10. López C., M.C., Velásquez, A.A., and Toro, G.E., 2002, Transpresión activa en el Valle del Cauca-Colombia, in III Jornadas de Sismología Histórica— Serie Técnica No. 1, 2002: Caracas, Venezuela, p. 222–225. López C., M.C., Velásquez, A.A., Toro, G.E., Audemard M., F.A., Meyer, H., and Hermelín, M., 2003a, Avances de las investigaciones paleosísmicas en el Valle del Cauca, in IX Congreso Colombiano de Geología. Conocer la Tierra para un Futuro Mejor: Medellín, Colombia, Ingeominas, p. 121.
López C., M.C., Velásquez, A.A., Toro, G.E., Audemard M., F.A., Meyer, H., and Hermelín, M., 2003b, Evidence of Holocene compression in the Valle del Cauca, along the western foothills of the Central Cordillera of Colombia [abs.], in XVI International Union for Quaternary Research Congress: Session S17. Paleoseismology in the Twenty-First Century: A Global Perspective (Posters): Reno, Nevada. López C., M.C., Audemard M., F.A., and Velásquez, A.A., 2004, Compresión Holocena en el Valle del Cauca, Colombia, in Monsalve J., H., and Vargas J., C.A., eds., I Seminario Latinoamericano de Sismología, II Congreso Colombiano de Sismología: Armenia, Colombia, Grupo Quimbaya— Universidad del Quindío, Grupo de Geofísica—Universidad Nacional del Colombia, p. 1–17. López C., M.C., Audemard M., F.A., and Velásquez, A.A., 2005, Evidencias geomorfológicas y estratigráficas de compresión Holocena en el Valle del Cauca, Colombia, in X Congreso Colombiano de Geología: Bogotá, Colombia, Ingeominas, Universidad Nacional de Colombia, Sociedad Colombiana de Geología, abstract 112, 11 p. López C., M.C., Moreno-Sánchez, M., and Audemard M., F.A., 2009, Deformación tectónica reciente en los piedemontes de las Cordilleras Central y Occidental, Valle del Cauca, Colombia: Boletín de Geología Universidad Industrial de Santander, v. 31, no. 1, p. 11–29. Lowell, D.J., 1985, Structural Styles in Petroleum Exploration: Tulsa, Oklahoma, Oil & Gas Consultants International, Inc., 460 p. MacDonald, W.D., Estrada, J.J., Sierra, G.M., and González, H., 1996, Late Cenozoic tectonics and paleomagnetism of North Cauca Basin intrusions, Colombian Andes: Dual rotation modes: Tectonophysics, v. 261, p. 277– 289, doi:10.1016/0040-1951(95)00184-0. Marín, W., and Romero, J., 1988, Proyecto Neotectónica del Sur Occidente Colombiano: Cali, Colombia, Ingeominas. Maya, M., and González, H., 1995, Unidades litodémicas en la Cordillera Central de Colombia: Boletín Geológico: Ingeominas, v. 35, p. 43–57. McCourt, W.J., 1984, The geology of the Central Cordillera in the department of Valle del Cauca, Quindío and NW Tolima: British Geological Survey Report Series, v. 84, p. 8–49. McCourt, W.J., Aspden, J.A., and Brook, M., 1984, New geological and geochronological data from the Colombian Andes: Continental growth by multiple accretion: Journal of the Geological Society of London, v. 141, p. 831–845, doi:10.1144/gsjgs.141.5.0831. Medwedeff, D.A., and Suppe, J., 1997, Multibend fault-bend folding: Journal of Structural Geology, v. 19, no. 3–4, p. 279–292. Mejía, J., and Meyer, H., 2004, Modelo detallado preliminar de la sismicidad en el occidente de Colombia, in Monsalve J., H., and Vargas J., C.A., eds., I Congreso Latinoamericano de Sismología, II Congreso Colombiano de Sismología: Armenia, Colombia, Grupo Quimbaya—Universidad del Quindío, Grupo de Geofísica—Universidad Nacional del Colombia 12 p. Meyer, H., 1983, Reports from Hansjurgen Meyer Regarding Macroseismic Studies and Geophysical Data for the Calima III Project in “Seismic Hazard Evaluation Calima III Project Colombia”: San Francisco, California, Woodward-Clyde Consultants para Consorcio Integral-Planes-CVC, 116 p. Meyer, H., and Mejía, J., 1995, On the convergence related faulting in the North Andean block; new details from regional seismic observations, in IASPEI (International Association of Seismology and Physics of the Earth’s Interior), ed., Subduction Zones: Structures, Dynamics and Magmatism, Volume Session S17 (Posters): 29th Assembly: Thessalonoki, Greece. Montes, C., Restrepo-Pace, P.A., and Hatcher, R.D., Jr., 2003, Three dimensional structure and kinematics of the Piedras-Girardot fold belt: Surface expression of transpressional deformation in the northern Andes, in Bartolini, C., Buffler, R.T., and Blickwede, J., eds., The Circum–Gulf of Mexico and the Caribbean: Hydrocarbon Habitats, Basin Formation, and Plate Tectonics: American Association of Petroleum Geologists Memoir 79, p. 849–873. Montes, N., Osorio, J.A., Velandia, F., Acosta, J., and Núñez, A., 2005a, Caracterización sismogénica de la Falla Ibagué, Colombia, in Colombia, U.N.d., ed., X Congreso Colombiano de Geología: Bogotá, Colombia, Ingeominas, Universidad Nacional de Colombia, Sociedad Colombiana de Geología, abstract 168, 10 p. Montes, N., Osorio, J.A., Velandia, F., Acosta, J., Nuñez, A., Audemard M., F.A., and Diederix, H., 2005b, Evaluación Paleosísmica Los Gomos, Falla Ibagué, Colombia, in X Congreso Colombiano de Geología: Bogotá, Colombia, Ingeominas, Universidad Nacional de Colombia, Sociedad Colombiana de Geología, abstract 169, 10 p.
Holocene compression at Tuluá, along the western foothills of the Central Cordillera of Colombia Nivia, A., 2001, Mapa Geológico del Departamento del Valle: Cali, Colombia, Ingeominas, scale 1:250,000. Nivia, A., Galvis, N., and Maya, M., 1997, Mapa Geológico de Colombia. Geología de la Plancha 242-Zarzal. Memoria Explicativa: Bucaramanga, Colombia, Ingeominas, scale 1:100,000. Ouchi, S., 1985, Response of alluvial rivers to slow active tectonic movement: Geological Society of America Bulletin, v. 96, p. 504–515, doi:10 .1130/0016-7606(1985)96<504:ROARTS>2.0.CO;2. Paris, G., Marín, W., Romero, J., and Wagner, J.J., 1989, Evidencias de actividad neotectónica en el Suroccidente Colombiano, in V Congreso Colombiano de Geología: Bucaramanga, Colombia, Sociedad Colombiana de Geología, p. 193–213. Paris, G., Machette, M., Dart, R., and Haller, K., 2000, Map and Database of Quaternary Faults and Folds in Colombia and Its Offshore Regions: U.S. Geological Survey Open-File Report 00-0284, map at 2,500,000 scale and report, vol. 2004, 66 p. Philip, H., 1983, Structural analysis and interpretation of the surface deformations on the Asnam earthquake of October 10, 1980: Tectonics, v. 2, p. 17–49, doi:10.1029/TC002i001p00017. Philip, H., Rogozhin, E., Cisternas, A., Bousquet, J.C., Borisov, B., and Karakhanian, A., 1992, The American earthquake of December 7, 1988; faulting and folding, neotectonics, and paleoseismicity: Geophysical Journal International, v. 110, p. 141–158, doi:10.1111/j.1365-246X.1992 .tb00718.x. Ramírez, J.E., 1975, Historia de los Terremotos en Colombia: Bogotá, Los Andes, 236 p. Schwinn, W.L., 1969, Guidebook to the Geology of the Cali Area, Valle del Cauca, Colombia: Bogotá, Colombian Society of Petroleum Geologists and Geophysicists, 29 p. Shimazaki, K., 1986, Small and large earthquakes: The effects of the thickness of seismogenic layer and the free surface, in Shamita, D., Boatwright, J., and Scholz, C.H., eds., Earthquake Source Mechanics: Washington, D.C., American Geophysical Union Maurice Ewing Series, no. 6, p. 209–216. Slemmons, D.B., 1977, State-of-the-Art for Assessing Earthquake Hazards in the United States: Faults and Earthquake Magnitude: Vicksburg, Mississippi, U.S. Army Engineer Waterways Experiment Station, Soils and Pavements Laboratory, 129 p.
107
Suter, F., Sartori, M., Neuwerth, R., and Gorin, G., 2008, Structural imprints at the front of the Chocó-Panamá indenter: Field data from the North Cauca Valley Basin, central Colombia: Tectonophysics, v. 460, p. 134–157, doi:10.1016/j.tecto.2008.07.015. Taboada, A., Rivera, L.A., Fuenzalida, A., Cisternas, A., Phillip, H., Bijwaard, H., Olaya, J., and Rivera, C., 2000, Geodynamics of the northern Andes: Subductions and intracontinental deformation (Colombia): Tectonics, v. 19, p. 787–813, doi:10.1029/2000TC900004. Toussaint, J.F., and Restrepo, J.J., 1987, Límites de placas y acortamientos recientes entre los paralelos 5°N y 8°N, Andes colombianos: Revista Geológica de Chile, v. 31, p. 95–100. Trenkamp, R., Kellogg, J.N., Freymueller, J.T., and Mora, H.P., 2002, Wide plate margin deformation, southern Central America and northwestern South America; CASA GPS observations: Journal of South American Earth Sciences, v. 15, p. 157–171, doi:10.1016/S0895-9811(02)00018-4. Van der Hammen, T., 1958, Estratigrafía del Terciario y Maastrichtiense continentales y tectonogénesis de los Andes Colombianos: Boletín Geológico: Servicio Geológico Nacional (Colombia), v. 6, p. 67–128. Vergara, H., 1988, Rasgos Neotectónicos en el Noreste del Departamento del Tolima: Bogotá, Ingeominas, Boletín Geológico, v. 30, p. 21–42. Wells, D., and Coppersmith, K., 1994, New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement: Bulletin of the Seismological Society of America, v. 84, p. 974–1002. Wesson, R.L., Helley, E.J., Lajoie, K.R., and Wentworth, C.M., 1975, Faults and future earthquakes, in Borchardt, R.D., ed., Studies for Seismic Zonation of the San Francisco Bay Region: U.S. Geological Survey Professional Paper 941A, p. 5–30. Wilcox, R.E., Harding, T.P., and Seeley, D.R., 1973, Basic wrench tectonics: American Association of Petroleum Geologists Bulletin, v. 57, p. 74–96. Woodward-Clyde Consultants, 1983, Seismic Hazard Evaluation Calima III Proyect: Cali, Colombia, Consorcio Integral–Planes, Ltda. Ingenieros Consultores, Corporación Autónoma Regional del Cauca (C.V.C.), 116 p.
MANUSCRIPT ACCEPTED BY THE SOCIETY 7 DECEMBER 2010
Printed in the USA
The Geological Society of America Special Paper 479 2011
Style and timing of late Quaternary faulting on the Lake Edgar fault, southwest Tasmania, Australia: Implications for hazard assessment in intracratonic areas Dan Clark* Geohazards Division, Earthquake Hazard and Neotectonics Project, GPO Box 378, Canberra, ACT 2601, Australia Matt Cupper* Mike Sandiford* School of Earth Sciences, The University of Melbourne, Melbourne, Victoria 3010, Australia Kevin Kiernan* School of Geography and Environmental Studies, University of Tasmania, Hobart, Tasmania 7001, Australia
ABSTRACT Geomorphic analysis of the ~30-km-long Lake Edgar fault scarp in southwestern Tasmania suggests that three large surface-rupturing events with vertical displacements of 2.4 m to 3.1 m have occurred in late Quaternary time. Optically stimulated luminescence (OSL) age estimates from a sequence of three periglacial fluvial terraces associated with faulting constrain these events to ca. 18 ka, ca. 28 ka, and ca. 48–61 ka. A similar amount of vertical displacement during each faulting event suggests that surface-breaking earthquakes on this fault are characteristically of magnitude MW 6.8–7.0. Estimates for the average slip rate calculated over two complete seismic cycles range from 0.11 to 0.24 mm/yr, which is large for a stable continental region fault. This sequence represents the first recurrence data for surface-rupturing earthquakes on an eastern Australian Quaternary fault, and one of only a few for the entire Australian continent.
INTRODUCTION Quantitative seismotectonic and seismic hazard analysis of stable continental regions is hampered by an absence of good constraints on the recurrence intervals of large earthquakes, which are infrequent in these regions compared to plate-boundary
regions. Paleoseismological investigations provide the only viable avenue to obtain such constraints (e.g., Machette, 1998) in a useful time frame. The success of paleoseismological investigations has greatly increased with recent developments in Quaternary dating techniques for sedimentary deposits, such as thermoluminescence (TL) and particularly optically stimulated luminescence
*E-mails:
[email protected];
[email protected];
[email protected];
[email protected]. Clark, D., Cupper, M., Sandiford, M., and Kiernan, K., 2011, Style and timing of late Quaternary faulting on the Lake Edgar fault, southwest Tasmania, Australia: Implications for hazard assessment in intracratonic areas, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, p. 109–131, doi:10.1130/2011.2479(05). For permission to copy, contact
[email protected]. © 2011 The Geological Society of America. All rights reserved.
109
110
Clark et al.
(OSL; e.g., Burbank and Anderson, 2000). In this paper, we use the OSL technique to constrain the timing of large earthquakes on the Lake Edgar fault in Tasmania, Australia (Fig. 1). Australia is a relatively active stable continental region, with an estimated seismic-moment release of ~9 × 1023 dyn cm/yr (Johnson et al., 1994). The historical seismic record suggests that earthquakes of moment magnitude (MW) 6.8 occur about once every 20–30 yr across the Australian continent. A longer record of large earthquakes is indicated by an abundance of surfacerupturing faults of presumed late Quaternary age (Crone et al., 1992, 1997, 2003; note that late Quaternary is taken to mean younger than marine isotope stage 5e, or younger than ca. 125 ka). Crone et al. (1997, 2003) obtained recurrence data for several of these surface-rupturing faults in Western Australia, South Australia, and the Northern Territory. These authors demonstrated episodic rupture characteristics on these faults with interseismic intervals of the order of 10 k.y. to 100 k.y. A prior study of the Lake Edgar fault by McCue et al. (2003) provided evidence for multiple Quaternary surface-rupturing events but lacked firm time constraints. The objective of our new work is to obtain new dating control for the timing of the earthquakes. We begin by presenting the results of field mapping and dating of well-constrained sandy sediment samples. These data are then used to construct a history of Quaternary displacement on the Lake Edgar fault, from which significant insight into the behavior of intracratonic faults may be gained. The results therefore have particular bearing on seismic hazard assessments in Australia, and for intracratonic areas worldwide. LAKE EDGAR FAULT AND ITS REGIONAL CONTEXT Regional Geological Context and Pre-Quaternary History of Movement The northerly trending, west-side-up Lake Edgar fault scarp lies on the boundary between the Adamsfield-Jubilee and Tyennan tectonic units (Fig. 1; see Seymour and Calver, 1995). The scarp marks the surface trace of the Lake Edgar fault, which has been mapped over a strike length of ~45 km (Turner et al., 1985; Seymour and Calver, 1995). The fault terminates to the north in the Adamsfield ultramafic complex and links into a southeasterly trending fault fronting the Arthur Range to the south. Drilling prior to construction of the Edgar dam (Fig. 2) suggested that the fault dips ~65°–70° west (Roberts et al., 1975). The fault is thought to have accommodated ~12 km of left-lateral displacement during the Paleozoic (Roberts et al., 1975; Calver et al., 1990). A narrow, possibly discontinuous, sliver of limestone has been dragged into the fault trace for a distance of 1.2 km south of the western Edgar dam abutment (Roberts et al., 1975). The fault juxtaposes mudstone, siltstone, orthoquartzite, and carbonates of the Proterozoic Clark Group against mudstone, chert, sandstone, and conglomerates of the Cambrian Island Road Formation and Ragged Complex (Brown et al., 1995). Island Road Formation and Ragged Complex rocks are only present on
the western side of the fault, whereas Clark Group rocks are on both sides of the fault. The ranges to the east of the fault, which direct west-flowing drainage across the fault in the Lake Edgar region, are mainly composed of orthoquartzite. Large areas of Quaternary sediment are localized to the eastern side of the fault (Brown et al., 1995; Kiernan, 2001; Fig. 2) owing to east-sidedown movement on the fault. Regional Geomorphology and Evidence for Quaternary Deformation The geomorphology of the region surrounding the Lake Edgar fault is dominated by processes and landforms relating to the Huon Plains, through which the scarp cuts, and the Mount Anne Massif, which lies to the east of the fault (Fig. 2). The vegetation consists of heathy shrubland and button grass plains (Figs. 3A–3C), with scrub and forest locally along drainage courses (Balmer and Corbett, 2001). Large-scale erosion, transportation, and deposition of coarse clastic sediment appear to have been very limited during the Holocene (Kiernan, 2001). The morphology of fans, which are prominent along the foot of the Mount Anne Massif, is the result of Pleistocene periglacial fluvial and alluvial sedimentation (Kiernan, 1990). These sediments are typically mantled by peat and organic-rich silt. Predam aerial photographs suggest that there has been a minor amount of lacustrine sedimentation within Lake Edgar and its companion lake (Kiernan, 2001; Fig. 4). Deposits and landforms relating to three or four distinct glaciations have been identified in the Lake Edgar region (Kiernan, 1990). The youngest is late Pleistocene (younger than 25 ka), whereas the oldest may be ca. 2 Ma in age. The intervening one or two glaciations are considered to be middle Pleistocene in age. The Lake Edgar fault scarp was first recognized as the surface expression of a recently active fault by Carey and Newstead (1960). Periglacial fans emanating from the ranges to the east are clearly cut by the fault, both to the north (Figs. 3D, 3E, and 5) and south (Figs. 3A, 3B, and 6) of Edgar Dam. Variability in the fault’s scarp height where it cuts the fan (2.5 m to 6 m; McCue et al., 2003, their figure 5) has been attributed to episodic incision of the scarp by fluvial processes. McCue et al. (2003) suggested that at least two surface-rupturing earthquake events, each resulting in ~2.5 m of west-side-up vertical displacement, had occurred in the Quaternary. McCue et al.’s (2003) log of a trench excavated by the Hydro Electric Commission (Roberts et al., 1975) clearly shows the trace of the most recent faulting event in fan gravels inferred to be late Pleistocene in age. The fault plane dips steeply to the west (60°–70°) but flattens toward the surface, becoming almost horizontal at its intersection with the ground surface. Folded diagenetic (or postdepositional) quartz laminae in the silty strata beneath the fan gravels truncate against the contact with the overlying (overthrust) gravel sheet. This deformation and truncation of strata inferred to be initially horizontal were attributed to a penultimate event (McCue et al., 2003).
Implications for hazard assessment in intracratonic areas
111
Figure 1. Location map and regional geology in the area of the Lake Edgar Fault (compiled from Brown et al., 1995; Kiernan, 2001; Seymour and Calver, 1995; Turner et al., 1985). Lithologic contacts are shown as dashed lines, while fault contacts are shown as solid lines. Lake Edgar fault is marked by a heavy line. Key to tectonic units: RC—Rocky Cape; D—Dundas; S—Sheffield; T—Tyennan; AJ—Adamsfield-Jubilee; NE—NE Tasmania; TB—Tasmania Basin.
Glacial outwash fans cut by the Lake Edgar fault have been interpreted to date from the early Pleistocene Eliza glacial stage (Kiernan, 1985, 1990). However, as mentioned already, McCue et al. (2003) suggested that the fault also truncates deposits relating to the most recent glaciation, but this was based upon
an earlier model (i.e., Carey and Newstead, 1960) that only recognized a single late last glacial stage glaciation in Tasmania. The younger of their two proposed events was inferred to have occurred subsequent to the Last Glacial Maximum (LGM; i.e., within the past 25,000 yr), based upon several lines of evidence.
112
Clark et al.
Figure 2. Map of the Lake Pedder area (gray pattern is permanent water) showing the location of the specific study sites mentioned in the text (modified after McCue et al., 2003). Areas of significant Quaternary sediment accumulation (>1 m) are shown by stipple pattern. Triangular ticks on the fault indicate westerly fault dip. The surface of Lake Pedder is ~310 m above mean sea level. Abbreviations for surface displacement: U—up; D—down.
First, the scarp is a very sharp and prominent feature in the landscape, significantly more so than scarps associated with several historic surface ruptures in Australia (e.g., Meckering—Clark and McCue, 2003; Tennant Creek—Crone et al., 1992). Second, geologically recent movement on the structure is inferred from
the observation that drainages ponded against the fault scarp have not yet been filled with sediment. This is despite the fact that southwest Tasmania receives ~2500 mm of rain annually (Carey and Newstead, 1960). Furthermore, the unweathered nature of the gravels displaced in the trench excavation contrasts
Implications for hazard assessment in intracratonic areas
113
Figure 3. Series of photographs illustrating the geomorphology of the Lake Edgar fault scarp. (A) Fluvial fans south of Edgar dam (looking south along the fault scarp), (B) detail of terrace 1 (looking south along fault scarp), (C) detail of terrace 2 (looking south), (D) sag pond north of Harlequin Hill (looking east toward the Condominium Creek catchment), (E) scarp profile north of the sag pond (looking south along fault; location shown on Fig. 5), and (F) gravel on terrace 2, south of Edgar dam; 33 cm pick for scale. For A, field of view of the middle ground at the position of the arrows is ~4.2 km. There is a 1.8-m-tall person standing on the Terrace 1 bench (between the arrow heads) in C.
with the highly weathered gravels characteristic of Eliza glacial stage fans. McCue et al. (2003) discussed a second fault scarp along the Gell River, some 50 km north of the Lake Edgar fault (Fig. 1). The Gell River scarp is shorter (~10 km long) and more subdued than the Lake Edgar scarp. This relationship was interpreted to reflect a greater time having elapsed since the last event on the Gell River scarp relative to the Lake Edgar scarp, although proof of fault movement has not been established, and preliminary field
observations suggest the possibility that the scarp could relate to the surface outcrop of a thin Permian gravel bed rather than to a fault (David Wilson, Hydro Tasmania, 2004, personal commun.). Although not coinciding with a mapped fault, the Gell River scarp is close to, and parallels, the boundary between the Tyennan and Adamsfield-Jubilee strato-tectonic elements, as does the Lake Edgar scarp. The two scarps are separated by the Adamsfield ultramafic complex, which McCue et al. (2003) speculated acts to concentrate strain in the Lake Edgar–Gell River region.
114
Clark et al.
Figure 4. Former Lake Edgar and its companion lake (to the north) prior to flooding beneath the Lake Pedder reservoir. The inferred maximum level of Lake Edgar, which is bounded by the fault scarp on the west, and the current impoundment level are indicated. Note that a westerly flowing drainage breaches the scarp on the southwestern margin of Lake Edgar. Tick marks are shown on the downthrown side of the scarp. (Tasmania South West Project, Run 1, Image T360-23, 16 February 1961; image used with permission.)
Figure 5. Aerial photograph of the sag pond to the north of Harlequin Hill, showing fan sources in the hills to the east. Location of Figure 10A is indicated. Tick marks are shown on the downthrown side of the scarp. (Tasmap 1:25,000, South West, Run 65, Image 008, 28 January 1988; image used with permission.)
Implications for hazard assessment in intracratonic areas
Figure 6. (A) Portion of aerial photograph showing study area location detail for profiles to the south of Edgar dam. The fan forming the source for the terraces investigated can be seen emanating from the hills to the east of the scarp. Locations of insets B and C are shown. (B) Detail of the geomorphology of the dissected fan showing location of profiles, transverse (scarp top) profile, and trench. White circles mark the ends of E-W profiles, and shaded circles mark the ends of the scarp top profile. Black arrows show the slope direction. (C) Map of geomorphic units (mainly terraces) identified in this study, with sample localities. Tick marks are shown on the downthrown side of the scarp. Flow directions on streams are shown by open arrow heads. (Tasmap, 1:25,000, South West, Run 68, Image 023, 28 January 1988; image used with permission.)
115
116
Clark et al.
As is typically the case for stable continental region faults (e.g., McCue, 1990; Clark and McCue, 2003; Crone et al., 2003; Clark and Bodorkos, 2004), geomorphic evidence for Quaternary deformation in the Lake Edgar region is not complemented by a record of historic seismicity. Only two events of magnitude greater than three have been recorded in the region of Figure 1 (Shirley, 1980; Geoscience Australia online earthquake database, http://www.ga.gov.au/oracle/quake/quake_online.jsp, accessed November 2005). While both epicenters relating to these plot within 10 km of the Lake Edgar fault, the associated location uncertainties are large. For example, the larger event, of ~M 5.5, occurred in 1880 and was located using an isoseismal map (Michael-Leiba and Gaull, 1989). The position errors are likely to be of the order of 50 km or more. Results: Scarp-Related Geomorphology Although McCue et al. (2003) provided evidence for multiple Quaternary surface-rupturing events on the Lake Edgar fault, they did not obtain time constraints for the events. Here, we outline the geomorphic setting of samples for which we obtained chronological data relevant to the surface-rupture history of the Lake Edgar fault. We focused on two objectives for the study (Fig. 2): (1) dating the depositional ages of the sediments on the raised terraces south of Edgar dam (see McCue et al., 2003); and (2) dating the depositional history of sediment within a small sag pond located to the north of Harlequin Hill. These studies are described in detail subsequent to a discussion of the general scarp geomorphology. General Scarp Geomorphology The Lake Edgar fault scarp is clearly recognizable in aerial photographs (e.g., Figs. 4–6). Although it can be traced for almost 30 km (McCue et al., 2003), the southern half is the most pronounced. South of the Edgar dam abutment, the scarp is prominent on a periglacial fluvial fan that originates in the hills to the east (Fig. 3A). McCue et al. (2003) presented a model for the evolution of the geomorphology of this area that involved at least one surge of fan deposition having eroded the scarp before the fan was cut by subsequent faulting, thus forming a series of raised terraces on the hanging-wall block (Figs. 3B and 3C). These authors also suggested that till from the oldest glaciation in the area (ca. 2 Ma Eliza glaciation), which presumably underlies a higher terrace, is also cut by the fault. This area is discussed in more detail in the next section, in light of our recent work. Aerial photography acquired in 1961 prior to the filling of the Lake Pedder impoundment (hereafter referred to as “Lake Pedder”) reveals that the former Lake Edgar and its smaller companion lake formed by ponding of westerly flowing streams against the upthrown western side of the fault scarp (Fig. 4). The texture of the land surface revealed in the aerial photography suggests that the lakes once formed part of a larger lake, which has either drained due to breaching of the scarp (a prominent breach
is labeled in Fig. 4) or drier climate, or it has been partially silted up (McCue et al., 2003). This area, totaling ~4 km of scarp length, is now submerged beneath Lake Pedder. North of Harlequin Hill (Fig. 2), the scarp traverses a lowlying plain supporting a number of small streams originating on the Mount Anne/Mount Eliza massif to the east. The largest of these is Condominium Creek. Several of the streams enter a 600-m-long by 150-m-wide swampy pond, which drains through a 1.5–2-m-high breached scarp where the pond meets Condominium Creek (Figs. 2, 3D, and 5). The western boundary of the pond is defined by the fault scarp, suggesting that the pond owes its existence to reorganization of drainage following scarp formation, in a similar fashion to Lake Edgar. This pond is discussed in more detail in the section “Boring/Coring within the Fault-Bounded Pond.” The hanging wall near the pond is modified by erosion. Adjacent to the pond, the scarp is sharp and well defined, rising 1.5–2 m above the pond surface level over a distance of less than 5 m. West of this initial rise, the land surface continues to rise in a staggered fashion before reaching a saddle ~5 m above the pond some 150 m west of its western margin. The land surface then drops off gently to the north, down to the shores of Lake Pedder. The lowest point in the saddle occurs where Condominium Creek has breached the scarp. This geomorphology suggests that a significant topographic barrier existed prior to the formation of the 2-m-high scarp that now bounds the pond. This barrier was largely removed by fluvial erosion focused along Condominium Creek prior to the formation of the most recent scarp. The scarp becomes more prominent north of Condominium Creek, rising ~4.5–5 m over a horizontal distance of 10 m. The faulted scarp material is partly exposed along a track (see Fig. 5 for location), where it comprises an upper 200-mm-thick layer of peaty silt overlying sandy boulder gravel (Fig. 3E). Gravel clasts are as large as 150 mm in diameter and are highly weathered, some crumbling under hand pressure. The gravel lithology is mainly quartzite, with minor amounts of phyllite. Further north, the scarp is lost within a cover of dense vegetation that bounds the Huon River. Aerial photography shows that the river course is remarkably straight for some 10 km north of this point, suggesting a continuation of the fault, if not the scarp. Fluvial Terraces South of Edgar Dam Three E-W–trending topographic profiles (profiles labeled Fan, Three terrace, and South in Fig. 6) were obtained south of Edgar dam to complement the three leveling profiles acquired by McCue et al. (2003). These were tied together with a profile along the top of the scarp. The location of the profiles with respect to the scarp and Scotts Peak Road is shown on Figure 6. A scarp-parallel profile across the footwall was not attempted. The morphology of the footwall was extrapolated from the three points on the E-W profiles where they reach the footwall fan. The E-W profiles (Figs. 7B–7D) confirm the general westerly slope of the fan on both the hanging-wall and footwall blocks.
Figure 7. Leveling profiles over fluvial terraces on the upthrown western side of the Lake Edgar scarp. Locations of profiles are shown on Figure 6. (A) Scarp top traverse; (B) south traverse; (C) Three terrace traverse; (D) fan traverse.
117
118
Clark et al. The profiles also clearly show the two terraces on the hangingwall block that were first identified by McCue et al. (2003). A third higher terrace is also clearly visible in the “Scarp top” traverse and the “Three terrace” traverse. South of the “Scarp top” traverse, the land surface on the up-thrown block continues to rise gradually. A shallow soil pit excavated ~200 m south of the profile encountered bedrock beneath a thin peaty veneer, and quartzite bedrock outcrops at the crest of the rise, some 500 m south of the profile. This indicates that the fan gravels give way to bedrock within a short distance south of the profile, and the scarp thereafter becomes bedrock controlled. Shallow pits excavated on each terrace and on the footwall fan show an upper layer of peaty/sandy silt, 200–300 mm thick, overlying sandy fan gravels (Fig. 3F). The gravel clasts are 10 mm to over 150 mm in diameter, angular to subrounded, and predominantly quartzite. The gravel clasts on terraces 1 and 2 are fresh in appearance and significantly resistant to fracture when struck with a hammer, supporting the supposition of McCue et al. (2003) that they relate to glaciations younger than ca. 2 Ma. Gravel clasts on terrace 3 are significantly more weathered and can sometimes be crumbled in the hand, implying greater age, perhaps as old as the Eliza stage. The trench excavation, which is situated on terrace 1 (McCue et al., 2003; Fig. 8), reveals that the gravels at this location form a sheet ranging from 0.5 to 1.0 m in thickness. This sheet is underlain by organic-rich silty sand. Our soil pits did not penetrate through the gravels on terraces 2 and 3, but it plausible that they also form thin sheets mantling organic-rich silty sand and peat horizons. The fan deposits on the footwall block rise in elevation toward the south, roughly mimicking the hanging-wall terracing. The stratigraphy of the footwall block does not show the distinct terracing observed in the hanging-wall block. However, we presume that an interbedded gravel and organic silty sand stratigraphy exists, similar to that on the hanging-wall block. A colluvial fan has developed at one location, redepositing material from terrace 1, and the scarp face, onto the footwall fan (see Fig. 7D; cf. Figs. 6B and 6C). The terraced geomorphology described here is consistent with the evolutionary model of McCue et al. (2003), which involves erosional surges planing off the scarp before the fan was refaulted. However, our data suggest that three faulting events, not two, would have been required to produce the observed topography (Fig. 9). We found that the slot cut into the scarp following the penultimate faulting event is steep sided, whereas the profile of the fan prior to this was more U-shaped (Fig. 7). This might relate to the amount of time the fan had to establish a
Figure 8. Map of trench excavation in McCue et al.’s (2003) investigation. Sample locations for the McCue et al. (2003) study (thermoluminescence, pollen, and 14C) and optically stimulated luminescence sample sites from this study are shown.
Implications for hazard assessment in intracratonic areas
119
Figure 9. Schematic diagram depicting the history of events on the Lake Edgar fault, at the location shown in Figure 6A, as inferred from the terrace geomorphology (age ranges are discussed in the text). View direction is west. (A) Channel profile prior to the penultimate event 2 (PE2). (B) Channel profile displaced 2.4 m vertically across the fault by PE2. (C) Breaching and erosion of the PE2 scarp by a fan surge followed by gravel deposition. (D) Continuing fan activity resulting in additional gravel deposition. (E) Channel profile displaced 3.1 m vertically across the fault by the penultimate event 1 (PE1). (F) Breaching and erosion of the PE1 scarp by a fan surge followed by gravel deposition. (G) Channel profile displaced 2.4 m vertically across the fault by the MRE (most recent event). (H) Localized degradation of the composite scarp forming small alluvial/colluvial fans leading to the present-day configuration.
120
Clark et al.
U-shaped equilibrium profile in each instance. An implication of this is that the time between breach of the penultimate event scarp and the cessation of fluvial activity was relatively brief. Vertical displacement across the fault from each earthquake/ faulting event can be estimated from the difference in elevation between the terrace surfaces (Fig. 7). The maximum difference in elevation between the level of the footwall fan and the top of terrace 1 (youngest) is ~2.6 m. This could be an underestimate of the true displacement during the most recent faulting event, because sediments may have accumulated at the base of the scarp following drainage defeat. The difference in height between the top of terrace 1 and the base of the U-shaped cut relating to terrace 2 (middle) is ~3.1 m, and the difference in elevation between the base of the terrace 2 cut and the base of the terrace 3 (oldest) cut (measured at the location of the “Three terrace” traverse) is 2.4 m. The area investigated using scarp profiles during this study is located on the southern half of the scarp trace (cf. Figs. 2 and 6A). Because the along-fault displacement is generally characterized by displacement maxima near to the center of the trace (e.g., Hemphill and Weldon, 1999), it might be expected that the displacements measured here underestimate the maximum magnitude of displacement. The maximum slip is likely to have occurred in the region of the fault now submerged beneath Lake Pedder, perhaps 2.5 km south of Harlequin Hill. However, it is debatable whether the maximum relief would have been well preserved at this location prior to inundation, given that the Lake Edgar area was a focus for significant drainage concentration, and breaches of the scarp can clearly be seen in preflooding aerial photography (Fig. 4).
In the borehole closest to the scarp (hole 2), the basal dense clayey gravel is overlain by sandy clay and clayey/sandy gravel, which are interpreted as colluvium shed off the scarp. There is no evidence for cyclicity in the pond sediment that might indicate multiple cycles of faulting-induced sedimentation. This, together with the <3 m height of the fault scarp at this location, suggests that the present pond formed subsequent to the most recent faulting event. OSL Geochronology The study by McCue et al. (2003) included radiocarbon dating of samples from the trench locality on terrace 1 (e.g., Fig. 2). They dated samples from the peaty silt overlying the terrace 1 gravel sheet and also from silt underlying the gravel. The upper silt sample returned a modern age, whereas the lower silt samples gave ages older than 39,600 yr B.P., which exceeded the range of the radiocarbon technique. In light of this, and the geomorphic results presented here, we undertook an investigation using the
Boring/Coring within the Fault-Bounded Pond A small pond situated east of Huon Inlet and north of Harlequin Hill, and bounded on its western side by the fault scarp, was also investigated in detail (Fig. 2). The lake relates to westerly flowing drainages that were impeded by faulting, most notably Condominium Creek. The northwest corner of the pond has an outlet that breaches the scarp. Aerial photography suggests that the current lake area is about a third of the maximum attained previously (Fig. 5). Three core/boreholes were drilled within the lake using a combination of D-section and auger equipment to reveal the pond stratigraphy. The boreholes (Fig. 10) reveal a general stratigraphy consisting of an upper layer of dark-brown organic (peaty) mud overlying medium-grained white sand to clayey sand that becomes more clay-rich with depth. In turn, these sediments overlie dense clayey gravel, which is similar in character to the deposits on the uplifted fluvial terraces mentioned in the previous section. The clay matrix is interpreted as reflecting derivation of the sediments from the Mount Anne region, which is capped by dolerite, rather than from the quartzite ranges further south. The basal clayey gravel is therefore interpreted as prelake periglacial fluvial fan sediment. The auger would not penetrate the top of this horizon in hole numbers 2 and 3.
Figure 10 (Continued on following page). Detail of sag pond north of Harlequin Hill (location is shown on Fig. 5). (A) Borehole locations relative to the pond and scarp. Dashed lines represent streams. Flow directions on streams are shown by open arrow heads. (B) Stratigraphic sections revealed in boreholes. BWL—below water level; BOH—bottom of hole. For clarity, a schematic thickness of gravel has been added below the BOH in boreholes that terminated in gravel. Tick marks are shown on the downthrown side of the scarp.
Implications for hazard assessment in intracratonic areas
121
OSL technique to constrain faulting events by dating the gravelly sediments mantling the various terrace surfaces (see Appendix 1 for detailed sample context). The single-aliquot regenerative dose (SAR) OSL protocol used in this study (see Appendix 2) measures the luminescence signals from multiple grain subsamples, each composed of ~100–200 quartz grains that are 90–125 µm in size. The limitation of multigrain dating is that heterogeneity that might exist in the sample, for example, due to incomplete resetting or bioturbation, is not completely resolvable. The degree to which quartz-rich sediment is reset during transport under various conditions and in different environments is beginning to be quantified using single-grain OSL (e.g., Olley et al., 2004). However, irrespective of problems with resetting, a multiple grain age determination on a sample will provide a maximum estimate for the depositional age of the sediment. In the context of the present study, age determinations on the sandy gravels deposited on the various terraces provide a maximum estimate for the time of faulting (i.e., uplift of the sediment above the level of deposition).
Figure 10 (Continued).
OSL Results and Age Constraints on Field Relationships OSL dating of sandy gravel from near the surface of terrace 3 (EF-06, Appendix 1) produced a burial age of 61.3 ± 7.1 ka (Table 1; Fig. 11). This age provides an upper bound for the oldest recognized seismic event (the second penultimate event, PE2). An age of 28.0 ± 2.6 ka from a sandy gravel from terrace 2 (EF-17) provides a lower time bound for this event, and also an upper bound for the first penultimate event (PE1). Sandy gravels that mantle the lowest terrace (terrace 1) yielded burial ages of ca. 28 ka in the lower parts (EF-08, 28.8 ± 5.2 ka; EF-09, 28.1 ± 1.9 ka) and ca. 25 ka higher in the section, near the interface with the overlying peat (EF-11, 25.3 ± 4.0 ka). Although these ages are stratigraphically consistent, they are also statistically indistinguishable. They provide a lower bound for PE1 and an upper bound for the most recent event (MRE). The MRE is constrained as having occurred subsequent to the 25.3 ± 4.0 ka deposition of sample EF-11 sandy gravels and prior to the 18 ± 0.8 ka deposition of gravelly alluvium/ colluvium derived from terrace 1 (sample EF-03, 18.7 ± 1.3 ka; sample EF-15, 17.2 ± 1.0 ka). The OSL results indicate that the sandy sediment underlying the mantle of gravel on terrace 1 (Fig. 8) is significantly older than the gravel itself. This confirms the reverse nature of faulting, showing that older sediment (EF-10) has been thrust over younger sediment (EF-09). Sample EF-07, obtained from a sandy lens at the base of the gravel, has a deposition age of 102 ± 14 ka. Dark-brown fine silty sand from EF-10, sampled from a stratigraphic level ~250 mm below the base of the gravel, was saturated with respect to OSL signal, and gave a minimum estimate of burial time of 173 ka. Sample EF-10 was taken from immediately above sample EF-09 (28.1 ± 1.9 ka), across the main MRE fault strand.
