Since 2017, with the support of the China Seismic Experimental Site (CSES) of China Earthquake Administration (CEA), we have started to build the multi-scale community velocity model (CVM) in the Sichuan-Yunnan region, Southwest China. We are aiming to construct CVMs of four different scales in Sichuan and Yunnan, namely the regional block scale, fault zone scale, fault scale, and urban scale, with its lateral resolution about 10-50km, 1-5km, 0.1-0.5km and 1km, respectively. We first used the point-wise joint inversion of ambient noise surface wave dispersion, teleseismic Rayleigh wave ZH amplitude ratio and P-wave receiver function data from 114 permanent stations to a 3-D shear wave velocity model of the crust and uppermost mantle in the Sichuan-Yunnan area (Yang, Yao et al., 2020)  (https://github.com/yanyangg/SW_China_Vs_model). Then based on this initial 3-D reference model, we further constructed the first version of the regional CVM in Sichuan & Yunnan, that is, SWChinaCVM-1.0 (https://github.com/liuyingustc/SWChinaCVM), from joint inversion of dense body wave and surface wave travel times. The lateral resolution of this model can reach 50km in most regions. 

 With the dense arrays in the fault zone regions and basins, we are now starting to build high resolution models of other scales, in order to form the real multi-scale CVM in Sichuan and Yunnan.

A novel method is implemented to invert Rayleigh and Love wave dispersion curves of all paths jointly for 3-D shear wave velocity and radial anisotropy parameter simultaneously without intermediate steps. This new approach is based on our previous method of direct inversion of 3-D isotropic Vs model (DSurfTomo, Fang, Yao, et al., 2015). Conventional approach requires the computation of division Vsh over Vsv after inversion, which is sometimes unstable. The new parameterization in our approve allows for direct inversion of the 3-D radial anisotropy parameter (Vsh/Vsv). Therefore, spatial smoothing can be directly applied to  (Vsh/Vsv), which ensures a more stable inversion of (Vsh/Vsv). We use the method to derive high-precision crustal shear wave velocity and radial anisotropy models around the eastern Himalayan syntaxis using ambient noise dispersion data (5-40s). Results show the crust can be divided into several subregions with different rigidity and preferred mineral alignment orientation depth-dependently. 

The DRadiSurfTomo code, surface wave dispersion data, and the resulting model files in our paper (Hu, Yao, Huang, 2020, JGR) can be downloaded online (https://doi.org/10.5281/zenodo.3592528). 

 Azimuthal anisotropy retrieved from surface waves is important for constraining depthvarying deformation patterns in the crust and upper mantle. We present a direct inversion technique for the three‐dimensional shear wave speed azimuthal anisotropy based on mixed‐path surface wave traveltime data. This new method includes two steps: (1) inversion for the 3‐D isotropic Vsv model directly from Rayleigh wave traveltimes and (2) joint inversion for both 3‐D Vsv azimuthal anisotropy and additional 3‐D isotropic Vsv perturbation. The joint inversion can significantly mitigate the trade‐off between strong heterogeneity and anisotropy. With frequency‐dependent ray tracing based on 2‐D isotropic phase speed maps, the new method takes into account the ray‐bending effect on surface wave propagation. (Reference：Liu, C., Yao, H.*, Yang, H.‐Y.*, et al, 2019, JGR, https://doi.org/10.1029/2018JB016920)

DAzimSurfTomo package Download: https://github.com/Chuanming-Liu/DAzimSurfTomo

1. We have developed a new joint inversion methods for better imaging 3-D Vp and Vs crust structure using direct inversion of surface wave travel times (based on frequency-dependent ray tracing) and body wave travel times. (Fang, Zhang, Yao, et al., 2016). With a new parameterization approach, we have proposed a new method of directly invert for more reliable 3-D Vp/Vs models using both body and surface wave travel times (Fang, Yao , Zhang et al., 2019, GJI).

