Impact of Density Information on Rayleigh Surface Wave Inversion Results

Julian Ivanov,Georgios Tsoflias,Richard D. Miller,Shelby Peterie,Sarah Morton,Jianghai Xia

DOI: https://doi.org/10.1016/j.jappgeo.2016.09.011

IF: 1.845

2016-01-01

Journal of Applied Geophysics

Abstract:We assessed the impact of density on the estimation of inverted shear-wave velocity (Vs) using the multi-channel analysis of surface waves (MASW) method. We considered the forward modeling theory, evaluated model sensitivity, and tested the effect of density information on the inversion of seismic data acquired in the Arctic. Theoretical review, numerical modeling and inversion of modeled and real data indicated that the density ratios between layers, not the actual density values, impact the determination of surface-wave phase velocities. Application on real data compared surface-wave inversion results using: a) constant density, the most common approach in practice, b) indirect density estimates derived from refraction compressional-wave velocity observations, and c) from direct density measurements in a borehole. The use of indirect density estimates reduced the final shear-wave velocity (Vs) results typically by 6–7% and the use of densities from a borehole reduced the final Vs estimates by 10–11% compared to those from assumed constant density. In addition to the improved absolute Vs accuracy, the resulting overall Vs changes were unevenly distributed laterally when viewed on a 2-D section leading to an overall Vs model structure that was more representative of the subsurface environment. It was observed that the use of constant density instead of increasing density with depth not only can lead to Vs overestimation but it can also create inaccurate model structures, such as a low-velocity layer. Thus, optimal Vs estimations can be best achieved using field estimates of subsurface density ratios.