From the teleseismic P-waveform data recorded at the dense mega seismic array deployed in the western Sichuan area by the State Key Laboratory of Earthquake Dynamics,Institute of Geology,China Earthquake Administration,we investigate the crustal anisotropy beneath the stations using waveform correlation method and weighted stacking method. As a preliminary result,we measured the fast polarization azimuth and time delay of the Ps converted wave in the receiver functions at 4 stations on both sides of Longmen Shan Faults. Our results show:1)The waveform correlation method is better than the weighted stacking method and it turns out not only the fast polarization azimuth,but also the time differences between the fast-and slow-wave; The results obtained by using the weighted stacking method are something undetermined due to that the symmetric axis of the crustal anisotropy medium is unclear previously; Application of both methods will be in favor of judging the reliability of the results. 2)The fast polarization azimuths are consistent each other at the stations in Sichuan Basin,suggesting the crust beneath Sichuan Basin has well integrality and a weak lateral deformation. 3)Taking the epicenter of the Wenchuan earthquake as a boundary,the fast polarization azimuth is parallel with the Longmen Shan Faults on the north side of the Sonpan-Ganzi block,and perpendicular to the faults on the south side. This suggests that under the obstruction of the Sichuan Basin,the soft lower crust beneath the north side of the Sonpan-Ganzi block has a NE direction extended deformation along the Longmen Shan Faults,and the crust on its south side is in the status of compressive deformation perpendicular to the faults. Our results can be used for interpreting the single-side rupture of the Wenchuan earthquake and the aftershock evolution.

In the western Sichuan(100°～105°E,26°～32°N),a mobile array consisting of 297 broadband seismic stations has been deployed by the State Key Laboratory of Earthquake Dynamics,Institute of Geology,China Earthquake Administration since October of 2006.Until June of 2008,a total of 690 teleseismic events(m_b>5.0,30°≤Δ≤90°)have been recorded.The May 12 Wenchuan earthquake(M_S8.0)provides an opportunity to test the western Sichuan array.The preliminary data analysis of the May 12 Wenchuan earthquake and its larger aftershocks has been carried out in this study.Our results show: 1)The event parameters of the May 12 Wenchuan earthquake sequence need to be modified and their location error reaches to 8～24km.A more reasonable estimation of the location of the main shock is possibly at the depth of 19km.2)the wavefield analysis of the Lixian earthquake(M_S5.9)of May 16,2008 manifests that the surface waves of this event are not fully developed,and thus its source depth should not be very shallow.The peak values of the ground-motion velocity on the vertical and horizontal component have an abnormal increase by 4 times and more of the normal attenuation,which is related closely to the faults within the range of 200～250km,when the topography and site effects are not considered.3)The preliminary analysis of the crustal and upper mantle structure beneath the Sichuan basin and the Songpan-Ganzi block manifests that the crust beneath the Sichuan basin thickens along the western direction and its lower crust displays the hard structure.The crustal thickness in the northeast of Chengdu City reaches 46km.The crustal structure beneath the Songpan-Ganzi block has complex lateral variations.The crustal thickness in the Wenchuan earthquake source region reaches 52km.In the depth range of 14～20km,its crust has a complex high-velocity structure with the averaged velocity larger than 4.0km/s.The Wenchuan earthquake should be located within the area with high-velocity medium.In the lower middle crust,a low-velocity layer exists with the S-wave velocity of～3.6km/s,which could provide a relaxed boundary condition for the upper crust movement-deformation.This observation is consistent with the abnormal attenuation of ground motion with the epicenter distance obtained from the wavefield measurements.

Up to now,all of the methods of isolating the receiver function from teleseismic P waves are based on the data recorded at a single station. However,the estimation of the receiver function based on the so called source equalization scheme will not be valid for broadband data,when a sharp discontinuity exists within the crust. The outlet for solving this problem is to avoid using the source equalization assumption. In this study,we propose a method for isolating the 3 component receiver function from multi channel seismic data based on multi channel maximal likelihood deconvolution princple. The source equalization scheme is avoided in our method,and this is valuable for improving the estimation of the receiver function in case of complex crustal structure. The numerical test by synthetics proves our method to be effective. Using this new method,we obtain the 3 component receiver functions at stations along the passive seismic profile across the Dabieshan orogenic belt from Daqipu (30°20′N,115°03′E) nearby Daye,Jiangxi Province to Cuilin(34°40′N,114°49′E)nearby Lankao,Henan Province. In comparison with the radial components of the receiver function from the data of a single station,it is indicated that the results given by both methods have un negligible differences.

The newly developed Digital Seismic Network in Capital Circle consists of 107 stations, the seismic monitoring and epicenter determination capabilities of which need to be verified. In epicenter determination, we should bring the superiority of digital seismic record into full play, and improve the reliability and precision of epicenter determination by using the corresponding digital processing technique. In this study, several digital processing techniques, including digital filtering, polarization analysis, station scanning and phase tracing, are used for improving the phase identification in the epicenter determination. Totally, 7 microearthquakes(M_L<2.0)recorded by the Digital Seismic Networks in Capital Circle in 2,000 were located precisely by using the Geiger method. In addition, the reliability and precision of the method and program have been verified by determining the epicenters of artificial events. It is shown that the error of epicenter location is less than 2.0km and the error of focal depth determination is less than 3.0km. The results of this study can be used for evaluating the seismic monitoring capability of the Digital Seismic Network in Capital Circle.

In 1997, 7 strong earthquakes ( M >6.0) occurred in Jiashi region, Xinjiang Uygur Autonomous Region. In order to gain an insight into this seldom seen seismic event, a passive broadband seismic array experiment was carried out by the Institute of Geology, China seismological Bureau in 1998. During the experiment two events, the Binamu and Lonkou earthquakes ( M ≥6.0), and their aftershock sequences were recorded by the portable seismic array. In this paper, these two earthquake sequences are relocated by using the Geiger method and the Joint Hypocenter Determination (JHD) method, respectively. The results show that: 1) the JHD method can be used to locate the relative position of an aftershock sequence more precisely than the Geiger method; 2) the Binamu and Longkou aftershock sequences delineate two nonparallel SWW trending faults, the occurrence of which are roughly consistent with those given by the moment tensor solution of the corresponding main shocks, and hence can be used for determining the focal plane of the Jiashi events; 3) the Binamu and Longkou earthquakes occurred at the gradient zone of the crust mantle boundary and at the convergence of the crustal high and low velocity media nearby the deep faults; this may indicate that the genesis of these earthquakes bears a close relation to the deep geological environment within the crust; 4) the tensional rupturing feature of the Jiashi earthquake sequence might be resulted from the heterogeneous lateral deformation of the northern edge of the Tarim basin.