1.4
1.1
0.9
0.5
0.4
1.3
0.8
EF-08
EF-09
EF-10
EF-11
EF-12
EF-14
EF-15
40.0 ± 5.0
40.0 ± 5.0
15.0 ± 2.5
15.0 ± 2.5
30.0 ± 5.0
15.0 ± 2.5
15.0 ± 2.5
20.0 ± 2.5
1.24 ± 0.01
1.23 ± 0.01
0.47 ± 0.01
0.50 ± 0.01
1.75 ± 0.02
0.75 ± 0.01
0.59 ± 0.01
0.64 ± 0.01
0.02 ± 0.01
7.13 ± 0.06
7.31 ± 0.06
1.56 ± 0.02
1.41 ± 0.02
7.33 ± 0.06
2.48 ± 0.02
1.62 ± 0.02
3.04 ± 0.03
1.00 ± 0.01
2.13 ± 0.02
1.63 ± 0.02
2.34 ± 0.02
0.54 ± 0.02
0.33 ± 0.03
2.15 ± 0.02
0.91 ± 0.01
0.35 ± 0.03
1.01 ± 0.01
0.37 ± 0.02
0.73 ± 0.02
U (ppm)
§
0.03 ± 0.01
0.03 ± 0.01
0.03 ± 0.01
0.03 ± 0.01
0.03 ± 0.01
0.03 ± 0.01
0.03 ± 0.01
0.03 ± 0.01
0.03 ± 0.01
0.03 ± 0.01
α radiation –1 (Gy k.y. ) #
0.87 ± 0.05
0.93 ± 0.05
0.38 ± 0.02
0.37 ± 0.02
1.27 ± 0.07
0.62 ± 0.03
0.44 ± 0.02
0.54 ± 0.02
0.08 ± 0.02
0.35 ± 0.02
β radiation –1 (Gy k.y. )
0.57 ± 0.02
0.63 ± 0.02
0.16 ± 0.01
0.07 ± 0.01
0.23 ± 0.02
0.25 ± 0.02
0.22 ± 0.01
0.16 ± 0.01
0.03 ± 0.01
0.19 ± 0.01
γ radiation** –1 (Gy k.y. )
0.20 ± 0.02
0.18 ± 0.02
0.21 ± 0.02
0.21 ± 0.02
0.20 ± 0.02
0.20 ± 0.02
0.18 ± 0.02
0.20 ± 0.02
0.21 ± 0.02
0.22 ± 0.02
Cosmic-ray †† radiation –1 (Gy k.y. )
1.66 ± 0.08
1.77 ± 0.08
0.79 ± 0.04
0.69 ± 0.04
1.73 ± 0.11
1.10 ± 0.05
0.87 ± 0.04
0.93 ± 0.05
0.36 ± 0.03
0.79 ± 0.04
Total dose rate –1 (Gy k.y. )
19.5 ± 0.9
37.7 ± 5.6
17.4 ± 2.5
>300
30.9 ± 1.6
25.0 ± 4.4
95 ± 13
21.8 ± 1.9
14.8 ± 0.8
Equivalent §§ dose (Gy)
11.0 ± 0.7
47.7 ± 7.5
25.3 ± 4.0
>173
28.1 ± 1.9
28.8 ± 5.2
102 ± 14
61.3 ± 7.1
18.7 ± 1.3
OSL age (ka)
28.0 ± 2.6
0.7
EF-07
12.5 ± 2.5
0.36 ± 0.01
Th (ppm)
EF-17 0.3 12.5 ± 2.5 0.02 ± 0.01 1.26 ± 0.02 0.68 ± 0.02 0.03 ± 0.01 0.12 ± 0.01 0.12 ± 0.01 0.22 ± 0.02 0.49 ± 0.03 13.8 ± 1.0 *Estimated time-averaged moisture contents, based on measured field water values (% dry weight). † Obtained by instrumental neutron activation analysis (INAA; Becquerel Laboratories, Menai). § Assumed internal alpha dose rate. # Derived from INAA radionuclide concentration measurements, corrected for attenuation by water and beta attenuation. **Derived from field gamma spectrometry for EF-03–EF-12, INAA radionuclide concentration measurements for EF-14–EF-17, corrected for attenuation by water. †† Calculated using the equation of Prescott and Hutton (1994), based on sediment density, time-averaged depth and site latitude, longitude and altitude. §§ Including a ±2% systematic uncertainty associated with calibration of the laboratory beta source.
0.4
EF-06
15.0 ± 2.5
K (%)
17.2 ± 1.0
0.3
EF-03
Water* (%)
†
28.6 ± 1.1
Depth (m)
Sample
Radionuclide concentrations
TABLE 1. OPTICALLY STIMULATED LUMINESCENCE (OSL) SAMPLE PROPERTIES AND AGE DATA (ERRORS ARE 1σ)
122
Implications for hazard assessment in intracratonic areas
123
Figure 11. Ages from the optically stimulated luminescence samples with 1σ error bounds. Constraints on the timing of the three events are provided by the samples as shown. EF-10, at older than 173 ka, occurs off the top of timescale. Black arrows show the preferred ages of MRE and PE1. MRE—most recent event, PE1—first penultimate event, PE2—oldest recognized event.
The 11.0 ± 0.7 ka age obtained from sample EF-14 provides a maximum estimate for the timing of the change from clasticto organic-dominated sedimentation in the fault-bounded lake. Deposition within the lake did not involve a discernible organic component at the time of colluvial sediment accumulation (i.e., immediately after the MRE); in fact, there is no evidence of significant organic material accumulation until after 11 ka. DISCUSSION Timing and Magnitude of Paleo-Earthquakes on the Lake Edgar Fault Most Recent Event The timing of the most recent faulting event (MRE) is bracketed by the ages obtained on the faulted terrace 1 gravels (EF11, ca. 25 ka) and by the ca. 17–19 ka ages derived from the unfaulted colluvium derived from the uplifted deposits of terrace
1 (the average age of samples EF-03 and EF-15 is 18 ± 0.8 ka). This range, ca. 18–25 ka, spans the time of the LGM in the Australasian region, when average temperatures were ~9–10 °C lower than today (Miller et al., 1997; Barrows et al., 2001). The dominance of clean clastic sediment poor in organic peat content suggests that this was also a time of great aridity in the Lake Edgar region, with little or no vegetative cover on the valley floors. This interpretation is supported by the extent of colluvial deposits that were shed from the MRE scarp, despite the low angle of the MRE fault trace (Fig. 8), which is unlikely to have formed a free face subsequent to rupture. The accumulation of colluvium following the MRE may have been rapid compared to a fully vegetated region. Hence, we propose that the MRE faulting occurred immediately prior to the deposition of the colluvium shed off terrace 1 at ca. 17–19 ka (Fig. 11). McCue et al. (2003) argued that the MRE occurred in the very recent past based upon the prominence of the scarp, despite a current, relatively high average annual rainfall. Although our
124
Clark et al.
age data are incompatible with this interpretation, some explanation of the apparent youth of the landform is warranted. At present, there is almost no erosion and transport of clastic sediment because the landscape has been completely stabilized by vegetation (button grass plains, swamps, etc.). The geomorphological evidence suggests that the landscape has been in a stable form for the entire Holocene, and perhaps longer (Kiernan, 2001). The transition from clastic deposition (e.g., landscape instability and erosion) to organic deposition (landscape stabilized) probably occurred between 10 ka and 13 ka (this study; Michael Fletcher, 2003, personal commun.). Hence, there may have been a window of as little as 3 k.y. (to as much as 10 k.y.) between scarp formation and stabilization of the landscape. In this interval, it is likely that there was some vegetation acting to retard erosion. This, in combination with the low angle of the fault trace where it approaches the surface (we speculate that there was never a free face), is proposed to account for the preservation in the landscape of the youthful appearance of the scarp. It is difficult to estimate the magnitude of the earthquake that formed the 2.6-m-high MRE scarp with certainty since the length of the surface rupture is not well constrained, nor is fault dip known at seismogenic depth. The air photograph lineaments associated with significant Quaternary expression of faulting extend for ~20 km. However, a significant linear topographic break, which coincides with the inferred position of the Lake Edgar fault, and which controls the path of the Huon River, extends for another 10 km to the north of the scarp (Fig. 2). The empirical relationships among scarp height, fault length, and earthquake magnitude developed by Wells and Coppersmith (1994) suggest that the height of the MRE scarp relates to an earthquake of MW 6.8 ± 0.5, assuming that the fault dips steeply 65°–70° to the west at depth, and that slip is predominantly downdip. Such a magnitude is consistent with a scarp length of 25–30 km. If the dip of the Quaternary fault at depth is more shallow than the nearsurface drilling results suggest (Roberts et al., 1975), then a larger event would be required to generate the observed relief (because the downdip width of the fault would be considerably more). As mentioned in a previous section, the location of our scarp profiles on the southern portion of the fault trace also raises the possibility that the vertical relief figures calculated for each event underestimate the maximum values obtained. This would also lead to an underestimate of the magnitude.
transported at a later date onto the hanging wall without resetting their OSL systematics, but this hypothesis seems unlikely. However, the 47.7 ± 7.5 ka age obtained on sandy gravel sampled on the footwall fan adjacent to the midsection of the terrace 1 deposits (sample EF-12) indicates that sediments with a range of depositional ages are preserved on the footwall fan, and may have provided mixed age sources for the terrace gravels. Nonetheless, the ca. 28 ka age obtained on sample EF-17 from the top of terrace 2 provides a maximum time bound for PE1 (Fig. 11). There appears to be ~3.1 m of change in base level of the fluvial fan as a result of the first penultimate event. This single-event vertical displacement is slightly larger than that resulting from the MRE (2.6-m-high scarp), and it may have been generated by an event of MW 6.9 ± 0.6 (well within error of the estimate for the MRE). The relationships of Wells and Coppersmith (1994) suggest that this event may have ruptured the entire 45 km length of the Lake Edgar fault.
Penultimate Event 1 The near coincidence of ages from surfaces of terrace 1 (EF-08 and -09) and terrace 2 (EF-17) is consistent with the penultimate faulting event having occurred during active fluvial/ alluvial deposition ca. 28 ka (the average of the ages of EF-08, EF-09, and EF-17 is 28.3 ± 2.0 ka). This interpretation is consistent with unpublished data suggesting widespread landscape instability and eolian deposition in the lower Huon River at this time (P. McIntosh and K. Kiernan, 2004, personal commun.). An alternative is that fluvial deposition had ceased prior to PE1 faulting, and that footwall fan gravels and sands of ca. 28 ka age were
Older Quaternary Fault Activity There is no evidence in the trench, nor conclusive evidence in the geomorphology, for surface-rupturing earthquakes prior to PE2. Although the continued rise of the upthrown block to the south of the “Scarp top” traverse may relate to pre-PE2 seismic activity, we found no markers that might conclusively demonstrate this. Two explanations for the apparent lack of pre- PE2 events are plausible: (1) There were no Quaternary events prior to the PE2; or (2) any geomorphic expression relating to previous events has been planed off by ice, destroyed by other erosive processes, or buried by sediment accumulation.
Penultimate Event 2 The time of the second penultimate event, PE2, is broadly bracketed by the ca. 61 ka (61.3 ± 7.1 ka) age obtained on terrace 3 gravel and the ca. 28 ka (the average of samples EF-08, EF-09, and EF-17 is 28.3 ± 2.0 ka) age obtained on terrace 2 gravel. Further constraint is provided by the 48 ka (47.7 ± 7.5 ka) age obtained on a sample of the footwall fan gravel. Although terrace 3 has not been thoroughly sampled, the location of sample EF-06 near the center of the depositional channel that breached the scarp makes it highly likely that ca. 48 ka gravel would have been preserved had they existed. It is reasonable to assume that the ca. 48 ka gravel identified in the footwall reflects a phase of fluvial activity similar to those that resulted in the deposition of the ca. 61 ka and ca. 28 ka gravels, as this time coincides with a period (of probable landscape instability) between glacial advances identified in New Zealand (Williams, 1996; Barrows et al., 2002; Fink and Williams, 2003) and the Kosciuszko region of Australia (Barrows et al., 2001). The absence of gravel of this age on terrace 3 thus implies that PE2 predated ca. 48 ka. We therefore propose that the PE2 occurred between 48 ka and 61 ka (47.7 ± 7.5 ka to 61.3 ± 7.1 ka) (Fig. 11). The relief relating to PE2 is similar to that generated in the younger two events. A causative earthquake of a similar magnitude is implied.
Implications for hazard assessment in intracratonic areas Implications of the Data for Seismic Hazard Assessment The Lake Edgar fault is undoubtedly susceptible to reactivation under conditions imposed by the modern Australian intraplate stress field. The history of recurrence of large surfacerupturing earthquakes established herein, with the last (MRE) event having occurred in the past 20 k.y., requires that the Lake Edgar fault be classed as an active (or capable) fault (e.g., IAEA, 1991; USNRC, 1996; Machette, 2000). The data presented herein also provide fundamental insight into the behavior of intracratonic faults and have associated implications for hazard assessment, as discussed next. Fault Behavior The results of the geochronology and geomorphic analysis indicate surface-rupturing earthquakes on the Lake Edgar fault at around 17–19 ka (18.0 ± 0.8 ka), 26–30 ka (28.3 ± 2.0 ka), and 40–68 ka (47.7 ± 7.5–61.3 ± 7.1 ka). The interseismic intervals are thus 7–13 k.y. and 10–42 k.y., with ~18 k.y. having elapsed since the last surface rupture. While this close temporal clustering of large earthquakes is remarkable for an Australian stable continental region fault, it is possible that the last three events on the Lake Edgar fault are atypical of the long-term behavior of the fault. Investigations of stable continental region faults elsewhere in Australia and in the eastern United States (Crone and Luza, 1990; Crone et al., 1992, 1997, 2003) found earthquake recurrence behavior to be characterized by relatively short-lived periods of activity separated by long periods of quiescence. The inter-event times between successive earthquakes in the shortlived active periods can range from hours (e.g., the 1988 Tennant Creek earthquake sequence; Crone et al., 1992) to thousands of years (e.g., the Meers and Cheraw faults; Crone and Luza, 1990) to tens of thousands of years (e.g., the Hyden and Roopena faults; Crone et al., 2003). The periods of quiescence can be as long as millions of years or more (e.g., the Hyden fault; Crone et al., 2003). It is noteworthy that the three events identified on the Lake Edgar fault have resulted in the generation of similar amounts of vertical relief. The concept of a “characteristic event” (Schwartz and Coppersmith, 1984) might therefore be applicable to this fault (i.e., surface-breaking earthquakes are of similar magnitude). There is no evidence in the geomorphology, nor in the trench profile, indicative of smaller events between the three major surface ruptures. In addition, our OSL data effectively rule out the possibility of fault creep (or a series of small events) having formed the observed relief, as opposed to three large discrete events. If this were the case, the deposits on terrace 1 and terrace 2 would be of significantly different ages (at least 13 k.y. difference, assuming the upper bound slip rate of 0.24 mm/yr; see next section) rather than of similar age (i.e., ca. 28 ka). Slip Rate Providing an estimate of the slip rate in which we might have confidence is difficult for intraplate faults, where the recurrence
125
for surface-rupturing events is measured in thousands to many tens of thousands of years. Not only do interseismic intervals far exceed the record of historical seismicity, there is also the possibility that displaced geomorphic datums might be significantly altered by erosion. The problem is compounded if earthquakes exhibit temporal clustering, as may be the case for the Lake Edgar fault, because slip rates should then be averaged over a time period containing a large number of seismic cycles. Information from two complete seismic cycles on the Lake Edgar fault is preserved within the area of investigation. Given that the scarp remains a sharp feature in the landscape, it is not expected that erosion of the scarp (other than by localized fluvial activity) has significantly reduced the scarp height, subsequent to the last three events at least. Figure 12 presents a graphical summary of the timing of displacement events. The slip rate for the seismic cycle PE1–MRE is 0.25 ± 0.05 mm/yr, whereas that for the cycle PE2–PE1 is between 0.09 ± 0.02 and 0.16 ± 0.07 mm/yr. Estimates for the average slip rate calculated over the two complete seismic cycles range between 0.13 ± 0.02 mm/yr and 0.19 ± 0.05 mm/yr (i.e., a total range including uncertainties of 0.11–0.24 mm/yr). The average slip rate on the Lake Edgar fault appears large compared to those inferred for other intracratonic faults in Australia, even where significant fault-controlled relief is evident. For example, faults bounding the Mount Lofty and Flinders Ranges in South Australia, which are responsible for the generation of significant Quaternary relief (several tens of meters), have average slip rates estimated at 0.02–0.03 mm/yr for the past 3–5 m.y. (Sandiford, 2003a). The same author reported similarly low slip rates of 0.01–0.02 mm/yr from faults in the central Murray Basin and rates of as much as 0.1 mm/yr from the Otway Ranges in Victoria, all averaged over a similar time period (Sandiford, 2003a, 2003b). These lower slip rates are perhaps more consistent with the low strain accumulation rates that might be expected from a stable continental region, which might in general be considered incapable of sustaining long sequences of closely temporally clustered (e.g., recurrence intervals of thousands of years) large earthquakes. Intuitively, each successive earthquake in a temporally clustered sequence on a stable continental region fault might be expected to dramatically decrease the likelihood of a near-future recurrence. It is not possible to tell if the slip rate calculated from the last two complete seismic cycles on the Lake Edgar fault is representative of the long-term (i.e., hundreds of thousands to millions of years) slip rate on the fault because there is no evidence for events prior to PE2. While this discussion suggests that the last two seismic cycles may not be representative of the longer term, the possibility that the Lake Edgar region is anomalous in terms of strain accumulation rate must also be considered. As mentioned in the section on “Regional Geomorphology and Evidence for Quaternary Deformation,” it is plausible that strain is concentrated in the Lake Edgar and Gell River areas as a consequence of the rheology contrasts between the Adamsfield ultramafic complex and adjacent rocks abutting the contact between
126
Clark et al.
Figure 12. Slip rate estimates for the late Quaternary activity on the Lake Edgar fault. Uncertainties in ages are 1 standard deviation. Where several age determinations constrain the timing of an event (e.g., three samples date the first penultimate event [PE1]), the average of those measurements (with appropriate error propagation) was used for slip rate calculation. MRE—most recent event, PE2—oldest recognized event.
the Tyennan and Adamsfield-Jubilee tectonic elements (McCue et al., 2003; Fig. 1). Implications for Climate and Glaciation in SW Tasmania Cosmogenic-nuclide dating of the Timk and Judd moraines on Schnells Ridge (6 km east of the Lake Edgar fault) yielded ages of ca. 19 ka for the maximum extent of the most recent glaciation (LGM) that has affected the area (Kiernan et al., 2004). Dating elsewhere in Tasmania suggests that the late LGM occurred at around 20 ka, with retreat by 18 ka (Colhoun and Fitzsimons, 1990; Barrows et al., 2002; Colhoun, 2003). No firm age control exists on glacial deposits in the Lake Edgar region that predate the LGM. However, Kiernan et al. (2004) reported cosmogenic nuclide exposure ages of ca. 40 ka and ca. 70 ka for individual boulders from the composite Timk moraine, which correspond to periods of glaciation identified from moraines in southeastern Australian (39–46 ka and >59 ka; Barrows et al., 2001) and from southern New Zealand (41–45 ka and 65–75 ka; Williams, 1996; Fink and Williams, 2003). Glacial advances have also been documented in the South Island of New Zealand between 25 and 28 ka, at 19–21 ka, and at 14–17 ka (Fink and Williams, 2003). Given the correspondence between Timk and Judd moraine ages and the timing of glacial events in New Zealand, and the likeli-
hood that the New Zealand cold climate episodes reflect regional rather than local climatic influences, it might also be expected that cold climate episodes occurred in Tasmania between 28 and 25 ka and at 17–14 ka. The ages obtained from the fan gravels along the Lake Edgar fault suggest episodes of landscape instability at ca. 61, ca. 48, ca. 28–25, and ca. 17–11 ka. Considering the dating errors, these ages correspond well to cold climate episodes recognized in southeastern Australia and New Zealand. Therefore, we propose that the landscape instability, which leads to the deposition of sediments in fans in the Lake Edgar region, resulted from a reduction of vegetative cover as a consequence of cold climate. A probable climatic trigger for fan formation is best demonstrated by age of the youngest fans (ca. 17–19 ka), which formed immediately after the initiation of retreat from the LGM. The deposition of colluvial deposits in fans over the low-angle surface rupture following the most recent faulting event (ca. 17–19 ka) implies that vegetation did not effectively stabilize the landscape at this time. We believe that the age of 11.0 ± 0.7 ka (EF-14) obtained immediately below the sand-organics interface in the fault-bounded pond to the north of Harlequin Hill provides important time constraints on the recovery of vegetation after the end of the last glaciation. Hence, if precipitation had returned to values similar to today prior to the reestablishment of vegetation, then some
Implications for hazard assessment in intracratonic areas potential may have existed for slope instability, fan formation, and the shedding of sandy sediment into lakes. In a lacustrine setting, it might be expected that once vegetation restabilized the landscape, this influx of sandy sediment would be much reduced, and potentially be overwhelmed by organic sediment, producing the sediment profile revealed in the boreholes. The temporal relationship between the older periods of fan activity (pre-LGM) and cold climate episodes is not clear at the resolution of our age data. The reestablishment of relatively high postglacial rainfall levels prior to vegetation recovery provides a plausible mechanism for fan reactivation, as described previously. However, the older fans were far more extensive and energetic than the youngest fans and resulted in major localized erosion of the Lake Edgar fault scarp. We speculate that this reflects a higher annual rainfall budget than today, or more meltwater contribution to the Huon River catchment than today. CONCLUSIONS The main findings of this study are as follows: (1) Three recognizable surface-rupturing earthquakes have occurred on the Lake Edgar fault in the past 48–61 k.y., with the most recent event (MRE) occurring some 17–19 k.y. ago; (2) each event on the ~45-km-long fault is associated with ~2.4–3.1 m of vertical displacement at our study site, implying a characteristic moment magnitude event in the order of Mw 6.8–7.0, which is comparable to the largest earthquake events recorded in Australia; (3) the average slip rate calculated for two complete earthquake cycles ranges from 0.11 to 0.24 mm/yr, which is large compared to other stable continental region faults; and (4) the timing of latest Pleistocene fluvial fan deposition suggests previously unrecognized cold climate episodes in SW Tasmania at ca. 61 ka, ca. 48 ka, and ca. 28–25 ka. ACKNOWLEDGMENTS Mike Machette and Buddy Schweig are thanked for their constructive reviews of the manuscript. We thank Hydro Tasmania
for allowing us access to their lands and for comments relating to the original manuscript. We appreciate permission from the Tasmanian Department of Primary Industries, Water and Environment to conduct research in the Southwest National Park, and for field assistance. Matt Hayne of Geoscience Australia also provided valuable field assistance, and provided useful comment on the original manuscript. This work is published with the permission of the Chief Executive Officer of Geoscience Australia. APPENDIX 1. CONTEXT OF THE DATED SAMPLES
Sand and sandy gravel from five localities within the trench (Fig. 8), two locations within the sag pond (Figs. 10A and 10B), and four locations on various fluvial terraces (Figs. 6C and 7) were collected for single-aliquot regenerative-dose (SAR) OSL dating of quartz grains. The results of the analyses are presented in Table 1. Analytic procedures are presented in Appendix 2. The following abbreviations are used: HB—hanging-wall block; FB—footwall block. Global positioning system coordinates are provided in the WGS84 datum using a UTM projection (zone 55S) (Table A1). Upper Part of Gravel Mantle at Trench Locality, HB (Sample EF-11) 55G, 446654 mE, 5234465 mN
Sample EF-11 is from the south face of the trench excavation (Fig. 8). An account of the detailed trench stratigraphy is published in McCue et al. (2003) and is not reproduced here. However, the general stratigraphy consists of a 200–300-mmthick layer of peaty silt with fine to medium sand, which mantles the ground surface and overlies a 500–1100-mm-thick layer of fine to coarse gravel (clasts typically 10–40 mm, but up to 100 mm). The sampled gravel is poorly sorted and contains angular and subrounded clasts predominantly of quartzite and vein quartz. Clasts typically appear reasonably fresh and will ring if struck with a hammer. A soil pit excavated into terrace 1 next to
TABLE A1. COORDINATES OF LOCATIONS IMPORTANT TO THIS STUDY Easting Northing (mE) (mN) Lake Edga r 446900 5236400 Companion lake to Lake Ed gar 447100 5237200 Sag pond north of Harlequin Hill 468600 5242300 Fan traverse (west end) 446665 5234179 Three terrace traverse (west end) 446629 5234025 South traverse (west end) 4 4 6 6 67 52 33 8 00 Scarp top traverse (center) 446648 5234046 Trench 446654 5234465 Borehole 1 4 46 9 5 7 52 4 2 5 5 3 Borehole 2 4 46 8 3 6 52 4 2 6 2 9 Borehole 3 4 46 8 5 8 52 4 2 6 2 1 Gell River scarp 437000 5302000 Note: Coordinates are based on WGS 84 datum and UTM projection, zone 55S. Feature
127
Elevation (approx. m ) <302 <302 314 317 332 3 21 316 310 3 14 3 14 3 14 560
128
Clark et al.
the trench confirmed the competency of the gravel clasts, and that the upper surface of the gravel is the terrace 1 surface. The gravel layer overlies firm dark-brown silty fine sand that extends to the base of the trench. OSL sample EF-11 was taken near the top of the gravel sheet mantling the scarp on the hanging-wall block. The age of the sample provides a maximum time bound for the most recent faulting event, since the gravel mantle is cut by the most recent event (MRE) on the fault. Lower Part of Gravel Mantle at Trench Locality, HB (Sample EF-08) 55G, 446654 mE, 5234465 mN
Sample EF-08 is from the south face of the trench excavation (Fig. 8), within the lower half of the gravel that mantles the scarp on the hanging-wall block. The gravel unit thickens markedly at the sampling site compared to elsewhere in the trench. Stratigraphy within the underlying sediment is truncated upward at the gravel interface, suggesting that the irregular gravel thickness is in a channel. The sample was taken to provide an indication of the age structure of the gravel that is displaced by the most recent faulting event (MRE). Lower Part of Gravel Mantle at Trench Locality, FB (Sample EF-09) 55G, 446654 mE, 5234465 mN
Sample EF-09 is from the south face of the trench excavation (Fig. 8), within the lower half of the gravel that mantles the scarp, immediately beneath the fault plane on the footwall block. The footwall gravel appears to be identical to that on the hanging wall. The age of the EF-09 gravel provides a way to test this hypothesis. Subgravel Soil at Trench Locality, HB (Sample EF-07) 55G, 446654 mE, 5234465 mN
Sample EF-07 is from the south face of the trench excavation (Fig. 8) from within a lens of brown medium-grained silty sand. The lens is beneath the gravel mantling the scarp in the hangingwall block and is partly intercalated with the gravels. We suggest that the unit may have been ripped up from the underlying silty sand strata and redeposited. Hence, the age of the sampled material provides a maximum time bound for the deposition of the overlying gravel sheet. Subgravel Soil at Trench Locality, HB (Sample EF-10) 55G, 446654 mE, 5234465 mN
Sample EF-10 is from the south face of the trench excavation (Fig. 8), from within the dark-brown organic-rich sediment that underlies the gravel. The sample was taken from the hanging-wall block above sample EF-09. Samples EF-09 and EF-10,
which are on either side of the main MRE fault strand, therefore provide a test on the nature of faulting. Reverse motion should juxtapose older sediment above younger sediment. Colluvial Fan Derived from Terrace 1, FB (Sample EF-03) 55G, 446673 mE, 5234190 mN
Sample EF-03 is from a small hand-excavated pit dug into the upper surface of a localized colluvial fan shed off terrace 1 onto the footwall block fan (Figs. 6C and 7). The fan mantles the fault scarp and is not displaced by faulting, so this material must postdate the MRE. The stratigraphy at the sample site is composed of an upper 210-mm-thick layer of dark-brown peaty soil overlying a 110-mm-thick layer of gravelly coarse sand. This sandy layer underlies the surface of the colluvial fan. The sample was taken from this layer. The sand is underlain by a third unit, which is a brown gravelly clay that extends to the base of the pit (~450 mm). The gravel clasts are identical to those on terrace 1. Leveling data (Figs. 6C and 7) suggest that the gravelly clay is also part of the colluvial fan. Footwall Block Fluvial Fan, FB (Sample EF-12) 55G, 446739 mE, 5234182 mN
Sample EF-12 is from a small hand-excavated pit in the main fluvial fan, ~50 m east of the fault scarp (Figs. 6C and 7). Because the fan dips moderately to the west, it was judged that sediment from this location would not be contaminated by material derived from the upthrown block. The stratigraphy at the site of sampling is composed of an upper 250-mm-thick layer of peaty soil rich in roots that overlies sandy gravel, which extends beneath the soil pit at 500 mm depth. The gravel is poorly sorted and contains angular and subrounded clasts that range from 5 to 150 mm in size. Clasts are typically quartzite or vein quartz and appear to have undergone a similar amount of weathering to those on terrace 1 (see EF-11 description). The age of this sample will give an indication of the age structure of the footwall fluvial fan. Terrace 2, HB (Sample EF-17) 55G, 446649 mE, 5234046 mN
Sample EF-17 was taken from a small hand-excavated pit in the upper surface of terrace 2 along the line of the “Three terrace” traverse (Figs. 6C and 7C). The stratigraphy at the sampled location consists of (1) an upper 230-mm-thick peaty soil horizon, overlying (2) a 100 mm layer of well-sorted fine sandy loam, and (3) >100 mm of sandy/clayey gravel (sandy near the upper contact, becoming more clayey with depth). The clasts are composed of quartzite ~10 mm in diameter in the upper 50–70 mm of the layer, grading to cobbles as much as 200 mm in diameter near the base of the pit (e.g., Fig. 3F). The gravel clasts within the pit, and elsewhere on terrace 2, have similar competence as those on
Implications for hazard assessment in intracratonic areas terrace 1. The burial age derived from this sediment provides a maximum bound on the time of initial uplift of terrace 2 above the footwall fan level, during the penultimate event (PE1).
129
the potential to date the transition from sandy to organic sedimentation within the lake. APPENDIX 2. OSL PROCEDURES AND TECHNIQUES
Terrace 3, HB (Sample EF-06) 55G, 446618 mE, 5234021 mN
Sample EF-06 is from a small hand-excavated pit in the upper surface of a remnant (outlier) of terrace 3 along the line of the “Three terrace” traverse (Figs. 6C and 7C). The stratigraphy at the location of the sample is composed of 250 mm of peaty soil that overlies a gravelly medium sand that extends beneath the 400 mm depth of the pit. The gravel and cobbles appear much more weathered on terrace 3 than on terraces 1 and 2. In many instances, they will crumble when struck with a hammer, or even when handled. The clast composition and angularity appear to be similar to that on other terraces, however, implying a similar source region and transport distance. The sample was taken from the gravel layer. The age derived from this sample provides a maximum bound on the time of initial uplift of terrace 3 above the footwall fan level, during an older penultimate event (PE2). Fault-Bounded Pond, Colluvial Deposit Derived from MRE Scarp, FB (Sample EF-15) 55G, 446836 mE, 5242629 mN
Sample EF-15 is from the small fault-bounded lake north of Harlequin Hill (Fig. 5). Clayey to sandy gravel was retrieved from the borehole closest to the scarp (borehole 2, 5 m east of the scarp; Fig. 10), from within a unit that overlies the prelake fluvial fan deposits and underlies lake sand and mud. The gravel clasts are angular to subrounded quartzite and platy phyllite. This unit is interpreted to be colluvium shed off the fault scarp or terrace 1 after the MRE. The age of deposition of the sample therefore postdates the MRE. A soil pit was excavated into the toe of the scarp (terrace 1) adjacent to borehole 2 to see if the materials exposed matched those interpreted to be prelake fluvial fan sediments in the boreholes. The stratigraphy exposed in the pit consists of an upper 270-mm-thick layer of dark-brown organic-rich silty clay at the surface underlain by a light-brown gravelly clay. The clasts consist of dominant quartzite and phyllite to a diameter of 40 mm, very similar to that seen in the boreholes. Fault-Bounded Pond, Sand-Organics Transition, FB (Sample EF-14) 55G, 446858 mE, 5242621 mN
Sample EF-14 is from the small fault-bounded lake located near the center of the fault’s length. Clayey sand was retrieved from the borehole 25 m east of the scarp base (borehole 3; Fig. 10) from within a sandy unit immediately underlying organic muds. This unit is interpreted to be lacustrine in origin, and it provides
Background
Methods of dating sediment by stimulated luminescence rely on the net redistribution of charge that takes place when ionizing radiation interacts with an insulating crystal lattice, such as quartz (Aitken, 1977, 1998). When crystals in a sample are exposed to natural radionuclides in the environment, some electrons move out to the outer shells of atoms, where they may accumulate in meta-stable traps within the crystal lattice. As the net charge redistribution continues for the duration of the ionizing exposure (i.e., as long as sediment remains buried), the amount of trapped charge increases in proportion to both the duration and intensity of radiation exposure. Optical stimulation in the laboratory releases the stored electrons, emitting energy as light (luminescence). Optical energy can stimulate luminescence and, since the mid-1980s, optically stimulated luminescence (OSL) has been used to date crystals in sediment (Huntley et al., 1985). Optical dating provides a measure of time since sediment was last exposed to sunlight. Exposure at the surface may bleach (partial) or reset (total) the OSL signal of a sediment. Subsequent burial by further sedimentation causes the grains to accumulate trapped charges (measured as the radiation dose) from ionizing radioelements in the deposit, mainly from U and Th decay chains and 40 K, and cosmic rays. This accumulated energy is known as the paleodose or equivalent dose and is proportional to the age of the sample. Crystals are exposed to light in the laboratory to release their store of trapped energy. The age of the sample is the accumulated dose divided by the dose rate. Age (time) =
Dose (Radiation energy) . Dose rate (Radiation energy/time)
Technique
We collected samples for OSL dating by hammering 40-mmdiameter stainless-steel tubes into freshly cleaned sediment on vertical faces. Polymineralic samples were prepared under subdued, red light using standard procedures (Galbraith et al., 1999). The 90–125 μm quartz fraction was retained for dating. Equivalent doses were calculated using a single-aliquot regenerative dose (SAR) protocol (Murray and Roberts, 1998; Galbraith et al., 1999; Murray and Wintle, 2000). Small aliquots of ~100–200 grains were preheated at either 240 or 260 °C for 10 s and optically stimulated for 100 s at 125 °C by blue (470 nm) light from light-emitting diode (LED) arrays attached to an automated RISØ TL-DA-15 apparatus. Luminescence was detected using a Thorn-EMI 9235QB15 p.m. tube with a 7.5 mm Hoya U-340 filter. The grains were then given applied doses using a
130
Clark et al.
calibrated 90Sr/90Y beta-source and restimulated to record their regenerative OSL signals. Samples were given test doses of 2 Gy after each optical stimulation to monitor OSL sensitivity changes in the quartz crystals between the natural and regenerative cycles. OSL signals were integrated over the first 4.8 s period of illumination, with the mean signal from the final 20 s of illumination converted to the equivalent number of channels over 4.8 s and subtracted as background using Analyst version 2.12 software (Duller, 1999). The OSL data were corrected for any sensitivity changes, and dose-response curves were constructed using a minimum of six regenerative dose points. The equivalent dose was obtained from the intercept of the regenerated dose-response curve with the natural luminescence intensity. Frequency distributions of equivalent dose estimates from single aliquots were typically skewed by older equivalent doses. This suggests that most of the samples contained substantial populations of grains that had been inadequately exposed to sunlight prior to the last depositional event. Ages were generated using the weighted mean (central age; Galbraith et al., 1999) equivalent dose of the youngest quartile of aliquots to minimize the effect of partially bleached grains. Preheat tests produced flat plateaus over the range 160– 300 °C, indicating negligible thermal transfer in all of the samples. Recuperation tests showed no significant radiative recombination. These experiments, including recycling tests using duplicate regenerations of known dose, confirmed the reproducibility of the laboratory-induced luminescence signals. K, U, and Th concentrations were estimated by instrumental neutron activation analysis (INAA). These were converted to beta dose rates using the conversion factors of Adamiec and Aitken (1998) with a beta attenuation factor of 0.93 ± 0.03 (Mejdahl, 1979). Gamma dose rates were measured in the field for all except three samples using a portable spectrometer with NaI (Tl) crystal and converted to dry values by oven-drying sediment from the sample location. Gamma dose rates were estimated from the INAA radionuclide concentrations for those samples lacking in situ gamma spectrometry. Alpha dose rates were assumed to be 0.03 ± 0.01 mGy yr–1, a conservative estimate based on measurements of Australian quartz elsewhere (e.g., Bowler et al., 2003). Cosmic-ray dose rates were determined from established equations (Prescott and Hutton, 1994), allowing for sample depth, sediment density, and site altitude, latitude and longitude. Timeaveraged moisture content of the sediments was estimated from the oven-dried weight of OSL sediment samples and used to correct attenuation by water (Aitken, 1998). Present-day field water contents, which varied from 12.5% to 40% (Table 1), are considered representative of long-term averages. REFERENCES CITED Adamiec, G., and Aitken, M.J., 1998, Dose-rate conversion factors—Update: Ancient TL, v. 16, p. 37–50. Aitken, M.J., 1977, Thermoluminescence and the archaeologist: Antiquity, v. 51, p. 11–19. Aitken, M.J., 1998, An Introduction to Optical Dating: Oxford, Oxford University Press, 280 p.
