T13B-1935
Adjoint Tomography of the Crust and Upper Mantle in the Japan Subduction Zone
The adjoint tomography technique is an effective tool for using 3-D models as initial models and refining them by iteratively minimizing the misfit between synthetics and data. In this study, we use this technique to obtain a more detailed 3-D image of the descending slab in the Japan Subduction Zone and its neighboring regions. We have very dense station coverage of our study area with a total of 845 stations, including Hi-net (more than 600 stations), F-net and Global Seismographic Network (GSN) stations. We use Zhao et al.'s (1994) 3-D slab model embedded in Lebedev and Nolet's (2003) regional model as the initial model in the tomographic inversion and calculate synthetics for each event. According to finite-frequency theory, the sensitive region along the ray path is given by a 3-D 'banana-doughnut' kernel, and the overall spatial distribution of the sum of all available event-station kernels determines the resolvable volume in the inversion. Using the automated windowing code FLEXWIN, we select a set of 206 events. We processed the data and synthetics using two types of bandpass filters: 6--30 s for all the records and 24--120 s for F-net and GSN records. For the first iteration, the frequency-dependent traveltime misfit measurements between synthetics and data are made in 44,709 windows for the period range of 24--120~s and 119,376 windows for the period range of 6--30~s. The combined adjoint sources are thus constructed based on these traveltime misfit measurements. Given the adjoint sources, we use the adjoint spectral-element method to calculate banana-doughnut kernels for P, S and surface waves for the selected records. The weighted sums of the banana-doughnut kernels for all event-station pairs, with weights determined by the traveltime measurements, are used to construct misfit kernels. These gradients are then used in a non-linear conjugate gradient algorithm to further improve the existing 3-D models. We are currently at the first iteration of the 3-D models. The preliminary results indicate that seismic velocities of the Pacific slab need to be faster to reduce the misfit; the observed small-scale features requiring slower seismic velocities might be related to mantle wedge melts, but this needs further investigation in susequent iterations.
T13B-1936
Tectonic structure of Dokdo and adjacent area in the northeastern part of the Ulleung Basin of the East Sea using geophysical data
The northeastern part of the Ulleung Basin in the East Sea is composed of volcanic islands (Ulleungdo and Dokdo), seamounts (the Anyongbok Seamount, the Simheungtaek and the Isabu Tablemounts), and a deep pathway (Korea Gap). To understand tectonic structure and geophysical characteristics of Dokdo and adjacent area, We analysed geophysical potential data of KORDI(Korea Ocean Research and Development Institute), KIGAM(Korea Institute of Geoscience and Mineral Resources), and NORI(National Oceanographic Research Institute of Korea) around the Dokdo volcanic body except Ulleung Do because of empty data of its large island. Also, we eliminate the effect of water and sediments from the free-air gravity data to process 3D Moho depth inversion. 3D tectonic structure modelling of the study area was developed using Moho depth inversion result and sediment thickness data of NGDC(National Geophysical Data Center). The free-air gravity anomalies of the study area generally reflect bathymetric effects. Although the Dokdo seamounts have a similar topographic size, the decrease of free-air anomaly toward Isabu suggest that Isabu is oldest among the seaounts and have high degree of isostatic compensation. High Bouguer anomalies in the central part of the Ulleung Basin gradually decreases toward the Oki Bank. This feature suggests that the crust/mantle boundary is shallow in the central part of the Ulleung Basin. The complex magnetic pattern of Dokdo suggests that it might have erupted several times during its formation. The magnetic anomaly amplitude of Isabu is much smaller than that of Dokdo. Such low magnetic anomalies are attributed to a secondary change caused by the metamorphism or weathering of ferromagnetic minerals of the seamount during a long period of time after its formation. Analytic signals show high anomalous zones over volcanoes. Also, there are high analytic signal values in Korea Gap indicating magmatic intrusion in thick sediments. The power spectrum analysis of the free-air anomalies, the bouguer anomalies, and the magnetic anomalies indicate that the depth to the Moho discontinuity varies from -13.3 to -19.6 km. The inversion and 3D modelling suggest that the depth of Moho discontinuity ranged from -13 km to -20 km in the study area. They show the roots of seamounts in crust and the trending of mantle upwelling. The roots of volcanoes go deeper toward Isabu, indicating higher compensation. The sediments are very thick and the Moho depth is very shallow between Dokdo and Anyongbok seamount. The trending of mantle upwelling is northeast- southwest.
T13B-1937
P- and S- wave velocity structure in and around the Itoigawa-Shizuoka Tectonic Line (ISTL) fault system revealed by dense seismic array observations
The ISTL is a major tectonic structure that divides the Japanese Island arc into northeast and southwest parts. It was formed as a normal fault in the early Miocene and represents the southwestern boundary of the northern Fossa Mangna rift basin to the north, and the boundary between the Japanese arc accretionary prism units and the Izu-Bonin arc crust to the south. Previous studies have provided the sallow structure of the different ISTL fault segments, but the detailed crustal structure along the ISTL is yet to be revealed. The online permanent seismic station network in the area is not sufficient to accurately locate the earthquakes occurring in the area and also not dense enough to provide a detailed structure of the earth's crust. Over the past 3 years we have installed temporary seismic stations along the STL. We have deployed 60 stations in the southern, 58 stations in the central and 60 stations in the northern ISTL regions. We have combined the data retrieved from the temporary stations with the data available from the online permanent stations in the ISTL area and manually re-picked 63,275 P- and 68,847 S- wave arrival times from 1,945 events from the 5th August 2003 to 31st December 2006. The Double Difference tomography method (Zhang and Thurber, 2003) was used in order to accurately relocate the hypocenters and obtain a 3D P- and S- wave velocity (Vp and Vs) structure beneath the ISTL fault system. The relocated hypocenters in the southern ISTL coincide with the deeper extension of the active faults in the area. The relocated hypocenters are deeper than those reported by the Japan Meteorological Agency (JMA) in the northern ISTL and shallower at the central and southern parts. The average depth of the hypocenters is shallower in the northern ISTL (3 - 8 km) and gets progressively deeper towards the central (8-15 km) and southern (15-25 km) ISTL. The tomographic analysis has provided a detailed Vp and Vs image of the crust in the area below the ISTL. The 3D velocity model that we have acquired of the crustal structure in the area is in accordance with the geological boundaries. The northern tomograms fit accurately with the deeper extension of the Matsumoto basin and the central uplift zone geological units. In the central ISTL, the Yatsugatake volcano magmatic conduit was imaged. In the southern ISTL, we imaged the downwards continuation of the low grade metamorphic rocks that constitute the Chichibu-Shimanto belts of the southwest acrretionary prism of the Japanese arc, and of the igneous rocks that form the Izu-Bonin arc crust.
