G13A-0905
Postseismic self-healing of geoid undulations by water after the 2004 Sumatra-Andaman earthquake observed with GRACE
Earthquakes are accompanied with mass redistributions and cause changes in gravity field and shape of geoid, an equipotential surface coincident with the mean sea surface. Such coseismic changes were detected by satellite gravimetry after the 2004 Sumatra-Andaman Earthquake, but little has been known on what happens on geoid after the earthquake. Here we report slow postseismic recovery of coseismic geoid depression from satellite measurements. This cannot be explained with simple afterslip or viscous relaxation of Maxwellian upper mantle. It suggests the relaxation of coseismic dilatation and compression by the diffusion of supercritical H2O abundant in the upper mantle. Such a self-healing system of coseismic geoid undulations, a brand- new role of water in mantle, would significantly reduce the amount of permanent shifts of the Earth's rotation axis by earthquakes.
G13A-0906
GRACE Satellite Gravity Data Captures Coseismic and Postseismic Deformation from the Sumatra-Andaman Earthquake
Recently released gravity solutions from the Gravity Recovery and Climate Experiment (GRACE) provide much improved spatial resolution, and are now of sufficient quality to observe effects of coseismic and postseismic deformation due to the Mw=9.3 Sumatra-Andaman earthquake on December 26, 2004 and its companion Nias earthquake (Mw=8.7) on 28 March 2005. The gravity field disturbance extends over 1800 km along Andaman and Sunda subduction zones, and changes with time following events. These new solutions indicate significant sea floor uplift since the Sumatra-Andaman earthquake. The estimated uplift rate is about 4 cm/yr in a broad region northwest of the epicenter centered on the Andaman-Sunda fault zone. Postseismic uplift appears strongest along the southwest side of the Andaman-Sunda fault zone, while coseismic uplift is strongest near the epicenter. In the subduction zone (northeast of the Andaman and Sunda fault zone), both coseismic decreases in gravity and postseismic increases appear strongest within a few hundreds km of the fault. When interpreted as uplift, GRACE estimates appear consistent with predicted and geodetically observed viscoelastic relaxation effects.
G13A-0907
GRACE gravity observations associated with two great earthquakes in 2004
We report GRACE satellite observations of coseismic change and postseismic transients from two great earthquakes, the Sumatra-Andaman Island (thrust event; M 9.1) and North of Macquarie Island (strike-slip event; M 8.1) in December 2004. Instead of using global spherical harmonic solutions of monthly gravity fields, we estimated the regional gravity changes directly using over-flight satellite tracking data with regionally-concentrated basis functions (spherical Slepian basis) every 15-day interval. We found significant step-like (coseismic) and exponential-like (postseismic) behavior in a 3 year long time-series of the estimated coefficients for the spherical Slepian functions. We analyzed the coseismic results on the basis of the seismically-derived slip models in spatial and spectral domains. Also we evaluated alternate post-seismic mechanisms (afterslip, poroelastic rebound, and viscoelastic relaxation) with a particular interest in dilatation (sum of normal strain) change.
G13A-0908
Postseismic Displacement Following the Sumatra-Andaman Earthquake Detected by Continuous GPS Observation and the Effect of Viscoelastic Relaxation Using 3D- FEM
We have studied postseismic displacement following the Sumatra-Andaman earthquake of December 26, 2004 in Thailand and other Southeast Asian countries using continuous GPS observation. We will report the results of our GPS analysis from the beginning of 2001 to the end of October 2007. We have also constructed 3D-FEM to evaluate the effect of viscoelastic relaxation following the earthquake. We will also report this result. We used continuous GPS data from 6 sites operated by Chulalongkorn Univ. and Kyoto Univ. or JAMSTEC, 2 sites by Shizuoka Univ. and JAMSTEC, 3 sites by NICT in Thailand and Myanmar, 1 site by STE-Lab, Nagoya Univ., and IGS sites which are located in countries surrounding the Indian Ocean include Japan, China and Australia. Bernese 5.0 was used for the processing of 30 sec. sampling data to obtain static solutions. From our analysis, no significant motions were detected at each site until the day of the earthquake. Although postseismic displacements still have been detected at CHMI and SIS2 in northern Thailand, far from the epicenter, they seem to be decelerated. On the other hand, at SAMP and PHKT, close to the epicenter, where postseismic displacements also became smaller, but still may take a time to stop. An about 29 cm SW-ward motion was detected at PHKT from just after the Sumatra-Andaman earthquake to June 2007, which is larger than its coseismic displacement, about 26 cm. We have constructed 3D-FEM model to estimate how much viscoelastic relaxation affects postseismic displacements after the earthquake. We adopted a Maxwell viscoelastic body as well as Katagi et al. (2006), and modeled around the Andaman-Sea area using isoparametric hexahedral elements with 8 nodes (Zienkiewicz and Cheng, 1967). The Andaman-Sea is well known as a back arc basins, and its ocean floor is still spreading. Therefore, the mantle viscosity under the Sunda-plate may be smaller because of upwelling warm mantle. We are going to investigate and report the postseismic displacements with changing the value of mantle viscosity.