Balmer, J., and Corbett, E., 2001, The vegetation of the Lake Pedder area prior to flooding, in Sharples, C., ed., Lake Pedder: Values and Restoration: Centre for Environmental Studies, University of Tasmania, Occasional Paper 27, p. 67–86. Barrows, T.T., Stone, J.O., Fifield, L.K., and Cresswell, R.G., 2001, Late Pleistocene glaciation of the Kosciusko Massif, Snowy Mountains, Australia: Quaternary Research, v. 55, p. 179–189, doi:10.1006/qres.2001.2216. Barrows, T.T., Stone, J.O., Fifield, L.K., and Cresswell, R.G., 2002, The timing of the Last Glacial Maximum in Australia: Quaternary Science Reviews, v. 21, p. 159–173, doi:10.1016/S0277-3791(01)00109-3. Bowler, J.M., Johnston, H., Olley, J.M., Prescott, J.R., Roberts, R.G., Shawcross, W., and Spooner, N.A., 2003, New ages for human occupation and climatic change at Lake Mungo, Australia: Nature, v. 421, p. 837–840, doi:10.1038/nature01383. Brown, A.V., Calver, C.R., Corbett, K.D., Forsyth, S.M., Goscombe, B.A., Green, G.R., McClenaghan, M.P., Pemberton, J., and Seymour, D.B., compilers, 1995, Geology of Southwest Tasmania: Geological Atlas 1:250,000 Digital Series: Hobart, Tasmanian Geological Survey, scale 1:250,000. Burbank, D.W., and Anderson, R.S., 2000, Tectonic Geomorphology: Oxford, UK, Blackwell Science Inc., 288 p. Calver, C.R., Turner, N.J., McClenaghan, M.P., and McClenaghan, J., 1990, Pedder; Geological Survey Explanatory Report, Sheet 80: Hobart, Tasmanian Department of Resources and Energy, 105 p. Carey, S.W., and Newstead, G., 1960, Tasmania University Seismic Net: Geology Department, University of Tasmania, Publication 84, p. 9–14. Clark, D.J., and Bodorkos, S., 2004, Fracture systems in granite pavements of the eastern Pilbara craton, Western Australia: Indicators of neotectonic activity: Australian Journal of Earth Sciences, v. 51, p. 831–846, doi:10 .1111/j.1400-0952.2004.01088.x. Clark, D.J., and McCue, K., 2003, Australian palaeoseismology: Towards a better basis for seismic hazard estimation: Annals of Geophysics, v. 46, p. 1087–1105. Colhoun, E., 2003, Antipodal terrestrial deglacial events from Ireland and Tasmania (paper 83-1), in 16th International Quaternary Association Congress Program with Abstracts: Reno, Nevada, p. 221. Available at http:// gsa.confex.com/gsa/inqu/finalprogram/abstract_54762.htm (accessed 31 May 2011). Colhoun, E.A., and Fitzsimons, S.J., 1990, Late Cainozoic glaciation in western Tasmania, Australia: Quaternary Science Reviews, v. 9, p. 199–216, doi:10.1016/0277-3791(90)90018-6. Crone, A.J., and Luza, K.V., 1990, Style and timing of Holocene surface faulting on the Meers fault, southwestern Oklahoma: Geological Society of America Bulletin, v. 102, p. 1–17, doi:10.1130/0016-7606(1990)102<0001: SATOHS>2.3.CO;2. Crone, A.J., Machette, M.N., and Bowman, J.R., 1992, Geologic Investigations of the 1988 Tennant Creek, Australia, Earthquakes—Implications for Paleoseismicity in Stable Continental Regions: U.S. Geological Survey Bulletin 2032-A, 51 p. Crone, A.J., Machette, M.N., and Bowman, J.R., 1997, Episodic nature of earthquake activity in stable continental regions revealed by palaeoseismicity studies of Australian and North American Quaternary faults: Australian Journal of Earth Sciences, v. 44, p. 203–214, doi:10 .1080/08120099708728304. Crone, A.J., de Martini, P.M., Machette, M.N., Okumura, K., and Prescott, J.R., 2003, Paleoseismicity of aseismic Quaternary faults in Australia: Implications for fault behaviour in stable continental regions: Bulletin of the Seismological Society of America, v. 93, p. 1913–1934, doi:10.1785/0120000094. Duller, G.A.T., 1999, Analyst Version 2.12: Aberystwyth, UK, Luminescence Laboratory, University of Wales. Fink, D., and Williams, P., 2003, Timing of the Last Glacial Maximum in Fiordland, New Zealand—But was it the last advance? (paper 5-1), in 16th International Quaternary Association Congress Program with Abstracts: Reno, Nevada, p. 74. Available at http://gsa.confex.com/gsa/ inqu/finalprogram/abstract_54455.htm (accessed 31 May 2011). Galbraith, R.F., Roberts, R.G., Laslett, G.M., Yoshida, H., and Olley, J.M., 1999, Optical dating of single and multiple grains of quartz from Jinmium rock shelter, northern Australia: Part I. Experimental design and statistical models: Archaeometry, v. 41, p. 339–364, doi:10.1111/j.1475-4754.1999 .tb00987.x.
Implications for hazard assessment in intracratonic areas Hemphill, H.M.A., and Weldon, R.J.I., 1999, Estimating prehistoric earthquake magnitude from point measurements of surface rupture: Bulletin of the Seismological Society of America, v. 89, p. 1264–1279. Huntley, D.J., Godfrey-Smith, D.I., and Thewalt, M.L.W., 1985, Optical dating of sediments: Nature, v. 313, p. 105–107, doi:10.1038/313105a0. International Atomic Energy Agency (IAEA), 1991, Earthquakes and Associated Topics in Relation to Nuclear Power Plant Siting: International Atomic Energy Agency Safety Series 50-SG-S1 Revision 1, Code of Practice, 70 p. Johnson, A.C., Coppersmith, K.J., Kanter, L.R., and Cornell, C.A., 1994, The Earthquakes of Stable Continental Regions: Electric Power Research Institute Report TR102261V1, 309 p. Kiernan, K., 1985, Late Cainozoic Glaciation and Mountain Geomorphology in the Central Highlands of Tasmania [Ph.D. thesis]: Hobart, Department of Geography, University of Tasmania, 2 vols., 557 p. Kiernan, K., 1990, The alpine geomorphology of the Mt. Anne Massif, south-western Tasmania: The Australian Geographer, v. 21, p. 113–125, doi:10.1080/00049189008703008. Kiernan, K., 2001, The geomorphology and geoconservation significance of Lake Pedder, in Sharples, C., ed., Lake Pedder: Values and Restoration: Centre for Environmental Studies, University of Tasmania, Occasional Paper 27, p. 13–50. Kiernan, K., Fifield, L., and Chappell, J., 2004, Cosmogenic nuclide ages for Last Glacial Maximum moraine at Schnells Ridge, southwest Tasmania: Quaternary Research, v. 61, p. 335–338, doi:10.1016/j.yqres.2004.02.004. Machette, M.N., 1998, Contrasts between short- and long-term records of seismicity in the Rio Grande Rift—Important implications for seismic hazard assessments in areas of slow extension, in Lund, W.R., ed., Proceedings of the Basin and Range Province Seismic-Hazards Summit: Utah Geological Survey Miscellaneous Publication 98, p. 84–95. Machette, M.N., 2000, Active, capable, and potentially active faults—A paleoseismic perspective: Journal of Geodynamics, v. 29, p. 387–392, doi:10.1016/S0264-3707(99)00060-5. McCue, K.F., 1990, Australia’s large earthquakes and Recent fault scarps: Journal of Structural Geology, v. 12, p. 761–766, doi:10.1016/0191-8141 (90)90087-F. McCue, K.F., Van Dissen, R., Gibson, G., Jensen, V., and Boreham, B., 2003, The Lake Edgar fault—An active fault in southwest Tasmania, Australia, with repeated displacement in the Quaternary: Annals of Geophysics, v. 46, p. 1107–1117. Mejdahl, V., 1979, Thermoluminescence dating: Beta-dose attenuation in quartz grains: Archaeometry, v. 21, p. 61–72, doi:10.1111/j.1475-4754.1979 .tb00241.x. Michael-Leiba, M.O., and Gaull, B.A., 1989, Probabilistic earthquake risk maps of Tasmania: BMR Journal of Australian Geology and Geophysics, v. 11, p. 81–87. Miller, G.H., Magee, J.W., and Jull, A.J.T., 1997, Low-altitude cooling in the Southern Hemisphere from amino-acid racemization in emu eggshells: Nature, v. 385, p. 241–244, doi:10.1038/385241a0. Murray, A.S., and Roberts, R.G., 1998, Measurement of the equivalent dose in quartz using a regenerative-dose single-aliquot protocol: Radiation Measurements, v. 29, p. 503–515, doi:10.1016/S1350-4487(98)00044-4.
131
Murray, A.S., and Wintle, A.G., 2000, Luminescence dating of quartz using an improved single-aliquot regenerative-dose protocol: Radiation Measurements, v. 32, p. 57–73, doi:10.1016/S1350-4487(99)00253-X. Olley, J.M., Pietsch, T., and Roberts, R.G., 2004, Optical dating of Holocene sediments from a variety of geomorphic settings using single grains of quartz: Geomorphology, v. 60, p. 337–358. Prescott, J.R., and Hutton, J.T., 1994, Cosmic ray contributions to dose rates for luminescence and ESR dating: Large depths and long-term time variations: Radiation Measurements, v. 23, p. 497–500, doi:10.1016/1350 -4487(94)90086-8. Roberts, G.T., Cole, B.A., and Barnett, R.H.W., 1975, Engineering geology of Scotts Peak dam and adjacent reservoir: Watertightness: Australian Geomechanics Journal, v. G5, p. 39–45. Sandiford, M., 2003a, Neotectonics of southeastern Australia: Linking the Quaternary faulting record with seismicity and in situ stress, in Hillis, R.R., and Muller, D., eds., Evolution and Dynamics of the Australian Plate: Geological Society of Australia Special Publication 22, p. 101–113. Sandiford, M., 2003b, Geomorphic constraints on the late Neogene tectonics of the Otway Ranges: Australian Journal of Earth Sciences, v. 50, p. 69–80, doi:10.1046/j.1440-0952.2003.00973.x. Schwartz, D.P., and Coppersmith, K.J., 1984, Fault behaviour and characteristic earthquakes: Examples from the Wasatch and San Andreas fault zones: Journal of Geophysical Research, v. 89, no. B7, p. 5681–5698, doi: 10.1029/JB089iB07p05681. Seymour, D.B., and Calver, C.R., 1995, Stratotectonic Elements Map: Tasmanian Geological Survey, Tasmania Development and Resources, and Australian Geological Survey Organisation, scale 1:500,000, 1 sheet. Shirley, J.E., 1980, Tasmanian seismicity—Natural and reservoir-induced: Bulletin of the Seismological Society of America, v. 70, p. 2203–2220. Turner, N.J., Calver, C.R., McClenaghan, M.P., McLenaghan, J., Brown, A.V., Lennox, P.G., Boutler, C.A., and Godfrey, N.H.H., 1985, Pedder: Geological Atlas, 1:50,000 Series Sheet 8112 S (80): Hobart, Geological Survey of Tasmania, Department of Mines, scale 1:50,000. U.S. Nuclear Regulatory Commission (USNRC), 1996, Geologic and Seismic Siting Factors for Nuclear Power Plants: U.S. Nuclear Regulatory Commission, Section 100.23 10 CFR Part 100; available at http://www.nrc .gov/reading-rm/doc-collections/cfr/part100/part100-0023.html (accessed 31 May 2011). Wells, D.L., and Coppersmith, K.J., 1994, New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement: Bulletin of the Seismological Society of America, v. 84, p. 974–1002. Williams, P.W., 1996, A 230 ka record of glacial and interglacial events from Aurora Cave, Fiordland, New Zealand: New Zealand Journal of Geology and Geophysics, v. 39, p. 225–241, doi:10.1080/00288306.1996.9514707.
MANUSCRIPT ACCEPTED BY THE SOCIETY 7 DECEMBER 2010
Printed in the USA
The Geological Society of America Special Paper 479 2011
Multiple-trench investigations across the newly ruptured segment of the El Pilar fault in northeastern Venezuela after the 1997 Cariaco earthquake Franck A. Audemard M.* Fundación Venezolana de Investigaciones Sismológicas (FUNVISIS), El Llanito, Caracas 1073, Venezuela
ABSTRACT After the 9 July 1997 Ms 6.8 Cariaco earthquake in northeastern Venezuela, we undertook a multiple paleoseismic trench assessment on the newly ruptured portion of the dextral El Pilar fault. The surface rupture of that earthquake extended for 37 km from the seashore village of Villa Frontado to Río Casanay, along the onshore El Pilar fault section that runs between the gulfs of Cariaco and Paria (State of Sucre). This investigation intends to shed additional light on the past seismic history of that fault. For this, three backhoe-dug trenches were excavated between the towns of Cariaco and Río Casanay in early 1998, at the localities of Las Manoas, Carrizal de La Cruz, and Guarapiche. This effort was complemented by the evaluation of an outcrop already in existence in Terranova (7 km west of Cariaco). The three trench sites exhibit very different sedimentary settings. The Las Manoas site is an active papayacultivated sag pond. The Carrizal de La Cruz site is an active alluvial terrace, slightly down-faulted with a scarp facing against runoff, thus acting as a sort of shutter ridge. The third trench was cut at the foot of the northern slope of a pop-up structure, forming at a restraining overlap (as attested by 1997 Cariaco earthquake rupture mapping). All trenches were ~20 m long and 3–4 m deep. The main outcomes of this assessment are: (1) Over 10 earthquakes are common to the three trenches over a period of 5.6 k.y.; (2) the latest five events, including the 1997 event, clearly seem to recur roughly every 300 yr; (3) a minimum of 15 or 16 events can be deduced from colluvial deposits interfingered with the fault pond sequence at Las Manoas trench, averaging a repeat period of 350 yr over the longer 5.6 k.y. time span; (4) the predecessor of the 1997 event was the 1684 earthquake, for which chronicles could not provide a more precise determination previous to this study; and (5) the 1974 event might be present in the two easternmost trenches.
*
[email protected] Audemard M., F.A., 2011, Multiple-trench investigations across the newly ruptured segment of the El Pilar fault in northeastern Venezuela after the 1997 Cariaco earthquake, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, p. 133–157, doi:10.1130/2011.2479(06). For permission to copy, contact
[email protected]. © 2011 The Geological Society of America. All rights reserved.
133
134
F.A. Audemard M.
INTRODUCTION Following the mapping of the surface break of the Cariaco Ms 6.8 earthquake, which struck northeastern Venezuela on 9 July 1997, the Venezuelan Foundation for Seismological Research (FUNVISIS) decided to carry out a multiple paleoseismologic trench assessment across the ground rupture associated with the earthquake. This visible earthquake break was mapped onshore from the coast of Cariaco Gulf eastward for 37 km to the village of Río Casanay (Fig. 1; Audemard, 1997, 1999a, 2006; FUNVISIS, 1997; FUNVISIS et al., 1997; Baumbach et al., 2004), coinciding well with the mapped Quaternary trace of the El Pilar fault (FUNVISIS, 1994; Beltrán et al., 1996). We selected three sites along the surface rupture for excavation (labeled 1 through 3 in Fig. 1). This paper presents the last 5.6 k.y. of seismic history of this portion of the El Pilar fault determined from our study of these three trenches and an additional artificial outcrop found in the backyard of a house at Terranova (labeled 4 in Fig. 1). PRESENT-DAY GEODYNAMICS Northeastern Venezuela lies within the very complex plateboundary zone of the Caribbean, South America, and Atlantic plates (Fig. 2). From a geodynamic viewpoint, the Caribbean plate moves eastward with respect to South America (Bell, 1972; Malfait and Dinkelman, 1972; Jordan, 1975; Pindell and Dewey, 1982; Sykes et al., 1982; Wadge and Burke, 1983; Freymueller et al., 1993; Kaniuth et al., 1999; Pérez et al., 2001; Weber et al., 2001; among others). This Caribbean–South America plate boundary is particularly complex because it is a very wide active transpressional zone (Fig. 2; Soulas, 1986; Audemard, 1993, 1998; Beltrán, 1994; Singer and Audemard, 1997), in which a large fraction of the plate motion is accommodated by the right-lateral strike-slip Boconó–San Sebastián–El Pilar– Los Bajos/El Soldado–Warm Springs fault system (Molnar and Sykes, 1969; Minster and Jordan, 1978; Pérez and Aggarwal, 1981; Stephan, 1982; Aggarwal, 1983; Schubert, 1984; Soulas, 1986; Beltrán and Giraldo, 1989; Speed et al., 1991; Singer and Audemard, 1997; Pérez et al., 2001; Weber et al., 2001; Pindell and Kennan, 2007; Audemard, 2009). Currently, this region images both oblique subduction, which takes place under Trinidad and Paria Peninsula, and partitioned oblique collision (or transpression), occurring further to the west under the eastern Interior Range in Venezuela (Figs. 2 and 3). So, it re-creates the last and still active stage of the eastward migration of the Caribbean plate between the Americas throughout the Tertiary, characterized by shifting from oblique subduction to oblique collision (or transpression; Audemard, 2000, 2006). Los Bajos and El Soldado faults (Fig. 4), which should act as a sort of “lithospheric tear fault,” split apart those two concurrent geodynamic provinces (the partitioned transpressional boundary of the orogenic-float type on the west from the conventional type-B Lesser Antilles subduction zone on the east). This is imaged from distinct seis-
mic domains originally identified by Pérez and Aggarwal (1981), and further supported by Sobiesiak et al. (2002, 2005). At present, eastern Venezuela is the most seismically active area in the southern Caribbean plate boundary, as documented by the 1910-to-date instrumental seismicity catalog of FUNVISIS. This has been further shown by the 1997 Ms 6.8 Cariaco earthquake, and by historical reports of damage in this region due to earthquakes and tsunamis since the time of the Spanish conquest (Centeno-Graü, 1940; Gómez, 1990; Grases, 1990; Grases et al., 1999). The earthquakes of 1530, 1684, 1766, 1797, and 1853 are among the largest and most destructive historical events for the region. Different authors have related all of these events to the El Pilar fault, which is considered to be the second most important seismic source in eastern Venezuela after the southern end of the NW-dipping slab of the Lesser Antilles subduction zone. The latter zone lies under Trinidad and partly under Paria Gulf (Fig. 2). The associations of historical destructive earthquakes in eastern Venezuela with the El Pilar fault, except for those of the twentieth century with reported surface rupture (the 1929 Cumaná and 1997 Cariaco events), have been made by assumptions and do not rely on geological corroboration. Moreover, each of those events has been assumed to be associated with a particular segment of the fault based on intensity maps. A recent reassessment of the most significant historical events in eastern Venezuela associated with the El Pilar fault was presented by Audemard (2007) and Audemard et al. (2007). EL PILAR FAULT The El Pilar fault extends eastward for some 350 km (e.g., Soulas, 1986; Beltrán, 1993, 1994; Audemard et al., 2000), from the Cariaco Trough—located south of La Tortuga Island—to the gulf of Paria, exhibiting an ~80-km-long onshore portion in the State of Sucre between the gulfs of Cariaco and Paria (Fig. 3). The 1997 Cariaco event took place along this onshore section of the El Pilar fault, providing a visible 37-km-long surface rupture (Figs. 1, 3, and 4). In this region, the El Pilar fault roughly separates two very different geological provinces (e.g., Metz, 1968; Vignali, 1979). North of the fault, the eastern Cordillera de la Costa (CC in Fig. 2) stretches across the Araya and Paria Peninsulas (Figs. 2 and 3), which consist of Lower Cretaceous metasediments and igneous rocks accumulated in a tectonically and volcanically active environment. South of the El Pilar fault, SSE-vergent sheets of Cretaceous sedimentary rocks of the Interior Range of eastern Venezuela, piled up from the late Miocene, crop out extensively (Fig. 3). The slip direction of this fault has been unequivocally confirmed by the dextral character of the surface break associated with the 1997 Cariaco earthquake (Audemard, 1997, 1999a, 2006), and this is consistent with the focal mechanism solutions for the event (Pérez, 1998; Baumbach et al., 1999, 2004; Romero et al., 2002; Audemard et al., 2005). Lively controversy on this matter continued until very recently (e.g., Giraldo, 1996), irrespective of the numerous authors who consider the El Pilar fault
N
°
W
)
°
km
°
Figure 1. Location of excavated trenches (small black rectangles) reported on the surface faulting of the 9 July 1997 Cariaco earthquake. Specific trench-site maps are also provided in Figures 5 and 15. Coseismic slip measurements are reported along rupture, as well as the focal mechanism solution based on P-wave arrival times recorded by Fundación Venezolana de Investigaciones Sismológicas national seismologic network (after Audemard et al., 1999; Romero et al., 2002). This mechanism is off the trace because of the dip of the fault plane to the north at depth. The numbers in parentheses correspond to the number of measurements with that value.
°
°
135
Figure 2. Schematic geodynamic map of the southern Caribbean plate boundary zone (after Audemard et al., 2000). Legend: BF—Boconó fault, CC—Cordillera de La Costa range, EPF—El Pilar fault, OAF—Oca-Ancón fault, SSF—San Sebastián fault, WSF—Warm Springs (also Central Range) fault.
136
Trending across the El Pilar fault in northeastern Venezuela after the 1997 Cariaco earthquake
137
Margarita Island
N
Tortuga Island
Caribbean Sea
0m -20
CARIACO TR OUGH -200
m
Caigüire Hills Araya
Paria Pen
insula
Peninsula
EL PILAR FA
ULT Gulf of Paria
65.5°W
km
0
Venezuela
100
Eastern Interior Range 62.5°
63.5°
64.5°
Figure 3. The dextral El Pilar fault is shown in relation to the major physiographic and bathymetric features of eastern Venezuela (base map after Garrity et al., 2004).
W
Caribbean Sea N
N Margarita Island Tortuga Island c
SSF
El Pilar fa
d
ult b
a Caigüire Hills (pop-up)
Lo
El Pilar fault El So ld a
Gulf of Paria ajo s fa ult
sB
do
fa
ul
t
South America
Figure 4. Active transpressional tectonic setting of northeastern Venezuela, where the right-lateral strike-slip El Pilar fault plays a major role in strain partitioning (after Audemard et al., 2000). The four fault sections (a–d) into which the El Pilar fault has been subdivided are shown. SSF—San Sebastián fault.
138
F.A. Audemard M.
to be part of a major wrench system within the Caribbean–South America transpressional boundary. The active trace of the El Pilar fault has been mapped in detail on the basis of geomorphic evidence of tectonic activity by FUNVISIS (1994), and published by Beltrán et al. (1996). According to them, numerous geomorphic features along the active fault trace, such as right-laterally offset drainages, fault trenches, pop-ups, pressure ridges, fault saddles, sag or fault ponds, and fault scarps (for more details, refer to Figs. 1 and 2 in Beltrán et al., 1996), unambiguously establish the dextral sense of slip of this fault. These data support the right-lateral slip on the El Pilar fault, as proposed by both Metz (1968), relying on geologic data, and Pérez and Aggarwal (1981), on the basis of the few available focal mechanisms. Based on the neotectonic mapping, the El Pilar active fault trace was subdivided in four different sections (Audemard et al., 2000; Fig. 4): (1) a submarine trace west of Cumaná that bounds the Cariaco Trough (pull-apart basin) on the south and dies out at the Caigüire Hills, at Cumaná, in a restraining stepover; (2) a section that extends from the north side of the aforementioned stepover to the village of Río Casanay (location in Fig. 1 and section tip indicated in Fig. 4 by empty triangle); (3) a 30-km-long section that slightly diverges to the ENE and extends between Río Casanay and El Pilar; and (4) a fourth section, also east-west trending, that cuts across the swampy areas of the Sabanas de Venturini and runs offshore south of Paria Peninsula, before connecting in a transtensional stepover with, and transferring its kinematics to, the Warm Spring fault in Trinidad (Fig. 2). The El Pilar, Los Bajos, El Soldado, and Warm Spring faults all together, as mentioned earlier, are considered to be the easternmost portion of the major dextral plate-boundary fault system between the Caribbean and South America plates (Figs. 2 and 4). Further evidence of dextral slip along the El Pilar fault has been brought up by recent global positioning system (GPS) studies carried out in the region (Pérez et al., 2001; Weber et al., 2001). Weber et al. (2001) derived a rate of 20 ± 1–2 mm/yr for the Caribbean plate relative to the South America plate in a N89°E direction in the vicinity of the 1997 Cariaco main shock. According to results by Pérez et al. (2001), three additional important aspects can be deduced regarding the relative motion between Caribbean and South America plates in this region: (1) The elastic strain across this plate-boundary zone affects a region at least 110 km wide; (2) 68% of the 20.5 mm/yr right-lateral motion measured across most of the plate-boundary zone (almost 14 mm/yr) is elastically taken up by a 30-km-wide fault zone, which includes the El Pilar fault and other subparallel faults located north of it; and (3) although subordinate to the right-lateral strike-slip motion, compression is taking place along the plate-boundary zone as evidenced by those vectors located south of the main wrenching system. This confirms and supports the geologic data collected from neotectonics studies and compiled by Audemard et al. (2000; refer to Fig. 4). Furthermore, the magnitude and orientation of the GPS motion vectors are in good agreement with the slip rates derived from geologic criteria for most major active
faults in eastern Venezuela (Audemard et al., 2000). The slip rate of the El Pilar fault has been estimated to be 8–10 mm/yr, which is very similar to the rate derived by Weber et al. (2011) for the Central Range fault system in central Trinidad (WSF of Fig. 2), based on triangulation and GPS data merging. THE 1997 CARIACO SURFACE RUPTURE Audemard (1999a, 2006) described the 1997 Cariaco surface rupture in detail, showing that the ground break is composed of two very distinct sections. A main, very conspicuous, continuous, rather straight, N75°-trending alignment of en echelon surface breaks extends eastward from the seashore village of Villa Frontado (Muelle de Cariaco) to Las Varas (just east of Casanay) for some 30 km (Fig. 1). The rupture along this main section is generally <3 to 4 m wide and is typically expressed by a zone of N90°–100°-trending en echelon, right-lateral Riedel (R) shears. Mole tracks are present locally, typically within restraining overlaps, and this segment also exhibits tension gashes and en echelon folds. The western half of the rupture between Villa Frontado and Carrizal de La Cruz shows a dextral coseismic slip of ~25 cm but reaches its maximum value of 40 cm slightly northwest of Pantoño (3 km west of Casanay), and it abruptly ends ~5 km farther east (Fig. 1). The secondary rupture branches off northward within Casanay, discontinuously extending eastward up to Río Casanay, for some 10 km (Fig. 1). This secondary rupture, showing an overall convex-to-the-north shape and displaying a more modest (only few tens of centimeters wide) but similar surface expression, has a maximum coseismic slip of 20 cm at Guarapiche. West of this town, the rupture partly bounds both slopes of Guarapiche Hill (restraining stepover geometry, responsible for hill growth). Audemard (1997, 1999a, 2006) indicated that all structural features observed at various scales along both surface breaks confirm the right-lateral character of the El Pilar fault. In addition, aftershock activity has clearly defined the steep northward dip of the fault plane along the main rupture (Baumbach et al., 2004), which attests to tectonic inheritance on this major fault. This also explains why the epicenter of the 1997 earthquake lies north of the fault rupture (Fig. 1). Conversely, the dip of the secondary rupture tends to be vertical to steeply south-dipping, as revealed by aftershock hypocenters (Baumbach et al., 2004). This dip change must occur at or near the Guarapiche restraining bend. The 9 July 1997 Cariaco earthquake broke the eastern part of the second section out of the four defined by Audemard et al. (2000) (Fig. 4), described earlier in this paper. This event, in combination with the 17 January 1929 Cumaná earthquake, ruptured this entire fault section. With respect to the segmentation model proposed by FUNVISIS (1994), published by Beltrán et al. (1996), this 1997 rupture partly overlaps their adjoining sections b and c. This latter section of the El Pilar fault exhibits three main traces with different degrees of activity, based on their associated Quaternary landforms (FUNVISIS, 1994). The central fault trace, which appears to be the most active, consists
Trending across the El Pilar fault in northeastern Venezuela after the 1997 Cariaco earthquake of three to four different splays, one of which broke during the 1997 Cariaco event. TRENCHING ACROSS THE 1997 SURFACE RUPTURE After a detailed seismotectonic survey of the 1997 Cariaco earthquake rupture, we chose three sites between Cariaco and Guarapiche for trench assessment of the seismic history of the sector of the El Pilar fault newly ruptured during the 1997 Cariaco event. All three trenches were dug by backhoe at Las Manoas, Carrizal de la Cruz, and Guarapiche, which are located over a distance of 15 km along the rupture (black boxes across the surface rupture, labeled 1 through 3 in Fig. 1). The trenches were close to the epicenter of the 1997 main shock. Their dimensions were on the order of 20 m in length, 4 m wide at the top at most, and typically around 3 m in depth. All trenches were oriented N-S. Next, we shall describe each trench separately, from west to east. Las Manoas Trench The Las Manoas trench site is a few tens of meters north of National Road 10, linking Cariaco to Casanay, and 2.5 km east from its intersection with Road 9 (labeled 1 in Figs. 1 and 5). This 18-m-long trench was dug in a small Papaya tree field. The excavation cut through a small depression dammed behind a 2-m-high, west-east–elongated ridge (Fig. 6). The 1997 Cariaco rupture, displaying WNW-ESE–trending, slightly open cracks, occurred halfway up the southern flank of the shutter ridge that dams the active fault pond on the north (Fig. 7), as imaged in Figure 4A of Audemard (2006). The trench disclosed that the shutter ridge coincided with an underlying pop-up structure made of coarse angular colluvial deposits. Their degree of lithification by calcrete precipitation and their mottled yellowish-reddish alteration colors suggest that these deposits are not Holocene and are most likely Pleistocene in age (Fig. 8). In contrast, the other end of the trench exposed an active organic-rich, dark-brown clay-dominated sediment associated with the fault pond, resting atop of a whitish, calcretecemented colluvial bed (Fig. 6). This underlying clay-supported colluvium, which appears to be younger than that of the shutter ridge, contains large blocks, the number and size of which increase with depth (Figs. 6 and 8). The gentle inclination to the north of its upper contact suggests that the source of this colluvial unit is the north flank of the Interior Range, which lies a few hundreds of meters south of the trench site (Fig. 5). Squeezed between the fault pond and the shutter ridge deposits, another colluvial unit, less rich in matrix than that overlain by the brown fault pond clays, was exposed. Its clasts also exhibit better sorting and smaller average grain size, although equally angular to subangular. This colluvial unit showed a prevailing reorientation of the longer clasts that mimics the shearing trend of the El Pilar fault planes exposed in the trench walls (Fig. 8) along and near the contact between the fault pond and the shutter ridge deposits. In fact, the geometric disposition of all fault planes inside the
139
trench resembles a positive flower structure cut in half, fanning upward and increasing fault dip progressively from the fault pond edge toward the ridge top (Fig. 8). The most upright planes correspond to the synthetic R shears of the 1997 Cariaco rupture, which as an individual set also exhibits the upward-fanning shape in cross section (thicker lines in Fig. 8). This further supports the interpretation that the ridge is a pressure ridge or push-up resulting from transpression (compressive dextral motion). Here, the 1997 ground cracks at the surface also displayed some opening, creating fissures, as well as some vertical component of motion, because of being halfway up the south flank of the ridge (Fig. 7). Although a vertical displacement of only 6 cm is small, its implications in the paleoseismologic assessment of this trench are of high significance. A record of earthquakes is exposed in this trench. Evidence for past earthquakes includes organic-rich vertical wedges, resulting from the infill of open fissures (such as the one indicated by sample 37 in Fig. 8), and small colluvial wedges interfingered in the fault pond clay-rich sequence (any of those labeled 5 through 32 in Fig. 8). These colluvial wedges can eventually look like stone lines (e.g., those marked as 31 or 32 in Fig. 8). This interpretation seems to be supported by the fault behavior at the site during the 1997 event. Open cracks up to 5 cm across were visible at ground surface, which induced a few centimeters of vertical offset favoring the slope, similar to that of Figure 7. With time, degradation of the crack-free face fills the fissure but also leaves a vertical offset that runoff will regularize, thus remobilizing coarser materials downslope from the ridge into the fault pond. Some events produce both features, as is recorded in this exposure. Samples 35 and 36 date two filled crevasses, and they are coeval to two colluvial wedges (dated by samples 15 or 16 and 17 or 18; compare ages in Fig. 9). This tectonic-sedimentation interplay is also recorded within the sheared colluvial deposit. Several of the small colluvial wedges interfingered in the fault pond clays derive from former small, south-looking free faces in this colluvial deposit. These free faces coincide with the upward extension of fault planes defined by the reorientation of elongated clasts within the colluvial deposit (Fig. 8). In the west wall (Fig. 8), 13 colluvial wedges were sampled for radiocarbon dating, but 16 were recognized (two poorly developed wedges between samples 21 and 22, and another between samples 22 and 23). In addition, four fissure-fill deposits were sampled: one at the bottom of the fault pond (sample 33, not dated because of funding limitations) and three much younger ones near the top of the outcrop (samples 35 through 37, Fig. 8). We also collected organic samples from 16 colluvial wedges in the eastern wall, where identified, of which the 14 oldest were dated (see Figs. 9 and 10; Table 1). In this wall, a few filled open cracks were also identified within the shutter ridge sequence, but dating them was not possible due to the calcrete precipitation. Only a single age inversion was obtained between samples 11 and 12 (compare Table 1 and Fig. 9). To bracket the time of occurrence of a given earthquake on the basis of colluvial wedges, it is the common practice to
140
F.A. Audemard M.
1 km
.
. Las Manoas pop-up
f ault E l P .i l a.r . . . .
.
.
Aguas Calientes
Interior Range
Carrizal de La Cruz
Las Manoas
.
.
.
Road 10
A
N Cariaco
Lambertino’s Carrizal de La Cruz farmhouse
Las Manoas Las Manoas pop-up
.
.
.
.
.
.
.
.
1
Interior Range
.
2
B
Figure 5. Surface expression of the El Pilar fault between Cariaco and Aguas Calientes (Pantoño). The fault trace is subtly underlined by white dots, as well as by white arrows at the tips. (A) The fault roughly runs at the edge of the northern slope of the Interior Range of eastern Venezuela, crossing National Road 10 several times (reported on aerial photo 2203A of mission 172, taken in 1961). (B) Close-up of a portion of the previous image, showing the location of the trench sites of Las Manoas (1) and Carrizal de La Cruz (2) (reported on aerial photo 005 of mission 04021128). The dashed boxes on both images cover about the same area.
Detail in Fig. 7
Shutter ridge Figure 6. General ground view of Las Manoas trench cutting across active fault pond and shutter ridge (pressure ridge or pop-up) on the south and the north, respectively. Relative location of Figure 7 is also indicated.
Fault pond
1m
S
N
Trending across the El Pilar fault in northeastern Venezuela after the 1997 Cariaco earthquake
1997 open fissure Papaya trunk
W 6 cm
March 1998 Figure 7. Detail of an open fissure of the 1997 Cariaco earthquake, next to the Las Manoas trench (location on Fig. 6), showing not only the amount of opening but also the vertical offset because the fissure opens on inclined ground.
radiocarbon date the organic-rich layers deposited prior and subsequent to the wedge deposition. It is worth mentioning that a particular sampling approach was performed during this investigation for radiocarbon dating. These colluvial wedges, mainly formed of stone lines, show the clasts floating in a matrix made of fault-pond clays. Taking into account the high rainfall and the very fast rate of formation of organic soils in this wet tropical region, we assumed that the wedge matrix was close to, or just predated, the age of the event. Consequently, the organicrich clay matrix of each colluvial wedge was then sampled for radiocarbon dating, thus cutting the number of samples down by half. Consequently, the assumed age of each colluvial wedge is given by a single date. This was a substantial gain in cost reduction, when you consider that 90 radiocarbon dating analyses were required for this multiple-trench evaluation. Finally, including the 1997 earthquake, the occurrence of 15 earthquakes has been interpreted from evidence in this trench, with a recurrence
141
interval of 200–400 yr. An average repeat period of 350 yr for these large earthquakes, nearing Ms 7.0, can be calculated for the 5.6 k.y. time span covered by the fault pond sedimentation. Carrizal de la Cruz Trench The Carrizal de La Cruz trench site is just 40 m south of National Road 10, inside the very small village of the same name (2 in Fig. 5). The trench site, located in the Flores family’s property, is relatively flat, except for an E-W–oriented, 50-cm-high, south-facing scarp (Fig. 11) that cuts across the backyards of several houses. The trench was dug normal to the scarp, which faces natural runoff fed from the north flank of the Interior Range lying south of the trench site. The 1997 surface break occurred along the base of the small scarp (Figs. 11 and 12). The rupture is expressed as a set of roughly 1-m-long, WNW-ESE– to NW-SE–trending open fissures (Fig. 11). In some of these, the ground surface fell as much as half a meter inside them (Fig. 12). These features attest to local transtension, as the fault scarp suggests. This is in agreement with the slight trend change that the El Pilar fault undergoes near Carrizal de La Cruz (Fig. 5). The Carrizal de la Cruz trench walls exposed a very coarse, clast-supported, unconsolidated alluvial sequence in the upthrown block on the north, against which a finer, dark-brown, organicrich sequence is deposited on the south (Fig. 12). Beds within the very coarse alluvial deposit dip a few degrees to the north, showing the same dip as the ground surface (Fig. 12). This unit almost crops out at the crest of the scarp, implying scarp degradation and regression. From aerial photo-interpretation, this unit appears to correspond to an alluvial fan that has been progressively shifted to the east (Fig. 5). The clayey sequence, confined to the downthrown block, displays interbedded gravelly alluvial beds (Fig. 12). As a whole, this finer-grained sequence pinches out against the fault scarp. We interpret the occurrence of small colluvial wedges fanning down from the scarp top as further evidence of scarp degradation (Fig. 12). The clasts of these colluvial wedges are atypical because they are subrounded to rounded, due to the original roundness of the parental material. The presence of very fine materials at the foot of the scarp implies that it has intermittently dammed runoff for significant periods of time, allowing water impoundment. Snail shell horizons within the finer beds in the uppermost section of this sequence also attest to dammed water (Fig. 12). At the south end of the trench, the same very coarse alluvial sequence as that exposed north of the fault is overlain, in fault-bounded contact, by the same dark-brown clays. All this suggests that a large (6–8-m-wide) graben-like feature has formed within the coarse older alluvial deposit (Fig. 12). The past fault behavior of the El Pilar fault at this locality seems to have repeated during the Cariaco earthquake. Large open fissures formed, preserving the materials degraded from the scarp by runoff. Evidence for past earthquakes in this exposure includes scarp-derived colluvial wedges and/or fining-upward fissure infill, similar to the Las Manoas trench.
Distance (m)
Depth (m)
Figure 8. Log of west wall of the Las Manoas trench: entire log (top) and closeup of area of interest (bottom). Thick lines depict the fault planes that ruptured in 1997. Rectangle indicates portion of log being blown up, which is inserted below for clarity. Labeled black boxes indicate radiocarbon sampling. Ages (absolute, 1σ, and 2σ calendar dates) are given in Table 1 and are all graphically summarized in Figure 9. Please note that Figures 8, 10, 12, 14, and 18–20 are to scale, so the grain sizes of sediments within each figure are proportionally accurate.
142
AD AD
BC
BC
Figure 9. Chart of all samples dated in this multiple-trench evaluation. Dates reported are all 2σ calendar ages, younging from left to right. From top to bottom, each set of ages corresponds to individual-trench dates, which are displayed geographically from west to east, respectively.
143
144
F.A. Audemard M.
Height (m)
Distance (m)
Figure 10. Detail of east wall log of the Las Manoas trench, only showing the colluvial wedges derived from the shutter ridge into the fault pond. Labeled black boxes indicate radiocarbon sampling. Ages (absolute, 1σ, and 2σ calendar dates) are given in Table 1 and are all graphically summarized in Figure 9.
S Because these two types of features can be produced during a single earthquake, identification and dating of all individual earthquakes can become difficult. In addition, several fissures or large open cracks can open simultaneously during a single event, as observed along the 1997 Cariaco ground rupture (Audemard, 2006; Fig. 11). Consequently, some dozens of samples were sent in for 14C dating. Figure 9 shows that some events are dated and recognized several times. For instance, samples 46, 52, 58, and 70 are in the same age range (Fig. 9; Table 1). Samples 58 and 70 predate the same event, but they come from different fissure fills and imply that both fissures resulted from the same earthquake. Sample 46 dates the matrix of a colluvial wedge on the west wall (Fig. 13), corresponding to the same body dated by sample 70 but recorded in the other wall. Additionally, sample 52 should postdate a prior event.
N
DOWN
Runoff
South-facing scarp T-R open fissure
Figure 11. Surface rupture of the 1997 Cariaco earthquake at the Carrizal de la Cruz trench site. Open fissures appeared obliquely (oriented WNW-ESE to NW-SE) to the roughly E-W–trending fault scarp of the El Pilar fault. This ~0.5-m-high scarp faces against natural runoff, pounding finer sediments on the downthrown (southern) block, as evidenced by trench walls. T-R—Tension-Riedel shear.
UP
Trending across the El Pilar fault in northeastern Venezuela after the 1997 Cariaco earthquake
145
TABLE 1. RADIOCARBON AGES OF TRENCHES EXCAVATED ACROSS THE COSEISMIC RUPTURE OF THE 9 JULY 1997 CARIACO, VENEZUELA, EARTHQUAKE FUNVISIS sample no.
Beta analytic no.
Radiocarbon age (yr B.P.)
