T51D-0754
Deformation in the South Australian Craton: >1 Ga of intraplate activity
Active intraplate seismicity has often been related to either anomalously high heat flow within the crust or pre- existing zones of weakness. Located in the central part of the Australian plate, more than 1000 km from the nearest active plate-boundary, the South Australian Craton has a long history of intraplate deformation from the Neoproterozoic to the Quaternary. Ongoing intraplate deformation occurs in a zone of moderate seismic activity (magnitudes <6.0) that generally follows the trend of the Neoproterozoic Adelaide "Geosyncline". Anomalously high heat flow observations have been used to suggest that seismicity is concentrated where the crust is thermally weakened by high heat producing granites within the buried cratonic basement. Here we use surface geology together with gravity and magnetic data to produce cross-sections that constrain the nature of buried basement structures in this seismically active area. A spatial analysis of earthquakes suggests seismicity is localized on basement structures that juxtapose rocks with significantly different petrophysical properties. Hence seismicity is not solely related to high-heat producing granites in the South Australian Craton. For example, a splay of seismicity along the Kalinjala shear zone suggests that this ~1.7 Ga shear zone remains a zone of crustal weakness that is unrelated to elevated heat flow. Our observations highlight the strong link between the architecture of the basement and intraplate seismicity in cratonic areas.
T51D-0755
Intraplate deformation on north-dipping basement structures in the Northern Gawler Craton, Australia: reactivation of original terrane boundaries or later intra-cratonic thrusts?
Multiple intraplate orogenic events have deformed Neoproterozoic to Carboniferous sedimentary sequences that cover the Archean to Mesoproterozoic basement of the northern Gawler Craton, Australia. These intraplate orogenies reactivated north-dipping basement penetrating faults that are imaged on seismic reflection profiles. These north-dipping structures pre-date Neoproterozoic deposition but their relationships to significant linear magnetic and gravity anomalies that delineate unexposed Archean to Early Mesoproterozoic basement terranes are unclear. The north-dipping structures are either terrane boundaries that formed during continental amalgamation or later faults, which formed during a mid- to late-Mesoproterozoic transpressional orogeny and cross-cut the original lithological terrane boundaries. We model magnetic and gravity data to determine the 3D structure of the unexposed basement of the northern Gawler Craton. These models are constrained by drill hole and surface observations, seismic reflection profiles and petrophysical data, such that geologically reasonable models that can satisfy the data are limited. The basement structures revealed by this modelling approach constrain the origin and significance of the north-dipping structures that were active during the later intraplate Petermann, Delamerian and Alice Springs Orogenies. These results have bearing on which structures are likely to be active during present-day intraplate deformation in other areas, including, for example, current seismic activity along similar basement structures in the Adelaide "Geosyncline".
T51D-0756
Natural Seismicity in NW Australia: Another Look at Continental Intraplate Earthquakes
Northwestern Australia hosts a considerable number of smaller seismic events, but also Australia's largest earthquake on record (Meeberrie, 1941, M7.9). In 2005 we deployed a small network of seismometers to record natural seismicity in the region, to understand the amount, magnitude, nature and distribution of these events, and to interpret these events within a tectonic and neotectonic context. The region is largely underlain by Precambrian material, ranging from Archaean craton to Proterozoic mobile belts, and truncated to the west by the lithospheric-scale Darling Fault. This fault separates the Precambrian crustal elements from the younger Palaeozoic-Mesozoic Carnarvon Basin to the east, has a strike length of over 1400 km. It is thought to have been active since at least the Proterozoic. Estimates of slip on some southerly sections of this fault exceed 12 km. We believe that these various crustal elements (basins, cratons, mobile belts) respond differently to seismicity based on their rheology, thickness and deformation history. We supplemented data from our network with traces from some of the instruments of the national network of Australia. This data set proved adequate for the calculation of fault-plane solutions of some 13 mini- and micro- events in the region. The eventsrecorded occurred along or close to known faults, although field data for movement or reactivation on these faults is sparse. Most of our calculated events occur in close proximity to mechanical boundaries between crustal elements, indicating that the mechanical response of these elements controls the location of seismicity. The transitional tectonic setting of this part of the Australian margin---a passive margin sandwiched between two active margin segments---indicates that the term "intraplate" may be an inappropriate description of these seismic events.