G13A-0909
Three years of post seismic deformation in Central Andaman Islands
As of early 2007, GPS sites in the Andaman Islands had risen 15 to 26 cm following the 2004 Great Sumatra- Andaman earthquake. Near Port Blair, the uplift reversed co-seismic subsidence by 30%. The uplift, observed at eight GPS sites, has shifted the co-seismic neutral axis separating subsidence from uplift gradually eastward. Horizontal motions were 30 to 45 cm in a similar WSW to SW direction observed coseismically. The decay timescale for these motions is relatively short (~0.8 year), and the GPS time series at the westernmost sites now evidence a reversal in the north component of motion. Simulations of the GPS postseismic displacements as viscoelastic relaxation of coseismic stress change and as slip on the plate interface indicate that slip down- dip of the seismic rupture dominates near-field deformation during the first two years. The displacement gradients and decay timescale of GPS measurements cannot be duplicated with any reasonable parameterization of a viscoelastic model, and the relationship of horizontal- to vertical- displacements at the sites is opposite that predicted for poroelastic response. Postseismic slip beneath the Andaman Islands released moment equivalent to a magnitude Mw≥7.5 earthquake, and the distribution suggests deep slip in the stable frictional regime accelerated to catch up to the coseismic rupture. In this presentation we will update the GPS time series at nine sites in central Andaman to the present, assess possible implications of the fault slip patterns for properties of the Andaman subduction thrust, and discuss our plans to capture future deformation in which the viscoelastic signal should dominate.
G13A-0910
Postseismic Deformation and Stress Evolution Following the 1906 San Francisco Earthquake
Large earthquakes, like the M 7.8 1906 San Francisco earthquake, can have both short- and long-term effects on the strain field. Numerical simulation using finite element models is a key method to study time-dependent tectonic deformation mechanisms. Geodetic data are essential for testing model results for accuracy and realism. Previous GeoFEST (Geophysical Finite Element Simulation Tool) viscoelastic finite element model results indicate about 2-5 mm/yr of residual velocities in the 100 years following the 1906 San Francisco earthquake. These models combined simple, single fault viscoelastic finite element models featuring variable relaxation times in the lower crust with various elastic backslip models in order to create velocity profiles to compare with currently observed geodetic rates. This study will include new results from more sophisticated 3D GeoFEST models that incorporate multiple faults and imposed plate velocity boundary conditions and compare these results with GPS velocities recorded by the Bay Area Regional Deformation (BARD) network. In addition, the study will include results examining the evolution of the Coulomb stress changes in the viscoelastic model over the 100 years following the San Francisco earthquake.
