T43B-1330
Nano-particle Analysis of Fault Gouge in the Chelungpu Fault of Taiwan
The strain energy released during an earthquake is partitioned into radiated energy, heat and fracture energy during seismic rupture. The fracture energy can be estimated from grain size distributions of gouge in the slip zone of a fault core. However, the finer particle size is the most critical for calculating the total amounts of it. This study, thus, focuses on the issue of determining the lower cut-off grain size to estimate seismic fracture energy based on the analysis of nanosize particles. Fault gouge samples were slipped by the Chelungpu fault during the Chi-Chi earthquake was taken from an outcrop near Wu-Feng, central Taiwan and was separated it into different particle size ranges by centrifuge and filter. Mineral species were identified by XRD and the mircofabric of particles was investigated by SEM and TEM in different size ranges. The XRD results indicate that the dominant minerals include quartz, plagioclase, clay minerals (smectite and illite) and hydrated minerals (halloysite and vermiculite). The relative amount of quartz decreases with deceasing grain size but the relative amount of clay and hydrated minerals increase. Thus, we infer that quartz could be ruptured into 25~50 nm by faulting and the fault gouge on the surface outcrop was experienced weathering process.
T43B-1331
Analysis of Magnetic and Grain Fabrics across Fault Gouge in the Chelungpu Fault of Taiwan
Behaviors of gouge zones in the fault core of the 1999 Chi-Chi earthquake is a key to understanding its slip mechanism. Fault zone samples from two outcrops closed to the Chelungpu fault of the Chi-Chi rupture were densely collected and subjected to analysis of magnetic and grain fabrics. Symmetric elements, hanging wall, black gouge, foliated gouge, black gouge, and footwall, are consisted in the fault architecture. Spatial distributions of magnetic/grain lineation L, magnetic/grain foliation F and anisotropy degree show obvious drops in foliated gouges. Close to black gouges, L and F show different patterns: lineation L directly decreased, but foliation F increased dramatically in black gouges, then dropped in the foliated gouge. Ellipsoid shapes of magnetic susceptibility and grain change from prolate to oblate with increasing anisotropy from wall rocks to gouges. Directional plots of the magnetic susceptibility anisotropy show different types within and outside the fault zones. For wall rock, three principal axes are well clustered individually. However, in fault zones, only Kmin axes are well clustered at the area closed to the horizontal plane, the other two axes are scatteredly distributed along a girdle, especially for black gouges. Compression stress direction inferred from the magnetic fabric is similar to the regional compression orientation. Also, magnetic foliation planes of gouge samples show a 20 degree thrust and 20 degree clockwise rotation, compared with that of wall rocks. Orientation plots of grain ellipsoids across the fault zone are similar to the magnetic fabric analysis but have a systematic difference. The long-axis orientation of grains in wall rocks is parallel to the orientation of regional compression stress but rotates 29 degree clockwisely in gouge zones. Fabric analysis results elucidate that a strong relationship holds for grain and magnetic fabrics and the characteristic movement across the foliated fault zone should be tightly related to the formation of foliated gouges and black gouge of the Chi-Chi earthquake.
T43B-1332
Stress State Change Associated With the Chelungpu Fault in the Vicinity of the TCDP Borehole
The Taiwan Chelungpu Fault Drilling Project (TCDP) drilled two holes, 40-m apart, (hole-A, 2 km and hole-B, 1.35 km deep) penetrating the Chelungpu fault at where large surface displacements (~10 m) were observed in the 1999 Chi-Chi earthquake (Mw 7.6). Judging from various sources of data including continuous geophysical logs and core, the fault zone at log depths of 1109 m in hole-A and 1133 m in hole-B were interpreted to be the Chi-Chi rupture fault at depth (with a dipping angle of 30 degrees). In-situ stress state at the drill site can be inferred from 1) extended leak-off tests, 2) borehole breakouts and drilling-induced tensile fractures from borehole FMS/FMI logs and 3) shear seismic wave anisotropy from DSI logs. Further analyses of stress magnitudes incorporating information of tri-axial rock strength, width of breakouts, and shear stress drop during earthquake allow us to better constrain the possible magnitudes of the current principal horizontal stresses around the fault zone and showing that the current stress state is in strike- slip regime, which indicates the principal stress state has been changed from a reverse-fault regime as a result of slip on the Chelungpu fault.
T43B-1333
Holocene vertical tectonic movements of the Taipei Basin, northern Taiwan and its implications
Many geological data of the Taipei Basin, although, have been published by various studies in past decades, however, vertical tectonic movement rate of the Basin was not well understood so far. This study, therefore, used radiocarbon dates, obtained from fifteen boreholes in the Basin, to calculate the Holocene vertical tectonic movement rate. In addition to the derived tectonic movement rate, this study also discussed the causes of the tectonic patterns of the Taipei Basin. The Taipei Basin, located in the northern Taiwan, was a half graben subsided and extended along the western boundary, the Shangiao Normal Fault, of the Basin. The Holocene vertical tectonic movement rate of the Basin were calculated based on 94 radiocarbon dates in fifteen boreholes, the elevations of the radiocarbon dating samples, and the eustatic sea-level curve of the past 15 ka. The results showed the rate in the western part of the Basin, was -2.2 — -0.9 mm/yr (negative value indicates subsiding, and positive value indicates uplifting). In the central part of the Basin, the rate was ca. -1 — 1 mm/yr while in the eastern part of the Basin, the rate was 0.1 — 1.6 mm/yr. Along the Shiangiao Fault, the rate of the hanging-wall was ca. -1.6 — -0.4 mm/yr and the rate of the footwall was ca. 0 mm/yr. According to the results of this study, the present territory of the Taipei Basin was not actually consistent with the tectonic subsiding region. The vertical tectonic movement pattern demonstrated subsidence in the western part and uplift in the eastern part of the Taipei Basin. The subsidence of the western part was controlled by the extension of the Shangiao Faul. The uplift of the eastern part might be ascribed to the roll-over of the Fault. Another possibility is that the uplift of the east was controlled by the same behavior as the Western Foothills.Consequently, the deposition of the eastern part of the Basin, wass mainly related to the accommodations due to sea-level rise but not tectonic subsidence.
T43B-1334
A Study Of The Tomography And Seismicity In Taipei Basin And Tatun Volcano Regions, Taiwan And Their Structural Implications
The Taipei Basin with soft deposit in the north of Taiwan has been produced by the motion of the Sanchiao fault. Taipei, the capital of Taiwan, is located at the Taipei Basin and the Tatun volcano group which have volcanism is located adjacent to it. The local seismicity is very low, however, it is a very important topic to understand the seismotectonics of the Taipei Basin and Tatun volcano regions. We combined the seismic data from the records of the Central Weather Bureau Seismic Network (CWBSN) and Taiwan Strong Motion Instrumentation Program (TSMIP) to establish the tomography of Taiwan and then relocated those earthquakes in the study region. Totally, 17,548 events from 1974 to 2006 in the North Taiwan were used for Vp and Vp/Vs tomography. We focus our study on Taipei region that is bounded in latitude 24.91° to 25.31°N and longitude 121.3° to 121.8°E. A total of 339 seismic events with depth less than 100 km from 1977 to 2006 were relocated to study the seismotectonic structures. The pattern of high Vp/Vs ratio coincides with the Sanchiao fault which has made the Taipei Basin, and reveals that the effects of basin amplification affects about several kilometers. There is a zone with high Vp/Vs ratio beneath the Tatun volcano group at the depth of 5~10 kilometer and the strata with low Vp/Vs ratio above this zone appears like an arch. Besides, there are some earthquakes occurred surrounding the high Vp/Vs ratio zone. It may imply that there is a potential region for being full of fluid. It is worth to further analysis about the formation of these earthquakes.
