Shear Velocity Inversion Using Multimodal Dispersion Curves From Ambient Seismic Noise Data of USArray Transportable Array

We utilize an array method called the frequency-Bessel transformation method to extract the multimodal dispersion curves of Rayleigh waves from ambient seismic noise data recorded by the USArray Transportable Array. We observe as many as five overtones' dispersion curves of Rayleigh waves in the Midwestern United States, and four and three overtones' dispersion curves, respectively, in the U.S. Great Plain area and Northeastern United States. We employ a quasi-Newton method to invert the multimodal dispersion curves for the shear velocity. We find that the sensitivity of overtones to deeper structures is higher than that of fundamental mode in the same frequency range. The utilization of overtones significantly improved the non-uniqueness and convergence of the inversions, which make the final inversion results robust and reliable. Our inversion results for S wave velocity (Vs) model in the studied areas exhibit some differences compared with Shen and Ritzwoller's model (2016, https://doi.org/10.1002/2016JB012887). The Vs falls abruptly in the lower crust (21-33 km) in the U.S. Great Plain area and Northeastern U.S. areas. A high-velocity zone is observable in depth of 50-60 km in the U.S. Great Plain area and Midwestern United States, and all three models show larger Vs values below 50 km than those of Shen and Ritzwoller (2016).