Weighted stacking plus traveltime inversion: a proposed prestack P-S inversion

Maria Donati, and Nicolas W. Martin


This paper outlines an alternative method for obtaining P- and S-wave velocity models and V S /V P ratio, from P-S data. The method interactively minimizes an error function (observed - modeled P-S amplitudes) by using a trial V S /V P value as input. This minimization process consists of a modified least-squares method, which generates constraints on ΔV P /V P , ΔV P /V S , and V S /V P ratios. A P-S traveltime tomography scheme is followed for fitting the observed P-S traveltimes under the assumption that above constraints are honored. This traveltime inversion provides the final P- and S-wave interval velocity fields.

Although, this method could be more expensive than other published methods (Smith and Gidlow,1987; Stewart, 1990; Vestrum and Stewart, 1993; Ferguson, 1996, and Ferguson and Margrave, 1996), it is possible to gain a more robust P-S inversion algorithm combining amplitude (post-stack) and traveltime (pre-stack) informations at once, for getting a more realistic P- and S-wave velocity field with vertical and horizontal gradients, while P-S amplitudes are well represented. Additionally, this new approach does not need, in principle, any shear sonic log information or P-wave and S-wave velocity models in P-S time to get elastic parameters.

View full article as PDF (0.05 Mb)