T43B-1335
Evolution of large-scale paleo-landslides in volcanic area reconstructed based on LiDAR technology
The key information of landslide, including the range of landslide, morphological features, volume estimations, etc., is important when analyzing the triggering mechanism, hazard assessment and mitigation. For a paleo- landslide, however, these features could often be masked by the successive surface processes after sliding. To alleviate this difficulty, this research proposes a topological reconstruction method to enable the assessment of the key information based on the DEM derived from high-resolution LiDAR images. Two large-scale landslides in volcanic terrain were studied. Accordingly, the paleo-morphology of the slide area before landslide was reconstructed based on the originally undisturbed dome-shaped topography, by supposing the volcano kept an ideal conic shape and comparing with the nearby un-eroded hillslopes of the same volcanic cone. The morphology of the studied landslides is characterized by very gentle slope profile within the slide area and a steep cut-off slope situated around the toe indicating a surface singularity. By employing proposed reconstruction method, the landslide volume was then estimated from two different DEM, before and after the landslide. The total removed volume of these two paleo-landslides is about 820 ± 20 × 106 m3. The incision of creeks is accordingly obtained from the discrepancies between the current DEM and its envelope surfaces. By comparing to the current creek profiles, it shows that the maximum amounts of incision are situated in the middle to lower slope within the slide area, implying the effect of normal faulting as an important factor for triggering landslides and for topographic evolution. The maximum amount of incision could be as high as 15 m, which indicates the observed event(s) from the LiDAR topographic data are indeed ¡§paleo¡¨ landslide(s). As a result, this study demonstrates the LiDAR technique as a powerful tool for landslide analysis and also provides methods for analyzing paleo-landslide of various categories.
T43B-1336
A Study of Seismotectonics of the Miaoli-Taichung Area, Taiwan
The Miaoli-Taichung area, which is located at central-western Taiwan, is under full collision process between forearc and continent. During the last century, two large earthquakes occurred on the Tuntzuchiao fault (M7.1) and the Chelungpu fault (Mw7.6) in 1935 and 1999, respectively. In order to reduce the damage of the future earthquakes, it is urgent to map the potential earthquake sources beneath this region. The Miaoli-Taichung seismotectonics were studied based on analysis of recently published datasets of three-dimensional velocity structures, focal mechanisms and relocated catalog from 1991 to 2005. Results indicated that higher bedrock should exist beneath Taichung and may be a major structure in the study area. From the evidence of the seismicity, the Chelungpu and Shuangtung faults could be located in a same detachment. The west Changhua fault might also be located in the detachment even though very little seismicity was correlated to it during the study period. It is possible that the Changhua fault has been in a lock mode and could be a potential source. The Tuntzuchiao fault was also inactive from 1991 to 2005. However, it could extend northeastward after the 1999 Chi- Chi earthquake.
T43B-1337
Shallow Seismic Reflection Images of the Active Chingshui and Tachia Faults in Central Taiwan
Although there are many methods used for studying active faults, the seismic reflection method always plays an important role for imaging the fault and the subsurface structures. We could further correlate the shallow seismic data to the near surface geological condition, which is a key factor of forming a folding or faulting structure in an earthquake and resulting into different types of seismic hazard. In this paper, we will use shallow seismic reflection method to study an important fault system, Chingshui and Tachia faults in the deformation front area of western Taiwan. The Chingshui and Tachia faults are located at the western flank of the Tiehchanshan gas field in western Taiwan. Taiwan Petroleum Company has mapped the regional subsurface structure of the studying area by using many large-scale seismic lines. In the field, we may observe the vertical bedding of sandstone and conglomerate layers near the western front of the Tiehchenshan; however, the correlated faulting feature is not obvious according to the large-scale seismic data from the petroleum company at shallow depth. In other words, whether the observed geological features were related to a fault scarp or a folding scarp, or if the faults have been exposed to the surface was still not clear. The purpose of this study is to investigate that whether the near surface section of the two faults are actually blind or just not clearly observed in the regional large-scale seismic exploration data. A 48-channel seismic system was used in this study, a mini-impactor, JMS Mini65 was used as the source, and the data were collected with geophone arrays of three geophones. The preliminary result shows excellent quality of subsurface image. We will show the integrated interpretation of the seismic images with other results from geological survey, borehole logging, and geomorphic study.
T43B-1338
The transition zone between two structural domains in the active fold-and-thrust belt of central Taiwan.
The structural transition zones within fold-and-thrust belts usually contain complex structures and are often controlled by the basement geometry. In the western foothills of central Taiwan, most of the transition zones coincide with the high-angle normal faults within the basement, and these faults significantly influence the transitional structures. In this study, we focus on the northern Taichung basin, which is a transition zone between two structural domains in central Taiwan. They are folds-dominated domain in Miaoli and imbricate thrusts- dominated domain in Taichung. Series of large earthquakes occurred in 1935 in the Miaoli domain, and the devastating Chi-Chi earthquake occurred in 1999 in the Taichung domain, indicating the transition zone is possibly a highly active area. A 3-D geological model was established based on seven published balanced cross sections plus one cross section constructed in this study. The constructed cross section is based on the surface geological data and relocated earthquake hypocenters. The 3-D geological model indicated that the transition zone was initially developed as a north-south trending fold. The fold was then truncated by a south-dipping normal fault, which also separates the two structural domains at depth. The southern part of the truncated fold further acted as the footwall for a newly developed fault-bend fold, which is located underneath the Chi-Chi earthquake fault. The fault-bend fold structure may constrain the NW movement of hanging wall and cause large uplift for the northern section of the Chi-Chi earthquake fault. In addition, the surface trace of the Chi-Chi earthquake fault changed from the N-S direction to the E-W direction due to the geometric constraints of the fold.
T43B-1339
Ground Displacements and Fault-plane Geometry beneath: a case of 1999 Chi-Chi Earthquake (Mw 7.6) at Tsaotun in Central Taiwan
Long-term ground deformation recorded in deformed geomorphic surfaces is supposed to be the cumulative strain produced by associated active structure and certainly related to the subsurface structure geometry. Moreover, by time domain the deformation also can be divided into components: co-, post-, and inter-seismic. A case that may demonstrate the entire process is at Tsaotun in central Taiwan, where widely developed geomorphic surfaces have long been noticed and surface ruptures of 1999 Taiwan Chi-Chi earthquake (Mw 7.6) ran though. Landform investigation, geodetic work, sub-pixel comparison of aerial photos, and D-InSAR analysis are conducted to reconstruction the entire deformation process mentioned above. In our previous study, we have used sub-pixel correlation on high-resolution aerial photographs to obtain detailed horizontal displacement across the surface ruptures, which has revealed that the fault bends beneath greater and more rapid in the southern segment. Nevertheless, the coseismic ground displacement cannot entirely match the long-term surface deformation recorded in the geomorphic surfaces. However, by the post-seismic ground displacements obtained from InSAR a few other active structures, such as secondary strike-slip faults, may play a role to accommodate the tentative stress accumulation in a large earthquake. By available co- and post-seismic ground displacements, we successfully rebuild a fault-plane model at Tsaotun. Using this model we can satisfactorily explain the surface deformation in different time domain. The model also describe the details of when, where, and how much the rate of the ground displacements would occur. Our result is undoubtedly valuable to improve the building code and to assist urbanization plan for the purpose of seismic hazard mitigation.
