Crust and upper mantle velocity structure of the eastern Tibetan Plateau and adjacent regions from ambient noise tomography
Fan Wen-Yuan,Chen Yong-Shun,Tang You-Cai,Zhou Shi-Yong,Feng Yong-Ge,Yue Han,Wang Hai-Yang,Jin Ge,Wei Song-Qiao,Wang Yan-Bin,Ge Zeng-Xi,Ning Jie-Yuan
DOI: https://doi.org/10.6038/cjg20150510
2015-01-01
Abstract:The crust structure of the Tibet Plateau is still controversial. Whether there is a partial melting layer in the anomalously thick crust is critical to understand the dynamics and evolution history of the formation and on going process. We address the issue with robustly inverted surface wave and shear wave velocity models from the west China dense array data. Ambient noise tomography was applied to obtain the crust and upper mantle velocity models of the Tibetan Plateau. The method is based on extracting Rayleigh wave phase and group velocity dispersion curve from continuous waveform cross-correlation. Our data comes from the Peking University seismic arrays, INDEPTH IV passive seismic array, and four Chinese Provincial seismic networks. The usable data was collected from 05,2007 to 12,2009. The long time span assures high signal to noise ratio when the cross-correlation is performed. We extract the empirical Green's functions from the cross-correlation and perform a surface wave tomography inversion with the cross-correlation functions. With the phase and group velocity models from the previous step, a 3D shear wave velocity model is obtained. Our surface phase/group velocity models and inverted shear velocity model show strong heterogeneity and complexity over the whole studied region. Surface geological features are accurately depicted by the short period surface waves (6 s to 14 s, both phase and group velocity model). The Qaidam basin and Sichuan basin are well correlated to low velocity anomalies due to the thick sediments over these regions. Mountains, like Qilian, Kunlun, and Longmenshan are marked by high velocity anomalies. We also observe a continuous transition zone in between the Tibet plateau and Sichuan Basin at short period. For Rayleigh wave tomography models longer than 25 s, there are significant differences between Qaidam-Qilian region and the plateau south of Kunlun fault zone. The newly formed plateau area, i.e. Qaidam-Qilian region shares similar velocity structure signatures with the eastern boundary of the Tibetan Plateau and Sichuan basin. Both our phase velocity measurements and inverted shear wave velocity models show that the Tibetan plateau has a large-scale uniform low velocity layer in the middle crust (25 40 km in depth). The velocity at this depth of the region is much slower than the global average velocity of AK135. It is also shows distinct deviation from the eastern boundary of the Tibetan Plateau. The low velocity layer supports the channel flow model that explains the eastward material transportation of Tibet. From the intermediate period velocity maps, we observe a low velocity zone at Qilian Mountain region and it might be related to the elevated temperature in the middle and lower crust of this region. The variation of the temperature at this depth might correlate to the underplating of magma from the local mantle upwelling. These observations suggest that the on-going crustal thickening at Qilian Mountain area is possiblly due to the underplating process. Our detailed structure of Tibetan Plateau crust is critical to understand the on-going plateau formation process. The Rayleigh wave group/phase velocity models and 3D shear wave velocity model indicate there is a large-scale uniform low velocity within the plateau. The low velocity layer could explain the eastward material transportation of Tibet. The newly formed Qaidam-Qilian region, as a new plateau shows distinct velocity signatures in our velocity models. And the low velocity zone below Qilian Mountain is possibly an indication of the ongoing magma underplating process in that region.