G13A-0911
Stress Interaction Between the Cascadia Subduction Zone and the Northern San Andreas Fault
We evaluate the stress changes along the Northern San Andreas Fault (NSAF) associated with coseismic and postseismic deformation from Cascadia megathrust earthquakes in order to test the possibility that Cascadia earthquakes have triggered subsequent NSAF earthquakes. As a test case, we modeled the coseismic deformation from the 1700 Cascadia earthquake, the deformation from deep afterslip and sixty years of viscoelastic deformation prior to the ~1760 penultimate NSAF earthquake. We model the elastic and viscoelastic deformation on a layered spherical geometry, assuming that the entire Cascadia megathrust experienced a uniform fourteen meters of slip (~Mw 9.1) and seven meters of deep afterslip. The coseismic deformation increases Coulomb failure stress (CFS) on the NSAF by a maximum of about nine bars, in the section of the fault near Point Delgada, which may be enough to trigger a north-to-south propagating rupture. Postseismic afterslip both increases and reduces CFS along the NSAF, but its negative peak reduces the extent of the largest positive coseismic CFS by half. The CFS resulting from the estimated sixty years of viscous deformation leading to the penultimate NSAF earthquake does not contribute significantly to the total CFS on the NSAF. We compare our uniform-slip full-margin earthquake model with coseismic and postseimic CFS changes resulting from a southern Cascadia earthquake, with a uniform eight meters of slip (~Mw 8.6), and from a heterogeneous full-margin earthquake that includes less slip on the southern fault planes. Total CFS on the NSAF from both the southern Cascade earthquake and the heterogeneous full-margin earthquake peaks in the same northern location as from the homogenous full-margin model, however the CFS magnitudes are reduced by about a factor of two. Based on these results, it appears that the most likely nucleation point of a triggered NSAF event, from whichever of the three Cascadia source models, would be near Point Delgada. We also modeled CFS on the Cascadia receiver faults from a NSAF type earthquake using a distributed slip model for the 1906 earthquake. The maximum positive coseismic CFS on the Cascadia megathrust is about 20 bars at the southern tip of the margin, at depths between 12-16 km. This may be sufficient to trigger an earthquake on southern Cascadia or on the smaller upper plate thrust faults in the region, although the 1906 events failed to do so. The paleoseismic record however strongly favors the former case, with Cascadia events preceding the NSAF by ~50 years, whereas Cascadia earthquakes follow NSF events on average by 150 years.
G13A-0912
Spatial resolution of afterslip following the 2004 M6 Parkfield earthquake
A strong postseismic deformation signal was recorded by creepmeters and continuous GPS stations starting in the minutes to hours following the 2004 M6 Parkfield earthquake. Several studies have shown that the observed displacements were likely due to afterslip (aseismic slip due to coseismic stress changes), and the estimated moment release of this postseismic slip was anomalously large, exceeding the coseismic moment. Information regarding the spatio-temporal progression of afterslip has the potential to shed light on the underlying processes that give rise to postseismic deformation and aftershocks, as well as fault frictional properties. Preliminary results from a time-dependent inversion of Global Positioning System (GPS) data recorded at 14 continuously-operating and 13 survey-mode GPS receivers during the first 60 days of the postseismic period suggest that afterslip began at shallow depths but spread deeper and surrounded the region of peak coseismic slip estimated from geodetic data. The degree to which these results may be used to address questions relating to underlying processes depends on the degree to which features of the imaged slip distribution are well-resolved. Through a series of simulations and assessments of model resolution we have explored the extent to which it is possible to accurately recover the spatial distribution of slip at depth given the available station coverage at Parkfield. Resolution of slip at depth is known to be poor, and our results confirm that below 6 km slip cannot be reliably imaged. We applied these findings to construct a model fault geometry, comprised of variably-sized subfaults, that better reflects the resolving power of the data and lessens the computational burden by reducing the number of model parameters. Using the revised fault geometry in time-dependent modeling of the spatio-temporal evolution of afterslip, we can more realistically assess which features of the imaged slip history are robust and what insights into underlying processes the GPS data provide.
G13A-0913
Study of the postseismic deformation transient due to the Mw6 2004 Parkfield earthquake using a new 3-D semi-analytic model of stress-driven fault creep.
We developed a new semi-analytic model of time-dependent aseismic fault slip driven by coseismic stress changes. The model uses elastic Greens functions, expressed analytically in the Fourier domain, to solve for the displacement due to an arbitrary set of body forces in a 3-D computational grid representing a heterogeneous visco-elastic half-space. The numerical approach makes use of the convolution theorem and the fast Fourier transform, with number of operations scaling as N Log N (N being the grid size). Strike-slip as well as dip-slip faults of arbitrary orientation can be modeled using an equivalent body-force representation. Realistic fault geometries are easily included by taking advantage of the superposition theorem. Benchmark simulations show that our model compares favorably to known analytic solutions. We apply the model to investigate the postseismic transient following the Mw6 2004 Parkfield (California) earthquake. We use the slip model for the Parkfield event to calculate the coseismic stress changes that in turn are relaxed on the rest of the fault via aseismic creep. We use several rheologic relationships to study the evolution of postseismic afterslip. The model predictions are compared to data from continuous GPS stations in the near field of the rupture. We identify a range of dynamic (stress- driven) models that render a good agreement with both coseismic and postseismic deformation measurements.