T13B-1938
Spatial variation of lower limit of the shallow inland seismogenic layer and its seismic velocity structure in northeastern Japan
Northeastern (NE) Japan is located at a subduction of the Pacific plate beneath the North American plate. Variety of earthquakes can be observed in this region: earthquakes in the upper crust beneath the land, earthquakes along the plate interfaces and earthquakes within the descending oceanic slab. The volcanic front is formed parallel to the trench axis and associated with magma activity. It is believed that lower boundary of the shallow seismicity represents seismic-aseismic or brittle-ductile boundary in the crust and is inversely related to temperature. The spatial depth variation of lower limit of the seismogenic layer in NE Japan is in tight connection with seismic and magmatic activities, crustal deformation and crustal thermal regime and their influence on the crust strength. Reliable tomographic imaging and precise hypocenter relocation may help one to understand interaction of above mentioned tectonic processes in the highly compressed and deformed crust and their influence on seismogenic layer. Study area is covered by a dense nation-wide digital seismograph network composed of stations belonging to universities, JMA, Hi-net digital stations. Arrival time data from local earthquakes, recorded by the integrated seismic observation network during 1998-2008 have been used. Most of earthquakes are confined in a depth range of 3 to 25 km, and concentrated along the volcanic fronts. The tomography method by Zhao et al. (JGR, 1992) has been used in this study. The study area was divided into partially overlapped 2° by 2° size sub-areas for simplicity and accuracy in calculation. Numbers of events and stations vary depending on each sub-area. The 1-D model JMA2001 is taken as an initial model. We set all layers of grid-net up to the Moho discontinuity in the upper and lower crust with spacing of 5 km depths. The spacing between grid nodes is about 10 km in horizontal direction. The total number of grid nodes in one sub-area is 5819. Depth differences between initial and relocated events are mainly bounded in ±5 km range. Earthquakes occur predominantly at 7-8 km depth with ±3 km limit showing even and concentrated seismogenic layer beneath the Tohoku region and sparse and scattered one beneath Hokkaido, but commonly become shallower right beneath active volcanoes and deeper in-between them in NE Japan. Obtained 3D velocity structure images are in agreement with the results of previous studies. Our results show different structure patterns on both sides of the volcanic front at shallow depths. Volcanic front is a kind of evident fence between low velocity and high velocity in western part and eastern part, respectively. This picture is preserved up to the depth of 20 km. Earthquakes within the upper crust are located in high-velocity areas sandwiched between low-velocity zones as it can be seen on cross-sections. It is obvious that mantle upwelling with high temperature is one of the main causes of the shallow earthquakes in this region which plays a key role in lateral variation of the bottom of the seismogenic layer in NE Japan.
T13B-1939
Crust and upper-mantle structure beneath the Yellow Sea using Receiver function analysis
Korean peninsula consists of Precambrian tectonic blocks (Nangrim, Gyeonggi, and Yeongnam massifs) that are bounded by fold belts (Imjingang Belt and Ogcheon Fold Belt). A recent review of the tectonic evolution in the Korean Peninsula suggests that the Imjingang Belt represent the collision zone between the North China Block and the South China Block in the early Triassic. Later, a collision along the Ogcheon Fold Belt took place in the Jurassic to constitute the major tectonic framework of the Korean Peninsula. A controversial hypothesis, however, suggests that the Ogcheon Fold Belt form in the early Paleozoic as a boundary between the North China Block and the South China Block. Hypotheses on the tectonic history of the Korean Peninsula have focused on the location of the suture zone between the North China Block and the South China Block. Most hypotheses are proposed based on the observations of surface geology. Yellow sea is located in the center of controversial hypotheses, although no clear evidences have been found from the Yellow Sea. Scientists from Korea and China teamed up to resolve the long standing question of tectonic history of the region. 11 temporary broadband seismic stations are installed around the Yellow Sea to monitor seismicity in the region and to collect teleseismic waveform data. Receiver function technique is applied to the collected data to define 1-D representative crust and upper-mantle velocity models beneath the broadband seismic stations. 1-D velocity models are compared to identify any correlations between China and Korea. We present the preliminary results of the China-Korea international collaboration in the Yellow Sea area.
