Common-midpoint two-station analysis of estimating phase velocity using high-frequency ambient noise

Abstract

Two-station analysis is a frequently-used technique in seismology to determine interstation phase velocities of surface waves but it usually suffers from several drawbacks arising from 2π ambiguity and a narrow frequency band. This paper introduces a common-midpoint two-station analysis (CMP-TS) method to suppress the spurious energy due to 2π ambiguity and broaden the frequency band for high-frequency (>1 Hz) passive surface wave imaging. To demonstrate the advantages of the proposed method in estimating phase velocity, the authors compared results obtained by the CMP-TS method and the Multichannel Analysis of Passive Surface waves (MAPS) using synthetic data with homogeneous and inline noise-source distributions. For homogeneous noise-source distribution, phase-velocity dispersion curves obtained by both methods can match with theoretical curves, but the CMP-TS method could enhance the resolution of images of dispersion energy by about 50% compared with the MAPS method. As for inline noise-source distribution, phase velocities generated with the MAPS method can match with theoretical ones while phase velocities measured by the proposed method are underestimated. By modifying the source phase ambiguity term in the proposed method, the authors corrected the systematic measurement bias of phase velocities and concluded that the CMP-TS method is superior to the MAPS method in accuracy and resolution. Results from two field examples further demonstrated the capability of the CMP-TS method in estimating phase velocity with higher image resolution than the MAPS method. Finally, the results showed that it is reasonable for the proposed method to pick phase velocities using the one-wavelength criterion in near-surface applications.