T51D-0757
GPS-geodetic monitoring of the South West Seismic Zone (SWSZ) of western Australia: progress after two observation epochs in 2002 and 2006
The Australian south-west seismic zone (SWSZ) is a northwest-southeast-trending belt of intra-plate earthquake activity that occurs in the southwest of Western Australia, bounded by 30.5S to 32.5S and 115.5E to 118E. This is one of the most seismically active areas in Australia, with nine earthquakes over magnitude 5.0 that have occurred in the SWSZ between 1968 and 2002. The largest of these was the M6.8 Meckering earthquake in 1968. Since the SWSZ lies as close as ~150 km from the ~1.4 million population of the Perth region, it poses a distinct seismic hazard. However, little is currently known about the magnitude and orientation of this deformation, and whether there is any associated ongoing surface expression. It is also not known how this intra-plate activity compares with that observed elsewhere in Australia or elsewhere on Earth. Earthquake activity recorded by Geoscience Australia over the past four decades suggests that the SWSZ could be deforming with strain rates between 10-9/yr and 10-8/yr, or with displacements between 0.1 mm/yr and 1 mm/yr across the 200km width of the currently active SWSZ. This estimate is derived by applying the Kostrov formula for a moment release of 3.3x1019 Nm in the 34 years from 1968 to 2002. Early geodetic studies of the SWSZ that used both terrestrial and Global Positioning System (GPS) techniques were inconclusive, due mainly to the imprecision of the technologies used in relation to the likely small amount of any surface deformation. Therefore, in 2002 a new 48-point campaign-reoccupation GPS network was established across the SWSZ to attempt to detect surface deformation, using ground-level forced-centred monuments. The first two observational epochs were in May 2002 and May 2006. In both surveys, the dual- frequency carrier-phase GPS data were collected continuously at each monitoring point over a 5-7 day observation period. For the both campaigns, after excluding outlying sites (due to equipment malfunctioning), the estimated internal horizontal precision was 1.0 mm; the vertical precision was generally better than 4 mm. Comparison of these two repeat surveys shows that strain rates in the order of 10-9/yr can, in principle, ultimately be resolved if the deformation across the SWSZ occurs uniformly, but that the four-year time interval between the existing two surveys is at present not adequate to detect deformation with confidence given the noise in the estimated coordinates at each epoch. Further repeat surveys after, say, 8 and 12 years should begin to reveal any observable deformation significantly in excess of this observational error. These preliminary results suggest that the higher range of estimated of strain rates (10-8/yr) is unlikely.
T51D-0758
Finite-element models on spatiotemporal variations in intraplate seismicity caused by postglacial unloading and rebound: Implications for active normal faults in the Basin and Range Province
The actively extending Basin and Range Province was covered by numerous pluvial lakes and glaciers on several of the higher ranges during the Last Glacial Maximum (Osburn and Bevis, QSR, 2001). The largest lakes were Lake Bonneville and Lake Lahontan, located in the eastern and western parts of the Basin and Range Province, respectively. Regression of these lakes at the end of last glacial period caused significant isostatic rebound of the lithosphere (Bills et al., JGR, 1994; Bills et al., JGR, 2007). The rebound associated with the regression of Lake Bonneville has been shown, using two-dimensional numerical models, to affect the stress field of the lithosphere and to cause a slip rate increase on the Wasatch normal fault (Hetzel and Hampel, Nature 2005). Here we use three-dimensional finite-element models of normal fault arrays to investigate spatiotemporal variations in the regional stress field and in the rate of normal faulting caused by glacial-interglacial variations of the surface load. Our models indicate that regression of Lake Lahontan but also of smaller lakes and glaciers alter the regional stress field and hence may ultimately affect the intraplate seismicity. Paleoseismological data from faults in the east-central and northern Basin and Range Province seem to support the idea of an increase in seismicity after the Last Glacial Maximum (Friedrich el al., JGR, 2003; Stickney and Bartholomew, BSSA, 1987; Wesnousky et al., JGR, 2005).