T13B-1940
Opening of the South China Sea: New insights from magnetic data
Before drilling can be actually implemented in the South China Sea (SCS) ocean basin, geophysical investigations, data processing and interpretation, remain to be the best means to study the regional geodynamics. We made a systematic investigation on the major structures and tectonic units in the SCS basin based on magnetic data processing and interpretation. For enhanced magnetic data interpretation, we carried out various data reduction procedures, including upward continuation, reduction to the pole, 3D analytic signal and power spectrum analyses, and magnetic depth estimation. Magnetic data suggest that the SCS basin can be divided into five magnetic zones, each with a unique magnetic pattern. Zone A corresponds roughly to the area between Taiwan and a relict transform fault, zone B is a circular feature between the relict transform fault and the northwest subbasin, and zone C, zone D, and zone E are the northwest subbasin, the east subbasin, and the southwest subbasin, respectively. This complexity of magnetization suggests that the SCS has evolved from multiple stages of opening under different tectonic settings. Magnetic data reduction also fosters improved interpretation on continental margin structures, such as the Mesozoic and Cenozoic sedimentary basins and the offshore south China magnetic anomaly. Curie isotherm estimated from magnetic data is the shallowest in the deep basin, averaging at about 20 km in depth. This depth is substantially larger than the MOHO depth, and consequently it can be interpreted that the uppermost mantle of the deep basin is also magnetized. The southwest subbasin shows slightly shallower Curie isotherm than the east subbasin. A zone of shallow Curie isotherm depth is found existing in the continent-ocean transition zone along the northern continental margin. This zone is coincident in location with a fault zone and a high heat flow zone. A magnetic anomaly along the northern SCS continental margin shares some similarities with the North American east coast magnetic anomaly, both lining up along the continental margins and trending parallel to the coastline and tectonic trends. Different from North American anomaly, however, the SCS magnetic anomaly does not extend over the continent-ocean transition zone based on the studies of numerous seismic lines in the region, but primarily on the continental shelf. Magnetic depth estimations using Euler deconvolution imply that the sources of the SCS magnetic anomaly are mostly buried in the upper crust, with a possible depth ranging from a few kilometers to about 20 km. The magnetic sources may have variable shapes along the strike of the SCS magnetic anomaly, and some may extend to a depth of 40 km. The SCS magnetic anomaly is unlikely to be associated with Cenozoic magmatism concomitant with continental rifting and oceanic floor spreading.
T13B-1941
Geodynamic Evolution of the Northeastern South China Sea
We present a geodynamic evolution model of the northeastern South China Sea (SCS) updated the kinematic context based on a re-interpretation and analysis of all available magnetic data. Using the latest available multi-channel seismic data, two significant tectonic phases T1 and T2 were identified in the northeastern SCS. T1 is a slight tensional tectonic event and T2 is a major compressive event. In the absence of the drilling data in the deep basin, the identification of the magnetic anomalies, the age of major unconformities at ODP Sites 1146 and 1148 drilled on the northeastern SCS margin as well as a re- interpretation of the tectonic subsidence curve based on drilled holes in the west Taiwan basins provided age constraints for dating these two tectonic events. Tensional phase T1 is characterized by tilted blocks and fan-shaped deposits developed shortly (8-10 Ma) after the onset of SCS oceanic domain (37.8 Ma). It corresponds to the first ENE-WSW to E-W change in spreading directions, which occurred around chron C10 (~28.7 Ma). Tectonic phase T2 is characterized by the uplift of ENE-WSW rift features and NE-SW transverse features associated with the major plate reorganization in East Asia and a northeast shift of the Ryukyu subduction zone from the Luzon-Ryukyu transform plate boundary (LRTPB) to east of the present- day position of Taiwan, which occurred 17-18 Ma ago. As the T2 intra-plate compressive deformation only occurred south of the LRTPB, we suggest that the slab pull effect of the dead slab might have been transmitted to the oceanic crust south of the already inactive LRTPB.
T13B-1942
Seismic Tomography off SW Taiwan: a Joint Inversion From OBS and Onshore Data of 2006 Pingtung Aftershocks
On 26 December 2006, two Mw=7.0 events occurred offshore south of Pingtung; one is associated with normal-faulting and the other with strike-slip faulting. The area in which these earthquakes were located is not usually expected to have large earthquakes. We deployed 11 short period OBS over the source zone for one week and recorded a series of aftershocks; these same events were also recorded on land at the CWB network stations. The joint dataset made it possible for us to perform 3-D velocity tomography and earthquake relocation in this region, where the velocity structures were not well known and location of earthquakes with only land data was uncertain. The tomographic results show a prominent high Vp perturbation zone (HVPZ) that we consider as the uppermost mantle of the subducted plate dipping from SW to NE beneath southern Taiwan. Most of the relocated earthquakes are distributed just above the HVPZ or near and along the bottom of a relatively low velocity subducted crust. Our results show that the subducted and bent Eurasian plate off SW Taiwan could have been unbent and become an eastwards concave geometry for the upper 30 km. The main shock is near the bottom of the inflected surface. The distribution of the earthquake sequence generally displays in a NW-SE direction, coinciding with the plate convergence orientation between the Philippine Sea Plate and Eurasian Pltae. This orientation also follows a relatively low Bouguer gravity anomaly stripe that is due to a heavy loading of the Taiwan orogen on the east-dipping Eurasian Plate. Considering that the hypocenter of the first main-shock is near the bottom of the aftershocks, we suggest that the first normal faulting earthquake was caused by an unbending effect in the subducting crust and this event triggered the release of accumulated energy between the Philippine Sea Plate and Eurasian Plate. Thus, we suggest that the rupture surface of the Pingtung earthquake sequence had propagated upward and northwestward in the direction of plate convergence.