T43B-1340
Folding and Faulting at Tanliwun Along Leading Edge 1999 Chi-Chi Thrust
The 1999 Chi–Chi earthquake in Taiwan provided excellent opportunities for studying earthquake faulting and folding in escarpments 1 to 10 m high at the leading edge of the Chi-Chi thrust which extended for some 100 km. Trenches excavated across the escarpments showed well–preserved fault and fold structures at scales ranging from mm to m. We studied in detail relations between profile shapes of escarpments and subsurface structures exposed in Chushan trench at Tanliwun, near the south end of Chi–Chi rupture. We mapped faults and sedimentary bedding at scales of 1:10 or 1:20 in the trench walls. In order to understand the structures exposed in the trench walls, we supplemented the observational approach with theoretical modeling. For example, we used multi–fault dislocation modeling to show that the multiple small faults exposed in the trench walls could have produced the monocline–like escarpments observed there. The field observations and the modeling suggest that the structures that first appeared to be continuous monoclinal bending of layers are, in fact, discontinuous. The appearance of gross bending is a result of slip in the same sense along numerous small faults inclined to the sedimentary bedding and subparallel to the main fault in the near subsurface.
T43B-1341
Growth of Fault-cored Anticlines by combined Mechanisms of Fault Slip and Buckling
A primary goal of studies of blind faults underlying actively growing anticlines is assessment of earthquake hazard associated with slip on the faults. It is generally assumed that the amount of slip on the fault is directly related to the amplitude of the fold. Under this assumption, the potential for earthquakes on blind faults can be determined directly from fold geometry. However, anticlines grow over slipping reverse faults can be amplified by a factor of two or more by buckling of mechanical layering under horizontal shortening. Studies that attempt to estimate fault slip from fold geometry may therefore overestimate fault slip by a factor of two or more if the contribution to fold growth from buckling is ignored. We construct boundary element models to demonstrate that fault-cored anticlines in mechanically layered media subjected to layer-parallel shortening are not built solely by slip on the underlying fault. The amplitude of folds produced in a medium containing a fault and elastic layers with free slip and subjected to layer-parallel shortening are 2-5 times larger than the amplitudes of folds produced in homogeneous media without mechanical layering. We compare the model results with data from fault-cored anticlines in the western United States. Pitchfork Anticline on the western flank of the Big Horn Basin in Wyoming likely formed by the combined mechanisms of fault slip and buckling. Geometric features of Pitchfork Anticline such as a localized anticlinal dome shape with tight hinges and amplitude that increases away from the fault tip are characteristic features of buckle folds produced in our numerical simulations. The coseismic uplift pattern produced during the 1985 earthquake on a fault under the Kettleman Hills Anticline and subsurface fold geometry of the anticline inferred from seismic reflection images are consistent with folding produced by the combined mechanisms of fault slip and buckling.
T43B-1342
The tectonic stress field variation along Ryukyu subduction zone
The Ryukyu Trench where the Philippine Sea Plate (PSP) subducts beneath the Eurasian Plate (EP) is distributed from Taiwan to Japan. Behind the Ryukyu Trench, the Okinawa Trough is rifting actively. Hence the convergent and divergent plate boundaries have produced enormous earthquakes. The earthquake behavior is complicated due to the complexity of trench-arc-backarc tectonic system. To better understand the stress field along Ryukyu subduction zone, we apply the damped stress inversion technique (Hardebeck & Michael, 2006). The earthquake catalogue of F-net broadband Seismograph Network of Japan will be used in this study. To avoid the uncertainty of deep earthquakes induced by subduction slab, only earthquake shallower than 35 km deep will be adopted. The preliminary result shows that both σ1 and σ3 stress fields display anisotropy along Ryukyu system. Most of σ1 azimuths in the Ryukyu Trench region have high plunge angles (~ 30° to 60°) and are oblique to trench orientation, however, have near horizontal plunge angles and are parallel to trench orientation at 129.5°E, 24.5°N. Across the Ryukyu Trench, most azimuths of σ1 in Okinawa Trough display high plunge angle, but also have a perturbation close to 126°E, 26.5°N. Based on the GPS data, both horizontal and vertical velocity fields show that crustal deformations behave different between southern and northern Ryukyu Arc. The combined stress field and crustal velocity fields imply that the Ryukyu subduction zone could be segmented into two parts.
T43B-1343
Neo-tectonic pattern of the southwestern tip of the Okinawa Trough backarc basin (Ilan Plain, Taiwan)
The Ilan Plain is located at the southwestern tip of the Okinawa Trough backarc basin which propagates westward into the Taiwan orogen. On May 15, 2002 and two times on March 5, 2005, three large earthquakes occurred in the near of the coastline of the Ilan Plain. To understand the relationship between the earthquake mechanisms and the geological context, we have collected seven seismic reflection profiles roughly parallel to the coastline of the Ilan Plain. A major normal fault, trending approximately ENE-WSW has been identified east of the locations of the two earthquakes of March 5, 2005, in the prolongation of the ENE-WSW trend of their aftershock seismicity. This feature has been named the Ilan Shelf Fault and might extend to the east along the northern trend of the Ilan continental spur. The centroid seismic moment tensor solutions of the March 5, 2005 earthquakes are both of left-lateral strike-slip faulting type and are consistent with GPS data, showing that the Ilan Shelf Fault is a normal fault with a left-lateral strike-slip component. Based on GPS data, we suggest that a pseudo-rigid block, limited in the west by the Lishan Fault and in the north by the Ilan Shelf Fault, has rotated clockwise due to the northwestward collision of the Luzon Arc with Taiwan. The clockwise block rotation is the mechanism causing the left-lateral strike-slip motion along the Ilan Shelf Fault. The Ilan Shelf Fault is a major tectonic and structural feature which corresponds to the southern boundary of the tip of the Okinawa Trough backarc basin.
T43B-1344
Fluid Mechanical Interactions In The Active Creeping Chihshang Fault Zone In Eastern Taiwan
The Chihshang thrust fault is one of the most active segments of the Longitudinal Valley fault (LVF) situated along the plate suture between the Philippine Sea plate and the Eurasian plate in eastern Taiwan. During the past two decades, different surface monitoring efforts have been undertaken across the Chihshang fault at different spatial and temporal scales. Some interesting phenomena were observed, revealing a close interactions between fluids, creep in the shallow fault segment and seismogenic zone at depth : (1) seasonal variation clearly influence Chihshang fault creep rate; (2) coseismic slip propagation was attenuated at shallow depth during the 2003 Mw=6.5 Chengkung earthquake and followed by a dramatic postseismic creep; (3) a decreasing creep rate was observed before the 2003 earthquake and the total crust shortening after Chengkung earthquake compensated the deficit of surface creep about 3-4 years before the earthquake. To better understand the role of fluids in the creeping of Chihchang fault, subsurface geophysical investigation, in-situ hydro-mechanical tests and long-term pressure, deformation and seismic monitoring were initiated within the frame of an integrated project called Chihshang Fault Monitoring Observatory.. Eight boreholes were drilled at a depth of 30-100 m through the Chihshang fault fault zone at the Chinyuan site. Pore pressure variations in hydraulic observation wells induced by artificial pumping and injections and natural seasonal variation were monitored, together with surface electrical 4D tomography. Creeping was monitored with both creepmeters and tiltmeters set in surface and TDRs set in boreholes. First results show a complex fault zone affecting several hydrogeological sedimentary units with a high variability of hydraulic properties, 6 10-4 to 2 10-8 of storativity values and 10-4 to 4 10-6 m2/s transmissivity values respectively. Groundwater flow is parallel to the deformation zone direction that is characterized by relatively impervious materials and act as a low-permeability barrier. Indeed, the fault zone bounds a confined aquifer in the foot wall block and a free aquifer in the hanging wall. Complex connections exist between the two aquifers, the foot wall aquifer being recharged by rainfall infiltration through the hanging wall aquifer. When recharge occurs, the fault movement is accelerated in relation with induced overpressures in the foot wall block. The size of the foot wall aquifer estimated of a few hundreds of meters, we suspect that rainfall-induced-effective-stress variations in that aquifer could influence the poro-elasto-plastic behaviour of the fault zone shallow segment.
