Time-average velocity estimation through surface-wave analysis: Part 2 — P-wave velocity

Abstract

Surface waves (SWs) in seismic records can be used to extract local dispersion curves (DCs) along a seismic line. These curves can be used to estimate near-surface S-wave velocity models. If the velocity models are used to compute S-wave static corrections, the required information consists of S-wave time-average velocities that define the one-way time for a given datum plan depth. However, given the wider use of P-wave reflection seismic with respect to S-wave surveys, the estimate of P-wave time-average velocity would be more useful. We therefore focus on the possibility of also extracting time-average P-wave velocity models from SW dispersion data. We start from a known 1D S-wave velocity model along the line, with its relevant DC, and we estimate a wavelength/depth relationship for SWs. We found that this relationship is sensitive to Poisson’s ratio, and we develop a simple method for estimating an “apparent” Poisson’s ratio profile, defined as the Poisson’s ratio value that relates the time-average S-wave velocity to the time-average P-wave velocity. Hence, we transform the time-average S-wave velocity models estimated from the DCs into the time-average P-wave velocity models along the seismic line. We tested the method on synthetic and field data and found that it is possible to retrieve time-average P-wave velocity models with uncertainties mostly less than 10% in laterally varying sites and one-way traveltime for P-waves with less than 5 ms uncertainty with respect to P-wave tomography data. To our knowledge, this is the first method for reliable estimation of P-wave velocity from SW data without any a priori information or additional data.