T51D-0759
Coastal land loss and gain as potential earthquake trigger mechanism in SCRs
In stable continental regions (SCRs), historic data show earthquakes can be triggered by natural tectonic sources in the interior of the crust and also by sources stemming from the Earth's sub/surface. Building off of this framework, the following abstract will discuss both as potential sources that might have triggered the 2007 ML4.2 Folkestone earthquake in Kent, England. Folkestone, located along the Southeast coast of Kent in England, is a mature aseismic region. However, a shallow earthquake with a local magnitude of ML = 4.2 occurred on April 28 2007 at 07:18 UTC about 1 km East of Folkestone (51.008° N, 1.206° E) between Dover and New Romney. The epicentral error is about ±5 km. While coastal land loss has major effects towards the Southwest and the Northeast of Folkestone, research observations suggest that erosion and landsliding do not exist in the immediate Folkestone city area (<1km). Furthermore, erosion removes rock material from the surface. This mass reduction decreases the gravitational stress component and would bring a fault away from failure, given a tectonic normal and strike-slip fault regime. In contrast, land gain by geoengineering (e.g., shingle accumulation) in the harbor of Folkestone dates back to 1806. The accumulated mass of sand and gravel accounted for a 2.8·109 kg (2.8 Mt) in 2007. This concentrated mass change less than 1 km away from the epicenter of the mainshock was able to change the tectonic stress in the strike-slip/normal stress regime. Since 1806, shear and normal stresses increased at most on oblique faults dipping 60±10°. The stresses reached values ranging between 1.0 KPa and 30.0 KPa in up to 2 km depth, which are critical for triggering earthquakes. Furthermore, the ratio between holding and driving forces continuously decreased for 200 years. In conclusion, coastal engineering at the surface most likely dominates as potential trigger mechanism for the 2007 ML4.2 Folkestone earthquake. It can be anticipated that the mainshock nucleated at shallower depth (<500 m) near the Paleozoic surface a) where differential stresses are generally maximum and b) because earthquakes in aseismic regions are generally overestimated by 88% due to sparse instrumental coverage. The latter was suggested by recent research on shallow seismicitiy (<10 km) in SCRs in northeastern USA and eastern Canada. Data of the focal mechanism provided by the British Geological Survey (BGS) confirm fault zone orientations of 326°/74° (strike-slip fault component) and 71°/48° (normal fault component). http://www.earthquakes.bgs.ac.uk/earthquakes/reports/folkestone/folkestone_28_april_2007.htm
T51D-0760
Precambrian mylonitic belts and present-day stress field: the role played in the reactivation of the Pernambuco lineament, NE Brazil
The Pernambuco lineament is a steeply dipping continental-scale ductile shear zone rooted within the Precambrian lithosphere of intraplate northeastern Brazil. It formed during the Brasiliano orogeny ~600 Ma and reactivated during the Pangea breakup in the Cretaceous, when it controlled fault propagation and sediment accumulation in several rift basins. The region is now under an ~E-W-trending horizontal compression and ~N-S- trending extension, and faulting occurs predominantly in a strike-slip regime. We investigated two aftershock sequences and the preexisting tectonic fabrics along the Pernambuco lineament in order to evaluate the role of these fabrics in the coseismic reactivation of continental-scale structures. In 1991 and 2002 coseismic reactivations nucleated along an ~E-W and NE striking mylonitic branchs. Both fault segments reactivated the mylonitic foliation and form part of a major system fuelled by an ~E-W-trending compression and ~N-S-trending extension. We conclude that the interplay between the present-day stress field and preexisting fabrics controls seismogenic fault location, attitude, and kinematics. The Pernambuco lineament is an example of a long-lived continental scale structure, where selective reactivation has occurred. Other shear zones in the region also show a long history of brittle reactivation and present similar orientation in relation to the present-day stress axes. We suggest they are dormant structures prone for reactivation under the present-day stress field. http://www.dfte.ufrn.br
T51D-0761
Stress Channelling and Partitioning of Seismicity in the Charlevoix Seismic Zone, Canada
The Charlevoix seismic zone in the St. Lawrence valley of Québec is historically the most active in eastern Canada. The structurally complex region comprises rift faults formed during the opening of the Iapetus Ocean, superimposed by a 350 Ma meteorite impact structure, resulting in a circular highly fractured zone. Although seismicity is localized along two steeply dipping planar rift-parallel zones, previous work indicates that most of the large-scale rift faults appear to bound seismicity rather than generate earthquakes themselves. In order to gain insight into the mechanics of the partitioning of this seismicity, a simple two-dimensional model of the Charlevoix seismic zone was built using the finite difference code FLAC. The rift-related faults are represented by discontinuities, which are assigned various frictional strength parameters. The heavily fractured impact structure is represented by an elastic continuum of reduced modulus. Boundary displacements are used to generate a regional stress field in the direction of tectonic loading. Given a high strength, the rift faults have little effect on the stress patterns. Stress trajectories naturally flow around the region of reduced elastic modulus, leaving the fractured area with lower stresses than the background level. However, when the rift faults have a low strength, they are unable to support stress trajectories inclined to them, due to the resolved shear stress exceeding their strength. This prevents trajectories from diverting out of the rift, effectively channelling higher magnitude stresses into the region of the impact structure between the faults than would naturally occur. Low- strength bounding faults can thus explain the localization of seismicity into linear bands, rather than distributed seismicity throughout the impact structure. It also explains how the rift faults act as boundaries to seismicity. These results indicate that the interplay between faults of varying strength and zones of differing elastic modulus can give rise to complicated stress patterns, and can explain many of the seismicity patterns observed in the Charlevoix seismic zone.
