Depth-Domain Direct Inversion Based on Depth-Variant Wavelet Under Geo-Structure Constraints

Seismic post-stack impedance inversion is the most commonly used reservoir characterization method. Recently, pre-stack depth migration (PSDM) has been widely used to generate depth-domian seismic images. Compared to time-domain imaging, depth-domain imaging has more accurate structural information and imaging quality. For reservoir characterization of depth-domain seismic data, the conventional method is to convert the depth-domain seismic data to the time domain, and then convert the results to the depth domain after time-domain inversion. In this paper, we use a method that combines the S-transform of wave number decomposition with a depth-domain non-stationary convolution model for estimating depth-variant seismic wavelets, thus enabling direct inversion of depth-domain seismic data. And a complete depth-domain reservoir inscription workflow is proposed, including depth-domain initial model building guided by relative geological time (RGT), depth-domain well-seismic matching based on Dynamic Depth Warping (DDW), and depth-domain inversion with structural information constraints. Finally, we show the effect of applying this method to a certain real data, and the results show that the depth-domain inversion has higher accuracy for reservoir characterization than the traditional time-domain inversion.