Efficient and Accurate PS Model Building with Joint PP/PS Tomography

Abstract

Historically, velocity model building for PS imaging has been done sequentially, starting with estimation of P-velocity and anisotropic parameters using PP data alone, followed by updates of S-velocity to match PS and PP depth images. A better option is to include both PP and PS data in model building from the start and find the model that satisfies both data types, producing flat events on gathers and PP and PS images with consistent depths. We implemented a new method for joint PP/PS tomography in which moveout picks on PP and PS depth gathers are combined with floating event constraints between PP and PS images to simultaneously solve for Vp, Vs, and anisotropic parameters (Mathewson et al, 2013). Our method is an improvement on previous approaches that require separate treatment of PP and PS data in the early stages of model building, allowing the use of combined PP and PS data throughout the model-building process. Benefits of the joint PP/PS tomography include improved efficiency, reduced uncertainty, and better resolution of anisotropy.