2. We have developed a joint inversion method (using the Neighborhood Algorithm) that combines dispersion data and Rayleigh wave ZH ratios for the inversion of crustal Vs and Vp/Vs.  (Yuan, Yao, Qin, 2016, Chin. J. Geophys.). In addition, we have developed an iterative linearized inversion approach for 1-D crustal Vs structure as well as interfaces using surface wave dispersion, Rayleigh wave ZH ratio and P wave receiver functions. (Zhang P. & Yao, 2017, Earthquake Science)

3. We have developed a linear array ambient noise adjoint tomography strategy for imaging high-resolution crustal Vs structure. (Zhang C., Yao, et al., 2018, JGR)

We proposed a method to invert surface wave dispersion (traveltime) data directly for 3-D variations of shear wave speed, that is, without the intermediate stepof phase or group velocity maps, using frequency-dependent ray tracing (Fang, Yao*, et al., 2015, GJI). A fast marching method is used to compute, ateach period, surface wave travel times and ray paths between sources and receivers. This avoidsthe assumption of great-circle propagation that is used in most surface wave tomographicstudies, but which is not appropriate in complex media. We represent the 3-D shear wave speedmodel by means of 1-D profiles beneath grid points and the wave speed model is estimated with an iterativelyreweighted least squares algorithm, and upon iteration the surface wave ray paths and thedata sensitivity matrix are updated using the newly obtained wave speed model. This methodhas been applied to Taipei Basin in Taiwan (Fang, Yao, et al., 2015, GJI), Hefei urban area (Li, Yao et al., 2016, SRL), fault zone regions (Gu et al., 2019, PAGEOPH), a shale gasproduction field in SW China (Liu et al., 2018, JAG), Taiwan Strait (Zhang, Yao et al., 2018 JGR), and many other regions. 

(DSurfTomo Package Download: https://github.com/HongjianFang/DSurfTomo)

1. We developed compressive sensing(CS) techniques in the frequency domain (Yao et al., 2011, GRL; Yin & Yao, 2016, GJI) and wavelet domain (Fang, Yao, Zhang, 2018, GJI) to image frequency-dependent seismic radiation and ruptureprocesses of great earthquakes. For the four largestmegathrust earthquakes in the past 10 years. Our resultsreveal generally low-frequency radiation closer to the trench atshallower depths and high-frequency radiation farther from the trench atgreater depths (Yao et al., 2011, GRL; Yao et al., 2013, PNAS; Yin et al., 2016, GRL, ...). Together with coseismic slip models andearly aftershock locations, our results suggest depth-varyingfrictional properties at the subducting plate interfaces.

2. We developed an iterative backprojection methodwith subevent signal stripping to determine the distribution of subevents(large energy bursts) during the earthquake rupture.We also relocate thesubevents initially determined by iterative backprojection using the traveltimeshifts from subevent waveform cross-correlation, which provides moreaccurate subevent locations and source times. Subevents location results for the 2011 Mw 9.0 Tohoku-Oki Earthquake (Yao et al., 2012, GJI, PDF download)

We have developed the joint ambient noise and earthquake surface wave tomographic method (Yao et al., 2006; 2008; GJI) for better constraining the crust and upper mantle structure using seismic array data. In this method, we combine shorter period phase velocity dispersion curves from ambient noise cross-correlation functions and longer period dispersion data from earthquake surface wave two-station analysis. Then 2D phase velocity maps are constructed and 3-D Vs model can be obtained by point-wise inversion using the Neighborhood Algorithm (Yao et al., 2008, GJI). We have also investigated radial and azimuthal anisotropy of the lithospheric structure from surface waves (e.g., Yao et al., 2010, JGR; Huang, Yao, van der Hilst, 2010, GRL), which provides essential information for understanding the deformation patterns in the crust and upper mantle. The current study regions include the Tibetan Plateau, Southwest China and Vietnam (Qiao, Yao et al., 2018, Tectonics), North China Craton, Southeastern mainland China and Taiwan (Zhang, Yao et al., 2018, JGR), the equatorial eastern Pacific Rise, etc.