T43B-1345
Application of the digital photogrammetry to obtain near-fault surface displacement field, eastern Taiwan
Measuring earthquake surface deformation provides information of both general kinematics of earthquake faulting and secondary deformation features, which usually reflect fault behaviors and surface geological heterogeneity. Recently, technology provided increasing variety of methods to measure surface deformation of earthquakes. For instance, photogrammetic analysis, especially sub-pixel correlation of remote sensing imagery, which we focused on in this study, is useful to reconstruct the co-seismic surface displacement field. Photogrammetry represents the process of deriving topographic information about an object through measurements made on aerial images of the object. Aerial photographs acquired using a high-resolution digital images and ground survey has been to monitor changes in topographic features and surface deformation, and to obtain the digital elevation models (DEMs). To rectify the image data, it is necessary to obtain geometrical parameters including camera type, frame focus, and aerial height. Furthermore, it is important to achieve a reasonable accuracy without a great number of ground control points (GCPs) when dealing with large-scale imagery. In order to obtain better ground controls for geometric calibration, we carried out RTK-GPS measurements at each ground points with a resolution of a few centimeters in horizontal and vertical components. Therefore, GCPs provide strong geometrical constraints for triangulation and orthorectification of images. This measurement is crucial because otherwise it is almost impossible to obtain orthorectified photographs with a precision in pixel level as good as 10-20 cm. The particle image velocimetry (PIV) method is an application of an optical image (e.g., aerial photographs and SPOT images) correlation technique, in which the horizontal displacement field is measured by comparison of images acquired at two different times. This method is based on a sub-pixel correlation of orthorectified images, using sliding windows. Each image is divided into several small areas (windows). The cross-correlation function is a pattern- matching routine that determines the relative displacement between images, which are shifted according to the best overlap. The residual offsets of image pairs are computed from the phase shift of the Fast Fourier Transform of the sliding window. This value corresponds to the horizontal displacement vector of the window. The correlation between two images can be efficiently performed as long as their texture is similar. The creation of DEMs from oblique and non-metric imagery using automated digital photogrammetry can be difficult. Recently, technology provided increasing variety of methods, especially seamless DEMs are generated for entire blocks or for any sub-block or polygon area. In this study, we focused to put forward a survey method to monitoring morphological change through reconstruct the high precision DEMs and measure surface displacement.
T43B-1346
Coseismic Changes of groundwater level due to the 2003 MW6.8 Chengkung Earthquake
Changes of groundwater levels induced by the MW6.8 Chengkung earthquake on 10 December 2003 were recorded in 123 out of 519 monitoring wells in Taiwan. The earthquake resulted from displacement of Chihshang fault along the Longitudinal Valley, which is the boundary between the Eurasian Plate and Philippine Sea plate in the eastern Taiwan. The 123 wells were placed primarily in the western Taiwan, ranging in depth from 12 m to 306 m. They are clustered at 90 stations located approximately 89 to 168 km from the epicenter. Among the 90 stations, 43 recorded coseismic rises, 37 recorded coseismic falls, and 10 recorded both coseismic rises and falls. The distribution of coseismic water-level changes in the Pingtung plain shows that coseismic rise predominated in most of the southwest coastal area, whereas coseismic fall prevailed in the northwest side of the plain. Base on the poroelastic theory, coseismic water-level rise is induced by compression, while coseismic water-level fall is induced by extension. During the chengkung earthquake, we found that the coseismic strain revealed by the coseismic changes of groundwater level were opposite to the perennial principal strain rates derived from the GPS data. Whether the phenomenon reflects the coseismic rebound of crustal deformation remains to be investigated. Although the peak ground acceleration decreases exponentially with hypocentral distance, the coseismic change of groundwater level does not show any trend of increase or decrease with hypocentral distances.
T43B-1347
Pre-seismic Change of Groundwater Level Prior to the 2000 ML6.7 Earthquake in Taiwan
Groundwater-level changes may reflect co-seismic tectonic stress redistribution and crustal strain. Earthquake- related groundwater level changes have been monitored by more than 600 wells in Taiwan. The Pingding (PD) well is one of the monitoring wells, located at the northern tip of the Douliu Hill, and was installed in an aquifer consisting of semi-consolidated deposits. The site is right on the axis of an anticline extending from north to south. The 206 m deep well has been monitored since 1997, and the water-level is recorded at one hour-interval with a resolution of 1 cm. According to the monitoring data, the well water-level is quite complete, and usually changes very slowly, even during the rainfall. We analyzed the response of water-level in the PD well to earthquakes of magnitude greater than 5.5 in the Richter scale from 1997 to 2005. Co-seismic water-level changes were observed in 19 of 78 earthquakes; about half of the changes are water-level rises. An abnormal water-level change was found approximately one hour immediately before the co-seismic change during the 2000 ML6.7 earthquake. The phenomena imply that the well water-level in PD not only reflects the change of tectonic stress induced by earthquakes, but also responds to the pre-earthquake crustal deformation.
T43B-1348
Implications of Groundwater Level Changes Induced by Earthquakes
Earthquake-related groundwater level changes have been monitored by a dense network of multiple-well stations in Taiwan. The coseismic water level changes vary in wells of different depths at most stations. The variation of coseismic changes in the vertical direction implies possible inconsistency between the observed water-level changes and the coseismic strains calculated from simple dislocation models. The recovery rate of coseismic changes after earthquakes decreases with depth, suggesting the importance of aquifer confinement in the recovery process. Well water level data recorded in the vicinity of the seismogenic fault during the 1999 M7.6 Chi- Chi earthquake provide a preliminary framework of regional distribution of coseismic changes. We found that the magnitude of coseismic change is associated with characteristics, rather than depth, of the aquifer. The coseismic change in the footwall of the ruptured segment of the fault was much greater than that of the unruptured segment, suggesting that fault displacement is a controlling factor in the abrupt change of pore pressure. Coseismic rise or fall correlates well with hypocentral distance in the vicinity of the ruptured fault. Poor correlation, however, is found for such changes further from the earthquake epicenter.