T51D-0762
Deep Continental Crustal Earthquakes and Lithospheric Structure: A Global Synthesis
The distribution of earthquake depths within the continental crust defines the seismogenic thickness (TS), over which at least some part of crustal deformation is accommodated by rapid release of stored elastic strains. Intraplate continental seismicity is often thought to be restricted to the upper crust where TS is within the range of 15 to 20 km. This appears consistent with a lithospheric strength profile involving a weak, ductile lower crust located beneath a stronger, brittle upper crust. With the assumption of a strong uppermost mantle lid, this is often referred to the Jelly Sandwich model of lithosphere rheology. Studies in many places, however, document lower crustal earthquakes beneath continents in apparent disagreement with the model. We explore this and related issues through a survey of where and in what tectonic settings deep intraplate earthquakes are well documented in the continental crust. TS reaches Moho depth in many intraplate regions \--- Sierra Nevada, Colorado Plateau, East African and Baikal Rift Systems, North Island New Zealand, Tien Shan, and the Andean and Alpine forelands. A review of possible deformation mechanisms which could control continental earthquake depth and facilitate seismicity beneath the brittle-ductile transition suggests that the influence of fluids is the only mechanism capable of encouraging earthquake occurrence throughout the continental crust at any tectonic setting. Surface derived fluids can induce pore fluid pressure changes to depths of 25 km and melt-reactions can induce earthquakes at depths throughout continental crust. On a global scale, fluid-enhanced embrittlement is not limited by depth or tectonic environment. We find that deep crustal earthquakes occur where the lithosphere is in a transitional state between primarily stable (e.g., shields) and highly deformed (e.g., U.S. Basin and Range or Southern California). Observations of relative intensity of tectonic deformation and regional percent strain measurements indicate that lower crustal earthquakes exist within continental regions experiencing youthful (generally Neogene to present) tectonism typically with < 15 % strain. We propose a model that describes relative continental deformation and corresponding seismogenic thickness. In stable seismogenic regions, earthquakes are limited to depths where near-surface derived fluids can induce activity and therefore TS < 25 km. As continental lithosphere passes into an intermediate state of deformation, fluids from the near surface and melts formed in situ or derived from the mantle can act to seismically activate the entire continental crust. As deformation continues, percent strains can reach 100 % or more, mantle lithosphere is typically thinned or absent, and the lower crust via heating and/or sufficient weakening becomes aseismic.
T51D-0763
Evidence for the Influence of Fluids for Intraplate Earthquake Genesis: The 2001 Enola, AR Earthquake Swarm
In 2001, almost 20 years after the occurrence of 20,000 earthquakes in the 1982 sequence, the Enola, AR, earthquake zone reactivated with another 2-month long sequence of 2,500 earthquakes. Waveform cross- correlation and double-difference relative hypocentral location methods reveal that the 2001 sequence occupied a crustal volume of 8 km3, at depths between 4-6 km. The swarm volume hosts two clusters of earthquakes at different depths that developed during two separate seismic episodes. We analyzed 57 SV/P & SV/SH amplitude ratios including 15 events for which focal mechanisms were calculated using available P, SV and SH wave polarities. Earthquake clustering in space, time and in the amplitude ratio domain is used to observe a change in source properties from deeper oblique to shallower strike-slip focal mechanisms. The focal mechanisms of the early, seismically deeper cluster show a strong thrust component but the later and shallower cluster source solutions show predominant strike slip motion. We examine an effective stress model to explain the observed change of minimum stress axis direction from the vertical to the horizontal. This behavior is consistent with excess fluid pressures at deeper hypocentral depths. Like artificially-induced earthquakes caused by fluid-injection, it is likely that some intraplate earthquake zones are influenced by natural perturbations in hydrologic state of the hypocentral crust. http://www.ceri.memphis.edu/people/rabak/tmp/agu07/
T51D-0764
Possible Non-volcanic Tremor Discovered in the Reelfoot Fault Zone, Northern Tennessee
A swarm of ~80 microearthquakes was fortuitously detected in 20, 14 second-duration long-offset vibroseis shotgathers collected for a seismic reflection experiment near Mooring, TN, directly over the Reelfoot fault zone on the afternoon of 16 November 2006. These natural events show up in the shotgathers as near-vertically incident P waves with a dominant frequency of 10-15 Hz. The reflection line was 715m in length consisting of 144 channels with a sensor spacing of 5m, 8Hz vertical geophones, and recording using a Geometrics 24bit Geode seismograph. Small variations in event moveout across the linear array indicate that the seismicity was not confined to the same hypocenter and probably occurred at depths of approximately 10 km. The largest events in the series are estimated to have local magnitudes of ~-1 if at 10 km distance from the array. This is about 2.5 magnitude units lower than the threshold for local events detected and located by the CERI cooperative network in the area. The seismicity rate was ~1000 events per hour based on the total time duration of the shotgathers. The expected number of earthquakes of ML greater than or equal to -1 for the entire central United States is only 1 per hour. This detection of microseismic swarms in the Reelfoot fault zone indicates active physical processes that may be similar to non-volcanic tremor seen in the Cascadia and San Andreas fault zones and merits long-term monitoring to understand its source.