1σ
Terranova and Las Manoas trenches SU-01-98 119514 SU-02-98* 119515
960 ± 60 2240 ± 40
SU-04-98
119516
270 ± 60
SU-05-98
119517
4650 ± 80
B.C. 3515–3345
SU-06-98
119518
4490 ± 100
SU-07-98
119519
4000 ± 100
B.C. 3350–3015 B.C. 2985–2935 B.C. 2605–2400
SU-08-98*
119520
4020 ± 50
B.C. 2585–2470
SU-09-98* SU-10-98*
119521 119522
3900 ± 40 3640 ± 40
B.C. 2460–2310 B.C. 2030–1935
SU-11-98* SU-12-98 SU-13-98 SU-14-98*
119523 119524 119525 119526
3170 ± 40 3650 ± 110 2920 ± 110 2390 ± 40
SU-15-98 SU-16-98 SU-17-98
119527 119528 119529
1780 ± 120 1760 ± 70 990 ± 70
SU-18-98 SU-19-98 SU-20-98 SU-34-98 SU-35-98 SU-36-98 SU-37-98
119530 119531 119532 119533 119534 119535 119536
640 ± 90 101.4 ± 11%† 100 ± 0.1%† 110.6 ± 0.8%† 1490 ± 70 730 ± 80 270 ± 60
B.C. 1450–1405 B.C. 2145–1885 B.C. 1275–930 B.C. 485–465 B.C. 425–395 A.D. 110–410 A.D. 220–390 A.D. 995–1065 A.D. 1075–1155 A.D. 1285–1410 Recent Recent Recent A.D. 535–645 A.D. 1245–1305 A.D. 1525–1560 A.D. 1630–1670
Carrizal de la Cruz trench SU-40-98 SU-41-98 SU-42-98 SU-43-98
119539 119540 119541 119542
3010 ± 80 2850 ± 80 3300 ± 70 2180 ± 60
SU-44-98 SU-45-98 SU-46-98 SU-47-98
119543 119544 119545 119546
1780 ± 70 1660 ± 60 1280 ± 70 580 ± 60
SU-48-98
119547
490 ± 60
B.C. 1385–1120 B.C. 1120–905 B.C. 1660–1500 B.C. 360–280 B.C. 250–150 A.D. 160–370 A.D. 350–440 A.D. 670–855 A.D. 1310–1365 A.D. 1375–1420 A.D. 1410–1450
SU-49-98 SU-50-98 SU-51-98
119548 119549 119550
103.3 ± 0.7%† 2880 ± 70 2610 ± 60
Recent B.C. 1135–930 B.C. 815–780
SU-52-98 SU-58-98 SU-67-98 SU-68-98 SU-69-98
119551 119553 119554 119555 119556
1300 ± 60 1300 ± 80 1620 ± 80 1530 ± 80 430 ± 50
A.D. 665–785 A.D. 660–800 A.D. 380–550 A.D. 435–630 A.D. 1435–1485
SU-70-98 SU-71-98
119557 119558
1330 ± 60 1870 ± 60
A.D. 655–770 A.D. 85–235
A.D. 1015–1170 B.C. 375–330 B.C. 330–205 A.D. 1525–1560 A.D. 1630–1670
Calendar age (yr) 2σ A.D. 985–1220 B.C. 390–185 A.D. 1475–1685 A.D. 1740–1810 A.D. 1930–1950 B.C. 3635–3285 B.C. 3245–3105 B.C. 3500–3435 B.C. 3385–2900 B.C. 2875–2790 B.C. 2780–2205 B.C. 2845–2830 B.C. 2620–2450 B.C. 2475–2270 B.C. 2125–2065 B.C. 2060–1895 B.C. 1510–1385 B.C. 2325–1730 B.C. 1410–825 B.C. 745–700 B.C. 530–385 B.C. 5–A.D. 550 A.D. 110–430 A.D. 905–920 A.D. 950–1215 A.D. 1235–1440
A.D. 425–670 A.D. 1175–1405 A.D. 1475–1685 A.D. 1740–1810 A.D. 1930–1950 B.C. 1425–1000 B.C. 1260–825 B.C. 1735–1420 B.C. 380–45 A.D. 90–420 A.D. 245–550 A.D. 645–895 A.D. 1290–1440 A.D. 1325–1340 A.D. 1390–1495 B.C. 1265–855 B.C. 845–760 B.C. 670–550 A.D. 645–880 A.D. 620–895 A.D. 245–620 A.D. 390–665 A.D. 1420–1525 A.D. 1560–1630 A.D. 630–855 A.D. 25–265 A.D. 290–320 (Continued)
146
F.A. Audemard M.
TABLE 1. RADIOCARBON AGES OF TRENCHES EXCAVATED ACROSS THE COSEISMIC RUPTURE OF THE 9 JULY 1997 CARIACO, VENEZUELA, EARTHQUAKE (Continued) FUNVISIS sample no.
Beta analytic no.
Radiocarbon age (yr B.P.)
1σ
Carrizal de la Cruz trench (Continued ) SU-72-98 119559
860 ± 70
SU-75-98 SU-77-98
119560 119561
920 ± 70 4660 ± 70
SU-78-98 SU-79-98 SU-80-98
119562 119563 119564
3420 ± 60 2640 ± 60 2400 ± 70
SU-81-98 SU-82-98 SU-83-98 SU-84-98 SU-85-98 Guarapiche trench SU-100-98* SU-101-98 SU-103-98* SU-104-98
119565 119566 119567 119568 119569
2460 ± 70 2050 ± 50 1800 ± 70 1780 ± 80 1690 ± 70
B.C. 1760–1645 B.C. 825–790 B.C. 745–700 B.C. 530–390 B.C. 775–405 B.C. 100–A.D. 15 A.D. 135–340 A.D. 145–380 A.D. 260–430
119570 119571 119573 119574
1690 ± 40 4190 ± 110 7260 ± 40 3840 ± 70
A.D. 340–415 B.C. 2900–2590 B.C. 6135–6015 B.C. 2440–2175
SU-105-98 SU-106-98 SU-108-98 SU-109-98
119575 119576 119577 119578
5450 ± 100 3320 ± 70 1920 ± 60 1050 ± 80
SU-110-98 SU-111-98 SU-112-98 SU-113-98
119579 119580 119581 119582
1970 ± 70 119.8 ± 0.8%† 1060 ± 80 2060 ± 70
B.C. 4365–4225 B.C. 1675–1510 A.D. 45–145 A.D. 905–920 A.D. 950–1035 B.C. 35–A.D. 110 Recent A.D. 895–1030 B.C. 165–A.D. 25
SU-114-98
119583
3240 ± 80
B.C. 1605–1420
SU-115-98
119584
1740 ± 80
A.D. 225–410
SU-116-98 SU-117-98
119585 119586
1900 ± 70 200 ± 70
SU-118-98 SU-121-98 SU-122-98 SU-123-98
119587 119588 119589 119590
102.0 ± 0.8%† 660 ± 70 1170 ± 70 1020 ± 80
SU-124-98
119591
880 ± 60
SU-125-98 SU-126-98 SU-127-98
119592 119593 119594
650 ± 60 103.7 ± 0.7%† 4660 ± 90
SU-128-98* SU-129-98 SU-130-98*
119595 119596 119597
7680 ± 40 4730 ± 110 4710 ± 40
A.D. 55–220 A.D. 1650–1695 A.D. 1725–1815 A.D. 1920–1950 Recent A.D. 1285–1400 A.D. 785–975 A.D. 975–1045 A.D. 1105–1115 A.D. 1045–1105 A.D. 1115–1235 A.D. 1290–1400 Recent B.C. 3610–3590 B.C. 3525–3345 B.C. 6480–6440 B.C. 3975–3735 B.C. 3610–3590 B.C. 3525–3490 B.C. 3455–3375 B.C. 525–390 A.D. 790–980
A.D. 1055–1090 A.D. 1150–1260 A.D. 1025–1215 B.C. 3515–3350
SU-131-98 119598 2390 ± 70 SU-132-98 119599 1160 ± 70 FUNVISIS—Fundación Venezolana de Investigaciones Sismológicas. *Accelerator Mass Spectrometry dating. † Age reported as a percentage of the reference standard (100% being 0 B.P. = A.D. 1950).
Calendar age (yr) 2σ A.D. 1020–1285 A.D. 995–1265 B.C. 3630–3320 B.C. 3220–3180 B.C. 3165–3130 B.C. 1885–1535 B.C. 900–770 B.C. 780–370 B.C. 795–390 B.C. 180–A.D. 70 A.D. 75–410 A.D. 75–430 A.D. 220–540 A.D. 250–435 B.C. 3030–2470 B.C. 6165–5990 B.C. 2475–2110 B.C. 2090–2040 B.C. 4475–4035 B.C. 1750–1430 B.C. 35–A.D. 240 A.D. 855–1175 B.C. 115–A.D. 220 A.D. 800–1170 B.C. 330 B.C. 205–A.D. 90 B.C. 1685–1380 B.C. 1335–1330 A.D. 110–465 A.D. 475–515 B.C. 35–A.D. 260 A.D. 1520–1570 A.D. 1630–1950 A.D. 1250–1420 A.D. 690–1010 A.D. 880–1205 A.D. 1020–1275 A.D. 1270–1420 B.C. 3645–3100 B.C. 6535–6420 B.C. 4080–3650 B.C. 3630–3365 B.C. 775–365 A.D. 695–1015
Depth (m)
Figure 12. Log of east wall of the Carrizal de la Cruz trench: entire log atop and close-up of area of interest below, for purposes of clarity. Thick black lines depict the fault planes that ruptured in 1997. Rectangle indicates portion of log being blown up. Labeled black boxes indicate radiocarbon sampling. Radiocarbon ages (absolute, 1σ, and 2σ calendar dates) of samples are given in Table 1 and are all graphically summarized in Figure 9.
Distance (m)
147
148
F.A. Audemard M.
Height (m)
Distance (m)
Figure 13. Detail of west wall log of the Carrizal de la Cruz trench, imaging the area where radiocarbon sampling was conducted in the shear zone. Ages (absolute, 1σ, and 2σ calendar dates) are given in Table 1 and are all graphically summarized in Figure 9.
E Guarapiche
Guarapiche trench site Guarapiche pop-up Figure 14. Bird’s-eye view of Guarapiche Hill (pop-up), indicating the relative location of the Guarapiche trench site (photo taken by Audemard in March 1994).
W
Figure 15. Mapping of the secondary surface rupture of the 1997 Cariaco earthquake across the town of Guarapiche (after Audemard, 2006). Circle highlights the relative location of the Guarapiche trench site in relation to the eastern tip of Guarapiche Hill.
149
150
F.A. Audemard M.
N
S
slip at the Carrizal de la Cruz trench is definitely more significant. In other words, the available space for sedimentation is larger and the rate of sedimentation is higher here than at Las Manoas. As to encountered dating difficulties, a single age inversion was found between samples 41 and 42 (compare Table 1 and Fig. 9). Guarapiche Trench
25 cm
Figure 16. Ground rupture of the 1997 Cariaco earthquake, in the backyard of Mr. Martínez’s property, across which the trench was excavated. Fault plane was well exposed in trench wall (Fig. 17).
Radiocarbon ages show that the record of earthquakes exposed in this trench (~4 k.y.) is shorter than that of the Las Manoas trench, which covers a time window of 5.6 k.y., if sample 77 (which is dated at ca. 3.5 ka cal. B.C.) is not taken into account (Fig. 9). In this trench, there is a gap, almost 2 k.y. long, of no record (Fig. 9). This would imply that, although both trenches reached a similar final depth, the vertical component of tectonic
The Guarapiche trench site lies just north of the eastern tip of Guarapiche Hill or the pop-up structure (labeled 3 in Fig. 1 and pinpointed by an ellipse in Fig. 14). This property, currently owned by Mr. Abundio Martínez, is roughly 50 m west of the junction of the two roads that give access to Guarapiche from the west (Fig. 15). The trench was dug normal from the foot of Guarapiche Hill northward across the secondary rupture recognized after the 1997 Cariaco earthquake. The ground breaks were hardly visible due to their modest size (Fig. 16). However, they displayed the typical synthetic Riedel shear pattern described all along both the main and secondary ruptures by Audemard (1997, 1999a, 2006) and FUNVISIS et al. (1997). This rupture occurred ~10 m north of the base of the hill. The trench wall exposed a coarse, calcrete-bounded, alluvial sequence toward the north, and clay-rich deposits to the south, toward Guarapiche Hill (Figs. 17 and 18). Two colluvial units are interbedded within the clay-rich deposits (Fig. 18). While these colluvial units may have been deposited in response to earthquakes, or strong ground shaking, the evidence is not strong enough to make this interpretation. Similar to the Las Manoas trench, the El Pilar fault planes exposed in this trench also exhibit a sort of asymmetric positive flower structure, where the northern south-dipping petals seemed better developed (Fig. 18). The fault has both gently buckled and upheaved the alluvial terrace deposits. The finer reddish sediment accumulation, sitting near the positive relief to the south, seems to be controlled by this vertical deformation imprinted on the coarser deposits by the fault. Earthquake identification and characterization in this trench were also possible by means of the same two criteria recognized in the other two trenches. Radiocarbon sampling was also performed following the same procedure as in the two other trenches. In addition, the two walls were also sampled (Figs. 18 and 19). As for the Carrizal de la Cruz trench assessment, the older the events are, the more difficult they are to identify. In this case, intense weathering and carbonate precipitation obscure their recognition. Furthermore, this is exacerbated by the stratigraphic monotony exhibited by both the clayey package on the south and the coarse cemented alluvial deposit on the north. However, an almost 6-k.y.-long earthquake record could be interpreted (Fig. 9). In addition, this excavation revealed a shallow open crack in association with a 110°-striking subvertical fault plane, which contained buried pieces of plastic bags. This finding suggests the occurrence of a very recent event along this segment in the past 40 yr (“plastic era”), prior to the 1997 one. The 14C age of a sample collected from the underlying organic-rich soil (sample 111
Trending across the El Pilar fault in northeastern Venezuela after the 1997 Cariaco earthquake
151
1997 surface rupture
S Runo
ff
N
Figure 17. View to the southwest of the Guarapiche trench. Note the surface breaks (black and white bars) on the ground, on both sides of the trench, as well as the fault plane in the west trench wall. Square grids generated by ropes are 1 m × 1 m.
in Table 1) confirmed this. We conclude that this deformation is due to the 1974 Casanay earthquake, but no surface cracking was reported in that area. However, the very elderly Mr. Martínez, landowner of the Guarapiche trench site, confirmed that he spotted small open cracks on his property, near the toe of the hill, after the 1974 event. Auxiliary Outcrop at Terranova An ~1.5-m-high outcrop located south of National Road 9 and ~100 m west of the village main entrance (behind a house named Los Benitos; labeled 4 in Fig. 1) exposed two fault wedges filled with organic-rich soils (Fig. 20). These brownish wedges stand out prominently because they contrast well against the typical whitish color of the calcrete-rich sequence in which they were formed. These wedges seem to correspond to open crevasses associated with synthetic WNW-ESE–trending Riedel shears exhibiting some opening component (combination or R shears and T cracks). They show no relation with the 1997 ground rupture, since they are tens of meters north off its alignment. Topography is slightly depressed above the westernmost crack of the two, suggesting some dropping in an open fissure (Fig. 20). Three events were recognized and radiocarbon-dated at this site: two in the eastern wedge and one in the western wedge (see Fig. 9).
DISCUSSION Prior to this study, the prehistoric seismic activity of the El Pilar fault had been assessed by the Earth Sciences Department of FUNVISIS by means of a single trench excavation at Las Toscanas, between Guarapiche and Río Casanay, in 1994 (FUNVISIS, 1994; Beltrán et al., 1996, 1999). The 1994 study confirmed the Holocene activity of this fault by revealing the occurrence of four prehistoric events of magnitude ~7 in the last 7–8 k.y. In addition, the dextral sense of slip, determined from typical landforms of strike-slip faulting, was further confirmed from kinematic indicators measured on slickenside planes exposed in trench walls. Before excavating, it was understood that the 1994 assessment would only provide part of the seismic history of that portion of the El Pilar fault, due to the several active splays present between Casanay and Río Casanay. The 1994 study fell on a different fault splay to that ruptured in 1997. By comparing the seismic history from the 1994 and 1998 trench studies, the degree of dependency of activity between both fault splays (between Casanay and Río Casanay) could have been evaluated, but large uncertainties on the occurrence of the larger earthquakes in the Las Toscanas trench, due partly to the applied radiocarbon method (conventional dating on soils) and to the organic-matter availability for sampling, have precluded any reliable evaluation. In that sense, in this study, we have only
Figure 18. Log of west wall of the Guarapiche trench: entire log (top) and area of interest blown up (below). The 1997 rupture reutilized a steeply south-dipping plane exhibiting striation of 7°E pitch, highlighted by thicker black lines. These striae attest to reverse dextral motion on the El Pilar fault along the northern flank of the Guarapiche pop-up. Rectangle indicates portion of log being blown up, which is inserted below for clarity. Labeled black boxes indicate radiocarbon sampling. Sample ages (absolute, 1σ, and 2σ calendar dates) are given in Table 1 and are all graphically summarized in Figure 9.
Distance (m)
Depth (m)
152
153
Height (m)
Trending across the El Pilar fault in northeastern Venezuela after the 1997 Cariaco earthquake
Distance (m)
Figure 19. Detail of east wall log of the Guarapiche trench, displaying the locations of the radiocarbon sampling. Ages (absolute, 1σ, and 2σ calendar dates) of such samples are given in Table 1 and are all graphically summarized in Figure 9.
Height (m)
Figure 20. Simplified log of backyard outcrop behind a house (Los Benitos) at Terranova, exhibiting earthquakerelated open cracks filled by organic soils. Labeled black boxes indicate radiocarbon sampling. Radiocarbon ages (absolute, 1σ, and 2σ calendar dates) of samples are given in Table 1 and are all graphically summarized in Figure 9.
Distance (m)
AD AD
BC
BC
Figure 21. Event dating and correlation from this multiple-trench assessment. Each line represents an event determination. Different types of lines are represented, showing different degrees of correlation, from an event determined from one trench to that established from correlating all 4 trenches.
S
154
Trending across the El Pilar fault in northeastern Venezuela after the 1997 Cariaco earthquake correlated those earthquakes common to the three trenches and the Terranova outcrop. Our correlations, based partly on the excellent chronology provided by the Las Manoas trench, are shown in Figure 21. When interpreting this figure, it is important not to forget how samples were collected and how age determinations were derived. This correlation is not unique, but the option shown in Figure 21 tries to bracket earthquakes by satisfying the largest number of possible ages given by samples, taking into account the control provided by the Las Manoas results. In addition to the 1997 rupture, which is clearly observed in all three trench walls (Figs. 8, 12, and 18), several other earthquakes can be correlated among the four sites (Fig. 21). Still in the historical window (A.D. 1498 to present), three of the four studied outcrops equally exposed an event prior to 1997. This event was radiocarbon dated at 200–270 ± 60–70 yr B.P. in the Terranova outcrop and the two most distant trenches (samples 04, 37, and 117 in Table 1 and Fig. 21). The age of this event correlates well to one of the most poorly understood historic earthquakes of the entire seismic history in eastern Venezuela— the 4 May 1684 earthquake, the surface rupture of which was in an area that was probably unpopulated at that time (Audemard, 1999b, 2007). Our results suggest that the probable 1684 surface rupture extended over 25 km. This event appears to be the historical predecessor to the 1997 earthquake on this particular section of the El Pilar fault. Two other older events are also common to the four outcrops. These earthquakes occurred around A.D. 1150 and 400 B.C. (Fig. 21). Since the Terranova outcrop allowed the identification of only three events (Fig. 21), all other correlations have to rely only on the three trenches dug for this study. Nine other events are common to the three trenches, or to the two most widely separated trenches, with calendar ages around: 3550, 3200, 2850, 2450, and 1500 B.C., and A.D. 50, 400, 750, and 1400 (Fig. 21). This totals at least 12 events common to the two most widely separated trenches, in addition to the 1997 Cariaco earthquake. Two other events are only recognized in single trenches (2300 and 1150 B.C.). To these two events, we can add the Casanay 1974 earthquake, although it was not determined from this earthquake correlation, but from the trench assessment performed at Guarapiche. Recurrence intervals of these earthquakes, if only considering the 13 most common events to the El Pilar fault section under evaluation, would be in the order of 430 yr. Should we rely on the single assessment of the Las Manoas trench, where 16 colluvial wedges were easily recognized, the average return period would be close to 350 yr for earthquakes with magnitudes around Ms 7.0. This estimate is in good agreement with the earthquake recurrence derived by Audemard (1999b, 2007) for the Cariaco Trough segment of the El Pilar fault (Figs. 3 and 4), based on the two latest tsunamigenic earthquakes that affected Cumaná (A.D. 1530 and 1853). Regarding the magnitude assessment, no conventional paleoseismologic determination was possible. However, the Las Manoas trench did confirm that the prior events
155
should have had similar magnitudes to the 1997 earthquake, since vertical offsets measured on the 1997 surface features were much like those recorded in the trench walls. CONCLUSIONS Identification of sites favorable for paleoseismic study was much eased by the presence of the 1997 Cariaco surface break. Nonetheless, the most significant contribution brought up by the prior seismotectonic study of that rupture refers to the understanding of the control and/or modification of the local tectonic behavior by the El Pilar fault in specific environments. Recognition and dating of earthquakes in the Holocene sedimentary record of the three different sedimentary environments evaluated relied on the same two criteria: fill of open crevasses or fissures (fault wedges) and scarp-derived colluvial wedges. This multiple-trench assessment determined that the historical predecessor to the 1997 occurred in 1684, which is a poorly constrained historical event, only known from chronicles. Twelve events prior to the 1997 earthquake are common to the portion of the El Pilar fault assessed by the three trenches, which were spread some 15 km along the fault trace. Over a time window of almost 5.6 k.y., this implies that earthquakes of a similar magnitude, around Ms 7.0, recur every 430 yr on average. However, the Las Manoas trench displayed at least 16 colluvial wedges interfingered in an organic-rich fault pond, which would drop the average return period to 350 yr. This recurrence appears rather coincident with that of the El Pilar fault along its Cariaco Trough portion, where the two latest events occurred in A.D. 1530 and 1853. ACKNOWLEDGMENTS Lagoven S.A., a former affiliate company of Petróleos De Venezuela S.A., is thanked for their very important funding, and in particular, Hildebrando Martell, without whom trenching and dating would have not been possible. My appreciation goes to all landowners who allowed trench excavation free of charge: Angel Berroterán at Las Manoas, the Flores at Carrizal de la Cruz, and Abundio Martínez at Guarapiche. Special thanks are due to the Flores family who really made us feel comfortable in spite of the Sun actually getting unbearable sometimes. Help during fieldwork from Rogelio González is very much appreciated. I also wish to thank Marina Peña for her wonderful hand drafting. Finally, my appreciation is extended to all those who helped map the 1997 surface rupture and select the most promising trench sites, and who contributed in any other possible way during this research. We wish to thank Hotel Maigualida, particularly its manager and restaurant attendants, for a very pleasant stay and their warmth and hospitality, all typical of “Orientales.” Radiocarbon dating was performed by Beta Analytic Inc. (Miami, Florida). We are very thankful to Carol Prentice for a very thorough English revision, as well as for
156
F.A. Audemard M.
her comments and suggestions made to a former version of this contribution, although she remained very skeptical about our dating approach. My appreciation also goes to María Blumetti and Alessandro Michetti for their annotations. REFERENCES CITED Aggarwal, Y., 1983, Neotectonics of the southern Caribbean: Recent data, new ideas: Acta Cientifica Venezolana, v. 34, no. 1, p. 17. Audemard, F.A., 1993, Néotectonique, Sismotectonique et Aléa Sismique du Nord-Ouest du Vénézuéla (Système de Failles d’Oca-Ancón) [Ph.D. thesis]: Montpellier, France, Université Montpellier II, 369 p. + appendix. Audemard, F.A., 1997, Preliminary Geological Report on the Cariaco Earthquake July 09, 1997, Venezuela: International Center for Disaster-Mitigation Engineering Newsletter, v. 6, no. 2, p. 7. Audemard, F.A., 1998, Evolution géodynamique de la façade Nord Sud-Américaine: Nouveaux apports de l’histoire géologique du Bassin de Falcón, Vénézuéla, in Proceedings, 14th Caribbean Geological Conference, Trinidad 1995: Port of Spain, Trinidad, v. 2, p. 327–340. Audemard, F.A., 1999a, El sismo de Cariaco del 09 de julio de 1997, edo. Sucre, Venezuela: Nucleación y progresión de la ruptura a partir de observaciones geológicas, in Proceedings, Sixth Congreso Venezolano de Sismología e Ingeniería Sísmica: Mérida, Venezuela (CD-ROM). Audemard, F.A., 1999b, Seismic history of the El Pilar fault, northeastern Venezuela, from paleoseismic assessment across the surface rupture of the Cariaco 1997 earthquake: 1999 Spring Meeting, American Geophysical Union, Boston: Eos (Transactions, American Geophysical Union), v. 80, no. 17, supplement S227, abstract S42A-09. Audemard, F.A., 2000, Major active faults of Venezuela, in Proceedings, 31st International Geological Congress: Rio de Janeiro, Brazil (extended abstract; CD-ROM). Audemard, F.A., 2006, Surface rupture of the Cariaco July 09, 1997, earthquake on the El Pilar fault, northeastern Venezuela: Tectonophysics, v. 424, no. 1–2, p. 19–39, doi:10.1016/j.tecto.2006.04.018. Audemard, F.A., 2007, Revised seismic history of El Pilar fault, northeastern Venezuela, after the Cariaco 1997 earthquake and from recent preliminary paleoseismic results: Journal of Seismology, v. 11, no. 3, p. 311–326, doi: 10.1007/s10950-007-9054-2. Audemard, F.A., 2009, Key issues on the post-Mesozoic southern Caribbean plate boundary, in Gloaguen, R., and Ratschbacher, L., eds., Growth and Collapse of the Tibetan Plateau: Geological Society of London Special Publication 328, p. 569–586, doi:10.1144/SP328.23. Audemard, F.A., Romero, G., and Rendón, H., 1999, Sismicidad, Neotectónica y Campo de Esfuerzos del Norte de Venezuela: Caracas, Venezuela, Fundación Venezolana de Investigaciones Sismológicas report for PDVSACVP (Petróleos de Venezuela S.A.–Corporación Venezolana de Petróleo), 221 p. Audemard, F.A., Machette, M., Cox, J., Hart, R., and Haller, K., compilers, 2000, Map and Database of Quaternary Faults in Venezuela and Its Offshore Regions: U.S. Geological Survey Open-File Report 00-0018, 79 p. + map. Audemard, F.A., Romero, G., Rendón, H., and Cano, V., 2005, Quaternary fault kinematics and stress tensors along the southern Caribbean from microtectonic data and focal mechanism solutions: Earth-Science Reviews, v. 69, no. 3–4, p. 181–233, doi:10.1016/j.earscirev.2004.08.001. Audemard, F.A., Beck, C., Moernaut, J., De Rycker, K., De Batist, M., Sánchez, J., González, M., Sánchez, C., Versteeg, W., Malavé, G., Schmitz, M., Van Welden, A., Carrillo, E., and Lemus, A., 2007, La depresión de Guaracayal, estado Sucre, Venezuela: Una cuenca en tracción que funciona como barrera para la propagación de la ruptura cosísmica: Interciencia, v. 32, no. 11, p. 735–741. Baumbach, M., Grosser, H., Romero, G., Rojas, J., and Sobiesiak, M., 1999, Aftershock studies of the July 9, 1997, Cariaco earthquake: American Geophysical Union 1999 Spring Meeting, Boston: Eos (Transactions, American Geophysical Union), v. 80, no. 17, supplement S226, abstract S42A-01. Baumbach, M., Grosser, H., Romero, G., Rojas, J., Sobiesiak, M., and Welle, W., 2004, Aftershock pattern of the July 9, 1997, Mw=6.9 Cariaco earthquake in northeastern Venezuela: Tectonophysics, v. 379, p. 1–23, doi:10.1016/j.tecto.2003.10.018.
Bell, J., 1972, Geotectonic evolution of the southern Caribbean area, in Shagam, R., Hargraves, R.B., Morgan, W.J., Van Houten, F.B., Burk, C.A., Holland, H.D., and Hollister, L.C., eds., 1972, Studies in Earth and Space Sciences: A Memoir in Honor of Harry Hammond Hess: Geological Society of America Memoir 132, p. 369–386. Beltrán, C., compiler, 1993, Mapa Neotectónico de Venezuela: Caracas, Venezuela, Fundación Venezolana de Investigaciones Sismológicas, scale 1:2,000,000. Beltrán, C., 1994, Trazas activas y síntesis neotectónica de Venezuela a escala 1:2,000,000, in Proceedings, 7th Congreso Venezolano de Geofísica: Caracas, p. 541–547. Beltrán, C., and Giraldo, C., 1989, Aspectos neotectónicos de la región nororiental de Venezuela, in VII Congreso Geológico Venezolano: Caracas, Venezuela, Barquisimeto, vol. 3, p. 999–1021. Beltrán, C., Singer, A., and Rodríguez, J.A., 1996, The El Pilar fault active trace (northeastern Venezuela): Neotectonic evidences and paleoseismic data, in Proceedings, 3rd International Symposium on Andean Geodynamics: Saint-Malo, France, p. 153–156. Beltrán, C., Rodríguez, J.A., Singer, A., and Rivero, C.A., 1999, La trinchera de Las Toscanas: Evidencias paleosismológicas de actividad reciente de la falla de El Pilar entre Casanay y Río Casanay, Venezuela, in Proceedings, Sixth Congreso Venezolano de Sismología e Ingeniería Sísmica: Mérida, Venezuela (CD-ROM). Centeno-Graü, M., 1940, Estudios Sismológicos: Caracas, Venezuela, Litografía del Comercio, 365 p. Freymueller, J.T., Kellogg, J.N., and Vega, V., 1993, Plate motions in the north Andean region: Journal of Geophysical Research, v. 98, p. 21,853– 21,863, doi:10.1029/93JB00520. Fundación Venezolana de Investigaciones Sismológicas (FUNVISIS), 1994, Estudio Neotectónico y de Geología de Fallas Activas de la Región Nororiental de Venezuela: Proyecto Intevep 92-175: FUNVISIS report for Intevep S.A., 3 volumes. FUNVISIS, 1997, Evaluación preliminar del sismo de Cariaco del 09 de Julio de 1997, Estado Sucre, Venezuela (versión revisada): Caracas, Venezuela, FUNVISIS, 123 p. + 5 appendices. FUNVISIS, IMME (Instituto de Materiales y Modelos Estructurales), UDO (Universidad de Oriente), Venezuelan Academy of Sciences, CAV (Colegio de Arquitectos de Venezuela) and CIV (Colegio de Ingenieros de Venezuela)–Sucre, 1997, The July 9, 1997, Cariaco, Eastern Venezuela Earthquake: EERI (Earthquake Engineering Research Institute) Special Earthquake Report: Learning from Earthquakes, October 1997, p. 1–8. Garrity, C., Hackley, P., and Urbani, F., 2004, Digital Shaded-Relief Map of Venezuela: http://pubs.usgs.gov/of/2004/1322 (accessed 25 August 2004). Giraldo, C., 1996, Hipótesis acerca del desplazamiento de la falla de El Pilar, Venezuela nororiental, in Proceedings, 8th Congreso Venezolano de Geofísica: Maracaibo, Venezuela, p. 387–392. Gómez, J.M., 1990, Historia de las Fortificaciones de Cumaná (1st ed.): Cumaná, Venezuela, Talleres de Impresos Oriente, 145 p. Grases, J., 1990, Terremotos Destructores del Caribe: 1502–1990 (1st ed.): Montevideo, Uruguay, Orcyt-Unesco, 132 p. Grases, J., Altez, R., and Lugo, M., 1999, Catálogo de Sismos Sentidos o Destructores, Venezuela, 1530–1998: Academia de Ciencias Físicas, Matemáticas y Naturales/Facultad de Ingeniería Universidad Central de Venezuela, Editorial Innovación Tecnológica: Caracas, Venezuela, 654 p. Jordan, T., 1975, The present-day motions of the Caribbean plate: Journal of Geophysical Research, v. 80, no. 4, p. 433–434, 439. Kaniuth, K., Drewes, H., Stuber, K., Temel, H., Hernández, J.N., Hoyer, M., Wildermann, E., Kahle, H.G., and Geiger, A., 1999, Position changes due to recent crustal deformations along the Caribbean–South American plate boundary derived from CASA GPS project [abs.]: Birmingham, UK, General Assembly of the International Union of Geodesy and Geophysics (IUGG), poster at Symposium G1 of International Association of Geodesy. Malfait, B., and Dinkelman, M., 1972, Circum-Caribbean tectonic and igneous activity and the evolution of the Caribbean plate: Geological Society of America Bulletin, v. 83, no. 2, p. 251–272, doi:10.1130/0016-7606 (1972)83[251:CTAIAA]2.0.CO;2. Metz, H., 1968, Geology of the El Pilar fault zone, State of Sucre, Venezuela, in Proceedings, 4th Caribbean Geological Conference: Trinidad, p. 293– 298. Minster, J., and Jordan, T., 1978, Present-day plate motions: Journal of Geophysical Research, v. 83, no. 5, p. 331–335, 354.
Trending across the El Pilar fault in northeastern Venezuela after the 1997 Cariaco earthquake Molnar, P., and Sykes, L., 1969, Tectonics of the Caribbean and Middle America regions from focal mechanisms and seismicity: Geological Society of America Bulletin, v. 80, p. 1639–1684, doi:10.1130/0016-7606 (1969)80[1639:TOTCAM]2.0.CO;2. Pérez, O., 1998, Seismological report on the Mw = 6.8 strong shock of 9 July 1997 in Cariaco, northeastern Venezuela: Bulletin of the Seismological Society of America, Short Notes, v. 23, no. 2, p. 101–106. Pérez, O., and Aggarwal, Y., 1981, Present-day tectonics of southeastern Caribbean and northeastern Venezuela: Journal of Geophysical Research, v. 86, p. 10,791–10,804, doi:10.1029/JB086iB11p10791. Pérez, O., Bilham, R., Bendick, R., Hernández, N., Hoyer, M., Velandia, J., Moncayo, C., and Kozuch, M., 2001, Velocidad relativa entre las placas del Caribe y Sudamérica a partir de observaciones dentro del sistema de posicionamiento global (GPS) en el norte de Venezuela: Interciencia, v. 26, no. 2, p. 69–74. Pindell, J., and Dewey, J., 1982, Permo-Triassic reconstruction of western Pangea and the evolution of the Gulf of Mexico/Caribbean region: Tectonics, v. 1, no. 2, p. 179–211, doi:10.1029/TC001i002p00179. Pindell, J., and Kennan, L., 2007, Cenozoic kinematics and dynamics of oblique collision between two convergent plate margins: The Caribbean– South America collision in eastern Venezuela, Trinidad and Barbados, in Kennan, L., Pindell, J., and Rosen, N., eds., The Paleogene of the Gulf of Mexico and Caribbean Basins; Processes, Events, and Petroleum Systems: Proceedings, 27th Bob F. Perkins Research Conference, Gulf Coast Section of the Society of Economic Paleontologists and Mineralogists: Houston, Texas, Society of Economic Paleontologists and Mineralogists, p. 458–553. Romero, G., Audemard, F.A., Rendón, H., and Orihuela, N., 2002, Mapa de Soluciones Focales de Sismos Sentidos en Venezuela y Regiones Vecinas entre 1957 y 2002: Edición Conmemorativa XXX Aniversario de FUNVISIS: Caracas, FUNVISIS, scale ~1:2,450,000. Schubert, C., 1984, Basin formation along Boconó–Morón–El Pilar fault system, Venezuela: Journal of Geophysical Research, v. 89, p. 5711–5718. Singer, A., and Audemard, F.A., 1997, Aportes de Funvisis al desarrollo de la geología de fallas activas y de la paleosismología para los estudios de amenaza y riesgo sísmico, in Grases, J., ed., Diseño Sismorresistente: Especificaciones y Criterios Empleados en Venezuela: Academia de las Ciencias Naturales, Matemáticas y Físicas Publicación Especial 33, p. 25–38.
157
Sobiesiak, M., Alvarado, L., and Vásquez, R., 2002, Sismicidad reciente del Oriente de Venezuela, in Proceedings, 11th Congreso Venezolano de Geofísica: Caracas, Venezuela, 4 p. (CD-ROM). Sobiesiak, M., Alvarado, L., and Vásquez, R., 2005, Recent seismicity in northeastern Venezuela and tectonic implications: Revista de la Facultad de Ingeniería de la Universidad Central de Venezuela, v. 20, no. 4, p. 43–52. Soulas, J.-P., 1986, Neotectónica y tectónica activa en Venezuela y regiones vecinas, in Proceedings, 6th Congreso Geológico Venezolano (1985): Caracas, Venezuela, v. 10, p. 6639–6656. Speed, R., Russo, R., Weber, J., and Rowley, K.C., 1991, Evolution of Southern Caribbean plate boundary, vicinity of Trinidad and Tobago: American Association of Petroleum Geologists Bulletin, v. 75, no. 11, p. 1789–1794. Stephan, J.-F., 1982, Evolution Géodinamique du Domaine Caraïbe, Andes et Chaîne Caraïbe sur la Transversale de Barquisimeto (Vénézuéla) [Ph.D. thesis]: Brest, France, Université de Bretagne Occidentale, 512 p. Sykes, L.R., McCann, W.R., and Kafka, A.L., 1982, Motion of Caribbean plate during last 7 million years and implications for earlier Cenozoic movements: Journal of Geophysical Research, v. 87, no. B13, p. 10,656– 10,676, doi:10.1029/JB087iB13p10656. Vignali, M., 1979, Estratigrafía y estructura de las cordilleras metamórficas de Venezuela Oriental (Peninsula de Araya–Paria e Isla de Margarita): Caracas, Venezuela, Escuela de Geología y Minería, Universidad Central de Venezuela: GEOS, v. 25, p. 19–66. Wadge, G., and Burke, K., 1983, Neogene Caribbean plate rotation and associated Central American tectonic evolution: Tectonics, v. 2, no. 6, p. 633– 643, doi:10.1029/TC002i006p00633. Weber, J., Dixon, T., DeMets, C., Ambeh, W., Jansma, P., Mattioli, G., Saleh, J., Sella, G., Bilham, R., and Pérez, O., 2001, GPS estimate of relative motion between the Caribbean and South American plates, and geologic implications for Trinidad and Venezuela: Geology, v. 29, no. 1, p. 75–78, doi:10.1130/0091-7613(2001)029<0075:GEORMB>2.0.CO;2. Weber, J., Saleh, J., Balkaransingh, S., Dixon, T., Ambeh, W., Leong, T., Rodríguez, A., and Miller, K., 2011, Triangulation-to-GPS and GPS-to-GPS geodesy in Trinidad, West Indies: Neotectonics, seismic risk, and geologic implications: Journal of Marine and Petroleum Geology, v. 28, no. 1, p. 200–211, doi:10.1016/j.marpetgeo.2009.07.010. MANUSCRIPT ACCEPTED BY THE SOCIETY 7 DECEMBER 2010
Printed in the USA
The Geological Society of America Special Paper 479 2011
Lake sediments as late Quaternary paleoseismic archives: Examples in the northwestern Alps and clues for earthquake-origin assessment of sedimentary disturbances Christian Beck* Institut des Sciences de la Terre (ISTerre), UMR CNRS (Centre national de la recherche scientifique) 5275, Université de Savoie, Le Bourget du Lac cedex, 73376, France
ABSTRACT The late Quaternary sedimentary fills of several lakes of the northwestern Alps are revealed to be possible paleoseismological “archives” in a moderately active seismotectonic region. The strongest historically reported events can be correlated with specific layers having textures that result from different processes, such as: (1) mass failures of subaqueous slope deposits (especially delta foresets) evolving into hyperpycnal currents influenced by seiche effects and/or multiple reflections on lake basin slopes; (2) in situ liquefaction and flowage; and (3) microfracturing. Based on identification of the sedimentary signature of a well-documented historical earthquake, the paleoseismic interpretation can be extrapolated back to 16,000 yr B.P. with reconstruction of time series and textural identification of slope failure–related turbidites (the most frequent earthquake signature). The obtained time series are compatible with historical seismicity in terms of recurrence interval. The sedimentological approach developed for moderately seismotectonic environments appears to be valid for other large lake basins undergoing high-magnitude earthquakes.
