| The western Sichuan plateau is located on the eastern margin of Tibetan Plateau. In recent years, the western Sichuan has become the key region for understanding the dynamics of the Tibetan Plateau and the continental block boundaries, so that the geophysical explorations are very active in this area, because the high-resolution imaging of the crustal and upper mantle velocity structure beneath the western Sichuan region is very important for the dynamics of the Tibetan Plateau and active blocks in the continent.On the other hand, the western Sichuan plateau is a high-seismicity region in China’s continent. The high-resolution seismic imaging of the crustal and upper mantle velocity structure beneath this area also plays a key role in understanding the genesis of earthquakes in this region. However, limited to the station coverage, the resolution of the early published results is poor. In particular, the occurrence of the Wenchuan earthquake (Ms8.0) in May 12th of 2008 motivates again the great interest to explore the crustal structure beneath the western Sichuan region. Although a series of new results have been published about this issue, they are different given by different authors.Before the Wenchuan earthquake, the State Key Laboratory of Earthquake Dynamics, Institute of Geology, China Earthquake Administration carried out the deployment of a dense movable array composed of 297 broadband seismic stations in the western Sichuan (26°N-32°N,100°E-105°E) in October of 2006 (called the Western Sichuan Array hereafter). Its main task originally was to carry out high-resolution integrative seismic studies of the crust and upper mantle structure underneath the western Sichuan, as a part of the project entitled "Dynamic process and prediction of large earthquakes on active block boundaries" and funded by the National Basic Research Program (called 973 Program, for short). The Western Sichuan Array has collected a big amount of data, including the records before and after the Wenchuan earthquake. These data are great valuable for investigating the 3D high-resolution crust and upper mantle structure beneath the western Sichuan.In terms of the non-linear receiver function inversion technique, Liu et al. (2009) presented the 2-D S-wave velocity structure of crust and upper mantle and the average Poisson’s ratio over the crust along the 31°N profile through the Wenchuan earthquake epicenter from the Western Sichuan Array data. By using teleseismic P-wave traveltime tomography technique, Guo et al. (2009) presented the 3D P-wave velocity structure of the crust and upper mantle within the depth range of 400km beneath the western Sichuan region. To overcome the non-uniqueness of geophysical inversions, the investigation of the crustal velocity structure beneath a region from different data by using different methods will be the most efficient. Recent studies show that surface-wave Green’s function between two seismograph stations can be estimated from the long-time cross-correlation of ambient seismic noise. Ambient noise tomography is a method that obtains the surface wave dispersion by the long-time cross-correlation of ambient seismic noise between two seismograph stations and then yields the Earth’s interior velocity structure by the surface wave tomography technique. Shapiro et al. (2005) successfully extracted from the ambient seismic noise cross-correlation function by using group velocity dispersion. Since then, the technique of ambient noise tomography has been developing rapidly and become widespread concerned. In particular, ambient seismic noise has been widely used for surface wave tomography of shallow underground structure. Ambient noise tomography technology provides a new way to obtain the high-resolution underground velocity structure by use of the dense seismic array.In this thesis, the theory and method of extracting empirical Green’s functions (EGFs) and then obtaining Rayleigh wave phase velocity tomography through a long-time ambient noise cross-correlation are introduced. And also the neighbourhood algorithm (NA) as a global searching method for non-linear inversion is mentioned. By use of the image transformation technique of measuring interstation phase velocity dispersion presented by Yao et al. (2005), we have selected 11,358 of Rayleigh wave phase velocity dispersion curves. On this basis, we studied the high-resolution 3-D shear velocity of crustal structure in western Sichuan by using Rayleigh-wave phase velocity tomography and velocity structure inversion. Our results provide a new independent evidence to study the crustal structure of the western Sichuan plateau and the Sichuan Basin, and important constraints are given for further study.The method of measuring phase velocity dispersion curves from ambient noise is improved here as followed: 1) The Western Sichuan region has complex terrain, and both sides of the Longmen Shan faults have dramatic elevation differences. Therefore, the corrections of elevation are considered while calculating the distance between station pairs.2) In order to gain the phase velocity dispersion of stations at short range, we have improved the filter window for extracting phase velocity dispersion curves from the EGFs so that the strong amplitude around the zero point of EGFs can be more effectively suppressed. After the above process, the short-period dispersion curves are greatly improved.By means of ambient seismic noise tomography, our results can be summarized as follows:1) As