T43B-1349
Active Environment and Fault Segmentation of the Median Tectonic Line Active Fault System, in Japan
Understanding of rupturing patterns or fault segmentations for long active fault systems is one of the most intriguing research subjects to evaluate scale of cascade-type earthquakes. Many researchers have proposed segmentation models for long active fault systems, based on various fault activity data. But there are not sufficiently considered the stress condition through its entire length in these segmentation models. As a whole, many long active fault systems in the world might not indicate a uniform static stress condition through it's entire length, but heterogeneous. Therefore, we, at first, need to clear the heterogeneity of static stress condition of a long active fault system derived from the relative plate motion around it, and, then, divide into some stress condition segments based on the heterogeneous static stress condition, in order to consider segmentation models for long active fault systems. The Median Tectonic Line active fault system (hereafter MTLAFS), one of the longest and most active arc-parallel fault systems in Japan, is located in Southwest Japan, and the right-lateral strike-slip movement of this fault system is related to the oblique subduction of the Philippine Sea (PHS) plate. The MTLAFS has ruptured repeatedly during the last 10,000 years, and has high seismic potential both now and in the future. In recent, various kinds of geological and geomorphological researches are conducted for the MTLAFS, providing the detailed fault distribution and fault activity data. There evolve some extensional structures along the MTLAFS in the western Shikoku and Kyushu, while linear trend in the eastern Shikoku. Furthermore, the average horizontal and vertical slip rate along the MTLAFS become lower and higher from east to west, respectively. These facts suggest that the active environment of the MTLAFS changes through the eastern Shikoku to the western Shikoku and Kyushu. On the basis of geological and geophysical, we could divide the MTLAFS into three¡¡stress condition segments, East Shikoku, West Shikoku, and Kyushu segment. The East Shikoku segment and the Kyushu segment are under compressional and tensional stress conditions in N¡ªS direction, respectively. The West Shikoku segment locating between the East Shikoku segment and the Kyushu segment, is under a mixed stress condition (or a transitional stress condition) of compression and tension. Here, the MTLAFS corresponds to the northern tectonic boundary of the Nankai forearc sliver. The principle driving force of the Nankai forearc sliver is the NW subduction of the PHS plate. In addition to this, the SE pulling force derived from the back arc spreading in the Okinawa trough effects to the stress condition of the MTLAFS in the western Shikoku and Kyushu. As the results, the Nankai forearc sliver rotates counterclockwise relative to the Amurian plate, occurring this heterogeneous static stress condition of the MTLAFS. This regional stress condition and active environment around the MTLAFS is analogy with that of the North Anatolian Fault, however scale is different in comparison of these two fault systems.
T43B-1350
Fault model in the eastern margin of the Japan Sea based on the asymmetric topography profile of the oceanic ridges
The eastern margin of the Japan sea is one of the major convergent zone in East Asia. In a past few decades several M7-8 earthquakes (1964 Niigata, 1983 Japan Sea, 1994 SW-off Hokkaido, and 2007 off Mid-Niigata) occurred in this zone. They were often followed by large tsunami, which indicated a vertical seafloor movement by faulting. As a result of such crustal movement since 3Ma, many oceanic ridges with asymmetric topography profile have been formed along the coast. Recent seismic reflection surveys displayed the existence of active faults at the base of the ridges. However the fault parameters such as dip angle, width, top depth are still unknown. We introduce a method to estimate the unknown parameters paying attention to the asymmetric profiles as following 3 steps: (1) Calculation of surface deformation pattern for various combinations of fault parameters based on the dislocation theory. The calculated profiles of the surface pattern are stored to a deformation database. (2) Extraction of the ridge profile from observed seafloor topography. (3) Retrieval of the best fitting model to the observed profile from the deformation database. As a result of investigation using SEABEAM data, highly correlated surface deformation pattern were found for most oceanic ridges in the eastern margin of the Japan Sea. The region between the Okushiri Island and the Oga Peninsula is dominated by east-dipping faults with a high angle, whereas the regions on either side to the north and south tend to be dominated by west-dipping faults with a low angle (20 to 50 degrees). We will propose a fault model, which can be referred in a seismic and tsunami hazard assessment.
T43B-1351
Seismic Reflection Survey Across the Fukaya-Ayasegawa Fault System in the Northwestern Margin of Tokyo Metropolitan Area, Japan.
Several active faults, such as the Fukaya and Ayasegawa faults, have been traced fragmentarily from northwest to southeast in the northwestern margin of the Tokyo Metropolitan area, Kanto district. Because of the thick sediment cover in the Kanto plain, a considerable part of the fault system may be located as a blind fault. In 2005 the Earthquake Research Committee published a long-term evaluation for the Kanto-Heiya-Hokuseien (i.e. the northwestern margin of the Kanto plain) Fault Zone including these faults. According to the evaluation, the total length is about 82km, the average vertical displacement rate is 0.2-0.4m/ky, the displacement for one faulting event is 5-6m, and magnitude of a supposed event is M8.0. To clarify the detailed structure downto the pre- Neogene basement at the gapped area between the Fukaya and Ayasegawa faults in Saitama prefecture, we conducted a 10km-long seismic reflection survey which covers from the lowland in the northern Kounosu city to the eastern foot of the Yoshimi hills. The processed section shows clear images of the basement, some flexures and the two unconformities: the Kurotaki unconformity at about 3Ma and the Niwaya unconformity at about 15Ma. The basement, which crops out in the Yoshimi hills, is steeply dipping to the northeast in the western side of the section, whereas that in the eastern side is nearly horizontal and is situated at about 4km in depth. The flexural zone in the shallower part is totally 3km wide and shows cumulative deformation of reflectors since 3Ma at least. These flexures show no topographic expression, except the easternmost one which corresponds to the northern extension of the Ayasegawa fault. The long-term vertical displacement rate since 3Ma is estimated as about 0.1m/ky in the easternmost one. It is very interesting that the rate is equal to that estimated by using the 70ka terrace deposits. The displacement rate of the whole flexural zone since 3Ma is about 0.2m/ky which is comparable to or somewhat smaller than that of the Fukaya fault in the north.
T43B-1352
Identifying and dating blind thrusting events along the Biwako-seigan fault zone, central Japan, by dense geoslicing
We employed the ggeoslicerh technique to reveal the paleoseismic history of a blind thrust, for which traditional trenching is rarely useful or applicable because of broad surficial deformation and common shallow ground water at potential paleoseismic sites. A geoslicer is a relatively new paleoseismic tool for acquiring a subsurface geological sample; a thin sheet or slice of sediments is obtained to a depth of several meters by using a crane-hung weighted vibrator. Dense geoslicing and drilling across the southern Biwako-seigan fault zone (BSFZ), central Japan, together with a total of 109 radiocarbon dates, successfully revealed that the most recent blind thrusting event on the fault occurred between AD 1060 and 1260, most likely around AD 1170, producing a broad surficial warp more than 200 m wide with a vertical displacement of 5 m or more. Historical accounts strongly support that this event correlates with the 1185 M 7.4 Genryaku Kyoto earthquake. The geoslicing data also suggest four blind thrusting events during the last 18 kyrs and an average earthquake recurrence interval of 4300-5800 years, assuming characteristic behavior. Our results show that dense geoslicing can be a powerful paleoseismic tool for investigating blind thrusts buried at moderate depths, although paleoseismology of deeply buried blind thrusts still presents a big challenge.