INTRODUCTION As combined seismology and seismotectonics progressively led to the development of concepts and models dealing with the “seismic cycle” or “stick-slip behavior,” it became essential to estimate recurrence time intervals (e.g., Burridge and Knopoff, 1967; Shimazaki and Nakata, 1980). For this purpose, long-lasting paleoseismic records along major active fault zones, representing time series longer than historical data sets, became essential data. Detailed studies, such as along the North Anatolian fault, have
presented time and space distributions of major events and their driving mechanisms (e.g., Ambraseys and Finkel, 1991; King et al., 1994; Armijo et al., 1999). In parallel, the effects of sudden earthquake-induced increases in pore-water pressure in unconsolidated sediments have been investigated both for geotechnical needs (Mulder and Cochonat, 1996; Mulder et al., 1994) and as potential means of recording ancient earthquakes. These investigations concern both subaqueous marine or lacustrine settings and continental areas involving groundwater. The connection between coseismic ruptures and/or liquefaction is investigated through natural outcrops or trenches studies and then is combined
*
[email protected] Beck, C., 2011, Lake sediments as late Quaternary paleoseismic archives: Examples in the northwestern Alps and clues for earthquake-origin assessment of sedimentary disturbances, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, p. 159–179, doi:10.1130/2011.2479(07). For permission to copy, contact
[email protected]. © 2011 The Geological Society of America. All rights reserved.
159
160
C. Beck
with modeling (e.g., Kuenen, 1958; Seilacher, 1984; Sieh, 1978, 1996; Obermeier, 1989; Obermeier et al., 1991; Tuttle and Seeber, 1991; McCalpin, 1996; Caselles et al., 1997; Moretti et al., 1999). Empirical relationships between three groups of parameters—magnitude/intensity, distance between affected sediments and epicentral area, sediment texture—have been established and are progressively being refined with new data (e.g., Allen, 1986; Pederneiras, 1991; Vittori et al., 1991; Audemard and De Santis, 1991; Hibsch et al., 1994). Concerning marine or lacustrine subaqueous processes, contributions from different geodynamic settings and periods (from Proterozoic to Holocene) have progressively converged to offer better interpretations of several types of sedimentary features and layering as earthquake signatures (e.g., Macar and Antun, 1950; Plaziat et al., 1988; Van Loon et al., 1995; Ringrose, 1989; BenMenahem, 1976; Beck et al., 1992; Piper et al., 1992; Roep and Everts, 1992; Syvitski and Schafer, 1996; Calvo et al., 1998; Pratt, 1994, 1998; Alfaro et al., 1997; Nakajima and Kanai, 2000). Biochemical effects of earthquake-induced fluid release also have been mentioned (e.g., Tweddle and Crossley, 1991). Estimations of paleo-intensities of recorded earthquakes have been proposed based on properties and volumes of remobilized sediments (e.g., Séguret et al., 1984; Hibsch et al., 1994; Rodríguez-Pascua et al., 2002, 2003). In addition, time distributions of paleoearthquakes have been proposed for extended periods (in Holocene and older accumulations), both in marine and in lacustrine settings (e.g., Adams, 1990; Sims, 1975; Marco and Agnon, 1995; Beck et al., 1996; Weidlich and Bernecker, 2004). In several cases, different kinds of evidence—surface faulting, rockslides, lake sediment disturbances—can be used together to characterize the same major event (Becker et al., 2002, 2005) and thus better assess the seismic origin. When widespread and visible in different outcrops of the same unit, disturbed layers (interpreted as seismites and thus representing an isochronous surface) may offer reliable criteria for lateral correlation (Beck et al., 1998). Repeated earthquake-induced liquefaction events have also been invoked by Francis (1971) to explain the geometry of terrigenous accumulations in a deep subduction trough. Because the seismic profiles he analyzed showed horizontal nondeformed layering in a structural position where horizontal shortening could be expected, Francis (1971) related this apparent anomaly to liquefaction and reworking of sediments during major earthquakes; he considered these process to be responsible for erasing previous structures within the sedimentary pile. Alternatives to an earthquake-induced origin have to be discussed for each individual case within its own context. Some of the most frequently contentious situations are: 1. Marine coastal deposits, where exceptional wave amplitudes are believed to induce repeated pore-water pressure increases (Molina et al., 1998), or where the tidal bore in estuarine settings has been suggested to form “ball-andpillow”–like structures (Tessier and Terwindt, 1994); and 2. Glacial-periglacial conditions, where cryoturbation produces convolute bedding as a result of glaciotectonic de-
formation of marine or lacustrine sediments, as well as ice-block falls and related waves or catastrophic failure of damming moraines affecting proglacial lacustrine sediments (Clague and Evans, 2000). In deep-marine realms such as the Central and Eastern Mediterranean Sea, thick Holocene homogeneous mud layers, called “unifites” (Stanley, 1981) or “homogenites” (Kastens and Cita, 1981; Cita and Rimoldi, 1997; Rebesco et al., 2000), have been related either to a purely sedimentary process (separation of suspended-load part of major turbidite) or to a major tsunami (with volcanic and/or seismic origin). The purpose of this paper is to describe deformed beds, homogenites, and turbidites in the Holocene section of lakes in the northwestern Alps in order to integrate the processes of seismically induced ground motions and seiches and distinguish them from “normal” sedimentary processes such as hemipelagictype fallout and flooding. Lacustrine sedimentation generally offers, with respect to shallow-marine environments, better accumulation continuities and higher chronological resolutions; with respect to the deep oceanic environment, lacustrine basins register, without latency, the evolution and events occurring in their own watershed, avoiding the time lags and diluted effects of a vast realm. Isolated marine marginal basins may also act as lakes (e.g., Gorsline et al., 2000). If considering seismic shocks as triggering mechanisms for gravity reworking, lacustrine fan deltas, and/or Gilbert deltas are usually suitable sites for mass wasting (e.g., Postma et al., 1988; Nemec, 1990). Although slope failures may initiate on very slightly dipping surfaces (less than 1°; Field et al., 1982), the delta foresets, with steep slopes (up to 20° to 25°), are particularly sensitive to cyclic ground accelerations, which can induce failures evolving into slumps and subsequent density (hyperpycnal) currents (e.g., Kelts and Hsü, 1980; Siegenthaler and Stürm, 1991). Large water mass oscillations, such as seiches, represent another specificity of lacustrine basins (generally steep); although seiches and internal waves may have a nonseismic origin (landslides, for example), several historical cases point to earthquake genesis of strong seiches (Siegenthaler et al., 1987). Huge hyperpycnal flows, with high kinetic energy, may follow complex paths on lake bottoms with successive reflections on basin edges, as described in a fjord by Pickering and Hiscott (1985). Since the correlation established by Sims (1973, 1975) between disturbed layers and surveyed recent earthquakes in an artificial lake, different lacustrine accumulations (in situ or outcropping) have been used to assess late Quaternary seismicity in active tectonic settings (e.g., Doig, 1985, 1991; Hempton and Dewey, 1983; El-Isa and Mustafa, 1986; Shiki et al., 2000). Starting with three detailed examples, the following sections are dedicated to discussing the different ways earthquake are recorded in three lakes or paleolakes undergoing moderate tectonics and seismic activities, and to extracting some criteria to improve paleoseismic assessment over long time intervals. First, I present one case where the earthquake origin of a specific layer
Lake sediments as late Quaternary paleoseismic archives is well documented; next, I discuss an extrapolation applied to a 16,000 yr succession for the same sedimentary and seismotectonic setting. Finally, I propose the use of microgranulometric criteria to assess the flood versus slump origin of some turbidites and homogenites, and their application to a highly seismic area. SIGNATURE OF A HISTORICAL EARTHQUAKE IN LAKE LE BOURGET The northwestern Alps foreland (Fig. 1) is considered to be still experiencing distal effects of alpine collision, resulting in both horizontal and vertical relative displacements. Based on seismological and geodetic surveys, detailed patterns of active faulting (including subsurface décollements, blind ramps, and deeper crustal thrusts) have been proposed (e.g., Thouvenot et al., 1990, 1998; Jouanne et al., 1994, 1995), underlining the importance of NW-SE left-lateral strike-slip offsets, such as those along the Vuache fault (e.g., the 1996 Epagny event; Thouvenot et al., 1998; Figs. 1 and 2). In parallel with this tectonic evolution, the last glaciations-deglaciation cycles (isotopic stages 1 and 2) contributed to develop large and overdeepened lacustrine basins represented either by outcropping formations or by still subaqueous accumulations. The cross section of Lake Le Bourget (Fig. 2) illustrates this mixed heritage of alpine tectonics (Miocene growth strata in front of ramps), glacial erosion, and interglacial lacustrine accumulation (remnants of the last major cycles). After a detailed high-resolution seismostratigraphic study of the post– Last Glacial Maximum (LGM) fill (Van Rensbergen, 1996; Van Rensbergen et al., 1999; Chapron, 1999; Figs. 3 and 4), a set of 38 short gravity cores has been dedicated to the analysis of recent sedimentation. Chronicles concerning the strongest historical earthquake (1822 event with a VII–VIII epicentral MSK [Medvedev/Sponheuer/Karnik] intensity; Rothé, 1946) mention significant impact on Lake Le Bourget, suggesting a seiche effect and huge degassing (Chapron et al., 1999). Thus, the current investigation was focused on the possible signature of this major seismic event. Characterization of the 1822 Earthquake in Lake Le Bourget Sediments According to surface and subsurface structural data, the NW-SE Col du Chat fault and Culoz fault (CCF and CF on Fig. 2) represent tier faults connected to N-S–trending ramps. Seismological data point out rather superficial displacements within the Mesozoic–Cenozoic sedimentary cover or at the basement-sediment interface (Thouvenot et al., 1998). These two faults completely or partly cross Lake Le Bourget (largest natural French lake); three high-resolution seismic surveys (700 km of sparker and boomer profiles; Van Rensbergen et al., 1999) show clear evidence of direct, continuous slight deformation of the post-LGM sediment pile, which locally reaches 200 m in thickness. Sections A, B, and C in Figure 4 illustrate this deformation; on section A, across a high terrigenous feeding
161
area, successive avulsion and deepening of channel-levee systems represent indirect effects of the Culoz fault activity. The Col du Chat fault (section C) locally displays flower structure– like deformation. In addition to this mean long-term continuous deformation, significant but moderate seismic activity, related to these faults, is recorded. Among historical data, the strongest event— February 1822 (VII–VIII MSK intensity)—was reported by the Archbishop of Chambéry (in Rothé, 1941). Among the mentioned consequences, a “several feet lake level rise” and “lake water boiling” were interpreted (Chapron et al., 1999, p. 87) as the combination of seiche effect and huge degassing (methane escape). Based on the macroseismic inquiry, the epicentral area was located close to the northern termination of Lake Le Bourget in the so-called Chautagne Swamp (Fig. 1). The recent sedimentary record was investigated through a set of 38 short (1-m-long) gravity cores, which represent a three- to six-century archive, if the present-day mean sedimentation rate is assumed. Chronological Constraints Because these deposits represent the interference of environmental (climatic, anthropogenic) imprint and seismotectonic instability, different proxies and textural/physical properties were measured using the shortest sampling intervals in order to deconvolve the sedimentary “signal.” The chronology of this high-resolution archive was based on radioactive decays (AMS [Accelerator Mass Spectrometry] 14C measurements on terrestrial plant fragments, nonsupported 210Pb, 137Cs from atmospheric nuclear weapon tests and Chernobyl event). Seasonal lamination (observed on split core surfaces and in thin sections of impregnated sediments) and the identification of major floods reported in historical chronicles provide additional control. The whole data set allows a quasi-annual precision in the upper part of the core (see Figs. 5 and 6) and a 5–7 yr resolution for the lower part (Chapron et al., 1999). All the cores retrieved from the deep lacustrine basin (Fig. 3) show the same lithostratigraphic succession, displayed in Figures 5 and 6. In particular, a conspicuous set of three clay-rich levels can be used for lateral correlation; they are the signature of historical floods of the Rhône River (around A.D. 1734). In the interval corresponding to the 1815–1830 period (using previously mentioned chronological arguments), all the cores show a specific whitish homogeneous mud (called “homogenite” in the following paragraphs, using the first word in a descriptive way), which is attributed here to the 1822 seismic event, according to the evidence described in the following. Specific Texture of the 1822 Homogenite The thickness (measured on split core) of the homogeneous clayey-silty layer reaches 15 cm, with a silty laminae at the base (grain size of core LDB-11-A, Fig. 5; close-up of core 2001-02, Fig. 6). Different measured parameters such as grain-size distribution, magnetic susceptibility, and composition show the same strong homogeneity along the whole layer (Fig. 5). Although
A
162
B
J U
LAUSANNE
M ol a ss ic
GENEVA
P
SS
P
Annecy
Lake Anterne
A
L
Bresse
grabe n
R
BERNE
in as
I N N E
Chambéry
R
LYON
N
Cenozoic Mesozoic Pre-Mesozoic basement
5 50 km
Epagny, 1996 ML 5.3
Annecy Lake Annecy Chautagne, 1822 MSK VII-VIII
Lake Le Bourget
N Chambéry 40 km Major active fault traces associated with historical seismicity
Figure 1. Simplified structural geological map of the northwestern Alps, and location of the studied lakes on satellite photograph.
Lake sediments as late Quaternary paleoseismic archives
163
6E Lake Léman Quaternary GENEVA Tertiary
“Subalpine” Mesozoic “Jurassian” Mesozoic Pre-Mesozoic igneous -metamorphic basement
A
V.F.
46 N RCMG Seismic Line LDB-91-20
R
WSW
ENE
U
Elev 1000 (m)
Lake Le Bourget
ANNECY
PP
J
C.F.
0
SS
Lake Annecy
AA
LL
Lake Le Bourget Aix les Bains
Late Glacial-Holocene lacustrine sediments
1 km
CCF
8 7
R
E.F.
Riss/Würm interglacial lacustrine sediments
6
CHAMBERY
E
NORTH
Würmian glacial sediments (morainal and lacustrine subglacial)
T
5 4
U
10 km Main active fault traces associated with historical seismicity
Middle-Upper Miocene marine siliciclastics ("molasse") Oligocene-early Miocene freshwater siliciclastics ("molasse") "Urgonian" limestones (Barremian-Aptian)
3 Triassic to Neocomian sediments
O
2
V.F. Vuache fault
C.C.F. Col du Chat fault
E.F. Epine fault
C.F. Culoz fault
Paleozoic-Precambrian continental basement 1
Figure 2. Simplified geological map of the northwestern Alps and structural setting of Lake Le Bourget. NW-SE left-lateral strike-slip faults represent active structures and locations of the last significant earthquakes. RCMG—Renard Center of Marine Geology.
A 2 km
B
LDB 7
LDB-17
(-120 m)
0
LDB-16/01-03A
LDB 17 0
(-130 m)
LDB 12
LDB 11 A
LDB 15
LDB 16
(-145 m)
(-146 m)
(-146 m)
(-143 m)
20 Depth in core (cm)
LDB-01-01/ LDB 10 LDB-11A LDB-12
N
40 1822 HMG
40
60
60
LDB-7 80 m 100 th a isob
C
80
Depth in core (cm)
20
LDB-15
Figure 3. Geometry of the 1822 homogenite and locations of seismic lines shown in Figure 4.
164
C. Beck
Figure 4. High-resolution seismic profiles across Lake Le Bourget, showing evidence of active deformation of late Quaternary fill (locations in Fig. 3). Figure is after Van Rensbergen (1996), Van Rensbergen et al. (1999), and Chapron (1999). No line drawing is added to these nonmigrated seismic sections; the superficial trace of the Culoz fault appears either as a neat rupture surface (section B) or as progressive migration, deformation, and subsidence of channel-levee systems (section A) in an area with high terrigenous supply. The Col du Chat fault (section C) also crosses an area with high terrigenous flux and is marked by flower structure–like deformations. These profiles also show mass wasting (sl). TWT—two-way traveltime.
Lake sediments as late Quaternary paleoseismic archives
165
Core LDB-12
0
Grain-size distribution (%) 100 1 0 50
CuKPa 2 3
Core LDB 15 0
4
Core LDB-11-A
magnetic lineation
40
40
60
60
Depth in core (cm)
1822 HMG
1822 HMG
40
15 20
2.8
25 30 35
Grain-size distribution evolution
Core LDB-0101/LDB 10 60
magnetic foliation
Skewness index 0.6 0.7 0.8 0.9
1882 HMG
20
20
Depth in core (cm)
Depth in core (cm)
20
Mean grain-size Sorting (μm) index 2 6 10 12 2.4 2.6
(-128 m)
lds 65
s
Depth in core (cm)
s
80 80
100
80 silt + very fine sand
70
m 1822 event
75
rlds
clay
Undrained shear strength
s
100
1.00 1.01 1.02 1.03 1.04 1.05
80
lds
lds: Laminated deep sediments rlds: Reworked laminated sediments s: Medium to coarse sand m: Marl
Figure 5. Specific textural and physical properties of the 1822 homogenite measured in cores and close-up of the correlated disturbed layers.
nondrained shear-strength tests were performed on the split cores with 10 cm spacing, the same profiles were observed to have an anomalously low value in the “homogenite.” I underline the anisotropy of magnetic susceptibility profile: the high magnetic foliation usually characterizes deeper sediments modified by compaction. In the present case, the particular value is attributed to specific particle architecture, basically phyllite content (20%–30% of the whole mud), and a specific settlement process is inferred. In addition to having a particular texture, the “homogenite” also shows a specific general geometry. Lateral Correlations, Three-Dimensional Distribution, and Origin of the Reworked Material Based on lateral correlation between cores (five of them represented in Fig. 3), the general geometry of the “homogenite” can be reconstructed. This deposit appears to be clearly “ponded” in the deepest part of the central basin and thins outward in all directions. Another characteristic is the increase—toward the lake edges—of the basal coarse laminae, thickening (a few millimeters) and coarsening (medium sand). This indicates an increase of bed-load transport and involved bottom-water current velocity. Close to the western slope, the “homogenite” is laterally
connected to a gravity-reworked layer (core LDB-0101/LDB 10; Figs. 3 and 5, close-up). The latter includes slumped deep laminated sediments (interflow deposits; Chapron, 1999), debris flow with mud clasts of shallow-water lime mud (bench deposit), and siliciclastic coarse sand also coming from shore deposits. The same three-dimensional (3-D) geometry (smoothing of deepest negative reliefs, lateral transition to large slumps) has been documented at a different scale, deeper in Lake Le Bourget fill (at ca. 12,000 yr B.P.); a 4-m-thick homogeneous layer was mapped based on high-resolution seismic reflection data (Van Rensbergen, 1996; Chapron et al., 1996; Van Rensbergen et al., 1999). These observations illustrate the problem of the origin of the “homogenite” material: either it represents complete in situ reworked deep sediment (liquefaction, resuspension) or, on the other hand, completely allochthonous material (distal part of a turbidite initiated by mass failure evolving into hyperpycnal current). The discussion summarized in Figure 7 favors the second interpretation, based on two arguments. If we consider the “homogenite” to be a previous in situ deposit, it represents a significant time interval of sedimentation (~1 mm/yr), and the corresponding part of the depth-in-core versus age curve keeps the same slope as the rest of the succession (Fig. 7, hypothesis 1,
1822 A.D. event
Core LDB-11-A
Depth below lake floor (cm)
90
80
70
60
50
40
30
20
al"
rm
Historical Rhodanian floods
Anthropic eutrophication (ca. 1950/2000 A.D.)
1733/1734 A.D.
ime nt
sed
Homogeneous mud (sustained suspended load) Bas al Bas (bed-lo silty lam ad s e of i e nae 182 2 A. g r e g a t i Lam on D. h ina omo ) ted gen "no ite
Core 2001-02
1m
B
0.04
Depth in core 0.08
0.12
0.16
0.20
Frequency
Hypothesis 1
Time
fe
0.04
0.08
0.12
Hypothesis 2
Time
fe:
0.16
0.20
Historical flood event
hmg: 1822 homogenite
Frequency Time series: illite/smectite ratio in clay fraction, from core LDB 15; 1 cm sampling interval.
fe
P r es en t
Figure 7. Two models for the origin of homogenite material (allochthonous versus in situ reworked), based on time-depth relations (A) and their influence on time-series analysis (B). According to hypothesis 1, the homogenite represents in situ reworking of previous deposits; time equivalent of the latter is thus included in the data. According to hypothesis 2, the homogenite represents an instantaneous addition of sediments; its time equivalent is zero, and the true depth-time curve is the dashed line. A regularly spaced (1 cm) sampling was done for mineralogical analysis; the time series obtained for this parameter was used for spectral analysis, following the two hypotheses (1—including all values; 2—eliminating the samples belonging to the homogenite interval). With hypothesis 2, spectral separation is improved.
A hmg Power spectra
10
Figure 6. Views of split short gravity cores (LDB-11-A and LDB 2001-02) showing deep basin sedimentation during the past five centuries and the 1822 homogenite.
Depth (cm)
hmg Power spectra
P r es en t
Depth in core
0
166
Lake sediments as late Quaternary paleoseismic archives graph A). If we consider the “homogenite” to be allochthonous material added during a short episode of discharge, the time equivalent is negligible, and the curve shifts toward younger ages (Fig. 7, hypothesis 2, graph A). Hypothesis 2 implies a different age for other dated layers (dashed part of the curve), and especially for the historical flood events (fe on Fig. 7, graph A); mean or instantaneous sedimentation rates also change. Hypothesis 2 appears to fit with the whole set of chronological constraints (Chapron, 1999). In parallel, paleoclimatic investigations were performed on the cores, in particular, a search of high-frequency fluctuations using spectral analysis (Chapron et al., 2002). For the latter, the most accurate depth-age curve is needed to avoid bias caused by thick instantaneous events; thus, I checked the influence of including versus eliminating the values from these “event layers.” Among the proxies used, the clay mineral content (X-ray diffraction analysis) was analyzed using the illite/smectite ratio in the clay fraction, with 1 cm sampling space. The results are presented here as an additional argument to discuss the two hypotheses (Fig. 7, graph B). According to hypothesis 1, the “homogenite” represents a mixture of sediments with different ages, and the corresponding part of the sampling is taken into account for a time series analysis. According to hypothesis 2, this sampling is taken out of the time series. The spectral analysis results (see also Chapron et al., 2002) show a better discrimination of main frequencies using hypothesis 2 and also favor the latter. Summary and Explanatory Model The particular layer analyzed and mapped in the recent fill of Lake Le Bourget is thus interpreted as the signature of the 1822 earthquake, based on the following criteria: 1. The age interval in which the “homogenite” is observed;
Sliding/slumping
A
(translational, rotational)
2. The lateral transition to an erosion surface with bed-load transport of deposits; 3. The lateral transition to slump and debris flows involving lime mud and clasts coming from shallow areas; and 4. The specific particle settling with high segregation of fine grained (suspended load) from coarser (silt, sand) material, which implies a long-lasting bottom-current effect independent of any large river flood. Apart from huge degassing, for which preserved sedimentary evidence was not found, the observations and measurements can be explained by the model in Figure 8. The earthquake is considered as triggering sediment failures that evolve into mass wasting or slumps, which turn into hyperpycnal (turbidity) currents with bed load and suspended load. Although only one slump is figured, several synchronous slumps occurred in different situations around the deep lake basin. The seismic shock is also supposed to induce an oscillatory movement of the whole lake water mass (seiche phenomena), which progressively dampens; this should be the major origin for long-lasting erosive (to-and-fro) bottom currents. An alternative is to consider this particular bottom circulation as related to several reflections on the steep slopes of the lake basin of a hyperpycnal flow with high kinetic energy (following the “contained turbidite” model from Pickering and Hiscott, 1985). Apart from this main 1822 event, the two whitish layers, which represent fine-grained marly turbidites within an upper eutrophicated level (Fig. 6, close-up of the upper part of core 2001-02), could be correlated to the 1958 and 1964 earthquakes. They had respective VI–VII and V–VI MSK intensities, and their epicentral areas were close (2 km) and below the lake (Chapron, 1999). The interplay between fluid escape and sediment failure has been recently discussed for Lake Le Bourget (Chapron et al., 2004).
Widespread degassing
Figure 8. Proposed mechanism for the 1822 homogenite genesis and associated phenomena. Earthquake-induced acceleration triggers slope failures and mass wasting, which evolve into turbidite (for the 1822 events, several coalescing coeval slumps have been detected). Synchronous earthquake-induced seiche effects increase segregation of the fine-grained fraction and sustain its suspension. Laterally, on slopes, the seiche-related to-and-fro bottom currents concentrate coarser particles in the bed load.
Fluidized suspension
Seiche effect
B
167
Progressively damped oscillation Concentration of "cloud" in deepest (axial) part
Homogenized deposit
Reworking of silty/sandy fraction
168
C. Beck
TIME DISTRIBUTION OF EARTHQUAKE-RELATED EVENTS ALONG A 16,000 YR RECORD IN LAKE ANNECY
Post-LGM Lacustrine Sedimentary Fill
Because Lake Annecy (location in Figs. 1 and 2) was elected by a multidisciplinary team for detailed paleoclimatic studies (Oldfield and Berthier, 2001; CLIMASILAC project), it benefited from a deep drilling/coring program. The site was chosen (Fig. 9) far from the main tributaries and where the nearby watershed did not show any superficial disturbance (landslide, rock avalanche). Previous detailed high-resolution seismostratigraphy (Van Rensbergen et al., 1998) contributed to optimize this location. Thus, a complete succession since the LGM-related till was retrieved (44 m long; Fig. 10).
Overlying latest till tongues, the lacustrine sediment pile reaches 150 m thickness, especially in the northern termination (Van Rensbergen, 1996; Van Rensbergen et al., 1998). A northward and westward thickening can be observed (seismic profile, Fig. 9) due to the combined effect of a major northern terrigenous feeding (Beck et al., 2001) and a vertical component of the Vuache fault activity (Figs. 1 and 2). The lithological content and the main sedimentation episodes are summarized in Figure 10. The whole lacustrine succession is made of fine-grained (clay and silt) material with variable carbonate content (up to 80%) and layering; the carbonate fraction is a mixture of terrigenous and bio-induced calcite (Manalt
sonogram axis NNW aligned po
ckmarks
SSE
WSW
30 ms TWT
ENE
1st multiple
1 km Portion of sonogram
Seismic-reflection profile
N
Figure 9. High-resolution seismic and side-scan sonar evidence for the impact of the active Vuache fault on Lake Annecy sedimentary fill; TWT—two-way traveltime.
Lake sediments as late Quaternary paleoseismic archives et al., 2001; Yu et al., 2001; Dearing et al., 2001). Following the detailed chronology (Beck et al., 1996; Brauer and Casanova, 2001; David et al., 2001), the true varves episode (between 31 m and 17.5 m) lasted from 16,440 to 15,900 cal. yr B.P.; the 10.2 m limit is dated back to 14,000 cal. yr B.P. In brief, the different analyzed and correlated proxies in Lake Annecy’s fill provide a precise “normal” sedimentation reconstruction, within which several kinds of “events” can be detected and dated. The respective sedimentation rates for the different sedimentary units allow a time log to be reconstructed (Beck et al., 1996; Brauer and Casanova, 2001; Manalt et al., 2001; see Fig. 10). Very few millimeter-size dropstones were observed in the lower pre–Late Glacial part of the core (below 31 m), indicating spring rafting of frozen surface fragments rather than true ice rafting starting from a glacier still in contact with the lake basin. Nevertheless, I did not take into account this oldest part of the record, in order to eliminate possible disturbances due to ice-block falls.
0 1
(water depth: 57 m)
cv-1
Bio-induced marls 10.2 cv-2 cv-3
17.5
cv-4 cv-5
N
mi ath
sob
CLIMASILAC DRILLING SITE
1 km
cv: close-up core view,
Depth below lake-floor (m)
50
Lake Annecy
Delta-dominated terrigenous filling Beginning of thermal stratification and lacustrine bioproduction
Cold lake varve s.s. terrigenous filling
cv-6 31
Subglacial to proglacial sedimentation
43.7
Till
displayed on Fig. 11
Figure 10. Simplified log of the deep CLIMASILAC (CLImat MAgnétisme SIsmicité LAC alpin) core; cv-1 to cv-6 are detailed core views shown in Figure 11.
169
Types of Sedimentary Events Tentatively Related to Major Earthquakes As for Lake Le Bourget, a possible long-term mean influence of the Vuache fault (left-lateral strike slip with a normal component) was investigated because its “onshore” main trace ends exactly at the northern termination of the lake (Philippe, 1996; Deville et al., 1994; Thouvenot et al., 1998; Fig. 1). This portion of the high-resolution seismic profile from Figure 9 shows evidence of progressive deformation related to the normal component of the Vuache fault. A complete side-scan sonar survey of Lake Annecy (performed before the 1996 M5.3 earthquake) shows a set of NNW-SSE–aligned pockmarks (Fig. 9) in the prolongation of the Vuache fault as mapped onshore. I relate these structures to fluid escape features due to seismic shocks, following the interpretation of Field et al. (1982). Concerning the seismic activity imprint, I first based my investigation on the conclusions from Lake Le Bourget, and added several other types of disturbances; examples of the whole are presented in Figure 11 (location in Fig. 10). The characterized “events” are: 1. (as in Lake Le Bourget) coarse-grain turbidites, debris flows involving coarse sand and marl clasts (cv-1, cv-3), and thin homogenites (cv-2); and 2. (not observed in Lake Le Bourget recent sediments) flowing structures (part of “ball-and-pillow” in situ liquefaction) (cv-4), and microfaulting (cv-5, cv-6). Time Distribution of Events Tentatively Related to Major Earthquakes Using the chronological constraints, these particular layers were plotted on a time log (Fig. 12A; Beck et al., 1996), which allows recurrence times intervals to be estimated (Fig. 12B) and time distribution to be analyzed (Fig. 12C). In Figure11, the clustered events (see cv-1) are plotted as a single event, and for Figure 10C, the upper part of the time series (with several thousands of years intervals) has not been taken into account. If an earthquake origin is extrapolated back to 16,000 yr B.P. for all these sedimentary “events,” three main points of conclusion may be discussed (see also Beck et al., 1996): 1. There is an abundance of disturbances during the first thousands of years after the deglaciation and lake genesis, but these disturbances are also synchronous with the highest sedimentation rates (involving terrigenous input), and this leads me to envisage interferences between an increased seismotectonic instability (due to rapid disappearance of ice loading, as mentioned by Mörner, 1996, for Fennoscandia) and the high amount of poorly consolidated sediments. 2. The dominant recurrence time interval is compatible with the historical data (taking into account M5–5.5 earthquakes at 50 km maximum distance from the lake; data from Lambert et al., 1996; Levret et al., 1996).
C. Beck
10 cm
10 cm
170
10 cm
hmg
cv-2
cv-3
Gravity-reworking events
Figure 11. Examples of sedimentary disturbances and depositional “events” in CLIMASILAC core.
cv-1 Microfractures Liquefaction
cv-6
cv-5
10 cm
10 cm
10 cm
cv-4
3. The cumulative distribution has an increasing duration roughly following a power law. In a recent study of lacustrine seismites, Rodríguez-Pascua et al. (2003) analyzed the time distribution of paleo-intensities (deduced from disturbance volumes) and concluded that there were strong similarities with seismologic historical data. Because gravity-reworking processes represent the major type of layers taken into account in the previous example, the latter raise the problem of the origin of turbidites. They may be all attributed to subaqueous failures of nonconsolidated sediments (and thus all tentatively triggered by earthquakes; Adams, 1990); nevertheless, turbidity (density) currents were first described and
defined by F.A. Forel in Lake Geneva, in relation to a strong, rather long-lasting, flood episode of the main tributary (Rhône River) through a fan delta. In the following, I attempt to decipher the two origins: “flood turbidites” versus “slump turbidites.” DISCRIMINATING FLOOD TURBIDITES FROM SLUMP-INDUCED TURBIDITES: DESTABILIZATION OF LACUSTRINE DELTAS AND HUGE RUNOFF IN A SMALL, HIGH-ALTITUDE LAKE Lake Anterne (location in Fig. 1) was selected both for paleoenvironmental and paleoseismic purposes, with respect to
Lake sediments as late Quaternary paleoseismic archives
171
B
including shortest intervals between clustered events
Number of intervals
0
2 5
4
14 12 10 8 6 4 2
0
6
100
200
300
400
500
600
20
00
40
00
60
00
Time (k.y. B.P.)
Recurrence time interval (yr) 100
8
A
12
C
50
2 2
14
the two largest intervals not plotted)
n/N (%)
10
N=40 (clustered events represented by a single point,
2 3
2
0
100
2
200
300 400 Recurrence time interval (yr)
500
Single event 2
Clustered events graphically grouped
Figure 12. Time distribution of the sedimentary “events” observed in the CLIMASILAC core. (A) Distribution along a time log; (B) histogram of time intervals between two successive events; and (C) cumulative distribution in increasing duration, roughly following a power law. The time intervals between successive “events” (such as the ones displayed in Fig. 11) were measured using seasonal lamination and radiocarbon dating (Beck et al., 1996; Brauer and Casanova, 2001). Circled numbers on graph refer to numbers of clustered events too tightly spaced to show individually. Some of these clusters (cv-1 from Fig. 11, for example) may represent a composite unique reworking.
its altitude (2000 m) and its sedimentary fill (exclusively terrigenous, thus related to superficial runoff; Lignier, 2001; Arnaud et al., 2002). As simplified in Figure 13, it represents a small simple lacustrine basin (12 m depth, 500 m wide), the southeastern boundary of which is a Gilbert delta with an almost horizontal topset and a steep foreset. Due to its altitude, it is frozen more than 6 mo out of the year, and its (seasonal) sedimentation is driven by spring–summer snowmelt and rain. A set of short cores was taken along a profile from the distal part to the delta foreset base. The latter cores show a rather chaotic succession with superimposed slumps, while the distal cores bear finely (millimetric) laminated sediments (clay and silt). In the central cores (ANT 99 02, 120 cm long), thick layers bearing normal graded bedding are intercalated within laminated sequences (Fig. 13); these thicker layers were all considered first to be turbidites in the broadest sense (i.e., the results of hyperpycnal flow deposits), although slight macroscopic differences could be observed.
Grain-Size Evolution in Lake Anterne’s Turbidites Because the mean grain size is generally smaller than 1 mm (medium sand), a detailed granulometric vertical evolution of each turbidite could be measured (with 2–5 mm spacing) using a MALVERNTM Mastersizer S laser microgranulometer. As for Lake Le Bourget and Lake Annecy’s short cores, the chronology was established with AMS 14C and 210Pb (Lignier, 2001; Arnaud et al., 2002) and seasonal lamination counting. The analyzed different layers appear to follow two distinct paths on several binary diagrams; the most discriminative are presented in Figure 14 (sorting [SO] vs. skewness [SK]) and Figure 15 (coarse 99th percentile vs. median; the Passega [1964] diagram). Both figures clearly show two distinct paths; the upper settlings of the suspended load appear superimposed on the SK/SO diagram, while they are distinct on Passega’s diagram. Lateral correlations established between several basal foreset slumps and turbidites
172
C. Beck Core ANT 99 02
10 cm
5 cm
"Flood turbidite"
"Slump turbidite"
"Normal" seasonal lamination
Core ANT 99 02
Major flooding event delta bypassing
Hyperpycnal flow
Delta foreset failure Hyperpycnal flow
Coring site
Figure 13. The two main origins of hyperpycnal flows and subsequent turbidite genesis in relation to a lacustrine Gilbert delta in Lake Anterne. (Location of short gravity cores and view of the three types of layering are also shown.)
1.2
0.8
0.6
Skewness index
1.0
Figure 14. Base-to-top textural evolution of the two types of turbidites detected in Lake Anterne fill. For two turbiditic layers, the granulometric spectra of regularly spaced samples (2 mm) are plotted as skewness versus sorting index values; measurements on laminated sediments are added for comparison. Two distinct base-to-top granulometric evolutions (summarized on insert) appear as two skewness versus sorting paths. The flood turbidite path ends out of the laminated sediments realm.
Vertical evolution of turbidites 1.5
2.5
4.0
0.4
Skewness
1.2 1.0
Slump turbidite Flood turbidite "Flood turbidite" "Slump turbidite"
0.4
Sorting Laminated interval area
Sorting index 1.5
2.0
2.5
3.0
3.5
4.0
4.5
Lake sediments as late Quaternary paleoseismic archives
173
1000
C99 (μm)
500
"Flood turbidite" "Slump turbidite"
(
Laminated interval Vertical evolution of turbidites Slump turbidite L a m in a te in te rv a d l
800
Figure 15. Grain-size base-to-top evolution of the two types of turbidites plotted on Passega’s (1964) diagram. Plotted values correspond to the same samples as those shown in Figure 14. The two layers (flood turbidite and slump turbidite) also appear as distinct base-to-top paths.
Flood turbidite
100 80
C
=M
50
30 M=1
M=10
M (C50) (μm)
from the central part (core ANT 99 02), and comparison between delta topset deposits and coarse base of turbidites lead me to relate one type of textural evolution to slope failure of the delta foreset, that is to say to “slump turbidites.” Their base is slightly coarser and less sorted; macroscopically (Fig. 13), the transition from coarse base to fine-grained top is shorter. Within annual normal laminae, the granulometric contrasts (Figs. 14 and 15) appear reduced; as for “flood turbidites,” they are related to concentration of watershed runoff in narrow channels bypassing the delta topset and foreset. In “normal” sedimentation, the distribution within the lake basin is produced by overflows/interflows and subsequent decantation (cold, nonthermally stratified, lake), while underflows (hyperpycnal currents allowed by coarser bed load and denser suspended load) explain the genesis of “flood turbidites.” Starting from the discrimination of flood turbidites from turbidites related to mass wasting evolving in a density current, the chronology of these different “events” may be discussed (Fig. 16A), assessing a seismic origin for delta foreset failures. Comparison of Time Distribution of Slump Turbidites with Local Historical Seismicity Considering earthquake-induced ground acceleration as the most frequent driving mechanism for soil-liquefaction and/or failure of nonconsolidated subaqueous sediments (Obermeier, 1989; Piper et al., 1992), I compared the four characterized “slump turbidite” occurrences with regional historical seismicity. With respect to distance between Lake Anterne and epicentral areas, and to earthquake intensities (following Vittori et al., 1991), four
M=100
reported events appear compatible with the analyzed site. Taking into account the chronological precision of the sedimentary archive (±10 yr), these four events can fit with the four “slump turbidites” (historical earthquakes [HE] on Fig. 14A): Emosson (1905; MSK VII–VIII; 15 km), Visp (1855; MSK IX; 65 km), Chamonix (1817; MSK VII; 10 km); and Brig (1755; MSK VIII– IX; 70 km). These characterized “slump turbidites” thus are considered to be the respective traces of these four regional major events (Lignier, 2001; Arnaud et al., 2002). Although they concern a short duration (300 yr), the results obtained for Lake Anterne are considered to validate both the sedimentological (vertical textural evolution, transport, and settling processes) and the paleoseismological approaches. They were obtained from a simple sedimentary system with rather constant mean sediment accumulation. The interplay between main parameters driving the quality of the archive is sketched in Figure 16B; competition between frequency and size of slumps, and deltaic growth velocity may lead to an incomplete paleoseismic record (minimum paleoseismic activity). TENTATIVE APPLICATION TO A LAKE UNDERGOING HIGH-MAGNITUDE EARTHQUAKES: LAKE ISSYK-KUL, KYRGYZSTAN Before extracting some general conclusions about the northwestern Alps study sites, where the present-day seismotectonic activity is considered to be moderate, I will briefly summarize an application of these investigations to a highly different case study: a very large and deep lake within strong seismotectonic activity.
