A depth domain adaptive seismic wavelets extraction method

Pre-stack depth seismic image becomes more and more widely used in industry, which also leads to increasingly needs of its inversion results for reservoir characterization. Conventional method usually transfers the depth domain dataset in to time domain before inversion, which could result in artifacts because of the inaccurate velocity or insufficient TD conversion algorithm. However, direct inversion of the depth domain seismic image could suffers from strong dispersion effect, which makes it hard to use a stationary wavelet. In this paper, we propose a depth adaptive wavelets extraction method to accommodate the frequency variation for inversion of the depth image. The depth adaptive wavelets are extracted according a dominant amplitude-frequency trend, computed from depth-wavenumber decomposition results.