G13A-0914
Spatio-temporal evolution of postseismic slip associated with the 2007 Niigataken Chuetsu- oki Earthquake (M6.8) as inferred from GPS data
The Mid Niigata Prefecture, which is called Chuetsu district, locates in the strain accumulated zone named Niigata-Kobe Tectonic Zone (NKTZ). The crust is contracting along NW-SE direction, and must be compressed in same direction around Chuetsu district. Historically, several damaging earthquakes have occurred reflecting this regional stress state in this area, for instance, the Mid Niigata Prefecture Earthquake in 2004 (M6.8) occurred on 23 October 2004. The Niigataken Chuetsu-Oki Earthquake in 2007 (M6.8) occurred on 16 July 2007 in this area, about 40 km northwest to the hypocenter of the Mid Niigata Prefecture Earthquake in 2004. The focal mechanisms determined by seismic wave analyses clarified that reverse faults whose strikes direct NE-SW generated this earthquake. we report the results of a GPS network near the epicenter of the Niigataken Chuetsu-Oki Earthquake in 2007 that includes GEONET's sites and GPS sites established by the Japanese UNiversity COnsortium of GPS research (JUNCO). We present the spatio-temporal evolution of the postseismic slip estimated by an inversion analysis based on observed GPS data. The result shows that postseismic slip occurred in and around the co-seismic fault rather than at downdip extension, and it decayed to insignificant level within two weeks after the main shock had occurred. Next, we developed and applied new time dependent inversion method in which the length of each time step is variable in order to express the process of decay of the postseismic slip more accurately because we can not change the length of time step in an analysis as far as we use the method formulated by Yagi and Kikuchi (2003). The whole analysis period can be optimally splitted based on ABIC, however it is necessary to calculate ABICs for all splitting cases. We overcame the difficulty of prodigious requirements for calculation cost by utilizing a genetic algorithm. We will present the resultant postseismic slip evolution estimated by the new inversion method.
G13A-0915
Investigation of Total Crustal Movements Using Terrestrial and GPS Measurements along the Western Part of North Anatolian Fault in the Marmara Region
In order to monitor crustal movements along southern branch of North Anatolian Fault Zone (NAFZ), Geodesy Department of Kandilli Observatory and Earthquake Research Institute (KOERI) initiated to establish microgeodetic networks around eastern Marmara Region. General Command of Mapping (GCM)-Istanbul Technical University (ITU) network was constituted of nine pillars, measured five times between 1941 and 2007 applying with respectively trialteration, triangulation and space geodetic methods. In years 2004 and 2007 GPS campaign processing was extended and re-evaluated by adding three new stations from Marmara Region Continuous GPS Observation (MAGNET) network. This study aimed to gather displacements of the GCM-ITU network and analyze the movements of the stations in order to monitor crustral deformation along the Iznik fault. By taking the stations located at the south of the fault as stabilized, the displacements of the remaining stations were evaluated. The movements of stations were found to be ranging between 19 cm to 2 mm according to the accuracy of observation method. The results demonstrated that the stations at both the south and the north of the fault have moved during the 1941-2007 period, and these movements were independent of the movement of the fault itself.