T51D-0765
FOCUS: Fault Observatory for the Central United States
The mid-continent has a long, complex history of crustal modification and tectonism. Precambrian basement rocks record intense deformation from rifting and convergence that precedes accumulation of a thick sequence of Phanerozoic and recent sediments that constitute the present-day Mississippi Embayment. Despite its location far from the active North American plate margins, the New Madrid seismic zone of central U.S. exhibits a diffuse yet persistent pattern of seismicity, indicating that the region continues to be tectonically active. What causes this intraplate seismicity? How does the intraplate lithosphere support local, regional and plate-wide forces that maintain earthquake productivity in this supposedly stable tectonic setting? These long-standing scientific questions are the motivation behind the proposed establishment of a borehole geo-observatory in the New Madrid seismic zone. FOCUS (Fault Observatory for the Central U.S.) would allow an unprecedented look into the deep sediments and underlying rocks of the Embayment. The proposed drill hole would fill a critical need for better information on the geophysical, mechanical, petrological, and hydrological properties of the brittle crust and overlying sediments that would help to refine models of earthquake generation, wave propagation, and seismic hazard. Measurements of strains and strain transients, episodic tremor, seismic wave velocities, wave attenuation and amplification, heat flow, non-linear sediment response, fluid pressures, crustal permeabilities, fluid chemistry, and rock strength are just some of the target data sets needed. The ultimate goal of FOCUS is to drill a 5-6 km deep scientific hole into the Precambrian basement and into the New Madrid seismic zone. The scientific goal of FOCUS is a better understanding of why earthquakes occur in intraplate settings and a better definition of seismic hazard to benefit the public safety. Short-term objectives include the preparation of an International Continental Drilling Program (ICDP) preproposal, initiation of site selection activities, communication with the international scientific community via scientific meetings and presentations, and development of funding sources and proposals targeted at key agencies. Intermediate-term objectives include developing an ICDP drilling workshop proposal, investigating holes of opportunity and scientific characterization of alternative drilling sites, establishing an intermediate depth observation hole into the Paleozoic bedrock (~1 km), and preparing for an observatory dedication (not completion) on the 200th anniversary of the 1811- 1812 earthquakes. FOCUS has initially organized into a steering committee and 5 subcommittees for Geophysics and Monitoring, Drilling/Holes of Opportunity, Geology/Tectonics, Geotechnical, and Community Partners/Outreach and Education. http://www.ceri.memphis.edu
T51D-0766
Is There a Connection Between Seismicity and Deformation in the New Madrid and Wabash Valley Seismic zones?
We compare geodetic and geophysical data for two spatially connected intraplate seismic zones in the central U.S.: the Wabash Valley Seismic Zone (WVSZ) of southern Indiana and Illinois and the New Madrid Seismic Zone (NMSZ) of the Mississippi Valley. In both cases, regional seismic and potential field data provide evidence for high-angle, basement-penetrating, faults that define narrow, elongate Precambrian grabens that lie beneath relatively undeformed Paleozoic or Mesozoic rocks. While only the NMSZ has experienced large-magnitude earthquakes in the historical record, both areas have a Quaternary history including numerous moderate to large- magnitude events. They are separated by an enigmatic tectonic zone characterized by basement uplift, major Precambrian strike-slip and normal fault zones, and Mesozoic and Cenozoic magmatism. We examine data from a 56-site campaign GPS geodetic network in the southern Illinois Basin to infer present-day deformation in the WVSZ. We combine newly acquired data in 2007 with that from five previous GPS campaigns from the period 1997-2002. Results for the 1997-2002 period provide little evidence for statistically significant velocities for individual sites in the southern Illinois Basin. Average strains for the entire network, however, show marginally significant strains, with an orientation rotated 45° from the overall direction of intraplate stress in the U.S. mid-continent. We are currently processing results from the 2007 campaign, and anticipate that the signal/noise ratio will be considerably improved by this five-year extension of the observation period. In addition, we examine models that test the effect of the 1811-1812 New Madrid earthquakes on the near- and far-field strain and seismicity rates in the region through the processes of instantaneous elastic deformation in the lithosphere and associated postseismic viscoelastic flow in the asthenosphere. Our results indicate that significant changes in strain and seismicity rates in the southern Illinois Basin can persist for several hundred years following the New Madrid earthquakes. The seismicity rate can increase by as much as a factor of seven over the background rate in the near-field, but by a much smaller amount in the far-field. However, the effect on the modeled seismicity rates is highly dependent on the choice of lower-crust viscosity. We also investigate the possibility that the New Madrid earthquakes could modify seismicity or strain in the WVSZ by producing triggered slip on a buried fault in the Illinois Basin region. Our initial results demonstrate that elevated seismicity and strain in the WVSZ could result from aseismic slip triggered by viscous relaxation in the lower crust long after the New Madrid earthquakes.
