A Prestack Linear Inversion Method Based on Depth Domain Seismic Records

Abstract

The present invention relates to a pre-stack linear inversion method based on depth-domain seismic records, including: Step 101: perform depth-domain well seismic calibration; Combine the interpretation results, and then interpolate the depth domain logging data, and use the sedimentary model as collaborative data to participate in the interpolation, and finally obtain the initial low-frequency elastic parameter data volume of the entire work area; Step 103: first time-domain seismic records of the entire work area Domain seismic wavelet extraction, and then based on the corresponding relationship between the point spread function and the time domain seismic wavelet, time is further converted into depth to obtain the point spread function of the entire work area; Step 104: use the obtained point spread function to construct the corresponding unsteady state Convolution matrix form; Step 105: Combine the unsteady convolution wavelet matrix with the linear approximation formula of the reflection coefficient, and use the Bayesian inference linear inversion method to invert the elastic parameters of the entire work area.

Description

technical field [0001] The invention relates to a pre-stack linear inversion method, in particular to a method for directly performing pre-stack linear inversion by using depth domain seismic records. Background technique [0002] Seismic records are often in the time domain, but it is theoretically insufficient to explain geological problems only from time-domain seismic records and does not directly provide explanations for related geological problems, mainly because: 1) From a mathematical point of view, time-domain seismic records There is inconsistency in the coordinate scale between the logging data and the depth domain, so the relationship between time and depth needs to be determined through well seismic calibration. In well seismic calibration, strong human-computer interaction, such as translation or stretching, greatly enhances the uncertainty of time-depth relationship. 2) The seismic attributes obtained from time-domain seismic records are often also in the tim...

AI Technical Summary

Problems solved by technology

[0004] In summary, the current research on depth domain inversion methods mainly has the following problems: ① Theoretical research is not sufficient enough to combine depth domain imaging and inversion organically, and depth domain inversion in the true sense has not been realized; ② ignore wavelet Inversion with long or sliding time windows will reduce the accuracy of the inversion results; ③The inversion method based on the transformation domain has unclear physical meaning of the transformation domain, and it is unavoidable to avoid Error accumulation; ④ The inversion method of time-depth conversion based on depth migration has low computational efficiency, and the conversion process has precision loss; Network structure design, efficiency and other issues, it is difficult to apply it to actual production applications

Images