T13B-1943
Studying the lithospheric structure and seismotectonics of the Three-Gorges Reservoir region, China
For many years there has been a rich accumulation of geologic investigations in the Three Gorges Reservoir (TGR) region of the Yangtze River that provided useful information to hydro-power generation, engineering construction, flood control and geohazard mitigation. However, there has not yet been a complete knowledge base on the TGR region's deep lithospheric structure, whose relationship with seismicity is still largely unknown. The successful completion of the main dam and full water impoundment of the 660-kilometer-long reservoir makes the safety and geohazard mitigation a top concern for the TGR and the entire central China region. To address such concern, we are conducting a study of the lithospheric structure and seismotectonics of the TGR region by mapping the 3D crust-mantle seismic velocity structure of Huangling anticline in the core TGR region. Since 2007 we have been collecting and analyzing seismic and gravity data of the TGR region, and conducting passive surveys of local and teleseismic earthquakes using a mobile array of 60 RefTek125A (TEXAN) seismographs. Our field tests indicate that the records from the TEXAN's are compatible with the records of permanent stations for teleseismic events. Using 5-Hz geophones, the TEXAN recorded reliable first break waveforms of events that are located 2000 km away. The data that we have recorded so far include dozens of small events in the TGR region as well as more than 100 aftershocks of the devastating 2008/05/12 WenChuan earthquake located about 450-600 km from the TGR region. Preliminary analysis of the array seismograms indicated good data quality for M2.5 or greater events that are located more than 60 km away from the stations and for M4.0 or greater events more than 400 km away. The quality and quantity of the data allow us to generate shot profiles from the WenChuan aftershocks and local events in the TGR region. We will present details of the data, analysis process, and preliminary results on the crust-mantle seismic structure and its relations with major faults, seismicity, and near-surface geohazards in the TGR region.
T13B-1944
Cenozoic Transformation of the Cratonic Lithospheres beneath the Chinese Continent
The Chinese continent has three Precambrian cratons: Yangtze, Tarim, and Sino-Korean (also called North China) craton. The subduction of the Pacific and Philippines oceanic plates in the east and the collision with the Indian plate in the southwest during the Cenozoic strongly influenced the Chinese continent, and dynamically shaped the contemporary tectonic framework and the geographic relief of the continent. Consequently, the three cratons were influenced by the events, which possibly result in cratonic lithosphere thinning. The present Sino-Korean "craton" is one of the most active continental blocks in the world, and has thin lithosphere and strong seismic and thermal activities. Except for the Sichuan basin on the west of the Tibetan plateau, the Yangtze craton also has a thin lithosphere. Our recent tomographic models show that the lithosphere thinning of the two cratons is resulted in by upwelling melting from stagnant subduction slab in the upper-lower mantle transition zone. Previous studies showed that the Tarim craton seems not to be destructed because its present lithosphere is cool and rigid. Recent geological evidences showed that the Tarim craton was part of ancient Western China craton, broken in the Cenozoic during the Indo-Asian collision. New evidences from our seismic and seismic-thermal models of the present uppermost mantle and of the lithospheric base topography showed that the lithospheric base of the Tarim has also been destructed and thinned during the Cenozoic.
T13B-1945
Crustal Structure of Cretaceous Accretionary Orogenic Belts of North-Eastern Asia Based on Data from the 2DV Geophysical Transect
North-Eastern Asia consists of the Siberian craton and its deformed passive margin and broad accretionary mosaics of terrains with diverse affinities (carbonate platform, island arcs, micro-continents, etc) intruded and overlapped by several belts of Jura-Cretaceous plutonic and volcanic rocks. Three main orogenic belts can be distinguished in the region: 1. Late Jurassic to Lower Cretaceous Yano-Kolyma belt, 2. Cretaceous Oloi- Chukotka belt, and 3. Cretaceous Okhotsk-Koryak belt. Geophysical transect 2DV extended from Magadan to Pevek and has imaged the deep structure of the crust beneath these main belts. The interpretation of this new seismic data allows the following insights: (1) The Oloi-Chukotka orogenic belt represents a fold belt with Archean to Early Proterozoic? crust near its margins transitioning to possible Paleozoic age (rifted) crust beneath its central part and accreted Precambrian basement beneath the Omolon and Ckukotka blocks. (2) Thickness of the crust increases across the paleo continental margins from 35-39 km (Uda-Murgal island arc and inner zone of Okhotsk-Chukotka volcanic belt-OCVB) up to 55-58 km (near the junction of OCVB with Yano-Kolyma orogenic belt and under margin of Omolon massif near Oloi volcanic arc). Thickening of the lower crust can be seen as imbrication. Seismic "transparency" and increasing velocities up to 5-7 km/sec characterize the middle crust to the Moho beneath the main igneous/ore-bearing zones of these orogenic belts, suggesting that fluid-magmatic processes were important in creating present-day crustal composition/structures (3) The South-Anyui suture was imaged and is interpreted as structures related to thrusting of Chukotka onto weakly deformed Oloi terrain from north to south. (4) Isotopic signatures of magmatic rocks shift to more depleted values towards Cretaceous paleo continental margins (beneath OCVB), in agreement with the rejuvenated nature of the lower crust beneath extensional provinces superimposed on accretionary belts, as supported by xenolith studies.
T13B-1946
Imaging Poisson ratio of uppermost mantle beneath China
We have obtained a Poisson's ratio image of the uppermost mantle beneath China by performing tomographic inversion of travel time differences between Sn and Pn. The arrival pairs were selected from the Annual Bulletin of Chinese Earthquakes from 1984 to 2007. The dataset includes 50,136 arrival pairs from 11,470 earthquakes recorded by 116 stations. The average Poisson's ratio is 0.26. The preliminary tomographic results show that 1) the Poisson's ratio is low in the stable cratons around the Tibetan Plateau such as the Tarim and Junggar basins, the Ordos craton, and the southern region of the Sichuan basin; and 2) the Poisson's ratio is high in the entire Tibetan Plateau, the North-South Seismic Zone, and North China; and 3) the high Poisson's ratio region in the Tibetan Plateau extends to the surrounding cratons, suggesting that a) the uplift of Tibetan Plateau results from a high Poisson's ratio or partially melted rocks beneath the plateau, and b) the soft rocks intruded into the bottom of surrounding cratons.
