Cascaded and Joint full-waveform Inversion of Rayleigh Wave and Body Waves from High-speed Train Seismic Data

Summary The high-speed train (HST) is a new type of moving combined source and can generate seismic signals with wide frequency bandwidth, strong low frequency and good repeatability. FWI using the HST seismic data is a promising tool to extract accurate subsurface model parameters along the railway. In this paper, we develop Rayleigh wave and body wave cascaded FWI (CFWI) and joint FWI (JFWI) methods for the HST seismic data. A joint misfit function is defined and a weighting factor is used to control the contributions of different waves in JFWI. The Rayleigh wave and body wave inversion is alternately implemented on the large scale and the body wave individual inversion is performed on the small and intermediate scales in CFWI. Numerical results reveal that CFWI has greater accuracy than JFWI, Rayleigh wave and body wave individual inversion with the same computational cost. The CFWI method for the HST seismic data can reconstruct plausible P- and S-wave velocity models below the subgrade.