G13A-0916
Coseismic and Postseismic Deformations From Great 2006-2007 Kuril Earthquakes Revealed by Regional GPS Observations
The 1200-km long Kuril arc is the last subduction zone never previously explored by space geodetic methods. In 2006, we installed the continuous GPS network (CGPS) over the whole arc and added several survey-mode stations (SGPS). In 2006-2007, the paired great earthquakes near the central Kurils happened several months after we installed the network: Mw 8.3, Nov. 15, 2006 underthrusting event, and Mw 8.1, Jan. 13, 2007 tensional outer-rise event. Although the earthquakes have prevented us from estimating reliable interseismic surface velocities for most of the Kuril arc, it has given us the chance to examine great earthquakes and their transient response in the region that was a seismic gap for a century. Two SGPS stations nearest to the hypocenters captured the largest observed offsets of about 0.6 m reflecting the superposed effect of both events. These offsets are mostly attributed to the Nov. 2006 event. More distant stations captured coseismic offsets caused by each event ranging from several mm to 60 mm. Significant transient signals associated with rapid postseismic afterslip in the rupture or with the relaxation in the viscous mantle were noticed for the Nov. 2006 event but not for the Jan. 2007 event. Large amount of afterslip was observed in the first 12 hours following the Nov. 2006 main shock. For both events, we inverted observed GPS offsets to evaluate the size and rake of the coseismic slip. In forward modeling, the PREM layered model of the spherical Earth was adopted (the method of F. Pollitz). The rupture dimensions and geometry were constrained by the spatial distribution of aftershocks, shallow seismicity, plate tectonics considerations, and CMT solutions. In case of the Jan. 2007 event, plate tectonic constraints are inapplicable. To ensure correct estimation of the Nov. 2006 coseismic offsets, we modeled postseismic transients by the logarithmic approximation in agreement with the rate-strengthening friction. For the Nov. 2006 event, our geodetic estimate of the slip with the preferred rupture geometry is 3.3 m. For Jan. 2007 event, we prefer the CMT fault plane with the strike parallel to the trench; however, the estimates of slip for both CMT fault planes are nearly identical: 3.2 and 3.5 m. Geodetically estimated rakes agree with those of CMT within 18 degrees. Geodetic coseismic moments were estimated with the commonly adopted crustal shear modulus of 30 GPa. For both earthquakes, the geodetic moments are about twice smaller than CMT moments. The only plausible way to make the geodetic and CMT moments agree is to assume a higher shear modulus. GPS time series during 7.5 month following the Nov. 2006 event show that the Kuril GPS stations are moving to SE, opposite to the direction of interseismic motion. Whether the observed drift reflects the postseismic afterslip or the viscous relaxation in the mantle can be determined from the future observations.
G13A-0917
Measurements of Early Afterslip using 30-sec GPS data: the Tokachi-Oki Earthquake
Thirty-second GPS are are used to infer the early phase of afterslip following the 2003 Tokachi-oki earthquake. GPS data from the GEONET network were analyzed using GIPSY-OASIS and standard models. No constraints are put on the position estimates, so the the 30-second data sense both the static offset of the mainshock, large aftershock, and postseismic deformation. The GPS time series at this stage of processing contain significant common-mode errors. The time-dependent inversion method [Segall and Matthews, 1997] is then employed to infer the space-time evolution of afterslip. Common-mode errors are simultaneously estimated with fault slip, thus significantly improving the resolution of the 30-second GPS positions. This analysis concentrates on two periods (i) between the main shock and the largest aftershock and (ii) 24 hours after the largest aftershock. The inferred slip for period (i) is significant in the southwest side of the main rupture region. The slip pattern is similar in the begining of period (ii), but the slip starts to propagate to the northeast. In contrast, afterslip in the first 30 days was marked both in the southwest and northeast sides of the rupture zone [Miyazaki et al., 2004]. Such a difference in the triggered slip time may be a manifestation of the spatial heterogenity in frictional properties.
G13A-0918
Long-term post-seismic deformation over Mongolia, a look from InSAR.
Western Mongolia has been the most seismically active intracontinental region in the world in the past century, with four earthquakes of magnitude 8 and greater since 1905. The Bolnay-Tsetserleg earthquake sequence of 1905, in particular, has released the largest amount of seismic energy ever observed inside a continent. GPS measurements (Calais et al., 2003) and paleoseismology (Ritz et al., 1995) are consistent with deformation being accommodated by left-lateral slip on the three major E-W trending faults slipping at rates still poorly determined, but possibly on the order of 1 to 3 mm/yr. The combination of a recent sequence of large earthquakes and low background interseismic strain accumulation makes Mongolia an ideal target for the study of long-term post-seismic deformation by viscoelastic relaxation and associated stress triggering. Vergnolle et al., [2003] and Pollitz et al., [2003] used GPS results to probe the lithospheric structure in Mongolia and subsequently show that stress trigerring due to viscoelastic relaxation within the earthquake sequence is a likely explanation for the Mongolia cluster sequence. We present new geodetic measurements that significantly improve the spatial resolution of the deformation field in the area of the M8.1 1957 Gobi Altai earthquake. We use satellite data from the ERS1 and ERS2 satellites to produce ground velocity and time series map of the deformation field. In conjunction with existing GPS velocities, the obtained velocity field is used to constrain post-seismic models of viscoelastic strain relaxation.