174
C. Beck
2000
1900
Age A.D. 1800
1700
0 Interplay of deltaic growth velocity with slope failures frequency
t1
Σ Δ Sl
20 "Slump turbidites"
Th(t)
t0
HE
Depth in core (cm)
40
dTh
Deltaic growth velocity
dTh dt
A 60 HE
ΔSl
80
HE
ΔSl=
f (sediment properties, horizontal and vertical acceleration)
100 HE
B 120
"Flood turbidites" Lake Anterne, Core 99 02
Figure 16. Depth distribution of inferred major floods and earthquake-related gravity-reworking events in Lake Anterne during the past three centuries. (A) Detailed time-depth curve; and (B) simplified delta foreset evolution undergoing seismicity. HE—reported major historical seismic event. The thickness of each slump turbidite is a function of the initial mass-wasting dimensions; the later ones depend on the sediment texture and on the earthquake characteristics.
An attempt is made to validate the different results obtained in the northwestern Alps (moderate seismicity and rather small size lakes) by applying a similar approach to Lake IssykKul, located within the Tien Shan mountain system. There, high-magnitude historical earthquakes are frequent (up to M8; Abdrakhmatov et al., 2002). Lake Issyk-Kul is 178 km long, 60 km wide, and has a 668 m maximum depth; it was investigated within the frame of a Belgian-Kyrgyz-Russian scientific cooperation project. A detailed high-resolution seismic reflection survey (De Batist et al., 2002) showed sedimentation influenced by huge gravity reworking (mass wasting, turbidites), especially in the central deep basin (Fig. 17), and also by local bottom currents. The survey also shows evidence of active deformation of the late Quaternary pile and interferences between the last glaciation-deglaciation cycles and the active seismotectonic settings (De Batist et al., 2002). Twenty-two short gravity cores (with 180 cm maximum length) were retrieved in different environments with respect to the bathymetry and the main tributary systems. The most continuous, well-layered sedimentation was found on the northern
“margin” (Giralt et al., 2002) in a situation of interflow-overflow deposits, with a mean sedimentation rate close to 1 mm/yr and mean bio-induced carbonate content close to 50%. On the other hand, the cores retrieved in the deep central basin and at the foot of the western delta foreset show highly disturbed and discontinuous sedimentation (Fig. 17). Core IK-98-19a (120 cm) seemed to be entirely made of a unique “homogenite.” Tentative correlations were proposed with core IK-98-08a (Lignier, 2001), which showed successive slumped layers mixing medium to coarse sand with marl. A detailed textural analysis of core IK-98-19a (Lignier, 2001) was performed; the grain-size analysis (2 mm sampling interval; 510 measurements) results are plotted on an SK versus SO diagram (base-to-top evolution) (Fig. 17). The mean grain-size vertical profile shows (as the magnetic susceptibility) a clear “break” between coarse sandy base (bed-load displacement) and homogeneous fine-grained mud (sustained suspended load), similar to the one observed in Lake Le Bourget. The SO versus SK path shows several “oscillations,” which I interpret as to-and-fro bed-load displacement. Either seiche effects (direct
60
50
40
30
120
110
100
90
80
70
60
50
40
30
20
10
10
20
0
hmg
50
0
100
0
77°E
IK-98-08a
0m
IK-98-19a
th 60
0m
Mean grain size (µm) 40 60 80 100
feeding
isoba
th 30
Sampling spacing: 2 mm
20
Major deltaic
isoba
Skewness index
0.5
1.0
1.5
2
78°E
2
Top
3
Homogenite level
(520 measurements)
4 Sorting index
Base
5
Skewness index vs Sorting index path (base to top) for core IK-98-19a
Lake Issyk-Kul, Kyrgyzstan, Tien Shan
6
Figure 17. Specific texture of a major earthquake-related deposit (1-m-thick homogenite) in Lake Issyk-Kul deep basin (Tien Shan, Kyrgyzstan). Mean grain-size vertical evolution is plotted as skewness versus sorting path; it differs from Lake Anterne’s plot (Fig. 14) due to the occurrence of a complex coarse base. Laterally, at the toe of the feeding edge (core IK-98-08a), it is represented by a thin homogeneous layer overlying a sharp erosion surface and mud clasts.
Depth in core (cm)
0
42°30' N
Depth in core (cm)
50 km
175
176
C. Beck
consequence of a major earthquake) and/or a hyperpycnal current reflected several times (cf. Pickering and Hiscott, 1985) may explain the specific textural evolution of the coarse fraction. High kinetic energy of such hyperpycnal currents (few m/s) and the shape of the central basin created this particular settling (Lignier, 2001). The mean normal (background) sedimentation rate allowed the earthquake responsible for the “homogenite” to be correlated with one of the two major historical events (1889 or 1911, with 8 and 8.7 respective estimated magnitudes; Abdrakhmatov et al., 2002). The volume of the implied reworked sediments (homogenite plus basal sand) is considered to be directly related to the high magnitude/intensity of the triggering seismic event (Lignier, 2001); lateral correlations among the whole set of cores point to a western origin (failure and mass wasting on the western deltaic foreset). CONCLUSIONS AND PERSPECTIVES According to the few examples detailed here, added to previous works published by different authors, the “paleoseismological” use of lacustrine basin sedimentary fills appears to be a useful tool for paleoseismological investigations, requiring several preliminary precautions: 1. A high chronological resolution in the possible different coring settings; 2. A detailed knowledge of transport and settling mechanisms (pure sedimentary processes, including reworking by nonseismic surface water waves; and earthquakeinduced processes) through analysis of textures and physical properties; and 3. Investigation of the lacustrine basin geometry, which plays a major role in water movement and in density current trajectories. Going into a more “paleoseismological” approach implies: 1. Estimation of the sizes of paleofailures and destabilized volumes (Séguret et al., 1984; Lignier, 2001); 2. Estimation of the thickness of disturbed bottom sediments (Hibsch et al., 1994; Lignier, 2001; RodríguezPascua et al., 2002); and 3. Use of several lakes that recorded the same event in order to estimate a paleo-epicentral area (Lignier, 2001). Similar investigations, in progress, have been dedicated to the Marmara Sea, within international projects concerning the seismic hazards along the North Anatolian fault (Armijo et al., 1999; Mercier de Lépinay et al., 2003). They have yielded preliminary results concerning the activity of the last 20,000 yr (Beck et al., 2007). The sedimentological approach presented here has also been conducted on moraine-dammed lakes crossed by a major active fault in the Mérida Andes (Carrillo et al., 2006, 2008). ACKNOWLEDGMENTS Investigations on the potentialities for alpine lake sediments to register seismotectonic activity began and were developed
thanks to the Renard Center of Marine Geology (University of Ghent); I wish to greatly thank Marc De Batist, Pieter Van Rensbergen, and Jean-Pierre Henriet. Through Ph.D. advising and analytical facilities, the Laboratory of Sedimentary Dynamics (University of Lille) also largely contributed to this research; I am grateful to Hervé Chamley, Jean-François Deconinck, Nicolas Tribovillard, Alain Trenteseaux, and Philippe Recourt. The Laboratory of Glaciology and Environmental Geophysics (University of Grenoble) and the European Centre of Environmental Research (CEREGE, University of Aix-Marseille) contributed by means of radiogenic and magnetic measurements; I acknowledge Michel Pourchet and Nicolas Thouveny for these precious contributions. Side-scan sonar surveys were planned and performed thanks to Pierre Cochonat and Claude Augris at IFREMER (Institut Français de Recherche pour l’Exploitation de la Mer). Financial support was provided by national and regional agencies, including CNRS (Institut National de Sciences de l’UniversINSU—programs, and ISTerre internal funding), Rhône-Alpes Regional Council, and Savoie and Haute Savoie Departmental Councils and Natural Parks. REFERENCES CITED Abdrakhmatov, K.E., Janzakov, K.D., and Delvaux, D., 2002, Active tectonics and seismic hazard of the Issyk-Kul basin in the Krygyz Tian-Shan, in Klerkx, J., and Imanackunov, B., eds., Lake Issyk-Kul: Its Natural Environment: NATO Science Series IV, Volume 13: Dordrecht, the Netherlands, Kluwer Academic Publishers, p. 147–160. Adams, J., 1990, Paleoseismicity of the Cascadian subduction zone: Evidence from turbidites off the Oregon-Washington margin: Tectonics, v. 9, no. 4, p. 569–583, doi:10.1029/TC009i004p00569. Alfaro, P., Moretti, M., and Soria, J.M., 1997, Soft-sediment deformation structures induced by earthquakes (seismites) in Pliocene lacustrine deposits (Guadix-Baza Basin, central Betic Cordillera): Eclogae Geologicae Helvetiae, v. 90, p. 531–540. Allen, J.R.L., 1986, Earthquake magnitude-frequency, epicentral distance, and soft sediment deformation in sedimentary basins: Sedimentary Geology, v. 46, p. 67–75, doi:10.1016/0037-0738(86)90006-0. Ambraseys, N.N., and Finkel, C.F., 1991, Long-term seismicity of Istanbul and the Marmara Sea region: Terra Nova, v. 3, p. 527–539, doi:10.1111/ j.1365-3121.1991.tb00188.x. Armijo, R., Meyer, B., Hubert, A., and Barka, A., 1999, Westward propagation of the North Anatolian fault into the northern Aegean: Timing and kinematics: Geology, v. 27, no. 3, p. 267–270, doi:10.1130/0091 -7613(1999)027<0267:WPOTNA>2.3.CO;2. Arnaud, F., Lignier, V., Revel, M., Desmet, M., Beck, C., Pourchet, M., Charlet, F., Trentesaux, A., and Tribovillard, N., 2002, Flood and earthquake disturbance of 210Pb geochronology (Lake Anterne, NW Alps): Terra Nova, v. 14, no. 4, p. 225–232. Audemard, F., and De Santis, F., 1991, Survey of liquefaction structures induced by recent moderate earthquakes: Bulletin of Engineering Geology and the Environment, v. 44, p. 5–16. Beck, C., Rochette, P., and Tardy, M., 1992, Interprétation en termes de paléoséismicité de niveaux destructurés dans des rythmites lacustres quaternaires des Alpes nord-occidentales: Comptes Rendus de l’Académie des Sciences Paris, v. 315, no. II, p. 1525–1532. Beck, C., Manalt, F., Chapron, E., Van Rensbergen, P., and De Batist, M., 1996, Enhanced seismicity in the early post-glacial period: Evidence from the post-Würm sediments of Lake Annecy, northwestern Alps: Journal of Geodynamics, v. 22, no. 1/2, p. 155–171, doi:10.1016/0264 -3707(96)00001-4. Beck, C., Deville, E., Blanc, E., Philippe, Y., and Tardy, M., 1998, Horizontal shortening control of middle Miocene marine siliciclastic accumulation (Upper Marine Molasse) in the southern termination of the Savoy
Lake sediments as late Quaternary paleoseismic archives Molasse Basin (northwestern Alps/southern Jura), in Masle, A., Puigdefàbregas, C., Luterbacher, H.P., and Fernández, M., eds., Cenozoic Foreland Basins of Western Europe: Geological Society of London Special Publication 134, p. 263–278. Beck, C., Van Rensbergen, P., De Batist, M., Berthier, F., Lallier, S., and Manalt, F., 2001, The Late Quaternary infill of Lake Annecy (northwestern Alps): An overview from two seismic-reflection surveys: Journal of Paleolimnology, v. 25, no. 2, p. 149–161, doi:10.1023/A:1008175728826. Beck, C., Mercier de Lépinay, B., Schneider, J.-L., Cremer, M., Çağatay, N., Wendenbaum, E., Boutareaud, S., Ménot, G., Schmidt, S., Weber, O., Eris, K., Armijo, R., Meyer, B., Pondard, N., Gutscher, M.-A., Turon, J.-L., Labeyrie, L., Cortijo, E., Gallet, Y., Bouquerel, H., Gorur, N., Gervais, A., Castera, M.-H., Londeix, L., de Rességuier, A., and Jaouen, A., 2007, Late Quaternary co-seismic sedimentation in the Sea of Marmara’s deep basins, in Bourrouilh-Le Jan, F., Beck, C., and Gorsline, D., eds., Sedimentary Records of Catastrophic Events: Sedimentary Geology, v. 199, Special Issue, p. 65–89. Becker, A., Davenport, C.D., and Giardini, D., 2002, Palaeoseismological studies on end-Pleistocene and Holocene around Basel, Switzerland: Geophysical Journal International, v. 149, p. 659–678, doi:10.1046/j .1365-246X.2002.01678.x. Becker, A., Ferry, M., Schnellmann, M., and Giardini, D., 2005, Multiarchive paleoseismic record of late Pleistocene strong earthquakes in Switzerland: Tectonophysics, v. 400, p. 153–177, doi:10.1016/j.tecto.2005.03.001. Ben-Menahem, A., 1976, Dating historical earthquakes by mud profiles of lakebottom sediments: Nature, v. 262, p. 200–202, doi:10.1038/262200a0. Brauer, A., and Casanova, J., 2001, Chronology and depositional processes of the laminated sediment record from Lake d’Annecy, French Alps: Journal of Paleolimnology, v. 25, no. 2, p. 163–177, doi:10.1023/A:1008136029735. Burridge, R., and Knopoff, L., 1967, Model and theoretical seismicity: Bulletin of the Seismological Society of America, v. 57, p. 341–371. Calvo, J.P., Rodriguez-Pascua, M., Martin-Velasquez, S., Jimenez, S., and De Vicente, G., 1998, Microdeformation of lacustrine laminite sequences from late Miocene formations of SE Spain: An interpretation of loop bedding: Sedimentology, v. 45, p. 279–292, doi:10.1046/j.1365-3091.1998 .00145.x. Carrillo, E., Audemard, F., Beck, C., Cousin, M., Jouanne, F., Cano, V., Castilla, R., Melo, L., and Villemin, T., 2006, A late Pleistocene natural seismograph along the Boconò fault (Mérida Andes, Venezuela): The morainedammed Los Zerpa paleo-lake: Bulletin of the French Geological Society, v. 177, no. 1, p. 3–17, doi:10.2113/177.1.3. Carrillo, E., Beck, C., Audemard, F.A., Moreno, E., and Ollarves, R., 2008, Disentangling Late Quaternary climatic and seismotectonic controls on Lake Mucubají sedimentation (Mérida Andes, Venezuela), in De Batist, M., and Chaprorn, E., eds., Lake Systems: Sedimentary Archives of Climate Change and Tectonics: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 259, Special Issue, p. 284–300. Caselles, J.O., Moretti, M., Alfaro, P., Canas, J.A., and Clapès, J., 1997, Estructuras sedimentarias de deformaciòn (sismitas) inducidas por licuefacciòn con un simulador de terremotos: Geogaceta, Sociedad Geológica España, v. 21, p. 67–70. Chapron, E., 1999, Contrôles Climatique et Sismo-Tectonique de la Sédimentation Lacustre dans l’Avant-Pays Alpin (Lac du Bourget) durant le Quaternaire Recent: Université Joseph Fourier/Grenoble 1, Géologie Alpine, Mémoire Hors-Série 30, 261 p. Chapron, E., Van Rensbergen, P., Beck, C., De Batist, M., and Paillet, A., 1996, Lacustrine sedimentary records of brutal events in Lake Le Bourget (northwestern Alps–southern Jura): Quaternaire, v. 7, no. 2/3, p. 155–168, doi:10.3406/quate.1996.2068. Chapron, E., Beck, C., Pourchet, M., and Deconinck, J.-F., 1999, 1822 AD earthquake-triggered homogenite in Lake Le Bourget (NW Alps): Terra Nova, v. 11, p. 86–92, doi:10.1046/j.1365-3121.1999.00230.x. Chapron, E., Desmet, M., De Putter, T., Loutre, M.F., Beck, C., and Deconinck, J.F., 2002, Climatic variability in the northwestern Alps, France, as evidenced by 600 years of terrigenous sedimentation in Lake Le Bourget: The Holocene, v. 12, p. 177–185, doi:10.1191/0959683602hl520rp. Chapron, E., Van Rensbergen, P., De Batist, M., Beck, C., and Henriet, J.-P., 2004, Fluid-escape features as a precursor of large sub-lacustrine sediment slide in Lake Le Bourget, NW Alps, France: Terra Nova, v. 16, no. 5, p. 305–311, doi:10.1111/j.1365-3121.2004.00566.x. Cita, M.B., and Rimoldi, B., 1997, Geological and geophysical evidence for a Holocene tsunami deposit in the Eastern Mediterranean deep-sea record:
177
Journal of Geodynamics, v. 24, no. 1/4, p. 293–304, doi:10.1016/S0264 -3707(96)00030-0. Clague, J.J., and Evans, S., 2000, A review of catastrophic drainage of morainedammed lakes in British Columbia: Quaternary Science Reviews, v. 19, p. 1763–1783, doi:10.1016/S0277-3791(00)00090-1. David, F., Farjanel, G., and Jolly, M., 2001, Palyno- and chronostratigraphy of a long sequence from Lac d’Annecy (northern outer Alps, France): Journal of Paleolimnology, v. 25, no. 2, p. 259–269, doi:10.1023/A:1008142810171. Dearing, J.A., Hu, Y., Doody, P., James, P.A., and Brauer, A., 2001, Preliminary reconstruction of sediment-source linkages for the past 6000 yrs at the Petit Lac d’Annecy, France, based on mineral magnetic data: Journal of Paleolimnology, v. 25, no. 2, p. 245–258. De Batist, M., Imbo, Y., Vermeesch, P., Klerkx, J., Giralt, S., Delvaux, D., Lignier, V., Beck, C., Kalugin, I., and Abdrakhmatov, K.E., 2002, Bathymetry and sedimentary environments of Lake Issyk-Kul, Kyrghyz Republic (Central Asia): A large, high-altitude, tectonic lake, in Klerkx, J., and Imanackunov, B., eds., Lake Issyk-Kul: Its Natural Environment: NATO Science Series IV, Volume 13: Dordrecht, the Netherlands, Kluwer Academic Publishers, p. 101–124. Deville, E., Blanc, E., Tardy, M., Beck, C., and Ménard, G., 1994, Thrust propagation and syntectonic sedimentation in the Savoy Tertiary Molasse Basin (Alpine foreland), in Mascle, A., ed., Exploration and Petroleum Geology in France: European Association of Petroleum Geologists Special Publication 4, p. 269–280. Doig, R., 1985, A method for determining the frequency of large-magnitude earthquakes using lake sediments: Canadian Journal of Earth Sciences, v. 23, p. 930–937. Doig, R., 1991, Effects of strong seismic shaking in lake sediments, and earthquake recurrence interval, Temiscaming, Québec: Canadian Journal of Earth Sciences, v. 28, p. 1349–1352. El-Isa, Z.H., and Mustafa, H., 1986, Earthquake deformations in the Lisan deposits and seismotectonic implications: Geophysical Journal of the Royal Astronomical Society, v. 86, p. 413–424. Field, M.E., Gardner, J.V., Jennings, A.E., and Edwards, B.D., 1982, Earthquake-induced sediment failure on a 0.25° slope, Klamath River delta: California Geology, v. 10, p. 542–546. Francis, T.J.G., 1971, Effects of earthquakes on deep-sea sediments: Nature, v. 233, p. 98–102, doi:10.1038/233098a0. Giralt, S., Klerkx, J., Riera, S., Julia, R., Lignier, V., Beck, C., De Batist, M., and Kalugin, I., 2002, Recent palæoenvironmental evolution of Lake Issyk-Kul, in Klerkx, J., and Imanackunov, B., eds., Lake Issyk-Kul: Its Natural Environment: NATO Science Series IV, Volume 13: Dordrecht, the Netherlands, Kluwer Academic Publishers, p. 101–124. Gorsline, D.S., De Diego, T., and Nava-Sanchez, E.H., 2000, Seismically triggered turbidites in small margin basins: Alfonso Basin, western Gulf of California, and Santa Monica Basin, California Borderland: Sedimentary Geology, v. 135, p. 21–35, doi:10.1016/S0037-0738(00)00060-9. Hempton, M.R., and Dewey, J.F., 1983, Earthquake-induced deformational structures in young lacustrine sediments, East Anatolian fault, southwest Turkey: Tectonophysics, v. 98, p. 7–14, doi:10.1016/0040-1951(83 )90294-9. Hibsch, C., Alvarado, A., Yepes, H., Perez, H., and Sébrier M., 1994, Holocene liquefaction and soft-sediment deformation in Quito (Ecuador): A paleoseismic history recorded in lacustrine sediments: Journal of Geodynamics, v. 24, no. 1–4, p. 259–280. Jouanne, F., Ménard, G., and Jault, D., 1994, Present-day deformation measurement in the French northwestern Alps/southern Jura mountains: Data from triangulation comparisons: Geophysical Journal International, v. 119, p. 151–165, doi:10.1111/j.1365-246X.1994.tb00919.x. Jouanne, F., Ménard, G., and Darmendrail, X., 1995, Present-day vertical displacements in the NW Alps and southern Jura: Data from levelling comparisons: Tectonics, v. 14, no. 3, p. 606–616, doi:10.1029/94TC03336. Kastens, K., and Cita, M.B., 1981, Tsunami-induced sediment transport in the abyssal Mediterranean Sea: Geological Society of America Bulletin, v. 92, p. 845–857, doi:10.1130/0016-7606(1981)92<845:TSTITA>2.0 .CO;2. Kelts, K., and Hsü, K.J., 1980, Resedimented facies of 1875 Horgen slumps in Lake Zurich and a process model of longitudinal transport of turbidity currents: Eclogae Geologicae Helvetiae, v. 73, no. 1, p. 271–281. King, G.C.P., Stein, R.S., and Lin, J., 1994, Static stress changes and the triggering of earthquakes: Bulletin of the Seismological Society of America, v. 84, p. 935–953.
178
C. Beck
Kuenen, P.H., 1958, Experiments in geology: Transactions of the Geological Society of Glasgow, v. 23, p. 1–28. Lambert, J., Levret-Albaret, A., Cushing, M., and Durouchoux, C., 1996, Mille Ans de Séismes en France: Nantes, BRGM/EDF/IPSN (Bureau de Recherches Géologiques et Minières/Electricité de France/Institut de Protection et Sûreté Nucléaire), Ouest Editions Presse Académiques, 70 p. Levret, A., Cushing, M., and Peyridieu, G., 1996, Etude des caractéristiques de séismes historiques en France, in Atlas de 140 Cartes Macrosismiques, Volume 2: Baneux, Institut de Protection et de Sûreté Nucléaire, Map 7. Lignier, V., 2001, Les Sédiments Lacustres et l’Enregistrement de la Palæoséismicité: Etude Comparative de Différents Cas dans le Quaternaire des Alpes Nord-Occidentales et du Tien-Shan Kyrghize [Ph.D. thesis]: Le Bourget du Lac, University of Savoie, 2 vols., 381 p. Macar, P., and Antun, P., 1950, Pseudonodules et glissements sous-aquatiques dans l’Emsien inférieur de l’Oesling (Grand Duché du Luxembourg): Annales de la Société Géologique de Belgique Mémoire, v. 73, p. 121– 150. Manalt, F., Beck, C., Disnar, J.-R., Deconinck, J.-F., and Recourt, P., 2001, Evolution of clay mineral assemblages and organic matter in the Late Glacial–Holocene sedimentary infill of Lake Annecy (northwestern Alps): Paleoenvironmental implications: Journal of Paleolimnology, v. 25, no. 2, p. 179–192, doi:10.1023/A:1008140114714. Marco, S., and Agnon, A., 1995, Prehistoric earthquake deformations near Masada, Dead Sea Graben: Geology, v. 23, no. 8, p. 695–698, doi:10 .1130/0091-7613(1995)023<0695:PEDNMD>2.3.CO;2. McCalpin, J., 1996, Paleoseismology: San Diego, Academic Press, 588 p. Mercier de Lépinay B., Labeyrie, L., Çagatay, N., and the MARMACORE Team, 2003, Interplay between recent sedimentation and active tectonics in Marmara Sea: Geophysical Research Abstracts, v. 5, p. 12,694. Molina, J.M., Alfaro, P., Moretti, M., and Soria, J.M., 1998, Soft-sediment deformation structures induced by cyclic stress of storm waves in tempestites (Miocene, Guadalquivir Basin, Spain): Terra Nova, v. 10, p. 145– 150, doi:10.1046/j.1365-3121.1998.00183.x. Moretti, M., Alfaro, P., Caselles, O., and Canas, J.A., 1999, Modelling seismites with a digital shaking table: Tectonophysics, v. 304, p. 369–383. Mörner, N.-A., 1996, Liquefaction and varve deformation as evidence of paleoseismic events and tsunamis: The autumn 10,430 BP case in Sweden: Quaternary Science Reviews, v. 15, p. 939–948, doi:10.1016/S0277 -3791(96)00057-1. Mulder, T., and Cochonat, P., 1996, Classification of offshore mass movements: Journal of Sedimentary Research, v. 66, no. 1, p. 43–57. Mulder, T., Tisot, J.-P., Cochonat, P., and Bourillet, J.-F., 1994, Regional assessment of mass failure events in the Baies des Anges, Mediterranean Sea: Marine Geology, v. 122, no. 1/2, p. 29–45, doi:10.1016/0025-3227 (94)90203-8. Nakajima, T., and Kanai, Y., 2000, Sedimentary features of seismoturbidites triggered by the 1983 and older historical earthquakes in the eastern margin of the Japan Sea: Sedimentary Geology, v. 135, p. 1–19, doi:10.1016/ S0037-0738(00)00059-2. Nemec, W., 1990, Aspect of sediment movement on steep delta slope, in Colella, A., and Prior, D.B., eds., Coarse-Grained Deltas: International Association of Sedimentologists Special Publication 10, p. 29–74. Obermeier, S., 1989, The New Madrid Earthquakes: An Engineering-Geologic Interpretation of Relict Liquefaction Features: U.S. Geological Survey Professional Paper 1336-B, 114 p. Obermeier, S.F., Bleuer, N.R., Munson, C.A., Munson, P.J., Martin, W.S., McWilliams, K.M., Tabaczynski, D.A., Odum, J.K., Rubin, M., and Eggert, D.L., 1991, Evidence of strong earthquake shaking in the lower Wabash valley from prehistoric liquefaction features: Science, v. 251, p. 1061–1063, doi:10.1126/science.251.4997.1061. Oldfield, F., and Berthier, F., 2001, The multi-proxy late-Pleistocene and Holocene record from the sediments of the Grand Lac d’Annecy, eastern France: Journal of Paleolimnology, v. 25, no. 2, p. 133–135. Passega, R., 1964, Grain size representation by CM patterns as a geological tool: Journal of Sedimentary Petrology, v. 34, p. 830–847. Pederneiras, R.G.G., 1991, Paleosismicidade e sedimentaçao—Evidências no compartimento Sul da Bacia do Recôncavo, Bahia: Geociencîas: PETROBRAS, v. 5, no. 1/4, p. 39–68. Philippe, Y., 1996, Transfer zone in the southern Jura thrust belt (eastern France): Geometry, development, and comparison with analogic experiments, in Mascle, A., ed., Exploration and Petroleum Geology in France: European Association of Petroleum Geologists Special Publication 4, p. 327–346.
Pickering, K.T., and Hiscott, R.N., 1985, Contained (reflected) turbidity currents in the Middle Ordovician Cloridorme Formation, Québec, Canada: An alternative to the antidune hypothesis: Sedimentology, v. 32, p. 373– 394, doi:10.1111/j.1365-3091.1985.tb00518.x. Piper, D.J.W., Cochonat, P., Ollier, G., Le Drezen, E., Morrison, M., and Baltzer, A., 1992, Evolution progressive d’un glissement rotationnel en un courant de turbidité: Cas du séisme de 1929 des Grands Bancs (Terre Neuve): Comptes Rendus de l’Académie des Sciences Paris, v. 314, p. 1057–1064. Plaziat, J.-C., Purser, B.H., and Philobbos, E., 1988, Diversity of Neogene seismites of the NW Red Sea (Egypt): A characteristic sedimentary expression of rifting: Tectonophysics, v. 153, p. 295, doi:10.1016/0040 -1951(88)90021-2. Postma, G., Babìc, L., Zupanìc, J., and Roe, S.L., 1988, Delta-front failure and associated bottomset deformation in a marine, gravelly Gilbert-type delta, in Nemec, N., and Steel, R.J., eds., Fan Deltas: Sedimentology and Tectonic Settings: Glasgow & London, Blackie, p. 91–102. Pratt, B.R., 1994, Seismites in the Mesoproterozoic Altyn Formation (Belt Supergroup), Montana: A test for tectonic control of peritidal carbonate cyclicity: Geology, v. 22, p. 1091–1094, doi:10.1130/0091-7613(1994 )022<1091:SITMAF>2.3.CO;2. Pratt, B.R., 1998, Syneresis cracks: Sub-aqueous shrinkage in argillaceous sediments caused by earthquake-induced dewatering: Sedimentary Geology, v. 117, p. 1–10, doi:10.1016/S0037-0738(98)00023-2. Rebesco, M., Della Vedova, B., Cernobori, L., and Aloisi, G., 2000, Acoustic facies of Holocene megaturbidites in the Eastern Mediterranean: Sedimentary Geology, v. 135, p. 65–74, doi:10.1016/S0037-0738(00)00063-4. Ringrose, P.S., 1989, Palaeoseismic (?) liquefaction event in late Quaternary lake sediments at Glen Roy, Scotland: Terra Nova, v. 1, p. 57–62, doi:10.1111/j.1365-3121.1989.tb00326.x. Rodríguez-Pascua, M.A., Calvo, J.P., De Vicente, G., and Gómez-Gras, D., 2002, Soft-sediment deformation structures interpreted as seismites in lacustrine sediments of the Prebetic zone, SE Spain, and their potential use as indicators of earthquake magnitudes during the late Miocene: Sedimentary Geology, v. 135, no. 1–4, p. 117–135. Rodríguez-Pascua, M.A., De Vicente, G., Calvo, J.P., and Pérez-López, R., 2003, Similarities between recent seismic activity and paleoseismites during the late Miocene in the external Betic Chain (Spain): Relationship between “b” value and the fractal dimension: Journal of Structural Geology, v. 25, p. 749–763, doi:10.1016/S0191-8141(02)00078-0. Roep, T.B., and Everts, A.J., 1992, Pillow-beds: A new type of seismite? An example from an Oligocene turbidite fan complex, Alicante, Spain: Sedimentology, v. 39, p. 711–724, doi:10.1111/j.1365-3091.1992.tb02148.x. Rothé, J.-P., 1941, La sismicité des Alpes Occidentales: Annales Institut de Physique du Globe de Strasbourg, v. 3, p. 26–100. Séguret, M., Labaume, P., and Madariaga, R., 1984, Eocene seismicity in the Pyrénées from megaturbidites of the South Pyrenean basin (Spain): Marine Geology, v. 55, p. 117–131, doi:10.1016/0025-3227(84)90136-1. Seilacher, A., 1984, Sedimentary structures tentatively attributed to seismic events: Marine Geology, v. 55, p. 1–12, doi:10.1016/0025-3227(84)90129 -4. Shiki, T., Kumon, F., Inouchi, Y., Kontani, Y., Sakamoto, T., Tateishi, M., Matsubara, H., and Fukuyama, K., 2000, Sedimentary features of the seismoturbidites, Lake Biwa, Japan: Sedimentary Geology, v. 135, p. 37–50, doi:10.1016/S0037-0738(00)00061-0. Shimazaki, K., and Nakata, T., 1980, Time-predictable recurrence model for large earthquakes: Geophysical Research Letters, v. 7, p. 279–282, doi:10.1029/GL007i004p00279. Siegenthaler, C., and Stürm, M., 1991, Slump-induced surges and sediment transport in Lake Uri, Switzerland: Verhandlungen International Verein. Limnologie, v. 24, p. 955–958. Siegenthaler, C., Finger, W., Kelts, K., and Wang, S., 1987, Earthquake and seiche deposits in Lake Lucerne, Switzerland: Eclogae Geologicae Helvetiae, v. 80, p. 241–260. Sieh, K., 1978, Prehistoric large earthquakes produced by slip on the San Andreas fault at Pallett Creek, California: Journal of Geophysical Research, v. 83, no. B8, p. 3907–3939, doi:10.1029/JB083iB08p03907. Sieh, K., 1996, The repetition of large-earthquake ruptures: Proceedings of the National Academy of Sciences of the United States of America, v. 93, p. 3764–3771, doi:10.1073/pnas.93.9.3764. Sims, J., 1973, Earthquake-induced structures in sediments of Van Norman Lake, San Fernando, California: Science, v. 182, p. 161–163, doi:10.1126/ science.182.4108.161.
Lake sediments as late Quaternary paleoseismic archives Sims, J., 1975, Determining earthquake recurrence intervals from deformational structures in young lacustrine sediments: Tectonophysics, v. 29, p. 141–152, doi:10.1016/0040-1951(75)90139-0. Stanley, D.J., 1981, Unifites: Structureless muds of gravity-flow origin in Mediterranean basins: Geo-Marine Letters, v. 1, p. 77–83, doi:10.1007/ BF02463322. Syvitski, J.P.M., and Schafer, C.T., 1996, Evidence for an earthquake-triggered basin collapse in Saguenay Fjord, Canada: Sedimentary Geology, v. 104, p. 127–153, doi:10.1016/0037-0738(95)00125-5. Tessier, B., and Terwindt, J.H.T., 1994, Un exemple de déformations synsédimentaires en milieu infratidal: L’effet du mascaret: Comptes Rendus de l’Académie des Sciences Paris, v. 319, no. II, p. 217–223. Thouvenot, F., Fréchet, J., Guyoton, F., Guiguet, R., and Jenatton, L., 1990, SISMALP: An automatic phone-interrogated seismic network for the western Alps: Cahiers du Centre Européen de Géodynamique et Séismologie, v. 1, p. 1–10. Thouvenot, F., Fréchet, J., Tapponnier, P., Thomas, J.-C., Le Brun, B., Ménard, G., Lacassin, R., Jenatton, L., Grasso, J.-R., Coutant, O., Paul, A., and Hatzfeld, D., 1998, The ML 5.3 Epagny (French Alps) earthquake of 1996 July 15: A long-awaited event on the Vuache fault: Geophysical Journal International, v. 135, p. 876–892, doi:10.1046/j.1365-246X.1998.00662.x. Tuttle, M., and Seeber, L., 1991, Historic and prehistoric earthquake-induced liquefaction in Newbury, Massachusetts: Geology, v. 19, p. 594–597, doi: 10.1130/0091-7613(1991)019<0594:HAPEIL>2.3.CO;2. Tweddle, D., and Crossley, R., 1991, Effects of an earthquake on demersal cichlid fishes of southern Lake Malawi: Journal of Fish Biology, v. 38, p. 305–308, doi:10.1111/j.1095-8649.1991.tb03116.x. Van Loon, A.J., Brodzikowski, K., and Zielinski, T., 1995, Shock-induced resuspension deposits from a Pleistocene proglacial lake (Kleszczow
179
graben, central Poland): Journal of Sedimentary Research, v. 65, no. 2, p. 417–422. Van Rensbergen, P., 1996, Seismic Stratigraphic Study of the Glacial and Lacustrine Infill of Lakes Annecy and Le Bourget [Ph.D. thesis]: Ghent, Belgium, University of Ghent, 2 vols., 320 p. Van Rensbergen, P., De Batist, M., Beck, C., and Manalt, F., 1998, Highresolution seismic stratigraphy of late Quaternary fill of Lake Annecy (northwestern Alps): Evolution from glacial to interglacial sedimentary processes: Sedimentary Geology, v. 117, p. 71–96, doi:10.1016/S0037 -0738(97)00123-1. Van Rensbergen, P., De Batist, M., Beck, C., and Chapron, E., 1999, Highresolution seismic stratigraphy of glacial to interglacial fill of a deep glacigenic lake: Lake Le Bourget, northwestern Alps, France: Sedimentary Geology, v. 128, p. 99–129, doi:10.1016/S0037-0738(99)00064-0. Vittori, E., Sylos Labini, S., and Serva, L., 1991, Palaeoseismology: Review of the state of the art: Tectonophysics, v. 193, p. 9–32, doi:10.1016/0040 -1951(91)90185-U. Weidlich, O., and Bernecker, M., 2004, Quantification of depositional changes and paleo-seismic activities from laminated sediments using outcrop data: Sedimentary Geology, v. 166, no. 1/2, p. 11–20, doi:10.1016/j.sedgeo.2003 .12.004. Yu, Y., Oldfield, F., Manalt, F., and Beck, C., 2001, The environmental significance of magnetic measurements of late Pleistocene and Holocene sediments from the Grand Lac d’Annecy, eastern France: Journal of Paleolimnology, v. 25, no. 2, p. 193–203.
MANUSCRIPT ACCEPTED BY THE SOCIETY 7 DECEMBER 2010
Printed in the USA
The Geological Society of America Special Paper 479 2011
Late Pleistocene–early Holocene paleoseismicity deduced from lake sediment deformation and coeval landsliding in the Calchaquíes valleys, NW Argentina Reginald L. Hermanns* International Centre for Geohazards, Norges Geologiske Undersøkelse, Leiv-Eirikssons vei 39, NO 7491 Trondheim, Norway Samuel Niedermann* Helmholtz-Zentrum Potsdam—Deutsches GeoForschungsZentrum, Telegrafenberg, D-14473 Potsdam, Germany
ABSTRACT Two earthquakes are recorded in lake sediments of a former rock-avalanche– dammed lake at the outlet of the Calchaquíes valleys, Argentina. The lake existed between 13,830 ± 790 and 4810 ± 500 a, as indicated by 10Be exposure ages of the landslide deposits that impounded that lake and caused the dam erosion. Two reverse faults, with buckle folds in the footwall and slump folds in the hanging wall, indicate that two earthquakes took place while the lake sediments were water saturated, i.e., during the lake phase. Two folds only a few meters apart occur within the same lake sediment sequence over a distance of 1.3 km on two layers. Within the same two layers, there are mixed zones of convolute bedding extending several hundred meters toward the center of the former lake, which are interpreted to be seismites. These disturbed zones occur also in other subbasins of the former lake that were not affected by faulting and folding. One seismite horizon was AMS (accelerator mass spectrometry) 14C dated to 7500 ± 70 cal yr B.P. by organic material. This age agrees with the 10Be surface exposure age of 7820 ± 830 a for a cluster of four landslides 40 km NNW of the outlet of this lake, suggesting that a strong earthquake occurred at this time.
*E-mails:
[email protected];
[email protected]. Hermanns, R.L., and Niedermann, S., 2011, Late Pleistocene–early Holocene paleoseismicity deduced from lake sediment deformation and coeval landsliding in the Calchaquíes valleys, NW Argentina, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, p. 181–194, doi:10.1130/2011.2479(08). For permission to copy, contact
[email protected]. © 2011 The Geological Society of America. All rights reserved.
181
182
Hermanns and Niedermann
INTRODUCTION
GEOLOGICAL SETTING
We carried out studies on the temporal and spatial distribution of large landslides in northwest Argentina (Hermanns and Strecker, 1999; Hermanns et al., 2000, 2001; Hermanns and Schellenberger, 2008; Trauth et al., 2000). An important result of those studies is that rock avalanches are restricted to tectonically active mountains where there is evidence for late Pleistocene faulting. Tectonically inactive mountains having the same topography, lithologies, and structure lack large landslides. These observations suggest that the landslides were triggered by earthquakes. Caution is required, however, because the coincidence of active faults and landslides does not prove that the failure was triggered by earthquakes on nearby faults. In addition, earthquakes large enough (M ~6) to trigger rock avalanches elsewhere in the world (Keefer, 1984) have not been recorded in the study area between 1960 and 2006 (Engdahl et al., 1998), nor are there any written accounts (Castano, 1997) of earthquakes of this magnitude in the area.1 However, coeval landslides with a 10Be surface exposure age of 7820 ± 830 a along a mountain slope in the area suggest that earthquakes did trigger them (Hermanns et al., 2004, 2006). Rock-avalanche deposits blocked valleys, forming lakes with surface areas up to 600 km2 in the late Pleistocene (Hermanns and Strecker, 1999; Trauth and Strecker, 1999). The lakes persisted ~8700 ± 1000 yr, into the Holocene (Hermanns et al., 2004; Hermanns and Schellenberger, 2008). Sediments deposited in these lakes are ideal archives for seismically induced deformations and were studied to test the hypothesis that some of the landslides described in the same area were seismically triggered (e.g., Hermanns and Strecker, 1999; Hermanns et al., 2006). In this paper, we describe deformation structures in the lake sediments to show that strong earthquakes occurred in the area in the late Pleistocene and early Holocene. The deformation features include faults, folds, mixed layers that extend normal to the folds, tension cracks, and liquefaction features such as sand volcanoes. Such soft-sediment deformation structures have been used in a variety of tectonic settings for paleoseismic reconstructions (e.g., Hempton and Dewey, 1983; Scott and Price, 1988; Clague et al., 1992; Marco et al., 1996; Aylsworth et al., 2000; Bowman et al., 2000; Rodriguez et al., 2000; Bowman et al., 2004; Begin et al., 2005).