T43B-1353
Coulomb stress change on the Xiaojiang and the Red River faults, southeastern Tibetan Plateau, from the 1970 Ms=7.7 Tonghai earthquake
The Xiaojiang and the Red River faults are two large-scale active faults in the southeastern Tibetan Plateau. Over the last 500 years, about 4 M >7.0 earthquakes, including the 1833 M~8 Songmin event, and 2 M~7.0 earthquakes occurred on the Xiaojiang fault and the Red River fault, respectively. In 1970, a devastating earthquake with Ms=7.7 happened near Tonghai of Yunnan province. This event raptures the Qujiang fault, a relatively small active fault in the corner where the right-lateral Xiaojing fault and the right-lateral Red River fault are conjugated. To investigate if the 1970 Ms=7.7 event would enhance the potential of seismicity along the two large-scale strike- slip faults, we explore quantitatively, in the framework of the Coulomb failure stress ( CFS), the idea that both coseismic and postseismic stress changes from this event may increase the likelihood of failure on some segments of the Xiaojiang and the Red River faults. A wide range of lithospheric rheology has been tested, associated with available ruptures of the 1970 event. Numerical results show that on some segments of the two active faults, the positive coseismic CFS is ~0.12-0.26 bar on the Xiaojiang fault, and ~0.24-0.52 bar on the Red River fault at 7-km depth. Relaxation of viscoelastic deformation of the lower crust and upper mantle increases the positive CFS to ~0.20-0.6 bar on the Xiaojiang fault, and ~0.32-0.84 bar on the Red River fault at the same depth after ~37 years. The modeled positive CFS correlates well with distribution of the instrumentally recorded seismicity between 1973 and 2003 around the two faults. This means that the 1970 event is presently increasing the CFS on the two active strike-slip faults. The Red River fault recorded little historic earthquake near the segment aside the 1970 event; along the Xiaojiang fault, more than ~37 strong earthquakes in the last ~16000 years have occurred with an average recurrence interval of M ~7.0 event merely ~200-500 years. The last strong earthquakes on different segments of the Xiaojiang fault are ~170 years ago (e.g. the 1500 M>7, the 1713 M~7, the 1733 M~7.8, and the 1833 M~8 earthquakes). Thus, the increase of CFS as our model shown, associating with the relatively short recurrence interval of strong earthquake should lead us to play a great attention on the potential of strong earthquake in the near future.
T43B-1354
Seismicity in the eastern Himalayan syntaxis
Previous studies of local seismicity in the Himalayan orogen and the Tibetan plateau have shown lateral variability in the depth of seismicity and in the fault types responsible for earthquakes. This study extends previous studies of seismicity with a local seismic network that was placed in the eastern Himalayan syntaxis. The syntaxis is a broad area of southeastern Tibet that marks the eastern edge of the Himalayan orogen and shows high gradients in GPS velocity, high rates of erosion, and at its core the syntaxis contains the young metamorphic massifs of Namche Barwa and Gyala Perri. The temporary seismic network was deployed for 15 months as part of the eastern syntaxis seismic experiment, a component of a larger interdisciplinary project researching the geodynamics of indentor corners. The array consisted of 70 seismometers in a regional array throughout much of the syntaxis with a focus on the massifs. Locations of earthquakes were determined from P- and S-wave arrival times, using a one-dimensional, three layer crustal velocity model. Initial results suggest that seismicity is primarily confined to the upper crust in the syntaxis, but the presence of a small number of deep earthquakes may be an indication that this distribution with depth is a tectonic rather than rheologic constraint. To the south of the syntaxis, the seismicity increases in depth, possibly related to subduction under the Burma arc. The area under Namche Barwa/Gyala Perri shows particularly high levels of seismicity, which may be a consequence of the rapid uplift of the massifs. Several other localized areas show elevated levels of seismicity, which is likely related to rifts and localized faulting. An area to the west of the massif, within the Lhasa block, has relatively little activity, which may have implications for tectonic models of deformation. To the east of the massif, seismicity correlates with river valleys, suggesting their locations may be fault controlled. Focal mechanisms show that strike-slip, normal, and thrust faulting are all legitimate styles of faulting in the region, and the area under the massifs shows particular variability in focal mechanisms. This range in fault types is an indication of the rapid spatial changes in tectonic style that occur within the syntaxis.
T43B-1355
Distribution of Active Faults and Recent Earthquake Ruptures along the Gobi-Altai Active Fault Zone, Southern Mongolia
The Gobi-Altai earthquake of 4 December 1957 in southern Mongolia is one of the largest recorded intracontinental earthquakes in the world. The surface ruptures associated with this M8.3 earthquake extend for about 260 km long and 40 km wide, involving strike-slip and reverse faulting (Kurushin et al., 1997). Earthquake ruptures appeared along a limited portion of the entire Gobi-Altai fault zone. West of the 1957 rupture zone, surface ruptures during historical and prehistrical earthquakes have been reported. Khil?fko and others (1985) reported surface deformation along the southern slope of the Bayan Tsagaan Mountains produced during the earthquake of 7 April 1958. They also identified two more Holocene rupture zones: west-northwest of the Bayan Tsagaan rupture near the village of Chandman, and west-northwest of the Chandman rupture near the village of Myangayn. However, previous studies do not discuss relationship between the distributions of active faults and earthquake ruptures along the entire active fault zone, because the exact location of the pre-existing active fault traces has not been mapped on large-scale maps. We have made distribution map of active fault traces based on interpretation of stereo-pair CORONA satellite photographs, which were taken between 1959 and 1972 for military intelligence during Cold War period. We also observed fault scarps in the field and made geomorphic profiles across the fault scarps. We were able to detect a lot of recent surface ruptures including the 1957 earthquake rupture, the Bayan Tsagaan, Chandman and Myangayn ruptures along the active fault traces. Fault scarps of these ruptures cross fluvial terraces of late Pleistocene as well as present riverbeds continuously. This indicates some of the active faults recently ruptured during a relatively short period. On the other hand, there exists an approximately 100-km-long portion between the 1957 and Chandman ruptures that has not displaced recent fluvial terraces, indicating that this portion did not recently rupture, probably during the 19th or 20th century.
T43B-1356
Deformation History And Paleoseismic Importance Of The Eupcheon Fault, SE Korea
The Korean Peninsula was widely regarded as being located at the relatively stable platform of the Asian continent. However, recently over 20 Quaternary faults have been discovered and reported in SE Korea. The concern on the activities of the faults is increasing, because new unclear power plants and low and medium nuclear waste disposal sites are decided to be located around this area. Especially, the Eupcheon Fault, one of the Quaternary faults, is very close to a nuclear power plant. We carried out tracing and trench surveys to know the characteristics and evolution of the fault. Our research is mainly focused on the fault contacts between Cretaceous sedimentary rocks and Tertiary volcanic rocks, and faulting events inferred from the trench sections. The fault is interpreted as a reactivated reverse fault (N20¢®¨¡E/40¢®¨¡SE) with ~ 6 m displacement. The fault may be developed as a normal fault in the Tertiary period under extensional regime, and it was repeatedly reactivated as a reverse fault at least three times during the Quaternary period under compressional regime, indicating inversion tectonics. Based on the data and interpretation, we suggested an evolution model for the area around the fault.