T13B-1947
Variations in Velocity, Gradient, and Anisotropy in the Upper Mantle of Eurasia using Sn Travel Times
Various researchers have modeled upper mantle velocity variations using Pn travel times. Based on Zhao (1993) and Zhao and Xie (1993), the effect of upper mantle gradient on arrival times can be approximated by a simple term that can be extended to 2D for tomographic studies. Previous Pn studies using this method suggest high gradients associated with stable and convergent zones, low gradients with extensional zones -- inversely related to temperature gradient. We invert Sn travel time data in Eurasia to solve for upper mantle slowness, gradient, and anisotropy. Using the Bondˆr et al. (2004) ground truth (GT) criteria as well as a priori knowledge, we have identified GT25 or better events available for tomography in the region. Data are culled based on source-receiver distance, minimum number of arrivals per event/station and residual outliers from highly regularized to optimally regularized models. The final data set includes ~2400 stations and ~14,500 events resulting in ~145,000 Sn arrivals. For the inversion, we assume a ray incident angle of 45 degrees into the upper mantle to accommodate the crustal legs, varying the crustal thickness. We solve for station and event terms to account for variations below the stations, event origin time bias and depth error, and overall pick errors. Initial results display a 56% RMS residual improvement over the constant slowness/gradient starting model and suggest that Sn velocity variations are consistent with Pn variations and tectonic activity. Sn gradient displays a "muted" pattern, relative to Pn . The overall Sn upper mantle gradient is low, with values of ~0.0005 1/s around cratons and zero to slightly negative gradients in Europe as well as eastern China. Small areas of higher gradient (~0.0025 1/s) are observed in eastern Tibet and Japan. Solving for anisotropy, in addition to slowness and gradient, removes some of the high frequency patterns in the resulting Sn slowness as well as the higher gradient pattern in Tibet, suggesting a possible tradeoff with the other parameters.
T13B-1948
The Crustal and Upper Mantle Structure of China From Teleseismic Receiver Functions
We collected and processed a large amount of high-quality broadband teleseismic waveforms recorded at 48 stations in the Chinese National Digital Seismic Network (CNDSN) across China to invert a fine 1-D crustal and upper mantle velocity model beneath each station by receiver function analyses. A cross-correlation based method was used to select mutually coherent receiver functions, which yielded over 200 traces for most of the stations. These high-quality receiver functions have been used to estimate the lateral variations of Moho depth and crustal Vp/Vs ratio (hence Poisson's ratio) in China by an advanced H-k domain search algorithm. The individual RZ (radial and vertical components) receiver functions were stacked for different ranges of back azimuths to represent their mean receiver functions for a given station. Because the geological and geophysical a prior constraints do not exist on all stations, we first adopted grid search to obtain a simple three-layer crustal velocity model selected from a complete model database by modeling stacked receiver functions on every station. The starting model with fine layers down to upper mantle was then constituted from this coarse model, the derived crustal thickness and Vp/Vs ratio, and velocity constraints at Moho from deep seismic soundings and other geophysical investigations for China. Finally this initial model was utilized in a linearized inversion to obtain a much finer earth structure model. The derived models do no change the main feature of Moho estimated from our previous studies, but reveal more detailed velocity structure across Moho than H-k domain search. Although we utilized the results obtained from Pn tomographic studies in projecting time to depth to constrain the velocity at Moho and the average crustal velocity, the crustal velocity structure, as well as the crustal thickness and Poisson's ratio obtained from receiver functions still show significant discrepancies with those inferred from Pn waves. The receiver function velocity images are characterized by block structures corresponding to geological features bounding by large fault systems, such as the North China Block, the South China Block, the Tianshan fold system, and the Tibetan plateau. The azimuthal variations, as well as complicated crustal and upper mantle structures such as sediment and LVZ beneath some stations have also been observed from receiver functions. These results will finally help to define a high-resolution earth velocity model for China, and hence better determine tectonic settings and identify crustal and upper mantle materials.
T13B-1949 INVITED
3D Shear Velocity Structure of Crust and Upper Mantle in China From Ambient Noise Tomography
We perform ambient noise tomography of China using the data from the China National Seismic Network and global and PASSCAL stations in the region. We obtain Rayleigh wave group and phase velocity dispersion maps at 1 by 1 degree grids for periods from 8 to 60 s. The results are combined with longer-period dispersion maps from global earthquake-based measurements. We then obtain the 3D shear velocity structure of the crust and upper mantle in China by inverting the dispersion curves at each grid. The inversion results show remarkable features for continental China and in particular the Tibetan Plateau (TP), including slow sedimentary layers of all the major basins at the shallow depth, striking east-west contrasts in Moho depth variation and lithosphere thickness, fast (strong) mid-lower crust and mantle lithosphere in major basins surrounding the TP (Tarim, Ordos, and Sichuan) (in contrast, Qaidam Basin does not have such a "deep root"). These strong blocks thus seem to play an important role in confining the deformation of the TP to be a triangular shape. The Moho changes from plateau to Tarim and Sichuan Basins are quite sharp. The India lithosphere seems to terminate around the Bangong Nujiang Suture as indicated by the fast-slow velocity contrast in the mantle lithosphere, but it seems to extend further north under E. Tibet. In northwest TP, slow anomalies extend from crust to great depth (200 km). A widespread, prominent low-velocity zone is observed in midcrust in the TP, which are generally connected and seem to reach to the surface near the margins of the TP, consistent with the notion of the growth of the TP by crustal channel flow and the extrusion of channel flow materials at the topographic fronts.
