| The Indian plate and Eurasian plate began to collide before 50 million years, and the formation of the qinghai-tibet plateau during the collision. In recent years, the western of SiChuan has been widely concerned for studying of thr dynamics of the interior continental. The region is divided into Longmen Songpan block, the Ganzi block and the Sichuan basin by Longmen Shan(LMS) fault and Xianshui(XSH) fault,and it is one of main areas of inland earthquakes in China. The Wenchuan earthquake,which took place in May 12, 2008, is located in the LMS fault in the region. After the earthquake, many geophysical scholars obtained rich achievements by deep exploration. By using receiver function inversion, ambient seismic noise and travel time tomography method showed that, there is distribution of high S wave velocity structure in the middle and upper of the beneath the Sichuan basin. The low-velocity material in the middle-lower crust of Songpan-Ganzi block moves then upward when blocked by the rigid Sichuan basin crust, causing the rupture of the LMS fault zone.Although the velocity structure was obtained by the above methods, no the interface information from those ways. Studying the discontinuities form crust and upper mantle of the Chuanxi area and getting the informotion of interface between different blocks in greater depth, help us know about dynamic of the southeastern margin of Qinghai Tibet Plateau.Converted wave sounding is one of the important ways to tudy the earth's crust and upper mantle velocity structure. Receive function put forward developing from the converted wave sounding. We used Common Conversion Point(CCP) stacking technique to image the crustal structure along the profile. We first calculate the ray-paths of the receiver functions using a background velocity model. The amplitude at each point on the receiver function, after being corrected for the incidence angle effect, is assigned to the corresponding location on the ray-path where the P-S conversion occurred, using its time delay with respect to the direct P. But receive function is interference by random noise, and migration imaging is influenced by noise, some datails of the imaging were covered by extraneous phase. In order to enhance the imaging effect and improve the image resolution, it is important toattenuat the noise of receiver function and improve the quality of the stack migration miaging.Curvelet transform is a multiscale directional transform to analysis signal, and it has developed rapidly in recent years. Compared with two-dimensional denoising method, based on curvelet transform to denoise has minimal damage to signal. On the directional identification, curvelet transform is better than the wavelet transform. It is especially suitable for detection and denoising the seismic phase. And using curvelet transform is effective in attenuation of the scattering noise induced by the laterally heterogrneous crust, and the signal-to-noise ratio of each receiver function and the spatial traceability of Ps phases as well as multiple reflections are significantly enhanced. After attenuation of noise in receiver functions using curvelet transform,we used the receiver function migration imaging to study the discontinuities structure form crust and upper mantle in the Chuanxi area.This work is based on the receiver functions data of 33 stations of the array along the profile of 31 ° N in the Chuanxi area. We used curvelet transform to denoise after moveout, and obtained discontinuities form crust and upper mantle at a depth of 0-800 km along 31°N profile. Our results show that:(1) Curvelet transform is a multiscale directional transform to analysis signal,and it is good at identifing direction. It is especially suitable for seismic phase detection and denoising. And using curvelet transform is effective in attenuation of the scattering noise induced by the laterally heterogrneous crust, improve the signal-to-noise ratio of data. Based on curvelet transform to denoise is different from setting the fixed threshold to denoise. Let C be a matrix that contains all of receiver functions in on station, And R is a matrix of the sames size as C. All of the rows of R are identical, as they all contains average receiver function which are sumed after move-out in one station. The curvelet transform provides a sparse and compact representation of R that allows the definition of a mask in the curvelet domain from the amplitude of R. This mask is then applied to C in the curvelet domain, and the filtered signals are transformed back into the time domain. It is proved that attenuat the noise of receiver function is obvious by using the method, and significantly improved the information of the converted phase.(2) After we used Common Conversion Point(CCP) stacking technique with the denoised receiver function, we not only got better effect with 4s or 5s of the receiverfunction, also obtained the higher imaging resolution by using 0.5Hz of the receiver functions. The obvious denoising effect can be observed in the picture of migration.Therefore, using the receiver function after curvelet transform can be used to improve the imaging effect and enhance the resolution of the image.(3) The migrated crustal receiver function image along 31 ° N profile we obtained have a same shape of mobo as S-wave velocity structure of the crust and upper mantle along the profile of 31°N by the joint inversion of P-receiver functions and(ambient noise) surface wave. Migration results show that, 10-20 km thick of positive velocity discontinuity under the moho of the eastern margin of the Qinghai Tibet Plateau. So the guess of double moho maybe ture.(4) Migration results show that, the upper mantle transition between the eastern margin of the Qinghai Tibet Plateau and the Yangtze block is 5-10 km thick than average thickness of the global transition. And 150 km under the LMS fault, the transition is more thickness, which we had obtained from tomography method. The results indicate that there may be the delamination of the shallow crust or upper mantle in the western of LMS fault. |
PDF Full Text Request