T43B-1357
Deep Structure and Tectonic Evolution of the Basins of Northwestern China
Northwest China contains a system of large sedimentary basins that include the Tarim, Qaidam, and Junggar. These basins are situated adjacent to the Tibetan Plateau and exhibit basements with contrasting evolutionary histories. Here we review three wide-angle seismic reflection/refraction profiles across these basins and their tectonic implications. The profiles include (1) a 1,400 km transect extending from the northern margin of the Tarim Basin to the eastern margin of the Qaidam Basin crossing the Altyn Tagh Range, (2) a 300 km transect extending from the northern to the southern margins of the Qaidam Basin, and (3) a 600 km transect extending from the northwestern to the southwestern margins of the Junggar Basin. The crustal structure of the Tarim Basin is interpreted as a typical stable continental platform. The seismic velocity boundaries between the felsic upper, intermediate middle, and mafic lower crust display clear divisions. Conversely, the Qaidam Basin, which lies at an elevation of ~3,000 m above sea level, is more similar to the soft deforming crust of the Tibetan Plateau. This crust is more felsic and lacks a high velocity mafic lower crust. Crustal structure of the Junggar basement includes a mix of oceanic materials and older blocks. Average crustal thicknesses are 55 km for the Tarim, 60 km for the Qaidam, and 50 km for the Junggar and the average seismic velocities are 6.0 km/s, 5.8 km/s, and 6.3 km/s (respectively). Of the three basins described, the Junggar is the thinnest and has the highest velocity.
T43B-1358
Crustal Structure Across the Three Gorges Dam from Wide-Angle Seismic Data
The Three Gorges Dam (TGD) is one of the biggest engineering projects in the world. In order to understand the crustal structure across the TGD, a wide-angle seismic profile was acquired from Fengjie, Sichuan Province, to Guanyindang, Hubei Province. The crustal velocity and reflectivity cross-section across the TGD shows lateral velocity variation and a northward-slipping, crustal-scale, strong reflector. The TGD is located at a transition in crustal thickness from about 45 km in the northwest to about 35 km in the southeast. The topography of China is characterized by three terraces. The Qinghai-Tibet Plateau is the first terrace with a surface elevation above sea level of >4500 m; the second terrace lies north and east of the Qinghai-Tibet Plateau at an elevation of 1000-2000 m; the third terrace is in the eastern part of China with an elevation of <500 m. Three cross sections along latitude 30, 31 and 32 from eastern Tibet to the East China Sea illuminate two steep topographic gradients at Longmenshan (97.5E), between eastern Tibet and the Sichuan basin, the TGD area (110E), and the flanking marginal seas (125E). The TGD area, named after the three gorges of Xiling Xia, Wu Xia, and Qutang Xia (Xia means "gorge" in Chinese), is located at the middle segmentations of the Yangtze River, extending from Fengjie, Sichuan Province, to Yichang, Hubei Province, with a total length of about 200 km. Due to the abrupt relief from the second to third topographic terrace of China, the unstopping, huge amounts of river stream provide the best place for the construction of the largest hydro power plant ever in the world. In order to provide a background of the crustal environment for the dam project, a wide-angle seismic profile along the Yangtze River was carried out. Tectonically, the studied area belongs to the Yangtze block and cut several sub-tectonic units: the Wanxian depression fold belt, the Zigui platform fold belt, the southwestern Hubei province complex fold belt, the Huangling Faulted Dome (HLFD), and the Jiangling Depression (JLD). We present the interpretation results of the crustal structure along the Three Gorges extension which shows: there are strong lateral variations both of crustal thickness and P-wave velocity, the Jiangling depression (the southeast of the profile) acts as the strong foreland, and the TGD experienced crustal thickening.
T43B-1359
Sag-ponding and its Significance in determining Paleo-seismic events along the active strike- slip fault
During the development of one active fault, we really want to know how it behaves and what it will do next. This mostly depends on the record and preservation of the information showing the action of the fault. Sparse young sediments or sediments with coarse grain along most of big strike-slip faults make it hard record and preserve the vestige of the paleo-seismic events. This extremely restricts the development of the Paleo-seismic research. Sag-ponding as well as the deposits in ponds, which are formed by the movement of the fault, can help settling the difficulty. Periodic sag-ponding is a feature to which should be paid more attention along the strike-slip fault, it can develop a pond to capture plenty fine sediments which well record the action of the faults. Sag-ponding can easily be found on the main active strike-slip faults in northern and eastern Tibet. By disclosing the sag-ponding depositions with 3-D excavations, sediment distribution and characters of relevant sag-ponds, and the relation between the sag-ponding and faulting were discussed. 1. Mechanism of the formation of the sag-pond When the valleys and ridges intersecting with the fault are displaced, the fault scarps will block the flow of the streams cut by the fault, or make the gullies develop ancon-like bend. This would form a space for water-storage, and thus a sag-pond comes into being. If the fault behaves like this many times, multi-sag-ponding will occur. 2. Rhythmic sag-ponding deposition features and stratigraphic sequence (1) Vertical characteristics. Observed from the stratigraphic profiles disclosed by the excavation, stratigraphic sequence shows good rhythms. There are several rhythms in each pond, and one rhythm is composed of the lower coarse layers and the upper fine layers. That is, the grains are coarser below and finer upward. (2) Transverse variation. In the direction parallel to the fault, the deposition center of each sag-pond appears regular movement, or migration, associated with the slipping of the fault. 3. Faulting pattern of the sag-ponding Observations indicate that the deposition rhythm is concerned with the periodic behavior of the faulting. During the long-term activity of the fault, when the strain accumulates to some extent, the fault will move suddenly and violently, then come into a relative quiet period. This leads to the grains courser downside and finer upside in one rhythm in the sag-pond. If the fault acts like this several times, it will form a sequence composed of several deposition rhythms. 4. Slip amounts estimated by transverse variations of the depositional rhythm. One rhythm in a sag-pond represents a sag-ponding process. As a result of the strike-slipping, the pre-formed deposition center migrates with the movement of the fault in a direction parallel to the strike of the fault trace. By measuring the location variation of each of the deposition centers, we can determine the amounts of horizontal displacements of sag-ponding sediments in each pond. 5. Paleo-seismic events reflected by vertical variations of depositional rhythms Every rhythm of the deposition is the product of one slip event of the fault. These rhythmic structures actually reflect the abruptness and periodicity of the movement of the fault, which are really paleoearthquake events and their reoccur intervals. It can be inferred that each sag-ponding rhythm corresponds to a paleoearthquake event, thus the sag-ponding deposition sequence can be discussed in contrast with the paleoearthquake event sequence.
