Wave-Equation Migration Velocity Analysis in the Surface Common Offset Domain: Application to Viking Graben Field Data

Abstract

The quality of subsurface seismic images depends greatly on the accuracy of the velocity model used for imaging, but the highly nonlinear relationship between recorded data and the velocity model makes estimating accurate velocities a nontrivial task. Two methods, full-waveform inversion (FWI) and wave-equation migration velocity analysis (WEMVA), are commonly used to estimate velocities. These methods, both iterative and wave-equation based, estimate velocities by minimizing an objective function to achieve a desired level of accuracy. Full waveform inversion (FWI) can fail in complex environments when the initial model is an inaccurate representation of actual subsurface velocities. Methods based on wave-equation migration velocity analysis (WEMVA), however, use pre-stack image focusing in order to converge to a more global solution regardless of the starting model. We develop WEMVA in the surface common-offset domain by parsing data into separate bins containing a limited range of offset and processing each bin independently. We circumvent cycle-skipping, which can lead to an erroneous result, by using a sufficient number of bins to limit the amount of residual moveout between neighboring traces. We present the results of WEMVA in the common-offset domain on a 2D marine field data set recorded offshore Norway in the Viking Graben area.