A Comprehensive Comparison Between the Refraction Microtremor and Seismic Interferometry Method for Phase Velocity Estimation

Document Type


Publication Date



Passive-source seismic-noise based surface-wave methods are now routinely used to investigate the near-surface geology in urban environments. These methods estimate the shear-wave velocity of the near surface, and two methods that use linear recording arrays are seismic interferometry (SI) and refraction microtremor (ReMi). These two methods process noise data differently and thus can yield different estimates of the surface-wave dispersion, the data used to estimate the shear-wave velocity. In this paper we systematically compare these two methods using synthetic data with different noise source distributions. We arrange sensors in a linear survey grid, which is conveniently used in urban investigations (e.g. along roads). We find that both methods fail to correctly determine the low frequency dispersion characteristics when out-line noise sources become stronger than in-line noise sources. We also identify an artifact in the ReMi method and theoretically explain the origin of this artifact. We determine that SI combined with array-based analysis of surface waves is the more accurate method to estimate surface-wave phase velocities because SI separates surface waves propagating in different directions. Finally, we propose a solution to eliminate the ReMi artifact that involves the combination of SI and the τ-p transform, the array processing method that underlies the ReMi method.


Date provided is the online early release publication date.