T43B-1360
Salt Kinematics after Earthquake
Salt extrusions are simple natural models. The shapes of salt extrusions are complex gauges of the forces extruding them [9]. The uplift rates of a part of the Namakdan is between 1-3 mm/y-1 at the rim to 3-6 mm/yr-1 in the interior [5]. The salt glacier flowed plastically during the brief annual rainy season [6].The salt flow faster with temperature rise and flow slowly when temperature fall [3]. The displacements with>10cm/y and >50 cm/y suggest that, the salt extruded with rate 82 mm a-1 [9] but spread cm to m per year [8].The recent InSAR study Near Namakdan had no explanations about high activity of salt after earthquake [10].The coseismic vertical displacements suggest reactivation of blind thrust [11] .Our recent field measurements reported here suggest that any fast flow in salt may related to the mild to strong Earthquakes and may caused by diapiric reactivation. The earthquake of 27Nov 2005 with Mw ~ 6 occurred in Qeshm Island has a distance of 65 km to Bandar Abbass.The base of our measurements illustrated by Aftabi[1,2]and suggested by Bailly[4].Two wooden stakes about 50 cm in length and (2x3 cm<) in section were hammered vertically into the surficial marly salt along a line on the SW slopes of the thin southern namakier of Namakdan. Two others hammered on the walls of the cave in the south western part of the Namakdan diapir. The distance between stakes was measured (+/- a mm and cm) using a meter scale and the azimuths between them were measured (+/- a degree) by compass. Between readings, the meter scale was stored and carried in an ice chest to minimize its thermal expansion or contraction. Readings were made immediately one and two day after in the same times after installation measurements. We expected repeat of the main shock as mild earthquake one year later as earthquake cycling, we therefore return there and measured distance between stakes in salt one year after main shock but in the same times.IIEES reported the 26 and 27th 06 mild earthquake with Ml 3.3[12]. The distances between stakes both lengthened between measurements demonstrating local extension strains in the tiny southern namakier of Namakdan inside the cave and out in the marly cover.We interpreted some big activities in our 05 InSAR image as brine movements in the rim of the salt.Some of the distances between stakes exhibited complete elastic recovery during one day [1]. Others time-dependant elastic-plastic recovery[1], while in the Namakdan diapir the stakes just extended rapidly in scale of 1cm after 24 hours indicate permanent non recoverable plastic strains[1]with high rate after earthquake.With special thanks to M.Madani, C.J.Talbot and E.Feilding. REF: [1]Aftabi, P., 2000[2]Aftabi, p. et al, 2005[3]Aftabi, P., 2006[4]Bailly, E.B, 1931[5]Bruthans, J et al,In Press [6]Talbot, C.J., & Rojers, E.A.1980[7]Talbot, C.J, 1998[8] Talbot, C.J et al., 2000[9]Talbot, C.J, &Aftabi, P., 2004 [10]Nilforoushan, F. et al., 2005[11]Niessen et al.,In press [12] www.iiees.ac.ir http://WWW.GSI.IR
T43B-1361
Three-dimensional Seismic Structure of the Locked-Sliding Transition on the Subducting Plate Boundary beneath the Southern Part of Kii Peninsula, Southwestern Japan
The Nankai trough region, where the Philippine Sea Plate is subducted beneath the SW Japan arc, is a well- known zone of interplate earthquakes. The most recent great earthquakes occurred in 1944 (Tonankai Earthquake, M=7.9) and 1946 (Nankai Earthquake, M=8.0). Detailed crustal and upper mantle structure of the subducting Philippine Sea Plate and the overlying SW Japan arc are important to constrain the process of earthquake occurrence. Recent seismic experiments reveal the relation between the crustal structure and the seismogenic zone. However, little is known about the deeper part of the plate boundary, especially the transition zone on the subducting plate. To reveal the detailed structure of the transition zone on the subducting plate, we conducted seismic array observations in the southern part of Kii Peninsula, southwestern Japan. We deployed a temporary seismic array in the southern part of Kii Peninsula. Thirteen 3-component portable seismographs were deployed from January to February 2004. Waveforms were continuously recorded. We also conducted a deep seismic reflection profile to image the deeper part of the plate boundary in the southern Kii Peninsula. Two hundred eighty seismometers were deployed on a 60-km-long line in the east-west direction with about 200 m spacing, and waveforms were continuously recorded during a four-day period in November 2004. Three explosive shots were fired on EW-line. Our seismic experiment was designed in conjunction with the 2004 Daidai-toku seismic experiment (NS-line) across Kii Peninsula (Ito et al., 2006). We recorded the explosive shots fired on the NS-line as well as local earthquakes. Arrival times of local earthquakes and explosive shots were used in a joint inversion for earthquake locations and 3-D Vp and Vp/Vs structures. To obtain the detailed structure image of the transition zone on the subducting plate, the data recorded on the EW-line and NS-line were processed using the seismic reflection technique. Seismic reflection image shows the lateral variation of the reflectivity along the top of the subducting Philippine Sea plate. A broad reflection band is present where the clustered tremors occurred. Clustered tremors are located in and around the low Vp and high Vp/Vs zone. The low Vp and high Vp/Vs generally suggests the existence of fluid (e.g., Zhao et al., 1996), and indicates the occurrence of the tremors may be associated with fluids. At the northern Cascadia margin, there is a change in the reflection character on seismic images from a thin sharp reflection where the subduction thrust is inferred to be locked, to be a broad reflection where aseismic slip is thought to be occurring (Nedimovic et al., 2003). These studies suggest that the lateral variation of the reflectivity along the top of the Philippine Sea plate indicates the change of the degree of plate coupling caused by fluid.
T43B-1362
Earthquake Geology of the Bogd Fault, Gobi Altay, Mongolia
The Gobi-Altay earthquake of December 4, 1957, in southern Mongolia, is one of the largest recorded intracontinental earthquakes and one of four great earthquakes that occurred in this region during the 20th century. The rupture pattern associated with this M8 (published magnitude determinations range from 7.8-8.3) earthquake is complex, involving strike-slip and reverse faulting on several distinct geological structures in a zone about 260 km long and 40 km wide. The principal fault involved in this earthquake is the Bogd Fault, a left-lateral, strike-slip fault that ruptured for a distance of more than 260 km, with reported offsets up to about 6 meters. Earlier workers mapped the surface rupture in detail and conducted preliminary paleoseismic research on several of the faults involved in the rupture. Our new work is focused on acquiring new slip-rate data for the Bogd Fault, determining the ages of prehistoric earthquakes, and analyzing topographic features associated with the rupture to better characterize the slip vectors of the 1957 and earlier earthquakes. We conducted a kinematic GPS survey to measure offset risers on a flight of fluvial terraces in the western part of the 1957 rupture, and collected samples for 10Be analysis to determine terrace ages. At another site farther to the east near Noyan Uul, we collected additional samples for 10Be analysis to constrain the age of an abandoned alluvial fan surface offset from its source by 235 ± 35 m. We also surveyed offset terrace risers near the Bitut River and collected OSL samples to constrain their ages. We conducted paleoseismic studies at four sites and found evidence for the penultimate event at three of these sites, and evidence for three pre-1957 events at two sites. Analysis of OSL and radiocarbon samples from these paleoseismic sites will allow us to constrain the ages of these three prehistoric earthquakes. Analysis of detailed topographic data at several sites suggests the amount of slip on the Bogd Fault has been characteristic on the different fault segments over several earthquake cycles. In addition, these studies will allow us to better constrain the slip rate and the average time interval between earthquakes on the Bogd fault.
T43B-1363
Surface deformation adjacent to the Hukou fault in Northwestern Taiwan detected by ENVISAT ASAR interferometry
The Taiwan Island, which is young, as revealed by its dense seismic activities and rapid surface deformation, is located at the convergent zone between Eurasia plate and Philippine Sea plate. Because of the continued northwestward movement of the Philippine Sea plate and the active extension of the Okinawa Trough, the northern part of Taiwan is now under deformation. Since the northern Taiwan is the most populated area in Taiwan, the tectonic activity and the potential geological hazard of this area is an important issue for Taiwan. In order to realize the surface deformation behavior of this area, we apply DInSAR-technique to engage this study. The Hukou fault, main focus of this study, is one of the major and active structures in northwestern Taiwan, along which some industrial parks and communities are well developed. The SAR images used in this study are all acquired from 2003 to 2007 by ENVISAT satellite, which is launched by the European Space Agency in 2002. Our preliminary interferometric results reveal that the surface deformation in the urban areas are much clear than that in mountainous and rural areas. In some areas, juxtaposed against the fault zones, clear deformation patterns are obviously observed indicates that the deformation of this area is still active. After stacking all our interferometric results, we obtain that the average slant range displacement (SRD) reaches to around 0.5 cm/yr near the fault area. Radar Interferometry can efficiently be applied to observe the land surface deformation, and further help us to interpret and predict the underground tectonics and potential natural hazard.