The Calchaquí valley and the Santa Maria valley are collectively referred to as the Calchaquíes valleys. They form a 300-km-long basin bounded by reverse faults that is drained by Las Conchas valley. The Calchaquíes valleys lie at the transition between the Cordillera Oriental and the northern Sierras Pampeanas and Puna Plateau tectonic provinces in the Central Andes (Fig. 1; Jordan et al., 1983). The Sierras Pampeanas and Puna Plateau are bordered by reverse faults and are composed of Late Precambrian and Paleozoic metamorphic rocks and Paleozoic granites (González Bonorino, 1951; Rapela, 1976). They are surrounded by basins containing thick Tertiary and Quaternary sediments (Strecker et al., 1989). The Cordillera Oriental, in contrast, is a fold-and-thrust belt (Mon, 1976) of Precambrian basement overlain by Cambrian to Tertiary sediments (Reyes and Salfity, 1973; Omarini, 1983). The Cordillera Oriental is cut by deep valleys that drain internal basins, as well as basins of the Sierra Pampeanas and the Puna Plateau to the south and west, respectively. Mountains reach elevations between 3000 and 5700 m and are barriers to moisture-bearing air masses originating in the Atlantic Ocean (Haselton et al., 2002). Hence, these valleys are arid with annual precipitation less than 200 mm (Bianchi and Yañez, 1992). Giant landslide deposits are common along the tectonically active basin margins and the outlet of the Calchaquíes valleys and have been interpreted to be secondary effects of strong earthquakes (Hermanns and Strecker, 1999; Hermanns et al., 2000, 2006). In El Tonco valley (Fig. 2), two rock avalanches and rock-fall deposits directly overlie the Cerro Paranilla ash. The landslides must have occurred shortly after tephra deposition and likely were coeval. They probably were triggered by an earthquake, because rainfall or other possible triggering mechanisms would have caused tephra redeposition before failure of mountain fronts (Hermanns et al., 2006; Hermanns and Schellenberger, 2008). The rock-avalanche breakaway scarps have been 10Be dated to 7820 ± 830 a ago (Hermanns et al., 2004). Three other rock avalanches, ~20 km south of the outlet of the Calchaquíes valleys, also occurred between 3.6 and 7.8 ka, as indicated by tephrostratigraphic data (Hermanns et al., 2000; Hermanns and Schellenberger, 2008), and may have occurred during the same event (Figs. 2 and 3). Andean deformation in Argentina is two-phased. An older episode of NW-SE contraction began between the early and late Miocene and ended after 0.98 Ma. It was followed by NE-SW contraction that began between the late Miocene and 4.17 Ma and that is still active (Marrett and Strecker, 2000). Seismicity in the study area is low (Fig. 1; Castano, 1997; Engdahl et al., 1998), but historic earthquakes are not a good measure of the largest events that can occur in the area. Recorded moderate earthquakes are restricted to the Calchaquíes valleys, one in the north at the margin of the Puna Plateau, and the other at the south end of the Cumbres Calchaquíes (Fig. 1). Most earthquakes in this region occur along the east front of the Andes in the Santa Barbara Province and the Subandean Belt in the
1
On 27 February 2010 an earthquake occurred outside our study area, but within the area outlined in Figure 1, in the vicinity of the city of Salta with a depth of 10 km and a magnitude of M 6.1. This earthquake was followed by several aftershocks, several of which had a magnitude higher than M 4. The earthquake produced widespread rock falls in the area, seriously damaging recent infrastructure projects. This chapter was accepted before that earthquake occurred, so we are not able to discuss details about it in this chapter. Please note that the recent earthquake does underline the main conclusion of the authors—that the seismic activity in the study area is higher than indicated by historic and instrumental records and that slope instability is strongly related to earthquakes in the region.
Late Pleistocene–early Holocene paleoseismicity, Calchaquíes valleys, NW Argentina
183
Figure 1. Map showing the distribution of mountain ranges, intramontane basins, narrow valleys, and the epicenters (circles) of instrumentally recorded earthquakes >Mb 4 between 1960 and 2006. Stars are epicenters of historical earthquakes with an estimated Modified Mercalli (MM) intensity >6. Box corresponds to location of Figure 2. Inset shows the tectonic provinces of the Central Andes after Jordan et al. (1983); box corresponds to location of map.
north or in the Sierras Pampeanas and within the Puna Plateau to the south. LAKE SEDIMENTS AT THE OUTLET OF THE CALCHAQUÍES VALLEYS Rio Las Conchas, the outlet of the Calchaquíes valleys, has been impounded by two rock-avalanche deposits (163 × 106 and 210 × 106 m3) on the east and south slopes of Cerro Zorrito (Hermanns and Strecker, 1999; Hermanns et al., 2000, 2004). Landslide dams formed at Casa de los Loros and El Paso (CL and EP, respectively; Fig. 4) at 13,550 ± 870 and 15,250 ± 1960 a, as indi-
cated by 10Be surface exposure ages of their deposits and breakaways, respectively, or perhaps they occurred simultaneously at ca. 13,830 ± 790 a (the weighted mean of both ages; Hermanns et al., 2004). The dams extended to an elevation of 1700 m (Hermanns and Strecker, 1999), impounding two connected lakes, a lower one (8 km2) in the La Yesera basin, and an upper lake, Santa Maria, in the Santa Maria basin (>600 km2) (Trauth and Strecker, 1999; Hermanns et al., 2004). A landslide fell into the lower lake from the south face of Cerro Zorrito at 4810 ± 500 a, as indicated by 10Be surface exposure ages of the deposits, triggering a tsunami that breached the Casa de Los Loros dam and drained the lake (Hermanns et al., 2004).
184
Hermanns and Niedermann
Figure 2. Simplified geologic map (after Hermanns et al., 2004) showing the distribution of paleolake and landslide deposits in the vicinity of the Calchaquíes valleys. Box shows location of Figure 4. E.P.—El Paso, C.L.—Casa de los Loros. Note that the El Paso landslide dam divided the paleolake into an upper and a lower basin.
Figure 3. Stratigraphic relations of rockavalanche deposits (triangles) and tephra beds (after Hermanns et al., 2000, 2004; Hermanns and Schellenberger, 2008).
Late Pleistocene–early Holocene paleoseismicity, Calchaquíes valleys, NW Argentina
185
Figure 4. Map of Cerro Zorrito showing the location of rock-avalanche deposits and preserved lake sediments in Las Conchas valley. Box shows location of Figure 5; asterisk is site of varved lake sediment profile.
The basins behind the landslide dams partially filled with laminated lacustrine sediments (Strecker and Marrett, 1999; Trauth and Strecker, 1999; Trauth et al., 2003). The maximum exposed thickness of the lacustrine sediments in both basins is 40 m (Trauth et al., 2003). Trauth and Strecker (1999) showed that the laminae in the lower basin are varves. Varve counts indicate that the lake persisted for at least 6700 ± 700 yr (Trauth et al., 2003), supporting the 10Be surface exposure ages of the landslides. Soft-sediment deformation structures are common in the lake sediments in the two basins (Fig. 5). Trauth et al. (2003) recognized eight deformed horizons in a section in the La Yesera basin, but they lie only 10 m away from the steeply inclined margins of the landslide deposit and are contaminated with pebble-size material derived from the landslide dam. We witnessed hyperconcentrated flows derived from the landslide deposit during a summer rainstorm and thus conclude that this site is not well suited for linking soft-sediment deformation with earthquakes. The larger, upper Santa Maria lake basin, with its gentle slopes, is better suited for such an analysis.
Within the upper basin, lake sediments are only partially preserved, with few outcrops at the margin and larger outcrops toward the center (Trauth et al., 2003). Preserved lake sediments more than 10 km from the center part of the lake are only a few meters thick, consist of sands and pebbles, and are poorly laminated. In the inner part of the basin, thicknesses of several tens of meters of sediments are exposed over distances of up to 1 km (Fig. 5; Trauth and Strecker, 1999; Trauth et al., 2003). These sediments comprise alternating sandy and silty layers a few centimeters to a few decimeters thick in more marginal outcrops, changing progressively to a silt and clay composition in the inner central part of the basin. Soft-sediment deformation structures are restricted to bedded sandy and silty deposits (Fig. 5), and so we focused on them. Soft-Sediment Deformation Structures Soft-sediment deformation structures in the inner part of the basin occur not only close to the landslide deposit that impounded the lake, but also more than 2 km away in the El Paso
186
Hermanns and Niedermann 65° 49´W
75
16
0
0 17
El Paso subbasin
50
16
25
s Con cha
Río Las
A
1575
26°S
160 0
16
1625
1650 1600
75
Thrust fault Anticline, overturned fold
0
1km Sediment samples
A Equal Area
Swamp deposit Lake deposit (with seismite horizon) Landslide deposit
1700
Las Conchas subbasin
1675
16
a
g
b
Strongly h deformed j layers
c P
T
Folded layers
k
Liquefied material ensheathing fold
f
l
Faulted layers
d
Undeformed continuation
e
Figure 5. Map of the El Paso and Las Conchas subbasins showing the distribution of lake deposits and softsediment deformation structures (after Hermanns et al., 2006). Symbols correspond to those in Figure 7. Inset: Kinematic analyses of slickensides on these faults (performed with the program Fault Kinematics) indicate SW-NE–directed shortening, parallel to the direction of active regional compression in NW Argentina (Marrett and Strecker, 2000).
and Las Conchas subbasins (Fig. 5). We mapped the deformation horizons, and deformation style and thicknesses of layers are described in detail in the following sections. Representative samples (500 g) from the deformed layers were collected for grain-size analysis for comparison with known seismites in other parts of the world. Method of Grain-Size Analysis of Soft-Sediment Deformation Structures Samples were split by quartering the bulk sample down in various steps, and representative fractions of 12–15 g of sediment were analyzed. Organic components were removed with H2O2, and the remaining material was wet-sieved in two fractions: >63 µm and <63 μm. The >63 µm fraction was wet-sieved in full phi steps, with phi = −2 being the coarsest fraction. Clay-size sediment was separated using Atterberg separation after Stoke’s law. The silt-size fraction was analyzed with a Micromeritics SediGraph 5100 in full phi steps. Soft-Sediment Deformation Structures in El Paso Subbasin Two thrust faults separated laterally by ~10 m occur at the west margin of the El Paso subbasin. They are associated with
buckle folds in the footwall and slump folds in the hanging wall of the faults, indicating that they formed when the lake sediments were saturated (Figs. 6A and 6B). The faulted sediment consists of ~75% clay (Fig. 7). The offset on the faults is probably only a few decimeters, but this is uncertain because thickness and color of laminae are fairly uniform and marker horizons are lacking. Kinematic analysis of slickensides using the program Fault Kinematics following Marrett and Allmendinger (1990) indicates SW-NE–directed shortening. Along the strike of these faults and ~500 m SSE, an overturned fold indicates only a few decimeter offset (Fig. 6C); this overturned fold varies in amplitude and can be traced for more than 500 m. At several places, one of the limbs of this fold is truncated (Fig. 8). A syncline several meters in amplitude has an axis parallel to the overturned fold (Figs. 5 and 9). The syncline is only a few meters lower in the stratigraphic sequence than the overturned fold and continues toward the south as an overturned fold for more than 1 km. The folds are composed of silty sediments and are mantled by mixed layers of more sandy composition (Fig. 8). Nondeformed, horizontal beds overlie the folded layers across an angular discontinuity, suggesting that the folded layers formed the lake floor during deformation.
Late Pleistocene–early Holocene paleoseismicity, Calchaquíes valleys, NW Argentina
187
Figure 6. Two reverse fault offsets in lake sediments in the El Paso subbasin: (A) slump folds in the hanging wall (letter l indicates site of grain-size sample) (hammer shaft is 32 cm long), (B) buckle folds in the footwall (pocket knife is 9 cm long), (C) reverse fault in sands ending in an overturned fold composed of silt.
Percentage by weight (%)
100
Clay
Silt
Sand
80 most liquefiable sands based on laboratory tests (Lee and Fitton, 1969)
60
(B) 40
(A) envelope of 19 curves of sands that liquefied during earthquakes in Japan (Kishida, 1970)
0 10
g h j
20
1
a b c e f d
100
k l
Mixed layers
Undeformed continuation of mixed layers Folded laminae
Material ensheathing the fold Faulted laminae
1000
Grain size (µm) Figure 7. Texture of deformed sediments in relation to liquefiable domains: Dashed lines indicate finest sediment that is “most liquefiable” (A) and “potentially liquefiable” (B) after Obermeier (1996a). Deformation structures interpreted to be seismites occur only in coarse silt and fine sand. Letters refer to sample sites shown in Figures 6, 10, and 13.
Perpendicular to the strike of the folds, there are 30- to 40-cm-thick mixed layers of sands and silts with complex convolute bedding and recumbent and overturned folds (Fig. 10A). These layers are composed of fine to very fine sands (Figs. 7 and 10). They continue for more than 100 m toward the center of the basin, decreasing in thickness to a few centimeters (Fig. 10B).
Tension cracks and sand volcanoes occur upslope of the folds. Sands fill openings in the cracks, and blocks of silty lamina have detached from the walls and slid down (Fig. 11). An upward motion of liquefied sand is indicated by the sand volcanoes. Silt laminae are bent upward along dikes feeding the sand volcanoes (Fig. 12).
188
Hermanns and Niedermann Lake sediments in the center of the El Paso basin are clays and silts, and deformation structures were not found, except close to the rock-avalanche dam, where a single, strongly folded 5-cmthick layer was traced for 200 m in a 12-m-thick lake sediment outcrop. Organic matter in the deformed layer was 14C AMS (accelerator mass spectrometry) dated at 7500 ± 70 cal yr B.P. (Hermanns et al., 2006). Deformation Structure in Las Conchas Subbasin Two deformation horizons also exist in the upper Las Conchas subbasin. They are separated by 11 m of undisturbed lake sediments. The lower layer (Fig. 13A) is continuous across the entire outcrop, for more than 400 m, as a 15- to 20-cm-thick layer characterized by complex convolute bedding with recumbent and overturned folds of silty composition lying within and topped by sand that fines upward (Fig. 13A). The upper deformed layer is 30–50 cm thick and can be traced in the outcrops for several hundred meters (Fig. 13B). However, it ends abruptly at one site near the center of the basin (Fig. 13C). Deformation involved “mixed layers” and “pseudonodules,” with folded and fragmented silt laminae and a gradation from a fragment-supported to a sand matrix–supported sediment. The grain-size distribution is bimodal (Fig. 13). The abrupt cutoff of the deformation is accompanied by an abrupt fining of the sediment to a silt and clay. DISCUSSION
Figure 8. Overturned folds in silty sediments. The folds are mantled by mixed layers of more sandy composition.
Figure 9. Syncline with an axis parallel to the strike of the faults. The syncline is truncated by overlying layers, indicating that the folding occurred at the sediment-water interface during deformation. Anticline in lake sediments has a wave length of 6 m.
Seismites are deformation structures in soft sediments produced by earthquakes (Seilacher, 1969). They have been described and interpreted in a variety of tectonic settings (Sims, 1975; Allen, 1986; Obermeier et al., 1990; Einsele et al., 1996; Marco et al., 1996; Rodriguez et al., 2000; Bowman et al., 2004). The same structures can also be produced by storm surges (Lowe, 1975) or by gravity loading and artesian pressure induced by a rapid fall in water level. Ten-meter-high storm waves with a frequency of 0.1 Hz cause cyclic loading that is capable of triggering failure in lowstrength sediments at water depths comparable to the wave height (Allen, 1982). Paleolake Santa Maria had a maximum surface area of 600 km2 and a maximum water level at 1700 m above sea level (a.s.l.). Waves no more than a few meters high are likely to form on such a lake. Soft-sediment deformation structures occur at an elevation of <1625–1650 and ~1650 m (Fig. 5), i.e., well below the maximum lake level. On this basis, we conclude that storms could not have formed the deformation structures. The spatial relations among the initiation of mixed layers and folds with several meters of amplitude on slopes <5°, thinning of the layers toward the center of the lake, the absence of coarse sediments in the mixed layers, and the preservation of delicate varve fragments indicate that the mixed layers did not form as a result of turbulent mass flows from slope failures or changes of artesian pressure.
Late Pleistocene–early Holocene paleoseismicity, Calchaquíes valleys, NW Argentina
189
Figure 10. (A) Layers of fine to very fine sand showing complex convolute bedding with recumbent and overturned folds. Letters refer to those in Figure 7. (B) Deformation structures of the same layer toward the center of the basin where layer thins out. Graphs represent grain size distribution of samples taken from lake sediment outcrop as indicated in boxes.
Tests for Seismic-Induced Deformation Criteria for recognizing “seismites” are (1) location in tectonically active areas, (2) sudden formation, (3) similarity to structures formed during earthquakes in other regions, (4) wide stratigraphic lateral extent of deformed units, (5) synchroneity of multiple deformation structures, (6) zoned distribution along a fault, (7) size, (8) depositional setting, and (9) correspondence in age of a seismite and another independent indicator of an earthquake (e.g., Seilacher, 1969; McCalpin and Nelson, 1996; Obermeier, 1996a, 1996b; Marco et al., 1996; Wheeler, 2002; Begin et al., 2005).
Location in a Tectonically Active Area The east front of the Andes is a few tens of kilometers to the east of the study area, but historical earthquakes with an intensity of MM > 6 (Modified Mercalli Intensity scale) have occurred within the Calchaquíes valleys (Fig. 1; Castano, 1997). In addition, faults with several meters of Holocene displacement have been reported from different sites within the valleys (Strecker et al., 1989; Hermanns et al., 2000, 2006). For example, the Las Bañadas thrust fault, with a maximum vertical offset of 5 m in Holocene deposits, lies at the west margin of the basin, 30 km SSW of the confluence of the Río Santa Maria and Río Calchaquí (Fig. 2; Strecker et al., 1989). A 4-km-long fault offset
190
Hermanns and Niedermann
Figure 11. Tension cracks containing fluidized sand and blocks of silty lamina that slid down the crack walls.
within late Pleistocene sediments is found adjacent to a rock-avalanche deposit that impounds Brealito Lake, 90 km NW of the outlet of the Calchaquíes valleys at the Puna margin (Hermanns, 1999). A 10 m fault offset of the El Paso ash on the range-bounding Sierra Aconquija fault (Fig. 1) at the southern end of the Calchaquíes valleys also indicates young deformation. This ash was originally reported to have a maximum 14C AMS age of 28,900 ± 150 cal yr B.P. (Hermanns et al., 2000) but recently was redated at 14C AMS 10,870 ± 290 cal yr B.P. (Hermanns et al., 2006; Hermanns and Schellenberger, 2008). Further fault offsets with more pristine morphology run parallel to this mountain-bounding fault (Casa, 2009). Sudden Formation Both fault offsets occurred suddenly, as indicated by slump folds in the hanging wall and buckle folds in the footwall. The slump folds are overlain unconformably by undeformed horizontal laminae. We could not prove that laminae in the upper basin are varves, but this interpretation is favored by the similarity in lake sediment thickness in the upper basin to that in the lower basin, which does contain varves (Trauth and Strecker, 1999; Trauth et al., 2003). If the sediments are varves, each of the deformation structures formed within a single year.
Figure 12. Liquefied sands erupted upward, folding silty lamina in this direction.
Similarity to Structures Formed during Earthquakes in Other Regions Mixed layers, layers with complex convolute bedding with recumbent and overturned folds, and sand volcanoes in other areas have been attributed to seismic activity (e.g., Marco et al.,
Late Pleistocene–early Holocene paleoseismicity, Calchaquíes valleys, NW Argentina
191
Figure 13. Two mixed layers with soft-sediment deformation structures separated by 11 m of undisturbed lake deposits. (A) The lower layer (left) is characterized by complex convolute bedding with recumbent and overturned folds of silt lying in and topped by fluidized sand. (B) The upper layer (center) is a mixed layer with folded and fragmented silt laminae grading from clast-supported to matrix-supported structure. (C) The upper layer terminates abruptly toward the center of the basin (right), where the sediments fine to silts. Graphs represent grain size distribution of samples taken from lake sediment outcrop as indicated in boxes. Lowercase letters in graphs (a–c) refer to those in Figure 7.
1996; Obermeier, 1996a, 1996b; Rodriguez et al., 2000; Bowman et al., 2004). The convolute beds and the mixed layers mantling the folds have textures similar to those of sediments that have liquefied during historic earthquakes and laboratory tests (Fig. 7; Lee and Fitton, 1969; Kishida, 1970; Obermeier, 1996a, 1996b, and references therein). In contrast, all sediments that behaved in a brittle (faulted sediments) or ductile (folded sediments) manner have textures unfavorable for liquefaction (Fig. 7). Wide Stratigraphic Lateral Extent of Deformed Units Although the fault in the El Paso subbasin could only be found in two outcrops ~500 m apart, overturned folds continue for a further 1.3 km to the end of the outcrop. Sediment deformation was observed perpendicular to the overturned folds 100– 300 m toward the center of the basin.
The lower deformed layer was traced throughout the entire outcrop in the Las Conchas subbasin. The upper layer was traced for several hundred meters. However, toward the center of this subbasin, the deformed nature ends abruptly, most likely controlled by the change of the grain-size distribution of the sediments toward a nonliquefiable composition (Fig. 7). Synchroneity of Multiple Deformation Structures Due to the absence of marker beds, it is impossible to correlate deformation horizons between the two subbasins. Furthermore, fossil plant material is scarce in these sediments (Trauth and Strecker, 1999; Hermanns et al., 2000), the luminescence signal of the sediments apparently was not reset prior to deposition (Trauth et al., 2000), and radiocarbon ages on carbonate shells are too old (Trauth and Strecker, 1999; Hermanns et al., 2004, 2006). However, the occurrence of the deformed layers in
192
Hermanns and Niedermann
the middle of the lake sediment sequences in both subbasins suggests that they may have formed contemporaneously. The coeval formation of faults, folds, and strongly deformed layers within the El Paso subbasin is shown by their spatial association with one another. The soft-sediment deformation layer associated with organic material dated at 7500 ± 70 cal yr B.P. (Hermanns et al., 2006; Hermanns and Schellenberger, 2008) is restricted to a single site, and it is not possible to ascertain whether this age is related to the upper or lower deformation layer at other sites.
The evidence collectively indicates that at least two, and possibly three, earthquakes occurred along this mountain front between 13,830 ± 790 and 4810 ± 500 a ago. The seismites occur near the middle of the lacustrine sequence, suggesting that the earthquakes occurred in the early Holocene. One seismite horizon was 14C AMS dated to 7500 ± 70 cal yr B.P., which is consistent with this interpretation and with the age of a cluster of coeval landslides 40 km to the NNW. The second event is undated. Magnitudes of Paleoearthquakes
Zoned Distribution along a Fault, Size, and Depositional Setting These criteria cannot be tested because the lake was a unique feature occupying a relatively small area. The only lakes of similar size in the region are salt pans in the intra-Andean Puna Plateau more than 100 km away, and they have not been studied. Correspondence in Age of a Seismite and Another Independent Indicator of an Earthquake Independent evidence of paleoearthquakes was found in the Tonco valley, 40 km NNW of the study area. In that valley, four landslides with volumes up to 70 × 106 m3 overlie in direct contact a primary air-fall tephra, indicating synchroneity. A major rainstorm, which is an alternative trigger of multiple contemporaneous landslides, would have redeposited the tephra layer before triggering the large landslides (Hermanns et al., 2006). These deposits were dated by 10Be surface exposure ages of their breakaway scarps to have formed 7820 ± 830 a ago. In conclusion, the soft-sediment deformation structures pass most of the tests for a seismic origin, and they are interpreted to be seismites. Recurrence of Seismic Events in the Late Pleistocene– Early Holocene Two episodes of faulting of the lacustrine sediments at the outlet of the Calchaquíes valleys date to between 13,830 ± 790 and 4810 ± 500 a. The two deformed layers identified in the two subbasins of lake Santa Maria are attributed to these earthquakes. We are unable to positively relate deformation structures between the two subbasins, but since deformation horizons in all outcrops occur in the middle of the sequence, it is likely that they formed contemporaneously. Similar sediments in the upper and the lower parts of the lake sediment profile in the Santa Maria basin lack soft-sediment deformation structures. Therefore, we are confident that the two deformed layers record the only earthquakes capable of producing such structures during the history of the lake. The lake existed for 8700 ± 1,000 yr and was formed by two large rock avalanches with ages that overlap, suggesting that they also may have formed simultaneously during an earthquake. However, the landslides occurred at a time when climate was wetter than today (Bookhagen et al., 2001; Hermanns et al., 2004), and so, alternatively, they may have been caused by fluvial erosion of toe slopes.
The threshold magnitude for seismically induced liquefaction is Mw 5 (Allen, 1986; Obermeier, 1996a). Rodríguez-Pascua et al. (2000) proposed a relation between the style of deformation of varved lake sediments and earthquake magnitude for deep and shallow lakes. They argued that mixed layers, such as the lower seismite layer in the Las Conchas subbasin, form during earthquakes of ~M 5.5–6.5, whereas clast- and matrix-supported mixed layers, such as the upper seismite layer in the Las Conchas subbasin, form during earthquakes of ~M 6.5–7.5. The magnitude of coeval landslides in the Calchaquíes valleys indicates a similar magnitude of seismic events. An empirical relation between landslide volume and earthquake magnitude proposed by Keefer (1994) suggests that the coeval landslides in the Tonco valley were triggered by earthquakes of magnitudes ~M 7. A 4 km segment of the Las Bañadas fault, 30 km SSW of the outlet of the Calchaquíes valleys, offsets Holocene alluvial-fan deposits, with a maximum displacement of 5 m (Strecker et al., 1989). These displacements are consistent with earthquakes of >M 7 (Wells and Coppersmith, 1994) or >MS 6.5 (Strom and Nikonov, 1997). In conclusion, the earthquakes that formed the seismites in the study area are likely to have had magnitudes of M 6–7. Earthquakes of this size have not been recorded in the area (Fig. 1) since 1960 (Engdahl et al., 1998), nor are there any written accounts of earthquakes of this magnitude at the outlet of the Calchaquíes valleys (Castano, 1997). Our data, which span a much larger interval, thus suggest that earthquakes in the late Pleistocene–early Holocene were significantly stronger than the historical and instrumental records indicate. No paleoseismic records for this area span the entire Holocene, but it is likely that earthquakes of similar magnitudes and recurrence intervals as in the late Pleistocene and early Holocene occurred during the rest of the Holocene. ACKNOWLEDGMENTS This project was financed by Deutsches GeoForschungsZentrum Potsdam and Sonderforschungsbereich (SFB) 267 “Deformation Processes in the Andes.” We thank B. Diekmann and U. Bastian for carrying out grain-size analyses at the AlfredWegener-Institut Potsdam and for their support. F. Hubberten helped with sample preparation. We thank A. Villanueva Garcia and A. Alonso for help in the field and for logistical support, M. Trauth and M. Strecker for fruitful discussions, and
Late Pleistocene–early Holocene paleoseismicity, Calchaquíes valleys, NW Argentina G. Borm, J. Erzinger, and B. Merz for their encouragement and help. John Clague and two anonymous reviewers provided helpful comments, which improved the manuscript. Support for this work was also provided in part by the Research Council of Norway through the International Centre for Geohazards (ICG). Their support is gratefully acknowledged. This is ICG contribution 308. REFERENCES CITED Allen, J.R.L., 1982, Sedimentary Structures: Their Character and Physical Basis: New York, Elsevier, 663 p. Allen, J.R.L., 1986, Earthquake magnitude-frequency, epicentral distance, and soft-sediment deformation in sedimentary basins: Sedimentary Geology, v. 46, no. 1–2, p. 67–75, doi:10.1016/0037-0738(86)90006-0. Aylsworth, M., Lawrence, D.E., and Guertin, J., 2000, Did two massive earthquakes in the Holocene induce widespread landsliding and near-surface deformation in part of Ottawa Valley, Canada?: Geology, v. 28, no. 10, p. 903–906, doi:10.1130/0091-7613(2000)28<903:DTMEIT>2.0.CO;2. Begin, Z.B., Steinberg, D.M., Ichinose, G.A., and Marco, S., 2005, A 40,000 year unchanging seismic regime in the Dead Sea rift: Geology, v. 33, no. 4, p. 257–260, doi:10.1130/G21115.1. Bianchi, A.R., and Yañez, C.E., 1992, Las Precipitaciones en el Noroeste Argentino: Salta, Instituto Nacional de Tecnología Agropecuaria, 383 p. Bookhagen, B., Haselton, K., and Trauth, M.H., 2001, Hydrological modelling of a Pleistocene landslide-dammed lake in the Santa Maria basin, NW Argentina: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 169, no. 1–2, p. 113–127, doi:10.1016/S0031-0182(01)00221-8. Bowman, D., Banet, D.D., Bruins, H.J., and Van der Plicht, J., 2000, Dead Sea shoreline facies with seismically-induced soft-sediment deformation structures, Israel: Israel Journal of Earth Sciences, v. 49, no. 4, p. 197– 214, doi:10.1560/GXHT-AK5W-46EF-VTR8. Bowman, D., Korjenkov, A., and Porat, N., 2004, Late-Pleistocene seismites from Lake Issyk-Kul, the Tien Shan range, Kyrghyzstan: Sedimentary Geology, v. 163, no. 3–4, p. 211–228, doi:10.1016/S0037-0738(03 )00194-5. Casa, A.L., 2009, Sistema de fallas Aconcagua, in Hermanns, R.L., Costa, C., Audemard, F., Jermyn, C., and Lara, L., eds., Proyectos Multinacional Andino: Geociencia para las Communidades Andinas, Atlas de Deformaciones Cuaternarias de los Andes: Servicio Nacional de Geología y Minería, Publicación Geológica Multinacional 7, p. 129–136. Castano, D.E., 1997, Teremotos históricos, sismicidad y tectónica en el noroeste Argentino, in VII Congreso Geológico Chileno: Antofagasta, Universidad Católica del Norte, p. 665–669. Clague, J.J., Naesgaard, E., and Sy, A., 1992, Liquefaction features on the Fraser Delta; evidence for prehistoric earthquakes?: Canadian Journal of Earth Sciences, v. 29, no. 8, p. 1734–1745. Einsele, G., Chough, S.K., and Shiki, T., 1996, Depositional events and their records; an introduction, in Shiki, T., Chough, S.K., and Einsele, G., eds., Marine Sedimentary Events and Their Records: Sedimentary Geology, p. 1–9. Engdahl, E.R., Van der Hilst, R., and Buland, R., 1998, Global teleseismic earthquake relocation with improved travel times and procedures for depth determination: Bulletin of the Seismological Society of America, v. 88, no. 3, p. 722–743. González Bonorino, F., 1951, Descripción Geológica de la Hoja 12e, Aconquija (Catamarca-Tucumán), v. 75: Buenos Aires, Dirección Nacional de Minería, p. 1–50. Haselton, K., Hilley, G., and Strecker, M.R., 2002, Average Pleistocene climatic patterns in the southern Central Andes: Controls on mountain glaciation and paleoclimate implications: The Journal of Geology, v. 110, p. 211– 226, doi:10.1086/338414. Hempton, M.R., and Dewey, J.F., 1983, Earthquake-induced deformational structures in young lacustrine sediments, East Anatolian fault, southeast Turkey: Tectonophysics, v. 98, p. T7–T14, doi:10.1016/0040-1951 (83)90294-9. Hermanns, R.L., 1999, Spatial-Temporal Distribution of Mountain-Front Collapse and Formation of Giant Landslides in the Arid Andes of Northwest-
193
ern Argentina (24–28°S, 65–68° W) [Ph.D. thesis]: Potsdam, Germany, Universität Potsdam, 123 p. Hermanns, R.L., and Schellenberger, A., 2008, Quaternary tephrochronology helps define conditioning factors and triggering mechanisms of rock avalanches in NW Argentina: Quaternary International, v. 178, p. 261–275, doi:10.1016/j.quaint.2007.05.002. Hermanns, R.L., and Strecker, M.R., 1999, Structural and lithological controls on large Quaternary rock avalanches (sturzstroms) in arid northwestern Argentina: Geological Society of America Bulletin, v. 111, no. 6, p. 934– 948, doi:10.1130/0016-7606(1999)111<0934:SALCOL>2.3.CO;2. Hermanns, R.L., Trauth, M.H., Niedermann, S., McWilliams, M., and Strecker, M.R., 2000, Tephrochronologic constraints on temporal distribution of large landslides in northwest Argentina: The Journal of Geology, v. 108, no. 1, p. 35–52, doi:10.1086/314383. Hermanns, R.L., Niedermann, S., Villanueva Garcia, A., Sosa Gomez, J., and Strecker, M.R., 2001, Neotectonics and catastrophic failure of mountain fronts in the southern intra-Andean Puna Plateau, Argentina: Geology, v. 29, no. 7, p. 619–623, doi:10.1130/0091-7613(2001)029<0619 :NACFOM>2.0.CO;2. Hermanns, R.L., Niedermann, S., Ivy-Ochs, S., and Kubik, P.W., 2004, Rock avalanching into a landslide-dammed lake causing multiple dam failure in Las Conchas valley (NW Argentina)—Evidence from surface exposure dating and stratigraphic analyses: Landslides, v. 1, no. 2, p. 113–122, doi:10.1007/s10346-004-0013-5. Hermanns, R.L., Niedermann, S., Villanueva Garcia, A., and Schellenberger, A., 2006, Rock avalanching in the NW Argentine Andes as a result of complex interactions of lithologic, structural and topographic boundary conditions, climate change and active tectonics, in Evans, S.G., Scarascia Mugnozza, G., Strom, A.L., and Hermanns, R.L., eds., Massive Rock Slope Failure: New Models for Hazard Assessment: NATO Science Series 49: Dordrecht, the Netherlands, Springer, p. 497–520. Jordan, T.E., Isacks, B.L., Allmendinger, R.W., Brewer, J.A., Ramos, V.A., and Ando, C.J., 1983, Andean tectonics related to geometry of subducted Nazca plate: Geological Society of America Bulletin, v. 94, p. 341–361, doi:10.1130/0016-7606(1983)94<341:ATRTGO>2.0.CO;2. Keefer, D.K., 1984, Landslides caused by earthquakes: Geological Society of America Bulletin, v. 95, p. 406–421, doi:10.1130/0016-7606(1984 )95<406:LCBE>2.0.CO;2. Keefer, D.K., 1994, The importance of earthquake-induced landslides to longterm slope erosion and slope-failure hazards in seismically active regions: Geomorphology, v. 10, p. 265–284, doi:10.1016/0169-555X(94)90021-3. Kishida, H., 1970, Characteristics of liquefaction of level sandy ground during the Tokachioki earthquake: Soils and Foundations, v. 10, no. 1970, p. 103–111. Lee, K.L., and Fitton, J.A., 1969, Factors affecting the cyclic loading strength of soil: American Society for Testing Materials Special Technical Publication 450, p. 71–95. Lowe, D.R., 1975, Water escape structures in coarse-grained sediments: Sedimentology, v. 22, no. 2, p. 157–204, doi:10.1111/j.1365-3091.1975 .tb00290.x. Marco, S., Stein, M., and Abnon, A., 1996, Long-term earthquake clustering: A 50,000-year paleoseismic record in the Dead Sea graben: Journal of Geophysical Research, v. 101, no. B3, p. 6179–6191, doi:10.1029/95JB01587. Marrett, R., and Allmendinger, R.W., 1990, Kinematic analysis of fault-slip data: Journal of Structural Geology, v. 12, no. 8, p. 973–986, doi:10.1016/0191 -8141(90)90093-E. Marrett, R., and Strecker, M.R., 2000, Response of intracontinental deformation in the Central Andes to late Cenozoic reorganization of South American plate motions: Tectonics, v. 19, no. 3, p. 452–467, doi:10.1029/ 1999TC001102. McCalpin, J.P., and Nelson, A.R., 1996, Introduction to paleoseismology; the scope of paleoseismology, in McCalpin, J.P., ed., Paleoseismology: San Diego, Academic Press, p. 1–32. Mon, R., 1976, The structure of the eastern border of the Andes in north-western Argentina: Geologische Rundschau, v. 65, p. 211–222, doi:10.1007/ BF01808464. Obermeier, S.F., 1996a, Use of liquefaction-induced features for paleoseismic analysis; an overview of how seismic liquefaction features can be distinguished from other features and how their regional distribution and properties of source sediment can be used to infer the location and strength of Holocene paleo-earthquakes: Engineering Geology, v. 44, no. 1–4, p. 1–76, doi:10.1016/S0013-7952(96)00040-3.