G13A-0919
The QuakeSim GeoFEST modeling system - new features inspired by San Andreas Fault motion
The NASA QuakeSim project modeling environment GeoFEST (Geophysical Finite Element Simulation Tool) has demonstrated features for rapidly scaling up to very large problems (of order 1e8 linear tetrahedral finite elements) by incorporating load-balanced parallel partitioning of the mesh, low-communication iterative solution of the sparse linear systems, and solution-driven strain energy based automatic parallel mesh refinement. Inspired by several San Andreas Fault modeling problems, new features have been added and are incorporated into a new public release. These include support for tectonic velocities at the boundaries within the adaptive mesh framework (giving support for repeat-event spinup), independent specified periodic motions on specified sets of fault nodes (supporting steady motion of faults at depth in same model with periodic crust events), buoyancy counterforces that arise from vertical deformation across density contrast surfaces. Support has been added for nonuniform slip along fault strike, as well as for nonplanar faults such as triangule faceted faults of the community fault model. This latter feature requires hand-crafting of the initial finite element mesh, using a meshing tool such as the Los Alamos Nationlal Laboratory's LaGriT package; GeoFEST support tools now translate LaGriT mesh information into the GeoFEST format. In addition we have implemented improvements to iterative convergence control and automatic refinement thresholds. These features are demonstrated by solutions obtained on geometric meshes over a range of simple to complex, including simple model comparisons to 2-D and analytic solutions for a repeat-event crustal fault overlaying a deep viscoelastic layer with steady dislocation applied (a simple model representing parts of the San Andreas fault system). http://quakesim.org
G13A-0920
The Role of Viscoelastic Relaxation in Triggering of Aftershocks, Inferences from two Californian Earthquakes
A number of studies indicate that a first order causal relationship exists between the location of aftershocks and the areas that have received positive coseismic static stress changes following a major earthquake. Some studies have also argued the influence of dynamic stress triggering on aftershocks for a short period of time (up to several weeks). Aftershocks, however, typically last several years, which may indicate the time dependent role of postseismic processes. In particular, the contribution of viscoelastic (VE) relaxation of the ductile lower crust and upper mantle to the aftershock triggering has not been studied in detail. Here we investigate the contribution of VE relaxation process for two Californian earthquakes, the 1992 Landers and 1994 Northridge events, where best possible crustal models and aftershock data exist. They also represent different tectonic regimes in which the stress transfer to the upper crust due to the relaxation processes may differ, and hence the VE contribution may be different. In our analysis the aftershocks are binned in 30-day periods and both coseismic and postseismic stress changes are resolved on the nodal planes of the focal mechanisms of aftershocks in order to see which model does better with time. For Landers, our preliminary results show that the model with total stress changes (coseismic + VE) explains the aftershock triggering about 5 percent better than coseismic alone at the start of the aftershock period, and this improvement increases over time. For Northridge, the VE contribution becomes only significant after 5 months following the mainshock.
G13A-0921
Coupled afterslip and viscous flow following the 2002 Denali, Alaska earthquake
We investigate the processes of postseismic deformation following the 2002 Denali Fault, Alaska earthquake using 4.5 years of continuous and campaign GPS data. Afterslip is modeled on a fault in an elastic lithosphere overlying a Maxwell (linear) viscoelastic asthenosphere. We assume afterslip is governed by a nonlinear velocity- strengthening friction law. Postseismic GPS time-series are best explained by a combination of two mechanisms: viscous flow in the lower crust and upper mantle with viscosity of about 1019 Pa s, and afterslip on the fault above 30-40 km depth. Models with afterslip only (no distributed viscous flow) underestimate displacements at sites more than 100 km from the fault. The rate-state frictional parameter a-b, is estimated to be in the range 10-3-10-2, consistent experimental values for granite at conditions near the transition from velocity weakening to velocity strengthening. It has been suggested previously that nonlinear rheology of the upper mantle is necessary to explain the observed evolution of surface displacement rates with time. However, the displacement rates at continuous GPS sites are reproduced remarkably well by our model with afterslip in a fault zone with nonlinear rheology and a linear viscous upper mantle. The Denali earthquake may have caused increased locking at the interface of the subducting Pacific plate south of the Denali Fault. Northeast directed horizontal surface velocities at GPS sites over 100 km south of the Denali fault increased following the earthquake. The magnitude of the acceleration at these sites in southern Alaska cannot be explained with our simple models of postseismic deformation associated with afterslip and viscous flow directly below the Denali fault. The Denali earthquake reduced the reverse-sense of shear stress on the subduction interface, promoting increased coupling on the interface. Simple spring-slider models with rate-state friction confirm the possibility of increased coupling of the interface following the earthquake. The spring-slider model predicts that a sudden decrease in stress on a fault sliding with friction parameters near the transition from velocity strengthening to velocity weakening can cause the slider to stop sliding for a period of 5-10 years.