T13B-1950
Enhanced Monte-Carlo Adaptive Moving Window Inversion and its Application in China
We developed an Enhanced Monte-Carlo Adaptive Moving Window Inversion (EMAMW) and applied it to China. High-resolution images of the lithosphere beneath the Chinese continent were obtained. High quality first arrivals extracted from the Annual Bulletin of Chinese Earthquakes (ABCE) were used. Compared to the previous Monte-Carlo adaptive moving window inversion (Sun et al., 2004), our results show higher horizontal and vertical resolutions and guarantee smooth transitions among adjacent locations. The P and S models obtained have good correlation with tectonic features, and the predicted travel times through the 3-D model matched well with the observed ones at local and regional distances. We validate both P and S velocity models by fitting seismograms and resulting in better agreement compared to the previous version.
T13B-1951
Insight into the origin of the Tengchong intraplate volcano in southwest China from local and teleseismic data
A high-resolution tomographic image of the crust and upper mantle under Yunnan Province in southwest
China was determined by using a large number of teleseismic data measured precisely from digital
seismograms as well as local earthquake arrival times recorded by the dense Yunnan seismic network. Our
resulting model shows a clear low-velocity (low-V) column extending from the surface down to about 400 km
depth under the active Tengchong volcano and some high-velocity (high-V) anomalies existing in the mantle
transition zone. Furthermore, the low-V anomaly extends horizontally toward the northeast at ~250-400 km
depths. We consider that the Tengchong volcano is a subduction-zone volcano caused by the slab
dehydration and corner flow in the mantle wedge, though the subducted slab is a continental plate (Burma
micro-plate). Our results show that the upwelling flow under Tengchong originates at ~400 km depth. A
prominent low-V anomaly along the Red-River fault zone extends down to the upper mantle, reflecting that
the Red-River fault zone may have cut through the crust to the upper mantle. The two Dayao earthquakes
(M 6.1 and M 6.2) in 2003 occurred on the margin of high-V anomalies and are underlain by a prominent low-
V anomaly in the lower crust and upper mantle, suggesting that the Dayao earthquakes may be associated
with the deep fluids released from the dehydration of the subducted India slab (or Burma microplate). These
results have significantly improved over the previous studies and provided new seismic constraints on the
dynamic processes of the India-Asia collision.
http://jianshelei.bokee.com
T13B-1952
Tomographic inversion of Sg arrivals reveals the upper crustal S-velocity structure beneath the INDEPTH IV Transect: NE Tibetan Plateau to Qaidam Basin
During the INDEPTH IV controlled-source experiment in June 2007, Sg arrivals were well recorded by the 20 broadband and 29 short-period three-component seismographs, from GIPP, Germany and SEIS-UK, deployed at 5-6 km station spacing along the 270 km long profile across the Kunlun mountains in NE Tibet. Based on these arrivals, Sg arrivals could also be safely identified on the vertical-component geophones attached to the 949 IRIS-PASSCAL Texans, spaced at 100-600 m along the profile. A tomographic inversion of the Sg arrivals recorded by these 998 instruments from 5 large shots (1000-2000 kg) and 100 small shots (80 kg) reveals the S-velocity structure of the upper crust down to about 10 km depth beneath the profile. The major lateral variation in upper crustal S-velocities along the profile is from lower velocities beneath the Qaidam basin in the north to higher velocities beneath NE Tibet in the south. Beneath NE Tibet, there are also lateral variations with lower S-velocities beneath the valleys (basins) along which the North Kunlun and South Kunlun Faults run and higher velocities in the mountainous regions between and to the north and south, where older rocks are generally exposed at the surface. From the S-velocity model, there is no evidence within the upper crust for major overthrusting of NE Tibet over the Qaidam basin.
T13B-1953
Anisotropy reveals the upper mantle transition zone beneath the center of the Tibetan plateau
The Tibetan Plateau is the result of the collision between India and Eurasia, which started approximately 50 Ma ago. One of the best ways to image deformation in the plateau's interior is through the detection and interpretation of seismic anisotropy. Such anisotropy can give us insight into the past and current mantle deformation that will provide essential clues to its dynamic evolution. Shear wave split measurement has been a popular method to investigate the deformation of the upper mantle and the plate movement. In our study, the data were recorded by a seismograph array operated by Institute of Tibetan Plateau Research, Chinese Academy of Sciences between Sept. 2005 and Oct. 2006. The array consisted of 70 broadband seismometers, deployed between latitudes 28N-34N and longitudes 85E-90E. Our array recoded SKS and SKKS data from relatively narrow ranges of back azimuth. But, most of the core phases have good quality. We used a modified code of William Menke and followed the cross-correlation method to calculate the splitting parameters, that is, the azimuth of the fast polarization direction and the delay time between the split phase arrivals. A Fresnel zone analysis suggests that the upper mantle (above 175 km depth) is the most likely source of anisotropy. In general, the direction of fast polarization is related to the directions of past and present plate motions. In northern part of our study region the relationship between the deformation of the upper mantle and the crustal movement is nearly coherent. While GPS campaigns provide evidence suggesting crustal flow in the study region, the result of anisotropy in southern parts of our study region argues against such flow in the upper mantle, i.e. the fast polarization directions are quasi perpendicular against known surface features and geodetic estimates of the crustal displacement fields. The changes of delay times reveal a pronounced different tectonic regime between south and north, which means an upper mantle transition zone existing near 30.5N. It divides the study area into two parts. In the southern part of the study region delay time is 0.4 s in average. It is smaller than the average 1.0 s in the northern part. We suggest that the change of observed delay times from south to north reflects a fundamental change in deformation regime across our study region. It may be related to the subduction of Indian plate to the plateau and suggest a transitional zone in the upper mantle.
