Abstract
Refraction microtremor (ReMi) and multichannel analysis of surface waves (MASW) are effective approaches to estimate shallow shear-wave velocity (VS) structure often needed to estimate ground motions using recent ground motion prediction relations. Interferometric MASW (IMASW) uses slowness-frequency slant-stack analyses combined with interferometric time-domain dispersion analyses to improve resolution of lower-frequency Rayleigh-wave dispersion to better constrain VS. Cross-correlation interferometry is used to obtain deterministic correlation Green’s function (CGF) IMASW seismograms from ambient-noise and/or active-source wave fields contained in ReMi and/or MASW data. The CGFs are processed using the multiple-filter technique to estimate phase and group dispersion. In the IMASW approach, active seismic sources ensure that the stationary-phase contributions to cross correlations dominate CGF responses. In a single IMASW profile, each geophone represents a virtual source, and the IMASW approach stacks CGF common-offset data from all virtual sources to obtain a single averaged forward- and reverse-record section. CGF time-domain and slowness-frequency phase-slowness estimates are combined with CGF time-domain group slowness estimates for a consistency check on dispersion picks. A multistate Monte Carlo approach is used to estimate mean slowness depth and slowness uncertainties. IMASW is evaluated with passive ReMi data from two sites and active-source IMASW at six sites with independent downhole velocity–depth logs. Comparison of six P-S suspension log–IMASW profile pairs across the Van Norman Complex in northern San Fernando Valley shows that, on average, 30-m-depth shear-wave velocity estimates between the two methods differed by <1%. At two sites where P-S suspension log measurements of VS were made at the IMASW profile midpoint, the IMASWVS depth inversions resolve 3-m thickness VS variations accurately to the bottom of one borehole at 40-m depth and to 100-m depth at a >200-m-deep borehole site.