T51D-0767
High-resolution Earthquake Location in the New Madrid Seismic Zone
We present high-resolution earthquake locations for the New Madrid Seismic Zone (NMSZ) using the double- difference method developed by Waldhauser and Ellsworth (2000). The NMSZ consists of three intersecting faults centered in the central United States and is one of the few places where earthquakes frequently occur away from a major plate boundary. The zone generates approximately 200 earthquakes per year, and double- difference relocation techniques prove well suited for this region because the distance between neighboring events is small and station coverage relatively dense. The initial dataset consists of 1394 earthquakes recorded between 2000 and 2006 by 197 stations. The catalog contains approximately 67,000 P-wave and 54,000 S-wave observations, which yields 480,000 differential times. Waveform cross-correlation of P and S-waves provides an additional 73,000 high-precision differential times. We use the 1D model generated for the NMSZ. One unique characteristic of this model is the high (>3.0) Vp/Vs ratio, which reflects the thick sediment coverage in the region. Tests using synthetic data indicate that this high ratio has little effect on locations. Following relocation, hypocenters align along individual segments of the seismic zone and provide a sharper image of the NMSZ faults. We establish the robustness of these results through the use of synthetic differential times. We will also present initial P-wave and Vp/Vs velocity models derived using the double-difference tomography approach of Zhang and Thurber (2003).
T51D-0768
Local Stress Concentrators in the Crust as the Cause of Intraplate Earthquakes
Ignoring induced earthquakes associated with anthropogenic activities, and low magnitude swarms (e.g. in Moodus, Connecticut, and Arkansas), we applied a three-fold approach to the study of the seemingly random occurrence of meaningful intraplate earthquakes (IPE), located within large rigid plates characterized by negligible rates of strain accumulation. Analysis of multidisciplinary data from 39 (M >= 5.0) earthquakes spanning 20 intraplate regions, mechanical modeling, and field observations support the contention that IPE are primarily associated with localized pockets of stress concentrators lying within the rigid plates. The commonly observed stress concentrators include intersecting faults with optimal geometry, intrusives having large rigidity contrasts with surrounding rocks, and stress pillows. Suitably oriented faults with kinks, intersections, jogs and step-overs, lying within crustal zones of weakness, account for nearly two-thirds of the larger IPE. Mechanical modeling and comparison with known examples of IPE show that fault systems suitably oriented with respect to SHmax are the most likely locations of meaningful IPE. Stress concentrations near shallow granitic plutons, observed in the South Carolina Piedmont, among other places, and supported by mechanical modeling, can account for M < 4.0 events. Stress pillows, which are large, anomalous, and dense bodies in the lower crust within rifts, are found to account for approximately 15 percent of the IPE. In each case, the regions of elevated stresses are very local, comparable to the dimensions of the stress concentrators, demonstrating the need for very dense (<5 km spacing), focused GPS observations, and the futility of the currently employed plate boundary-type configurations (approximately >10 km).
T51D-0769
Seismotectonic Models of the Three Recent Devastating SCR Earthquakes in India
During the last decade, three devastating earthquakes, the Killari 1993 (Mb 6.3), Jabalpur 1997 (Mb 6.0) and the Bhuj 2001 (Mw 7.7) occurred in the Stable Continental Region (SCR), Peninsular India. First, the September 30, 1993 Killari earthquake (Mb 6.3) occurred in the Deccan province of central India, in the Latur district of Maharashtra state. The local geology in the area is obscured by the late Cretaceous-Eocene basalt flows, referred to as the Deccan traps. This makes it difficult to recognize the geological surface faults that could be associated with the Killari earthquake. The epicentre was reported at 18.090N and 76.620E, and the focal depth at 7 +/- 1 km was precisely estimated by waveform inversion (Chen and Kao, 1995). The maximum intensity reached to VIII and the earthquake caused a loss of about 10,000 lives and severe damage to property. The May 22, 1997 Jabalpur earthquake (Mb 6.0), epicentre at 23.080N and 80.060E, is a well studied earthquake in the Son-Narmada-Tapti (SONATA) seismic zone. A notable aspects of this earthquake is that it was the first significant event in India to be recorded by 10 broadband seismic stations which were established in 1996 by the India Meteorological Department (IMD). The focal depth was well estimated using the "converted phases" of the broadband seismograms. The focal depth was given in the lower crust at a depth of 35 +/- 1 km, similar to the moderate earthquakes reported from the Amazona ancient rift system in SCR of South America. Maximum MSK intensity of the Jabalpur earthquake reached to VIII in the MSK scale and this earthquake killed about 50 people in the Jabalpur area. Finally, the Bhuj earthquake (MW 7.7) of January 26, 2001 in the Gujarat state, northwestern India, was felt across the whole country, and killed about 20,000 people. The maximum intensity level reached X. The epicenter of the earthquake is reported at 23.400N and 70.280E, and the well estimated focal depth at 25 km. A total of about 3000 aftershocks (M> 1.0) were recorded until mid April, 2001. About 500 aftershocks (M>2.0) are well located; the epicenter map shows an aftershock cluster area, about 60 km x 30 km, between 70.0-70.60E and 23.3-23.60N; almost all the aftershocks occurred within the high intensity (IX) zone. The source area of the main shock and most of the aftershocks are at a depth range of 20-25 km. The fault-plane solutions suggest that the main shock originated at the base of the paleo-rift zone by a south dipping, hidden reverse fault; the rupture propagated both NE and NW. The aftershocks occurred by left-lateral strike-slip motion along the NE trending fault, compatible with the main shock solution, and by pure reverse to right-lateral, strike-slip motion along the NW trending conjugate fault. Understanding these earthquake sequences may shed new light in on the tectonics and active faults in the source regions.
