Estimation of shear-velocity profiles by imaging of surface waves

Shear velocity profiles are imaged by wavefield transformations of Love and Rayleigh wave data. The processing sequence starts with slant stacking a common-shot profile to separate phase velocities followed by a 1-D Fourier transform over time-intercept to image dispersion curves in slowness-frequency space. Then, iterative imaging maps all the data from the slowness-frequency domain to the slowness-depth domain. The principle involved in the latter is similar to time-domain p-τ velocity estimation from body waves. Each frequency penetrates to a different depth, which is specified through computation of the dispersion curve for an estimated earth model. Differences between the input model and the output image are minimized by changing the model through convergent iterations. The method is illustrated with synthetic data and applied to two real dispersed-wave data sets from a crustal array experiment performed in Southwestern Oklahoma in 1985 by The University of Texas at El Paso and The University of Texas at Dallas.