T13B-1954
Crustal and uppermost mantle structures beneath the southeastern Tibet from 3D full- wavefield finite-frequency tomography using ambient seismic noise
Our previous studies on finite-frequency tomography using teleseismic body-wave traveltimes in southeastern Tibet show seismic anomalies that indicate the delamination of a thickened Eurasian mantle lithosphere and its causal relationship to the formation of the north-south trending rift in southeastern Tibet. However, due to the limitations of teleseismic body waves, the resolution in the crust and uppermost mantle is poor. In order to improve the models at shallow depths, we apply the finite-frequency methodology to the surface wave tomography using ambient seismic noise. Unlike previous studies that calculated the dispersion curves from the estimated Green's functions to solve for phase or group velocity maps, we have developed a new approach using the 3D finite-frequency sensitivity kernels for the Green's functions. We have collected continuous data for the stations in the Namche Barwa seismic experiment deployed in southeastern Tibet during 2003-2004, and computed the cross-correlations between all possible pairs of records among the stations. We constructed 3D sensitivity kernels for the estimated Green's functions derived from the cross- correlation of ambient noise, taking into account the effects of surface topography. Finite-frequency phase delays between the estimated Green's functions and 3D full-wavefield synthetics are then used to resolve the crustal and upper mantle structures.
T13B-1955
Crustal and mantle velocity models of southern Tibet from finite frequency tomography
While the formation of the Himalayan Range and Tibetan plateau has broad implications for earth sciences the exact mechanism of mountain-building processes, in particular the roles of the Tibetan mantle lithosphere and upper mantle in the rise of the plateau, remains poorly constrained. Using traveltimes of teleseismic body waves recorded by several temporary seismic networks in the region, we carried out finite-frequency tomographic inversions to image the three-dimensional velocity structure beneath southern Tibet. The results reveal a region of relatively high P- and S-wave velocity anomalies extending at a steep angle from the uppermost mantle to at least ~200 km depth beneath the Higher Himalaya. There is a strong low P- and S- wave velocity anomaly that extends from the lower crust to more than 200 km depth beneath the Yadong- Gulu rift, suggesting that rifting in southern Tibet is not simply a shallow crustal process decoupled from the mantle by a low-viscosity lower crust. An anomalously low Vp and low Vp/Vs ratio structure is located at about 75-160 km depths beneath the southern end of the Pumqu-Xianza rift. Its upper limit coincides with the intermediate-depth earthquakes and the corner of the "ramp-and-flat" Moho geometry. Based on mineral- physics constraints, we suggest that the anomalously low Vp and low Vp/Vs are due to the co-existence of granulite and eclogite phases in the mantle. Together with the steep high-velocity anomaly in the south, the observations suggest a partial removal of the Indian lower crust and the underlying mantle lithosphere south of the Indus-Yalu suture in the study area.
T13B-1956
Detailed Moho Topography Beneath the Southern Lhasa Terrane, Tibet by Receiver Functions
The collision of the Indian and Asian plates since about 55Ma has created the most gigantic plateau with the thickest crust of the world. There is, however, no general agreement on the modes of the crustal thickening. Even the crustal thickness of the Tibetan plateau still remains poorly determined. While it is generally accepted that the Tibetan crust is roughly double normal thick, and that it thins somewhat toward the north, individual observations of Moho depth differ spatially, and between different techniques at the same place, by greater than 20 km. In this work we compare P and S receiver functions at station LSA, located in the southern Lhasa terrane, to determine the crustal thickness beneath the station. Two significant interfaces can be clearly seen in the P receiver functions at depths of about 60 and 80 km (Moho), whereas the latter phase is absent in the S receiver function data, although the S receiver functions sample a similar region of the Moho as the P receiver functions do. Possible mis-interpretation of primary conversions of P receiver functions by crustal multiples can be excluded by distinct differences in moveout curves of phases. We can model the observed P and S receiver functions by a strong topography of the Moho dipping to NEE direction at an angle of 32°. This result may indicate that the Moho beneath Tibet is very complicated and has strong lateral variations, and is consistent with earlier wide-angle reflection and receiver function data showing an imbrication Moho architecture resulted from separated tectonic crustal thickening. The Moho dip direction is locally perpendicular to the Indian plate motion, suggesting that the lower crust flow is decoupled from the underlying Indian mantle lithosphere. The observation may also explain the different results of Moho depths previously made and suggest that a detailed map of Moho depth is only possible with 2-D dense- spacing seismic experiments.
T13B-1957
A Tectonic Link Between Sri Lanka and Cuddapah Basin: Geophysical Evidence
The Indian subcontinent is generally believed to be a mosaic of three Protocontinents, namely, Aravalli, Singhbhum and Dharwar. The Dharwar Protocontinent comprises a large portion of the southern peninsular India and Sri Lanka. Sri Lanka is thus presumed to be a midplate platelet and an extension of southern Indian mass. But the available gravity signatures of the south India and Sri Lanka, which differ significantly, do not provide support for such a presumption. While southern India exhibits a large number of lineaments characterizing well-known faults and shear zones, gravity image of Sri Lanka is devoid of any lineaments. Based on the gravity images it is difficult to visualize Sri Lanka as an extension of Southern India. The proterozoic Cuddapah basin occupying an area of over 35,000 sq.km., constitutes one of the most interesting geological features of South India on the east coast. The eastern coast in South India is also occupied by a major greenstone belt, which is characterized by a significant gap in the region east of the Cuddapah basin. In the plate tectonic scenario, if one were to move Sri Lanka into the Cuddapah basin it may be seen that Sri Lanka fits closely into this Proterozoic Cuddapah basin. The "gap" in the green stone belt gets filled up by the green stone belt of Sri Lanka exactly, suggesting that the green stone belts of Sri Lanka and those of India on the east coast once belonged to the same belt. Also, in this position the gravity image of Sri Lanka shows continuity with the rest of the south Indian land mass and merges as a single image. Based on the size, geophysical signatures and geological constraints, it is postulated that the ejection of the landmass constituting Sri Lanka from the Indian plate during early Proterozoic times resulted in the formation of the Cuddapah basin. The basin further evolved to its present stage due to subsequent orogenic and tectonic processes.