T51D-0770
Maximum magnitudes of earthquakes and the geothermal structure in Japanese inland area
Information on the potential maximum magnitude of earthquake in an area is valuable for not only the engineering aspect of seismology but also the scientific one. If we had a sufficient record of earthquake data, the upper limit of magnitudes in a given region might be determined simply by the record. Our historical record on large earthquakes does not have an enough length compared with the average interval of large crustal earthquakes. Although some previous studies estimated the possible maximum magnitude using historical earthquake catalogues, geological information, and active fault distributions, their results are not sufficient to discuss the spatial variation of the magnitude. We here report a possibility that the maximum magnitude in an area can be evaluated from the geothermal structure in the crust. The maximum width of the fault in the crust, which is constrained by the thickness of the brittle layer in the crust, gives the maximum earthquake magnitude. The thickness, in turn, should be generally related to the geothermal structure in the upper crust. Therefore, we can expect that the geothermal gradient in the upper crust is used to estimate the maximum magnitude of earthquakes. We examine a relationship between the maximum magnitude of earthquakes and the geothermal gradient in the crust. Two catalogs of large earthquakes (M ≥ 6.5) which occurred in the inland area of the Japanese Islands are used: (1) earthquakes occurring in 1927- 2005 and (2) historical earthquakes in 1500- 1927. The spatial distribution of the geothermal gradient is estimated by low (L ≥ 100km)-pass-filtering of gradient data observed in deep (D ≥ 1000m) boreholes. It is found that for each catalog there is a clear curve defining the upper-limit of the magnitude distribution as a function of the geothermal gradient. Combining the curves and the spatial distribution of the geothermal gradient, we depict maps of the maximum magnitude over the Japanese Islands for two different ranges of about 100 and 500 years.
T51D-0771
Earthquake relocation in the Ardenne (Belgium) : identification of active structures in intraplate context
The Ardenne region, although in intraplate context, presents a moderate seismic activity. Seismic pattern and structural knowledge of the region are not only keys to regional seismotectonic framework, but also to a better understanding of the regional seismic hazard prevention. In 1692 a magnitude 6¼ earthquake destroyed the city of Verviers (East of Belgium) and caused damages as far as Dover (UK). Still ongoing historical researches are getting the location of this event more and more precise. The research for the fault responsible for the 1692 quake has lead to preliminary surveys in the Verviers region. The recent study along the Vesdre river gives good confidence that its terraces are displaced by at least one major fault and maybe some more smaller faults. Recent subsurface geophysical study of the supposed fault trace gives clues to assess for its presence but no certitude for it, further analyses will be necessary. Elsewhere in the Ardenne, the precise location of every event of an earthquake swarm that occurred in 1989-- 1990 in the Hockai Faulted Zone (HFZ) allowed us to map a fault plane with great resolution. This shows the importance of having precise earthquakes locations. These two elements show the necessity to search for major active faults, which will benefit from a complete seismotectonic analysis. In this study, we focus on a region broader than the HFZ or the Verviers area, ranging 5° to 8° East and 49.5° to 51.5° North. The aim of this work is, on one hand, to identify seismically active structures and on the other hand, to analyse the focal mechanisms that will better constrain the regional tectonic setting. The active structures will be analysed later using different means : from aerial photographs or satellite imaging to geomorphological and geophysical surveying in the field. The belgian numerical data dates back to 1985. We recorded over 1100 earthquakes located in Belgium and the neighbouring regions for this period. Each event for the last 22 years has been routinely located using HYPO2000 (a local version of HYPO71). This dataset has never been relocated all at once. The relocation process is achieved by adapting and using different tools commonly used worldwide, among which HypoDD and COMPLOC. The main advantage is that all the data are processed at once, which greatly improves the relative location accuracy as the corrections for each station are computed for a whole dataset. We also evaluate the relative movement between each event's old and new location to search for a response that could be due to the seismic network geometry. Focal mechanisms are also calculated when possible, based on P-wave first motion and also based on P, SH and SV amplitudes. The newly located events evidence at least 5 interesting small-scaled zones which present well aligned events or at least earthquakes concentrations. Even though the analysis of focal mechanisms is still incomplete, we can already show the main trends and some heterogeneities in the present-day regional brittle deformation field.