a passive method without earthqukes, surface-wave tomography from ambient seismic noise can be free from the dependence of the natural seismic source parameters. Only if the continuous records of seismic noise are long enough, the high quality EGFs will be extracted by cross-correlation.2) The resolution of ambient seismic noise tomography by use of the dense seismic array is much better than the results from traditional surface-wave tomography. The intensive and uniform distribution of dispersion curves is very important to high-resolution surface-wave tomography.3) The seismic noise sources in periods of 10s-20s are mainly from sea waves, which lead to amplitude variations of cross-correlation relataed to seasonal changes. However, phase velocity dispersion curves will be still reliable, if they are extracted from the superposition of the positive and negative branches of EGFs.4) The method of ambient seismic noise tomography also has its own limitations known as the difficulties to obtain the information of periods above 40s, so that it is almost impossible to image the depth beyond the crust only by ambient noise data. Otherwise, the strong amplitude around the zero point of cross-correlation functions can disturb the signals if two stations are very close to each other (e.g., the distance of station-pair is less than 30km)By means of surface-wave tomography from ambient seismic noise, we obtained the Rayleigh wave phase velocity maps in periods of 2-35s divided by grids of 0.25°×0.25°. And then we achieved the 3-D S-wave velocity structure of the crust in western Sichuan from NA inversion of the 273 grid points of Rayleigh wave phase velocity distribution. Our results manifest the significant discrepancies between the crustal structures of the Chuandian block, Songpan-Garze block and Sichuan basin, which can be summarized as follows.1) The S-wave velocity structure of shallow crust in a depth range of 2km-8km (corresponding to the short-period phase velocity distribution of 2s-8s) is consistent with the surface structural features. As boundaries of the Chuandian block, Songpan-Garze block and Sichuan Basin, the Longmen Shan faults and Xianshuihe faults have performed significant control effects to them. The evident low-velocity structure of the foreland in the Sichuan basin suggests that thick sedimentary layer does exist as a thickness of about 8km-10km.2) The S-wave velocity structure of the upper-middle crust in a depth range of 8km-20km (corresponding to the intermediate period phase velocity distribution of 12s-18s) shows that the S-wave velocity structure of upper-middle crust in the Chuandian block and Songpan-Garze block exists obvious non-uniform lateral variations with sub-blocks of different scales and high-low velocity variation. The S-wave velocity of the middle crust in the Sichuan Basin as a whole is higher than the other two blocks.3) The S-wave velocity structure of the middle-lower crust in a depth range of 20km-40km (corresponding to the long period phase velocity distribution of 25s-30s) shows that the middle-lower crust of Songpan-Garze block, especially Chuandian block has a relatively wide range of low velocity distribution, which means they are relative weak in the lower crust. And the crust of the Sichuan Basin shows overall high-velocity characteristics, which means the Sichuan Basin has a relatively rigid middle-lower crust. Divided by the epicenter of Wenchuan earthquake, the crustal structure of the Longmen Shan fault shows that the high-velocity feature in the northern part and low-velocity feature in the southern part.4) Phase velocity maps of period above 12s showed that with the periods increased, the east side of the junction of the Longmen Shan fault, Xianshuihe faults and Anning fault appears gradually increasing high-velocity anomaly. It indicats the direction changes of the Chuandian block movement should be closely related to the hard crustal blocking from Sichuan Basin.5) Our results from phase velocity imaging and the S wave velocity inversion of the crust and upper mantle structure fit well with each other. Meanwhile, they are also basically similar with the results from receiver function’s inversion. The Rayleigh wave phase velocity dispersion from ambient seismic noise can provide important and valuable constraints for exploring the crustal structure beneath the Western Sichuan Array, especially for receiver function’s inversion.6) The 3-D S-wave velocity structure of the study area shows that the middle and lower crust of the Chuandian block have a large area of S-wave low-speed zone distribution, which can support the Channel flow conjecture in this area. But the S-wave velocity structure of crust in the Songpan-Garze block is relatively complex. Although a large area of S-wave low-speed zone distribution also exists in the middle and lower crust of the Songpan-Garze block, whether the Channel flow model can explain the crustal deformation or not still needs further research and higher resolution exploration. However, the middle and lower crust of the Songpan-Garze block is evidently weaker than the corresponding depth of the Sichuan Basin, which means the eastward escaped Songpan-Garze block is resisted by the hard middle and lower crust of the Sichuan Basin. Our results have provided the evidenced of the above deductions by a high-resolution 3-D crustal velocity structure and laid an important foundation for further study of the dynamics simulations. |
PDF Full Text Request