G13A-0922
Delineating Post-Seismic Transient Relaxation due to the Bhuj Earthquake on January 26, 2001
While observing post-seismic relaxation associated with major inter-plate earthquakes is irrefutable, it is rather difficult with intraplate earthquakes. The Mw 7.6 Bhuj earthquake on January 26, 2001 in Western India considered to be an intraplate event and provides a unique opportunity to examine post-earthquake relaxation processes far from plate boundaries. To study the characteristics of transient post-seismic deformation and distribution of strain accumulation rate, five GPS campaigns were made during 2001-2002 and one during 2007, at 14 sites. As the post-seismic relaxation was feeble and not wide spread, it could not be traced from 2001-2002 data. The sixth campaign was intended to facilitate delineation of post-seismic deformation. The displacement at a site in the vicinity of the rupture may contain contributions from at least three processes (i) the plate motion at the site, (ii) contribution from inter-seismic strain accumulation and (iii) contribution from post-seismic relaxation processes due to the earthquake. In order to remove the first two components, we considered the displacement with reference to the site AHMD, which is sufficiently away from the asperity zone. Logarithmic function fits well to six snap shots of position time series of EW and NS components. The best fit is seen at the site DHAM which is closest to the epicenter. Further, the distribution of measure of fit to the NS component portrays that the region of good fit positionally correlates with transect parallel to the rupture zone. It is inferred that the post-seismic relaxation is taking place - mainly in NS direction and the major relaxation (~\ 90%\) occurred with in one year after the earthquake occurrence. Rapidly decaying afterslip and poro-elastic (as also indicated by widespread coseismic liquefaction) mechanisms seem to be responsible for post- seismic relaxation.
G13A-0923
Postseismic deformation of the 2001 Bhuj earthquake from GPS, Gravity and InSAR data
Postseismic deformation phenomena have been observed following many major earthquakes and visco-elastic relaxation of the lower crust and upper mantle often dominates the intermediate-scale (one to several fault lengths) deformation patterns. Other relaxation processes, namely discrete afterslip and poroelastic rebound may also contribute to the observed deformation, especially in the near field and earliest phase of the postseismic transient. GPS measurements, observations of gravity field changes and InSAR range-change data collected following the Mw 7.6 2001, Bhuj earthquake provide a unique opportunity to probe lithospheric rheology in an intra-plate setting. The transient deformation rates following the Bhuj earthquake are substantially lower (< 3 cm of horizontal motion in first two years) than those observed following several recent inter-plate earthquakes, suggesting that the event is underlain by relatively strong lithosphere. We model the spatial deformation data assuming visco-elastic relaxation of the lithosphere, using constraints on geometry and rupture parameters on the source fault from earlier studies. We also explore contributions from and tradeoffs with afterslip and poroelastic rebound mechanisms. Our physical model of the visco-elastic deformation elucidates the nature of deep relaxation in the visco-elastic lower crust-upper mantle layers resulting from the Bhuj earthquake stress perturbation.
G13A-0924
Postseismic relaxation of the 2007 Central Peru earthquake
We present a comprehensive modeling of coseismic and postseismic effects of the 2007 Central Peru earthquake on deformation, gravity and stress fields obtained with a semi-analytical spherical self-gravitating Earth model. We discuss the expected signature of different stratification profiles including Maxwell and Burgers rheologies and find that measurements of short-term relaxation on horizontal deformations should provide a discrimination between different rheologies. We compute the time-dependent stress transfer of the 2007 earthquake on the 1868 and 1877 seismic gaps in terms of the Coulomb Failure Function (CFF). We find significant levels of stress transfer only on the 1868 seismic gap due to its proximity to the 2007 source, even if the sign of CFF variation is dependent on the depth of the receiver plane, which is not univocally fixed by geological evidences.
