(1) Full-wave seismic anisotropy tomography of Southern California
We applied a previously developed efficient approach to computing the sensitivity kernels of any seismic observables (such as traveltime and amplitude anomalies) to any model parameters (such as wave speed, attenuation and anisotropy parameters) to the teleseismic waveform records from a dense network of stations in Southern California and imaged the three-dimensional shear-wave azimuthal anisotropy structure. The anisotropic model we obtained reveals structural fabric in relation to surface geologic features such as the Salton Trough, the Transverse Ranges, and the San Andreas Fault. The depth variation of the shear-wave anisotropy does not favor a lithosphere-asthenosphere decoupling. At long wavelengths, the fast directions of anisotropy are aligned with the absolute plate motion inside the Pacific and North American plates (Lin et. al., 2014).
(2) Rayleigh-wave phase velocity tomography for various regions
Working with postdoc Cedric Legendre, IES colleague Bor-Shouh Huang and collaborators in other countries, we collected surface wave records from dense deployment of broadband stations in a number of regions including mainland China, north Vietnam, South Korea, Japan, India, and Anatolia. These waveform records are used to measure the inter-station dispersion curves which are then inverted for regional phase-velocity maps as well as maps of azimuthal anisotropy. These isotropic and anisotropic models are analyzed to help further our understanding of regional deformation tectonics and dynamics (Legendre et al., JGR & JAES, 2014; Legendre et al., EPS, 2015; Legendre et al., GJI, 2015; Legendre et al., Sci. Rep, 2015; Legendre et al., JAES, 2016; Tian et al., CJG, 2017).
(3) Improving the 1D model at the top of the outer core by SmKS differential traveltime inversion
We have also investigated the fine one-dimensional structure at the top of the Earth’s outer core by conducting a Bayesian inversion of the S3KS-SKKS differential traveltimes obtained from the cross-correlation of these two phases. We improved upon existing similar studies by accounting for the finite-frequency effect. The 1D model we obtained for the top of the outer core provides seismological constrains on the globally averaged wave speed profile which is more complicated than existing models such as PREM or AK135. The ensemble average model we obtained from the Bayesian inversion shows that the velocity gradient in the topmost part of the outer core has robust short-wavelength variations, which strongly supports the existence of chemically induced stratification immediately below the core-mantle boundary (Tang et. al., 2015).
(4) Inversion of earthquake source slip distribution in 3D structure using regional records
We have also developed a technique for point- and finite-source inversions of earthquakes in 3D structure using regional seismograms based on a strain Green tensor (SGT) database. To improve efficiency, the SGT database can be pre-calculated for a 3D regional seismic model with realistic topography (such as ETOPO1). The efficiency afforded by the SGT database enables us to carry out near real-time inversions of earthquake sources for seismic hazard assessment and mitigation purposes. We have applied this technique to moderate earthquakes in Taiwan (Hsieh et al., 2014; 2016) and in Yunnan, China (manuscript in preparation).
(5) Development of near real-time system for earthquake source slip distribution inversion
In collaboration with the Central Weather Bureau (CWB), we are in the process of developing a semi-automatic system to determine the source slip distributions of moderate and large earthquakes based on the fast finite-fault (FFF) inversion method of Ji et al. (2002). The purpose of this system is to enhance the earthquake early warning systems to include much more extensive information of earthquake sources than currently available, which is important for more rapid and reliable predictions of strong ground motions and the assessment of earthquake hazards following disastrous earthquakes.