194
Hermanns and Niedermann
Obermeier, S.F., 1996b, Using liquefaction-induced features for paleoseismic analysis, in McCalpin, J.P., ed., Paleoseismology: San Diego, Academic Press, p. 331–396. Obermeier, S.F., Jacobson, R.B., Smoot, J.P., Weems, R.E., Gohn, G.S., Monroe, J.E., and Powars, D.S., 1990, Earthquake-induced liquefaction features in the coastal setting of South Carolina and fluvial setting of the New Madrid seismic zone: U.S. Geological Survey Professional Paper 1504, p. 1–44. Omarini, R.H., 1983, Caracterización Litológica, Diferenciación y Génesis de la Formación Puncoviscana entre el Valle de Lerma y la Faja Eruptiva de la Puna [Ph.D. thesis]: Salta, Argentina, Universidad de Salta, 220 p. Rapela, C.W., 1976, El basamento metamorfico de la region de Cafayate, Provincia de Salta. Aspectos petrológicos y geoquímicos: Revista de la Asociación Geológica Argentina, v. XXI, no. 3, p. 203–222. Reyes, F.C., and Salfity, J.A., 1973, Consideraciones sobre la estratigrafía del Cretácico (Subgrupo Pirgua) del noroeste Argentino, in V Congresso Geológico Argentino, p. 355–385. Rodríguez-Pascua, M.A., Calvo, J.P., De Vincent, G., and Gomez-Gras, D., 2000, Soft-sediment deformation structures interpreted as seismites in lacustrine sediments of the Prebetic zone, SE Spain, and their potential use as indicators of earthquake magnitudes during the late Miocene: Sedimentary Geology, v. 135, p. 117–135, doi:10.1016/S0037-0738(00)00067-1. Scott, B., and Price, S., 1988, Earthquake-induced structures in young sediments: Tectonophysics, v. 147, p. 165–170, doi:10.1016/0040-1951(88)90154-0. Seilacher, A., 1969, Fault-graded beds interpreted as seismites: Sedimentology, v. 13, no. 1–2, p. 155–159, doi:10.1111/j.1365-3091.1969.tb01125.x. Sims, J.D., 1975, Determining earthquake recurrence intervals from deformational structures in young lacustrine sediments: Tectonophysics, v. 29, p. 141–152. Strecker, M.R., and Marrett, R.A., 1999, Kinematic evolution of fault ramps and role in development of landslides and lakes in northwestern
Argentine Andes: Geology, v. 27, p. 307–310, doi:10.1130/0091-7613 (1999)027<0307:KEOFRA>2.3.CO;2. Strecker, M.R., Cerveny, P., Bloom, A.L., and Malizzia, D., 1989, Late Cenozoic tectonism and landscape development in the foreland of the Andes: Northern Sierras Pampeanas (26°–28°S), Argentina: Tectonics, v. 8, no. 3, p. 517–534, doi:10.1029/TC008i003p00517. Strom, A.L., and Nikonov, A.A., 1997, Relations between the seismogenic fault parameters and earthquake magnitude: Physics of the Solid Earth, v. 33, no. 12, p. 1011–1022. Trauth, M.H., and Strecker, M.R., 1999, Formation of landslide-dammed lakes during a wet period between 40,000 and 25,000 yr BP in northwestern Argentina: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 153, no. 1–4, p. 277–287, doi:10.1016/S0031-0182(99)00078-4. Trauth, M.H., Alonso, R.A., Haselton, K.R., Hermanns, R.L., and Strecker, M.R., 2000, Climate change and mass movements in the NW Argentine Andes: Earth and Planetary Science Letters, v. 179, no. 2, p. 243–256, doi:10.1016/S0012-821X(00)00127-8. Trauth, M.H., Bookhagen, B., Müller, A.B., and Strecker, M.R., 2003, Late Pleistocene climate change and erosion in the Santa Maria basin, NW Argentina: Journal of Sedimentary Research, v. 73, no. 1, p. 82–90, doi:10.1306/061702730082. Wells, D.L., and Coppersmith, K.L., 1994, New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement: Bulletin of the Seismological Society of America, v. 84, no. 4, p. 974–1002. Wheeler, R.L., 2002, Distinguishing seismic from nonseismic soft-sediment structures; criteria from seismic-hazard analysis, in Ettensohn, F.R., Rast, N., and Brett, C.E., eds., Ancient Seismites: Geological Society of America Special Paper 359, p. 1–11. MANUSCRIPT ACCEPTED BY THE SOCIETY 7 DECEMBER 2010
Printed in the USA
The Geological Society of America Special Paper 479 2011
Rupture length and paleomagnitude estimates from point measurements of displacement—A model-based approach Glenn Biasi* Seismological Laboratory, University of Nevada, MS-174, Reno, Nevada 89557, USA Ray J. Weldon II* Department of Geological Sciences, University of Oregon, MS-1272, Eugene, Oregon 97403, USA Kate Scharer* Department of Geology, Appalachian State University, 572 Rankin Science Building, Boone, North Carolina 28608, USA
ABSTRACT We present a new method that allows paleomagnitude and paleorupture length to be estimated quantitatively given a measurement of earthquake rupture displacement at a point along a fault. Rupture displacement typically varies along a rupture profile such that a point paleoseismic displacement measurement constrains the average displacement only to within a factor of three or so. We used previously published results summarizing rupture variability and then applied a graphical method of identifying the relative likelihoods among a suite of magnitudes, one of which must have caused the measured displacement. Results were developed for displacement observations from 1 to 6 m using a magnitude range of 6.0 ≤ M ≤ 8.0. Probabilities of rupture lengths for a given displacement were developed at the same time. Although smaller earthquakes can cause ground rupture, we show that they would not strongly influence likelihoods for 1 m and larger observed displacements. Displacements less than 1 m are also of potential interest but will require extension of the method to include the declining probability that smaller magnitudes produce ground rupture. We also consider application of length distributions inferred from a displacement measurement to correlation of rupture evidence between sites. Dating evidence alone, even when excellent, does not provide a physical basis to relate rupture at one site to rupture at another. Ruptures, however, have an expected length, and thus do provide a physical basis for correlation. We present probability of correlation curves for given rupture lengths, which may be combined with probabilities of rupture length to obtain a probability of correlation given a point displacement. Applications for quantitative probabilities of magnitude and length given a paleoseismic displacement measurement include probabilistic seismic hazard analyses, where probabilities of magnitude and length must be assigned to branches in the analysis.
*E-mails:
[email protected];
[email protected];
[email protected]. Biasi, G., Weldon, R.J., II, and Scharer, K., 2011, Rupture length and paleomagnitude estimates from point measurements of displacement—A model-based approach, in Audemard M., F.A., Michetti, A.M., and McCalpin, J.P., eds., Geological Criteria for Evaluating Seismicity Revisited: Forty Years of Paleoseismic Investigations and the Natural Record of Past Earthquakes: Geological Society of America Special Paper 479, p. 195–204, doi:10.1130/2011.2479(09). For permission to copy, contact
[email protected]. © 2011 The Geological Society of America. All rights reserved.
195
196
Biasi et al.
INTRODUCTION There is presently something of a gap between the observable data gathered in paleoseismic excavations and the information needed to estimate seismic hazard. One reason for this gap is the point nature of paleoseismic investigations. Even for a socially significant fault, investigation is typically limited to a few discrete sites (e.g., Grant and Lettis, 2002; Weldon et al., 2005). Average surface rupture displacements from historical earthquakes are known to scale approximately with magnitude (Wells and Coppersmith, 1994; Scholz, 2002). However, slip varies along mapped surface ruptures by a factor of three or more, so that a measurement of displacement at one site (e.g., Sieh, 1984; Salyards et al., 1992; Weldon et al., 2002) really only constrains the average displacement for that earthquake to a broad range. Even less can be inferred about the earthquake rupture length. Unfortunately, earthquake moment is proportional to the product of the earthquake rupture length and average displacement. Consequently, despite a large amount of work spent finding and dating paleoruptures, paleoseismic data only broadly constrain paleomagnitude and seismic hazard. Despite these problems, practitioners recognize that a probabilistic estimate of paleomagnitude can be made, at least qualitatively, if a point estimate of rupture displacement can be obtained. For example, given a 2 m rupture displacement and no other information, one could assign greatest probability to an earthquake in the low M 7 range, lesser probabilities to smaller magnitudes (not expecting so large a displacement), and lesser probabilities to larger magnitudes (expecting a larger displacement). Similarly, a 2 m displacement observation allows an estimate of rupture length of tens of kilometers, with similar, lesser probabilistic expectations for longer or shorter rupture length. In detail, the probabilities assigned to length and magnitude estimates are informed by experience and data from other faults, but they have a significant qualitative component. The basic objective of this paper is to outline an approach for quantifying probability estimates of magnitude and rupture length given a point measurement of rupture displacement from a paleoseismic excavation. We also discuss some methods for correlation of events between paleoseismic sites and outline an approach for progress on the problem of event correlation. MODELING APPROACH FOR RUPTURE VARIABILITY A basic problem when using paleoseismic displacement measurements is the way to express the along-strike variability of rupture displacements without assuming the displacement profile of the rupture under study. Hemphill-Haley and Weldon (1999) approached this problem using data from 14 actual ruptures. First, they normalized the observed ruptures, dividing each by their individual length and average displacement, so that large and small ruptures could be compared directly. Second, they noted that while the shapes differed from rupture to rupture, the
variability itself was somewhat similar, and this was illustrated with two summary plots. In the first of these, they summed the individual profiles and renormalized the result into a single average. This average profile (Fig. 1A) has an approximately flat central third of its width and tapers to zero displacement in the outside thirds. It is plainly too regular to represent realistic surface ruptures. Averaging of normalized rupture profiles removes individual variability because points above average from one are matched by normalized displacements below average from another. In contrast, the histograms of individual normalized rupture measurements can be averaged and still retain realistic rupture variability (Fig. 1B). Histograms of slip have the virtue of specifying that variability occurs within the rupture without specifying how that variability is organized. That is, by drawing without replacement from the histogram of slip one may form any number of different, realistic looking ruptures. Importantly, the averaged histogram preserves exceptional measurements, such as where a length of rupture is three times the average, but weights all normalized slips in proportion to their relative frequency of occurrence in real rupture profiles. To model realistic rupture profiles, we begin from the idealized rupture shape suggested by Figure 1A and then add variability until it is similar to that of real ruptures. We start with the smooth shapes in Figure 2A, which are formed numerically by applying a cosine taper to rectangles scaled for length L and average displacement AD using the regression results of Wells and Coppersmith (1994): M = 6.93 + 0.82(log AD) ± 0.39,
(1)
M = 5.08 + 1.16(log L) ± 0.28.
(2)
and
Numerically, the following results depend on the particular regression coefficients in Equations 1 and 2, but the method requires only that a unique mapping exist between M, L, and AD. Slip variability is expressed in histogram form in Figure 2B for an M 7.4 earthquake rupture. Slips are concentrated in the highest displacement bin for any given profile because of its flat central portion. The smooth model of Figure 2A is included as an end member, but it would be unlikely in real life. As an approximation tool, however, Chang and Smith (2002) assumed a very similar profile shape for seismic hazard estimation along the Wasatch Front fault system of Utah. An important requirement in our inversion of slip variability is to include a broad enough range of magnitudes such that some earthquake in the range is considered certain to have caused the observed rupture. The range 6.0 ≤ M ≤ 8.0 is assumed in Figure 2A. To invert for the probability of event magnitude given an observed displacement, all possible magnitudes must be accounted for that could have led to that displacement. This accounting idea can be visualized by considering displacement observations between the horizontal lines of Figure 2A. For each
Rupture length and paleomagnitude estimates from point measurements of displacement
2
Displ. (m)
D/Dave
1.5 1
4 2
0.5
M7.4
0
50
100
0.2
0.4 0.6 0.8 Normalized length
1
Probability
0
200
250
300
350
0.4 0.2 0
B
0
0.5
1
0.05 1 2 3 Normalized displacement (D/Dave)
2
1
2.5
C
P(M|d = 2 ± 0.3 m) 0.5 0
0
1.5
Displ (m) Probability
0.1
B
P(d|M = 7.4)
0.6
0.15
0
150
SRL (km)
0.2 Frequency
A
6
0
0
Rupture profiles, 6 ≤ Mw ≤ 8
8
A
197
6
6.5
7
7.5
8
Moment magnitude
Figure 1. (A) Averaged rupture profile after normalizing 14 individual rupture profiles by length and average displacement. Averaging rupture profiles removes essential information about rupture displacement variability. (B) Averaging of individual normalized histograms of slip is different because large excursions from average are retained. All observations are weighted by the frequency with which they occur in the contributing rupture profiles. Based on the 14 contributing ruptures, displacements with D/Dave >2 are expected a few percent of the time. Both figures are redrawn from data of Hemphill-Haley and Weldon (1999).
Figure 2. (A) Model rupture profiles for M 6.0 to M 8.0 earthquakes in increments of 0.2 magnitude units. The basic shape is a tapered boxcar scaled for surface rupture length (SRL) and average displacement using data from Wells and Coppersmith (1994). Horizontal lines at displacements of 2 ± 0.3 m bound all the ways a 2 m displacement might be generated. (B) Histogram of normalized displacements for the M 7.4 example profile. This may be interpreted as the probability, P(d|M = 7.4), of some displacement being observed at random given the earthquake magnitude and slip model assumptions. (C) Probability distribution for magnitude given observed displacement d, P(M|d). The peak at M 7.4 corresponds to the large fraction of all 2 m displacements in the model M 7.4 rupture profile in A.
magnitude, one counts the length of rupture in the displacement range between the lines as its fraction of the total rupture length for that magnitude. Any earthquake is considered as likely as another to have produced ground rupture. For this (simplified) model, earthquakes of magnitudes 7.2 and smaller do not contribute displacements of 2.0 ± 0.3 m, magnitude 7.4 is the majority contributor because its flat maximum value falls in this range, and larger earthquakes have displacements near 2 m, but only on their tapered ends. Figure 2C shows the individual fractions of length near 2 m in the form of a probability distribution. For the slip model in Figure 2A, Figure 2C quantifies probabilities that any given magnitude event was responsible for the observation of a 2 m displacement. The next step is to add variability to the slip profile until it approximates the observation-based histogram of Figure 1B. We
add slip variability by assuming that it is normally distributed around the basic rupture shape of Figures 1A and 2A. In Figure 3A, variability is assumed to have a standard deviation of 0.15 times the average displacement. The variability histogram (Fig. 3B) is broader than in Figure 2B, but it is still too narrowly peaked around the average to be realistic. The probability of magnitude for a given displacement (Fig. 3C) is slightly broader, but it is still too focused on magnitude 7.4. Figure 4 shows rupture profiles and histograms with slip variability of 0.50 and 0.60 times the average displacement. Plainly, the resulting rupture profiles are too extreme to match realistic ruptures, but we emphasize that the histogram of variability (Figs. 4B and 4D), and not the rupture profile itself, is the figure of merit. The slip histograms of both are similar in shape and ratio of maximum to average displacement (Dmax and Dave, respectively) to the observation-based
198
Biasi et al.
A
10 5 0
0
50
100
150
200
250
300
350
B
P(d|M = 7.4)
2
0
20
0.1 0
40 60 80 Rupture length (km)
0.20 0
0.5
1
1.5
2
2.5
3
3.5
Displ (m) Probability
4
0
1
C
P(M|d = 2 ± 0.3 m)
Probability
Probability
SRL (km) 0.2
A
6 Slip (m)
Displ. (m)
Mw=7.4, s.d. = 0.5*Avg displacement
Rupture profiles, 6 ≤ Mw ≤ 8
15
100
Avg displacement: 2.19 m
B 0.10
0.5
0 0
6
6.5
7
7.5
0
8
Moment magnitude Figure 3. (A) Rupture profiles from Figure 2A with Gaussian variability of 0.15 times the average displacement. Negative slips are set to zero. Variability is calculated for every kilometer for ruptures over 100 km in length, and every 0.5 km for shorter ruptures. (B) Probability histogram of displacements for the M 7.4 case in A. (C) Distribution P(M|d) for d = 2 ± 0.3 m and rupture variability as in A. SRL—surface rupture length.
2
4 6 8 Slip (m) Mw=7.4, s.d. = 0.6*Avg displacement
Slip (m)
6
C
4 2 0
0
20
40 60 80 Rupture length (km)
Figure 4. (A) Rupture profile for the M 7.4 case in Figure 2A with variability of 0.50 times the average displacement. For clarity, only one rupture profile is plotted, but as for Figures 2 and 3, 6.0 ≤ M ≤ 8.0 ruptures would be used to calculate P(M|d). (B) Histogram of A. (C) Rupture profile with variability of 0.60 times average displacement M 7.4 case. (D) Histogram of C. Variability of 0.6AD was used for subsequent plots. Greater and lesser degrees of variability were considered, but those shown here gave the closest match to the histogram from real data.
Probability
0.15
100
Avg displacement: 2.19 m
D
0.10 0.05 0
0
2
4 Slip (m)
6
8
Rupture length and paleomagnitude estimates from point measurements of displacement
Probability
Probability
Probability
Probability
Probability
averaged histogram in Figure 1B. Rupture profiles with a standard deviation of 0.60 times the average displacement were considered a better match and are used in subsequent figures. Figure 5 presents probability distribution functions for paleomagnitude given observed displacements, p(M|d), in the range dobs = 1–6 m. Ten realizations for each case are plotted to show the impact of possible variations in the random component of the rupture profile. The probability distributions conform to the qualitative thought experiment described in the introduction. On
0.2 P(M| d
obs
= 1)
A
0.1 0 0.4
7 P(M|dobs = 2)
7.5
7 = 3)
7.5
7 P(M|dobs = 4)
7.5
7 P(M|dobs = 6)
7.5
8
B
0.2 0 0.4
P(M|d
8
C
obs
0.2 0 0.4
8
D
0.2 0 0.4
8
E
199
observing a 1 m displacement, a likely range of earthquakes of, perhaps, M 6.8–7.2 would be expected, while reserving smaller probabilities that the source may have been somewhat smaller or larger. The decline in probability for smaller earthquakes reflects the improbability that a random displacement measurement would happen to be taken near its peak. The decline at larger magnitude reflects the expectation that larger earthquakes should produce larger displacements and that a smaller fraction of the total would be near 1 m in length. Geologic or other considerations may eliminate some part of the maximum range of magnitudes in Figure 5. For example, the total length of a fault system may put an upper limit on magnitude. This can be accommodated easily and does not necessitate recomputation of Figure 5 p(M|d) plots. The case of 6 m dobs illustrates the consequences of truncating the distribution at M 8. Lacking larger magnitudes to which to ascribe the 6 m observation, the computations distribute the probability among magnitudes that are available, and therefore increase their individual weights. If a fault under study is not thought capable of earthquakes above some size, the probability distributions can be truncated at that point and the curves renormalized to arrive at revised probabilities among lower-magnitude events. PROBABILITY OF RUPTURE LENGTH Probabilities of rupture length are a derivative product of p(M|d) and are not a separate calculation from length versus magnitude data. Scaling of magnitude to length (Eq. 2) is applied to p(M|d) curves in Figure 5. Thus, probability distributions for the (log of) surface rupture length have the same shape as p(M|d) curves. To anticipate their use for correlation between sites, Figure 6 shows the probability of length as a complementary cumulative distribution. This allows them to be interpreted in terms of the probability that the rupture is at least the ordinate value in length. PALEOSEISMIC EVENT CORRELATION
0.2 0
7 7.5 Moment magnitude
8
Figure 5. Probability distribution functions for paleomagnitude given an observed displacement. Plots show 10 realizations of the random variation. Small “+” symbols mark the means by magnitude. One-σ ranges are visible where individual curve variations are large. (A) The probability distribution of magnitude given a measurement within 30 cm of 1 m, p(M|0.7 ≤ dobs ≤ 1.3). The range from M 6.8 to M 7.2 is most likely, but with decreasing probabilities, smaller and larger earthquakes might have caused the observed rupture. (B) Probability p(M|dobs ≈ 2 m). Compared to dobs ≈ 1 m, smaller magnitudes are predictably down-weighted. (C) Probability p(M|dobs ≈ 3 m). (D) Probability p(M|dobs ≈ 4 m). (E) Probability p(M|dobs ≈ 6 m).
To reconstruct the rupture history of a fault, paleoseismic studies are often conducted at more than one point along the fault. Earthquake dating with radiocarbon or other methods inevitably leads to uncertain estimates of event timing. One basis of correlation, which we term temporal correlation, associates an event at one site with an event at another on the basis of dating evidence (e.g., Fumal et al., 2002). A second basis for event correlation may be developed using the observation that ruptures generally have continuity for some length along the fault. When point displacements are available, the probabilities of rupture length calculated here can be combined with the probability that that rupture length spans the distance between the sites. This may be termed spatial correlation. Next, we examine the difficulty of correlation using dating evidence alone and then present the spatial correlation approach for event correlation made possible by P(L|d) results.
Probability
Probability
Probability
Probability
Probability
200
Biasi et al.
1 P(L|dobs = 1)
0.5 0 10
A 100
200
300
400
500
P(L|dobs = 2)
0.5 0 10
B 100
200
300
400
P(L|dobs = 3)
0.5 0 10
C 100
200
300
400
500
P(L|dobs = 4) 0.5 0 10
D 100
200
300
400
500
P(L|dobs = 6) 0.5 0
E 0
100 200 300 400 Surface rupture length (km)
500
Figure 6. Probability of rupture length given a paleoseismic displacement observation, P(L|d = dobs): (A) dobs = 1 m, (B) dobs = 2 m, (C) dobs = 3 m, (D) dobs = 4 m, and (E) dobs = 6 m. Numerically, P(L|d) is the cumulative probability of the distribution functions in Figure 5, with exponential scaling from magnitude to length from Equation 2. Other M-L regressions would adjust the scaling but not the shape of the curves.
Temporal Correlation Dating evidence alone does not provide a physical basis for correlation of paleoseismic rupture evidence between sites. Recent efforts have certainly added precision in traditional methods of dating events, by improving the precision of the dates themselves (Sieh et al., 1989), and by adding additional information such as stratigraphic ordering and estimated separations between layers (Biasi and Weldon, 1994; Bronk-Ramsey, 1995). Narrower event dates often work better to exclude correlations, e.g., where no overlap in ages remains. Noncorrelation is useful as a constraint on rupture extent but does not guide the affirmative case for correlation.
A variety of methods have been used in attempts to constrain event ages in the presence of dating overlap. These methods are summarized for the purpose of side-by-side evaluation in Figure 7. The first two methods (Figs. 7A and 7B) are based on the assumption that the true earthquake date is normally distributed and that the paleoseismic event dates are uncertain individual estimates of it. Figure 7A applies the well-known “Z” statistic (e.g., Davis, 1986) to assess the probability of contemporaneity (McCalpin, 1996). The Z-statistic is designed to test the hypothesis that two samples are different under the assumption that there is only one underlying population. The difficulty in applying the Z-statistic as a correlation tool is seen in its equation: Z = | T1 − T2 | / σ 21 + σ 22 , where T1 and T2 are mean ages of events at sites 1 and 2, respectively, and σ1 and σ2 are their respective standard deviations. Small values of Z mean that one cannot say that the two events are different. One can have a small value of Z if the mean ages coincide, regardless of the event dating (im)precision, or if the standard deviations are large compared to the difference in means. Under the second condition, the less one knows about the event dates, the larger is the probability of contemporaneity. From this, we conclude that the Z-statistic should not be viewed as the probability that the two earthquake dates are from the same event, but rather that (1 – P[contemporaneity]) should be used as the confidence that the events differ. Figure 7B illustrates a second approach based on the normal distribution, where the two event dates for correlation are assumed to be uncertain estimates of a single underlying distribution. Using established methods, one may combine the two estimates into a most likely, normally distributed event date (solid line). In regard to event correlation, a few observations are warranted. First, this approach assumes correlation, i.e., given that these are the same event, what do we say about the date of it? Second, the new mean estimate will be between the two contributing event means and may not give either contributing date much likelihood of actually being the right date. A paleoseismologist might prefer one event date over the other, for example, if one is dated by in situ organic samples and the other relies on detrital and reworked charcoal. Third, the resulting estimate is controlled primarily by the distance between the contributing means and the uncertainty of the poorly known date (Fig. 7B). The estimate is little improved by improved dating precision in the contributing dates. The maximum likelihood approach might be used to automate combination of event dates with similar uncertainties, but it should not be interpreted in terms of correlation, per se. The “event window” approach (Fig. 7C) is perhaps the most commonly used. Date ranges are plotted as bars, and the overlap is interpreted as the earthquake date range. This approach neglects any internal structure of the contributing probability distribution functions, but it has certain convenient properties. For example, when two event dates are being compared, the result is largely controlled by the more precise one. The approach also
Rupture length and paleomagnitude estimates from point measurements of displacement
Probability of contemporaneity
1 0.8 0.6 0.4 0.2 0 0
Annual probability
A
“Z” statistic Z = (T1 - T2) / sqrt(s 2 + s 2)2 Small Z, large overlap
1 2 “Z” statistic
3
B
0.04 0.03
We4
0.02
Comb V
0.01 0 1400
1450
1500
1550
1600
1650
1700
1750 1800
201
Figure 7. Graphical summary of methods for quantifying temporal correlation of paleoseismic earthquake dating evidence. (A) Overlapping Gaussian distributions, classical approach (e.g., McCalpin, 1996). The Z-statistic method is designed to test the hypothesis that the two events differ, not that they are the same. (B) Overlapping distributions, likelihood approach. Paleoseismic date distributions labeled V and We4 (example events from two adjoining sites of the San Andreas fault; Biasi et al., 2002) are assumed to be two uncertain estimates of the true date of a single earthquake. The width of the combined date estimate (solid line) is controlled most by the separation between mean dates of the contributors, and secondarily by their precision. (C) “Event window” method. Date ranges (say, 95%) are plotted as bars, and the overlap (shaded time range) is interpreted as the earthquake date range. (D) Ad hoc weighted overlap methods. An ad hoc event distribution is formed from the minimum overlap of two event probability distribution functions (PDF, dark central shaded region). Event probability distribution functions V and We4 in B overlap by 0.20. Note that the overlap tends to be smaller if one of the dates is well defined. A more formal approach would be to take the product of the two probabilities one bin at a time and interpret the sum of products as the probability that the events are the same. Of the methods presented here, this approach is closest to a quantitative correlation method. However, none of these temporal correlation strategies is without drawbacks.
Calendar year
Event 1 date range
C Event 2 date range
D Event 2 date PDF Event 1 date PDF
Minimum overlap (Min(E1, E2))
allows the combination of several dates, and it is simple to apply. The event window approach, however, assumes correlation, and thus is not a way of testing for it. Also, by assuming correlation, the overlapping portion may be in the low-probability tails of both contributors, meaning that the result may poorly represent the available dating evidence. Two estimates of a combined event date may be proposed that, unlike the aforementioned methods, do use the detailed shape of the event distributions (Fig. 7D). In the first, the event distribution of the single true event is taken one bin at a time to be the lesser value of the overlapping contributors. The minimum overlap estimate reflects the consistent portions of both dates, in analogy to set theory, and preserves some of the shape of the input event probability distribution functions. Numerically, it ranges from zero for no overlap to one for date functions of any
width that exactly overlie each other. Perhaps counterintuitively, the minimum overlap will be smaller if one of the event dates is precise than it would be for two imprecise dates with a wider time span of overlap. Thus, the minimum overlap method can provide a useful combined date distribution because it focuses on the date range consistent between individual contributors, but it assumes event correlation, and its theoretical justification is unclear. A second strategy for combining overlapping event probability distribution functions is to take their product. Intuitively, this is appealing because the product implements the logical AND condition between the two event dates—site 1, at time t1 and site 2 at the same time t1. While simple and attractive, some problems may be seen in an example. Consider two uniform event dates 60 yr apart, discretized in 10 yr bins and overlapping perfectly. There are six ways for t1 = t2, and 30 for t1 ≠ t2, so the probability of correlation would be 1/6. Numerically, this is much less than one might expect for date probability distribution functions that agree perfectly. This method of estimating correlation is also inconsistent, since the same 60-yr-apart events with 1 yr bins would have a 1/60 probability of correlation. Clearly, the probability should not depend on the discretization. It does, however, because the product of probabilities actually measures the area of the diagonal of a discrete joint probability as a fraction of the total area. The product means t1 is in the same bin with t2; so requiring one to be within 1 yr of another (the 60 bin case) is a higher standard (less probable) than requiring it to be within 10 yr. On the other hand, the product does identify times where the joint probability is highest. The product of probabilities thus has a better defined mathematical basis, but it does not directly address the problem of event correlation.
202
Biasi et al.
As this review shows, formalizing the probability of temporal correlation is problematic. The product of overlapping probabilities is best grounded in probability theory, but in a paleoseismic context, it is unlikely to ever yield high probabilities of correlation, even for precise event dates that overlie one another. This result points to the need to combine spatial and temporal evidence to go forward in developing event correlations and rupture histories. Spatial Correlation Spatial correlation is different from temporal correlation in that the continuity of rupture provides a physical basis for expecting that a rupture observed at a site will be seen some distance away from it. Using Figure 6, one may translate a rupture displacement measurement into probabilities of various rupture lengths. To convert a rupture length into a probability of correlation our approach is to assume that the paleoseismic site is located randomly within the rupture. Consider a rupture of length L and two paleoseismic sites separated by a distance x. The fraction of ways length L can include the first site and reach the second is (L – x)/L. This is the probability of correlation given rupture length, P(corr|L). For example, for a rupture length of 100 km, and an intersite spacing of 20 km, there is an 80% chance that the rupture will span both. There are then x/L ways the rupture could include the first site and not reach the second. Figure 8 plots P(corr|L) for site separations up to x = 100 km. Probabilities of correlation given an observed displacement are represented by the product P(corr|L)P(L|dobs).
1
20 km separation
Probability
0.8 0.6 0.4 100 km
0.2 0 0
100 200 Rupture length (km)
300
Figure 8. Probability of spatial correlation for a given rupture length. Curves are for intersite separations from 20 to 100 km in 20 km intervals. The actual probability of finding the rupture at the second site is somewhat lower because there is no guarantee that evidence was preserved there, and if it was, there is a finite chance that the second excavation will not develop the necessary exposures.
Some level of both spatial and temporal evidence for correlation will always be required for confident association of rupture observations between paleoseismic sites. The new aspect here is an improved ability to combine evidence from paleoseismic sites with different strengths. Specifically, some sites are well suited for the measurement of displacements, but they have weak dating control. Others with multiple organic-rich layers may provide better dating control but with less constraint on rupture displacement. The value for fault characterization of combining the two types of evidence was developed by Weldon et al. (1996). Quantifying P(L|dobs) and P(corr|L) are two steps forward in fulfilling that vision. DISCUSSION None of the randomized rupture histograms exactly matches the histogram of variability drawn from actual ruptures (Fig. 1B). On the other hand, had Hemphill-Haley and Weldon (1999) considered different earthquakes, a somewhat different histogram would have been produced. The essential quality for the model of rupture variability is that it covers the spectrum of possible D/Dave one may find in the field. By considering many runs (Figs. 5 and 6), the reader may gain some idea of the sensitivity of the results to rupture variability, or equivalently, to changes in the choice of ruptures that contribute to the histogram of observed D/Dave. The magnitude range we included, 6.0 < M < 8.0, is not the full range of possible earthquakes, but it was selected to encompass the range of ruptures likely to be found in practice. The way in which average displacement increases with magnitude for the largest earthquakes is controversial (Hanks and Bakun, 2002; Scholz, 2002; Romanowicz, 1992; Wesnousky, 1994; Shaw, 2009). With respect to our method, the essential quality is a one-to-one mapping of magnitude to average displacement and magnitude to length. Therefore, particular criticisms of the Wells and Coppersmith (1994) regressions and the state of the momentlength debate can be separated from the method itself. The lower magnitude range we consider is likewise not set at the smallest earthquake known to have ever produced ground rupture. Figure 5A shows, however, that even for a 1 m displacement, inclusion of smaller magnitudes would have little effect on the resulting probabilities. Below the magnitude at which its maximum displacement is less than dobs, that magnitude does not affect p(M|d) or P(L|d). The low-magnitude cutoff in each p(M|d) plot is related roughly to the magnitude one would infer from the magnitude–maximum displacement regression of Wells and Coppersmith (1994, their figure 10). Variability in Figure 4B and in the maximum displacement data set itself means the correspondence is not exact. Two complications arise if one were to attempt to extend the p(M|d) and P(L|d) plots to smaller-displacement measurements than 1 m. One complication is that with decreasing size, earthquakes become increasingly variable in their surface expression. For example, shallow events might reasonably match magnitude-length and magnitude–average displacement regressions,
Rupture length and paleomagnitude estimates from point measurements of displacement while deep events can preserve moment-area scaling (Hanks and Bakun, 2002) with a short surface expression. Average displacement is not perfectly correlated to surface rupture length at any average displacement, but below ~50 cm, the correlation is nearly uninterpretable (Wells and Coppersmith, 1994, their figure 13). A second complication is suggested in the first, namely, that not all earthquakes, especially in the M <~6.5 range, produce surface rupture. To include smaller observed displacements, the probability of producing ground rupture must be factored in. Numerically, it could be included as a reweighting of the p(M|d) curves in Figure 5, but doing so is beyond our present scope (see, however, Biasi and Weldon, 2006). The probability distributions for magnitude and average displacement were developed from the assumption that the observed displacement was drawn at random from within the rupture. This assumption is important because measuring at random within the rupture and sampling at random from the distribution of displacements for a particular magnitude (Figs. 4B and 4D) are mathematically equivalent, and they underlie the accounting concept illustrated in Figure 2 (horizontal lines). However, larger displacements are more likely to be preserved and stand out as targets for paleoseismic study. It is not obvious how to adjust our analysis to accommodate a dobs sample known to be biased. One can at least say that if the measured displacement is known to be above average, the magnitude from Figure 5 will be an overestimate. Additional knowledge about previous ruptures can also affect probabilities of correlation for a given length. The P(corr|L) model in Figure 8 assumes that the paleoseismic site with the displacement observation is randomly placed within the rupture. This assumption would have to be adjusted if one or other end of the rupture is known. Likewise, if a segmentation model is used to define lengths of possible ruptures (e.g., Chang and Smith, 2002), then the model underlying Figure 8 would not apply. In both cases, modifications to P(corr|L) could be suggested, but excepting perhaps the most recent event, it would be unusual to actually know enough about previous ruptures as to make it worthwhile. CONCLUSIONS We show that paleoseismic displacement observations can be used to develop quantitative estimates of paleomagnitude and paleorupture length. Estimates are in the form of probability distributions that encompass the intrinsic variability of rupture displacements. Probabilities of length can be combined with a simple model of probability of correlation given a rupture length to open the way to quantitative estimates of spatial correlation of earthquake ruptures between paleoseismic sites. Spatial evidence for correlation can be combined with temporal (i.e., dating) evidence to provide higher confidence of correlations than either method alone. Being able to combine evidence opens the possibility that the best site for characterizing a fault might be two sites, one where dating is accessible from layered organicrich stratigraphy and another where slip per event is available but dating evidence is scant.
203
ACKNOWLEDGMENTS This research was supported by the Southern California Earthquake Center (SCEC). SCEC is funded by National Science Foundation Cooperative Agreement EAR-0106924 and U.S. Geological Survey Cooperative Agreement 02HQAG0008. This is SCEC contribution 944. REFERENCES CITED Biasi, G., and Weldon, R.J., II, 1994, Quantitative refinement of calibrated 14 C distributions: Quaternary Research, v. 41, p. 1–18, doi:10.1006/qres .1994.1001. Biasi, G.P., and Weldon, R.J., II, 2006, Estimating surface rupture length and magnitude of paleoearthquakes from point measurements of rupture displacement: Bulletin of the Seismological Society of America, v. 96, p. 1612–1623, doi:10.1785/0120040172. Biasi, G.P., Weldon, R.J., II, Fumal, T.E., and Seitz, G.G., 2002, Paleoseismic event dating and the conditional probability of large earthquakes on the southern San Andreas fault, California: Bulletin of the Seismological Society of America, v. 92, p. 2761–2781, doi:10.1785/0120000605. Bronk-Ramsey, C., 1995, Radiocarbon calibration and analysis of stratigraphy: The OxCal program: Radiocarbon, v. 37, p. 425–430. Chang, W.L., and Smith, R.B., 2002, Integrated seismic-hazard analysis of the Wasatch Front, Utah: Bulletin of the Seismological Society of America, v. 92, p. 1904–1922, doi:10.1785/0120010181. Davis, J.C., 1986, Statistics and Data Analysis in Geology: New York, John Wiley and Sons, 646 p. Fumal, T.E., Rymer, M.J., and Seitz, G.G., 2002, Timing of large earthquakes since A.D. 800 on the Mission Creek strand of the San Andreas fault zone at Thousand Palms Oasis, near Palm Springs, California: Bulletin of the Seismological Society of America, v. 92, p. 2841–2860, doi: 10.1785/0120000609. Grant, L.B., and Lettis, W.R., 2002, Introduction to the special issue on paleoseismology of the San Andreas fault system: Bulletin of the Seismological Society of America, v. 92, p. 2551–2554, doi:10.1785/0120000600. Hanks, T.C., and Bakun, W.H., 2002, A bilinear source-scaling model for M– log A observations of continental earthquakes: Bulletin of the Seismological Society of America, v. 92, p. 1841–1846, doi:10.1785/0120010148. Hemphill-Haley, M.A., and Weldon, R.J., II, 1999, Estimating prehistoric earthquake magnitude from point measurements of surface rupture: Bulletin of the Seismological Society of America, v. 89, p. 1264–1279. McCalpin, J.P., 1996, Paleoseismology: San Diego, Academic Press, 588 p. Romanowicz, B., 1992, Strike-slip earthquakes on quasi-vertical transcurrent faults: Inferences for general scaling relations: Geophysical Research Letters, v. 19, p. 481–484, doi:10.1029/92GL00265. Salyards, S.L., Sieh, K.E., and Kirschvink, J.L., 1992, Paleomagnetic measurement of non-brittle coseismic deformation across the San Andreas fault at Pallett Creek: Journal of Geophysical Research, v. 97, p. 12,457–12,470, doi:10.1029/92JB00194. Scholz, C.H., 2002, The Mechanics of Earthquakes and Faulting: Cambridge, UK, Cambridge University Press, 471 p. Shaw, B.E., 2009, Constant stress drop from small to great earthquakes in magnitude-area scaling: Bulletin of the Seismological Society of America, v. 99, p. 871–875, doi:10.1785/0120080006. Sieh, K.E., 1984, Lateral offsets and revised dates of large earthquakes at Pallett Creek, California: Journal of Geophysical Research, v. 89, p. 7641–7670, doi:10.1029/JB089iB09p07641. Sieh, K., Stuiver, M., and Brillinger, D., 1989, A more precise chronology of earthquakes produced by the San Andreas fault in southern California: Journal of Geophysical Research, v. 94, p. 603–623, doi:10.1029/ JB094iB01p00603. Weldon, R.J., II, McCalpin, J.P., and Rockwell, T.K., 1996, Paleoseismology of strike-slip tectonic environments, in McCalpin, J.P., ed., Paleoseismology: San Diego, California, Academic Press, p. 271–329. Weldon, R.J., II, Fumal, T.E., Powers, T.J., Pezzopane, S.K., Scharer, K.M., and Hamilton, J.C., 2002, Structure and earthquake offsets on the San Andreas fault at the Wrightwood, California paleoseismic site:
204
Biasi et al.
Bulletin of the Seismological Society of America, v. 92, p. 2704–2725, doi:10.1785/0120000612. Weldon, R.J., Fumal, T.E., Biasi, G.P., and Scharer, K.M., 2005, Past and future earthquakes on the San Andreas fault: Science, v. 308, p. 966–967, doi:10.1126/science.1111707. Wells, D.L., and Coppersmith, K.J., 1994, New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface dis-
placement: Bulletin of the Seismological Society of America, v. 84, p. 974–1002. Wesnousky, S., 1994, The Gutenberg-Richter or characteristic earthquake distribution, which is it?: Bulletin of the Seismological Society of America, v. 84, p. 1940–1959. MANUSCRIPT ACCEPTED BY THE SOCIETY 7 DECEMBER 2010
Printed in the USA
Contents Introduction: Geological criteria for evaluating seismicity revisited: Forty years of paleoseismic investigations and the natural record of past earthquakes Franck A. Audema rd M . a nd Alessandro Ma ria M ichetti L Paleoseismicity of a low-slip-rate normal fault in the Rio Grande rift, USA: The Calabacillas fault, Albuquerque, New Mexico J.P. McCa lpin, J .B.J. Harriso n, G .W. Berger, a nd H .C. Tobin 2.. Late Quaternary earthquakes on the Hubbell Spring fault s ystem, New M exico, USA : Evidence (or noncharacteristic ruptures o(intrabasin faults in the Rio Grande rift Susan S. Olig, M artha C. Eppes, Steven L. Forman, Dav id W. Love, and Bruce D. Allen 3... Large-magnitude late Holocene seismic activity in the Pereira-Armenia region, Colombia Claudia Patric ia Lal inde P., Gloria Elena Toro, Andres Velasquez, a nd Franck A . Audema rd M.
4.. Evidence of Holocene compression at Tulud, along the western foothills of the Central Cordillera of Colombia M yriam C. Lopez C. and Franck A. Audemard M .
5... Style and timing o[late Quaternary faulting on the Lake Edgar fault, southwest Tasmania, Australia: Implications for hazard assessment in intracratonic areas Dan Cla rk, M att Cupper, Mike Sa ndifo rd, a nd Kevin Ki ernan
6... Multiple-trench investigations across the newly ruptured segment ofthe El Pilar fault in northeastern Venezuela after the I997 Cariaco earthquake Franck A Audema rd M
Z.. Lake sediments as late Quaternary paleoseismic archives: Examples in the northwestern Alps and clues for earthquake-origin assessment of sedimentary disturbances Chri sti an Beck
8.. Late Pleistocene-early Holocene paleoseismicity deduced (rom lake sediment deformation and coevallandsliding in the Calchaquies valleys, N W Argentina Reg ina ld L. Hermanns and Samuel N iedennann
2.. Rupture length and paleo magnitude estimates (rom point measurements of displacementA model-based approach Glenn Biasi, Ray J . Weldon II, and Kate Sc harer
. . THE GEOLOGICAL SOCIETY • OF AMERICA®
ISB N 978-0 -8 137-2479-9
3300 Penrose Place • P.O. Box 9140 Boulder, CO 80301-9140, USA
9 780813 7 2 4 7 9 9