G13A-0925
Post-Seismic Slip of the February 2004 Earthquake Swarm Detected by GPS in West Bohemia, Central Europe
In last decades earthquake swarms in West Bohemia, Central Europe, are monitored by a local seismic network WEBNET consisting of nine stations. In 2003 two GPS permanent observatories MARJ and POUS belonging to European permanent network EPN were established in NW Bohemia to detect possible inter-, co- and post- seismic slips of the earthquake swarms. Individual swarm events are often clustered and grouped sharply round a few tectonic fault zones. On 22 February 2004 a deep swarm of around 50 events in magnitude ML range -0.5 to 1.4 occurred in the Novy Kostel zone. Most of its events fitted in elongated cluster with -30° strike in the depth of 13.5 km differed from the main -11° strike of the Novy Kostel zone. The cluster lay 2 km apart from the main fault plane to west and preliminary focal event mechanisms displayed dip-slip regime. This two azimuth foci bifurcation most likely corresponds to two tectonic systems of this area. After the February 2004 swarm for almost two weeks the total post-seismic slip up to 10 mm was observed between time series of GPS observatories. The detected slip seems to match well to the deep-seated fault system with the -30° strike. It is remarkable that this system fits well just to pronounce gravity gradient zone existing in this area. Mutual relations among the slip movements, local earthquake occurrences and geological data will be displayed and discussed. The research was supported by the Czech Science Foundation (Project 205/05/2287), the Grant Agency of the Academy of Sciences CR (Project IAA300460507), the Support of Targeted Research programme (Project QS300460551) and the Ministry of Education, Youth and Sport (Projects LC506 and 1P05ME781).
G13A-0926
The Acceleration Gradient Tensor Field in Southern California Estimated using Continuous SCIGN GPS Data
We use the continuous SGIGN GPS network between the periods of late 1999 to 2004 (post-Hector Mine) to measure the acceleration gradient tensor field in southern California. We first interpolate the continuous time series with fourth-order polynomials, which have proven stable and capable of capturing first-order temporal variations within the time series. Annual and semi-annual signals are removed. We use Monte-Carlo simulations to estimate the model uncertainties in the time-dependent velocities and accelerations inferred from the continuous GPS time series. Once model continuous acceleration estimates are inferred for all stations, we interpolate the acceleration estimates within a 0.1° x 0.1° finite element grid in southern California using the method of Beavan and Haines [2001]. Boundary conditions of zero acceleration are used at the edges of the region considered. West of the Hector Mine rupture, during the first six months following the Hector Mine event, acceleration vectors show a large clockwise rotational pattern, with acceleration rates of 3 - 10 mm yr-2 for stations south of the San Jacinto Fault and east of the Los Angeles Basin. Significant rates of change of the strain rate tensor field in the first year following the Hector Mine event occur in the Hector Mine region, the region of Los Angeles Basin, and areas both south and west of the southern San Andreas Fault. Within the first year of the Hector Mine event, time rates of change of strike slip shear strain rates of order 1x10-7 yr-2 occur around the Hector Mine rupture, whereas rates of change of dilatational strain rate around the Hector Mine event (3-5x10-8 yr-2) occur over a much larger region surrounding the rupture zone. During this time period there are also large contractional rates of change of dilatational strain rates along the San Jacinto fault zone (- 3x10-8 yr-2) and both tensional and contractional rates of change of dilatational strain rate within the greater Los Angeles region (4-5x10-8 yr-2). The patterns of spatially variable accelerations following the Hector Mine rupture are related to the elastic response of the rupture but also suggest highly variable crustal properties that result in an inhomogeneous response and recovery associated with the Hector Mine event.
G13A-0927
An examination of using gravity as a proxy for strain accumulation in complex fault systems
The gravity signal contains information regarding changes in density at all depths and here we investigate its use as a proxy for strain accumulation in complex fault networks. Using a complex, strike slip fault system, such as the San Andreas Fault (SAF) network, and the seismicity modeled by Virtual California, we examine the rates of dilatational highs and lows for fault segment that initiate in a prescribed time period. The dilatational gravity signals are employed as they are well within in the range of portable instrumentation, and have the effects of topography removed; thus ensuring that we examine only those changes from density perturbations at depth.