T51D-0772
Geomorphic Evidences of Pleistocene Deformations in the Western Pyrenees (France)
Due to its intraplate position, the on-going tectonic activity of the Western Pyrenees is only revealed by a moderate and diffuse seismicity. This seismicity presents a general E-W pattern and is distributed along the northern flank of the chain in the western part. We focus here on the Arudy area that suffered from one of the major Pyrenean instrumental earthquakes (M=5,1) in 1980. An early Cretaceous normal fault of the Iberian margin is probably the seismic source of this event. The late Cretaceous inversion of the margin, first in a left-lateral strike-slip mode and then in a more frontal convergence, resulted in a shallow pop-up geometry near Arudy. This pop-up attests of the presence in depth of a crustal discontinuity. The present-day geodynamic arrangement might reactivate this accident in a right lateral mode. This reactivation leads to a strain partitioning between the deep crustal discontinuity and the shallow pop-up that achieved respectively the lateral and frontal components of the displacement. Folding of Quaternary terrace remnants above the thrusts limiting the pop-up, attests of the Pleistocene activity of the structure. Growth of alluvial depocenters in the footwall of these thrusts revealed by near-surface geophysical surveys (electric tomography and ground penetrating radar) attest too of this activity. This deformation resulted in ca 1500 m long and 5-10 m high folds during the middle to late Pleistocene. Using a strain partitioning model, this quantification of the near-surface deformation allows to estimate the displacement achieved by the blind crustal discontinuity. So the seismic hazard due to this fault is better constrained. Project funded by Région Aquitaine and TTI Production.
T51D-0773
Interference between thick- and thin-skinned tectonics along mountain fronts. Example of the Andean foothill (Neuquén basin, Argentina)
The Chihuido anticline (37°30'S-38°40'S and 69° W-70° W) in western Argentina underlines the eastern orogenic front of Andes. North-south-oriented, it is a crustal-scale anticline, 120 km long and 80 km wide. It culminates at 1500 m in elevation. It is limited to the west by the Agua Amarga syncline and by the deep-rooted Salado fault system late Cretaceous in age. The main river of the area, the Neuquén river, runs north-south behind the Chihuidos to the west in the Agua Amarga syncline. To the south, it bends to the east across the southern terminaison of the anticline. To the north, the northern end of the Chihuido had been cross cut by the Colorado river that currently flows 60 km farther to the north. Folding of terrace remnants of these rivers attests of a Pleistocene tectonic activity of the anticline. They appear clearly bended over a length of ca 30 km with an amplitude of 350 m at the apex. Behind the anticline above the Agua Amarga syncline, the rio Neuquén is depositing a strong thickness of alluvial deposits. Uplift of the anticline resulted in an increase of dip, to the west and to the east, of a decollement level made of the Huitrin evaporites Aptian in age. This tilt allowed decollement of pelicular shales and sandstones of the Rayoso formation and of the Cenamanian continental redbed clastics of the Neuquén group above it. This slide lead to the opening of valleys at the apex of the anticline, interprated as extrado tension gashes, and to the growth of superficial folds at the eastern toe of the Chihuido. These folds root in the Huitrin evaporites and achieve extension of the apex of the anticline. Farther to the west along the Salado fault system, vertical offset of Pleistocene alluvial fans with surface faulting attest of an on-going reactivation of the former mountain front. This reactivation is interprated as the consequence of the uplift of the Chihuido fold. The increase of dip of the decollement level beneath the former tectonic wedge allowed its reactivation because the wedge became again undercritics. Thus, the eastward migration of the mountain front (the Chihuidos) leads to a reactivation of the former inner front resulting in an out-of-sequence activity of thrusts. At the regional point of view, these deformations attest of a present-day compressive setting. Extension is only located at the apex of the crustal anticline and is pellicular. Project funded by TOTAL.