S018-0002
High-Resolution Lithospheric Structures Beneath Central California Revealed by Two-Dimensional Linear-Array Ambient-Noise Adjoint Tomography

Wednesday, 9 December 2020
Poster
Bin He1, Tianshi Liu2, Kai Wang3, Yingjie Yang3, Ting Lei2, Liang Ding1,4 and Qinya Liu1,2, (1)University of Toronto, Department of Physics, Toronto, ON, Canada, (2)University of Toronto, Department of Earth Sciences, Toronto, ON, Canada, (3)Macquarie University, ARC Centre of Excellence for Core to Crust Fluid Systems and GEMOC, Sydney, NSW, Australia, (4)Hohai University, School of Earth Sciences and Engineering, Nanjing, China
Abstract:
The Isabella anomaly, a high velocity upper mantle body in the upper mantle of southern Great Valley in Central California, is one of most well-known seismic velocity anomalies in North America. However, the origin of the Isabella anomaly is still under hot debate. The debate is mostly centred on two hypotheses. One considers the Isabella anomaly as the foundered mafic lower crust and mantle lithosphere of the southern Sierra Nevada batholith, whereas the other attributes it to a fossil slab connected to the Monterey microplate.

One of the purposes of the Central California Seismic Experiment (CCSE) is to distinguish these two hypotheses. Previous tomographic studies (Jiang et al 2018) highlighted a low-velocity anomaly in the crust and the high-velocity Isabella anomaly in the upper mantle. In this study, we apply the two-dimensional linear-array ambient-noise adjoint tomography method (Chao et al. 2017) to the quasi-linear CCSE array across the Isabella anomaly and perform a two-dimensional full-waveform inversions.

We process continuous seismic data recorded in 2013-2015 from the CCSE array following the standard procedures of Bensen et al. (2007) to obtain cross-correlation functions between all station pairs based on the NoisePy package (Jiang et al., 2020). Then, we use the stacked noise cross-correlation functions to extract the empirical Green’s functions (EGF) by following a 3D to 2D transformation to accommodate the differences between 3D and 2D Green functions. The multi-scale frequency-dependent travel-time misfits between EGFs and simulated 2D Green’s function are minimized at the four period bands of 20-50 s, 15-25s, 10-20s, and 6-15s based on the adjoint method (Tape et al. 2007) to update the S-velocity model iteratively. We finally obtain a high-resolution S-velocity model from the surface down to 60 km along the two-dimensional profile. The recovered model shows detailed crustal structures including low-velocity anomalies in the shallow crust and the high-velocity Isabella anomaly consistent with previous results and provides the basis for a more detailed geological interpretation.