Estimating urban subsurface models using metro seismic signals

Summary

Traffic-induced seismic waves provide valuable signals for characterizing the subsurface structures. The operation of metro systems generates powerful, broadband, and repetitive seismic signals that propagate in the urban subsurface. This repeatability mainly arises from scheduled metro operations on fixed tunnel routes with similar metro wagons, speed, and weights. Unlike other traffic signals that are generated by vehicles running on the free surface, the metro operates in subsurface tunnels. The use of metro-induced signals for the estimation of the subsurface model, however, remains unexplored. This paper proposes a metro seismic interferometry workflow for urban subsurface characterization. We analyse the characteristics of metro signals and derive a formula for the stationary-phase zone (SPZ) analysis of metro sources. Empirical Green’s functions (EGFs) are estimated by selecting segments of metro signals when the metro is in the SPZ. We performed a field study in Wuhan, China, by deploying a seismic survey line approximately 70° to a metro line. The EGFs reconstructed from multi-component metro seismic signals show distinct Rayleigh and Love waves. The signal-to-noise ratio (SNR) of the EGFs is further improved by stacking multiple metro events. The SNR and the accuracy of the dispersion curves estimated from the EGFs with SPZ analysis are much higher than those obtained without SPZ analysis, demonstrating the validity of our proposed method. The comparison with dispersion curves obtained from road traffic signals shows the reliability of the proposed method. In addition, the EGFs also contain abundant guided P waves with high SNR, enabling the reconstruction of the P-wave velocity model. This work shows that utilizing metro seismic waves provide a powerful, efficient, and cost-effective approach for the estimation and monitoring of urban subsurface models.