T13B-1958
Modeling Earth Deformation from Monsoonal Flooding in Bangladesh using Hydrographic, GPS and GRACE Data
The Ganges, Brahmaputra and Meghna Rivers, which combine in Bangladesh, together have a mean annual discharge second only to the Amazon. The vast majority of the flow occurs during the summer monsoon. The immense monsoon river discharge and rainfall causes widespread flooding in the Ganges-Brahmaputra Delta (GBD). In an average year, 1/3 of Bangladesh is submerged; it can reach 2/3 during an extreme flood. The mass of this impounded water represents a large load on the surface of the earth that can be readily observed in the GRACE gravity field and by GPS geodesy. It is the second largest seasonal anomaly in the GRACE gravity field, reflecting terrestrial water storage in SE Asia over the GBD. The 18 continuous GPS stations we have installed Bangladesh record a vertical signal of up to 6 cm that is inversely correlated to river level and discharge, in addition to rapid subsidence of up to 1.5 cm/yr from tectonics and sediment loading.. To calculate the water load throughout Bangladesh, use >300 river gages of the Bangladesh Water Development Board (BWDB). In addition, the GRACE gravity field provides estimates of the integrated water storage. We model the water loading as elastic deformation at this timescale using both the analytic code of Becker and Bevis (2004) and the finite-element code PyLith. We vary elastic properties beneath the flooded region to fit the observed GPS motions. We present results of the modeling including estimates of the elastic properties of the earth beneath the GBD.
T13B-1959
Imaging Crustal Structure Beneath Southern Australia: SoC Preliminary Results
We deployed in Southern Australia a broad-band seismic network of 28 three-component seismometers between December 2005 and August 2008 for a nine months period at most site. This seismic experiment called SoC (Structure of Cratons) should provide 3-D images of the lithosphere and constraints on the variation in crustal and lithospheric thickness through exploitation of recordings of distant earthquakes in a corridor extending from the Curnamona craton to the east coast of Australia, crossing the transition from Precambrian to Phanerozoic belts. We also analyze data from permanent broad-band stations: one GEOSCOPE station and eight stations from Geoscience Australia network. SoC is designed to be complementary of previous broad-band deployments in the region (SKIPPY and TASMAL). Including data from previous experiments we estimate if the lithosphere thins gradually toward the east and we also look at potential underplating of materials at the base of the crust. We present preliminary results obtained with teleseismic P receiver functions. We compare results from receiver functions computed in the ZRT coordinate system and receiver functions obtained from a rotation into the ray coordinate system L, Q and T. Two different inversion techniques are used to determine the crust and upper mantle structure that can explain the observed receiver functions. The 1-D velocity structure will be used to invert for seismic moment tensor of regional earthquakes. The crustal Vp/Vs ratio and the crust thickness can be estimated from the analysis of the travel times of waves converted at the Moho and of crustal multiples converted phases. The comparison between Moho topography and crustal Vp/Vs ratio give us the possibility to discuss the nature of the basement of the lithosphere beneath Mount Gambier and the Murray Basin: whether it is oceanic or continental.
T13B-1960 INVITED
Developing a Short-Period, Fundamental-Mode Rayleigh-Wave Attenuation Model for Asia
We are developing a 2D, short-period (12 - 22 s), fundamental-mode Rayleigh-wave attenuation model for Asia. This model can be used to invert for a 3D attenuation model of the Earth's crust and upper mantle as well as to implement more accurate path corrections in regional surface-wave magnitude calculations. The prerequisite for developing a reliable Rayleigh-wave attenuation model is the availability of accurate fundamental-mode Rayleigh-wave amplitude measurements. Fundamental-mode Rayleigh-wave amplitudes could be contaminated by a variety of sources such as multipathing, focusing and defocusing, body wave, higher-mode surface wave, and other noise sources. These contaminations must be reduced to the largest extent possible. To achieve this, we designed a procedure by taking advantage of certain Rayleigh-wave characteristics, such as dispersion and elliptical particle motion, for accurate amplitude measurements. We first analyze the dispersion of the surface-wave data using a spectrogram. Based on the characteristics of the data dispersion, we design a phase-matched filter by using either a manually picked dispersion curve, or a group-velocity-model predicted dispersion curve, or the dispersion of the data, and apply the filter to the seismogram. Intelligent filtering of the seismogram and windowing of the resulting cross-correlation based on the spectrogram analysis and the comparison between the phase-match filtered data spectrum, the raw-data spectrum and the theoretical source spectrum effectively reduces amplitude contaminations and results in reliable amplitude measurements in many cases. We implemented these measuring techniques in a graphic-user-interface tool called Surface Wave Amplitude Measurement Tool (SWAMTOOL). Using the tool, we collected and processed waveform data for 200 earthquakes occurring throughout 2003-2006 inside and around Eurasia. The records from 135 broadband stations were used. After obtaining the Rayleigh-wave amplitude measurements, we analyzed the attenuation behavior of the amplitudes using source- and receiver-specific terms calculated from a 3D velocity model of the region. Based on the results, we removed amplitudes that yielded negative average attenuation coefficients, and included an additional parameter in the inversion to account for the possible bias of the CMT moments. Using the high-quality amplitude measurements in a tomographic inversion, we obtained a fundamental-mode Rayleigh-wave attenuation- coefficient model for periods between 12 and 22 s for Asia and surrounding regions. The inverted attenuation model is consistent with the geological features of Asia. We observe low attenuation in stable regions such as eastern Europe, the Siberian platforms, the Indian shield, the Arabian platform, the Yangtze craton, and others. High attenuation is observed in tectonically active regions such as the Himalayas, the Tian Shan, Pamir and Zagros mountains.