A Modeling Method for Full Waveform Inversion of VSP Data Based on Surface Seismic Tectonic Constraints

Abstract

The invention provides a VSP data full waveform inversion modeling method based on ground seismic structure constraints. The method comprises the steps that S1 VSP data full waveform inversion is carried out based on ground seismic structure constraints to acquire an intermediate velocity model M1; and S2 for the intermediate velocity model M1, VSP data full waveform inversion without ground seismic structure constraints is carried out to acquire a final speed model M. A whole waveform inversion operator based on the fluctuation theory is used. Compared with a traditional VSP data modeling method, the method has higher modeling precision and can more clearly describe the structural details of a describing model. The introduction of a ground seismic structure constraint filter operator overcomes the problem of modeling noise caused by incomplete data of traditional full waveform inversion and the problem of local inversion extremum caused by low frequency loss. The problem of limited angle coverage of VSP data is solved to a certain extent. The modeling angle range is improved.

Description

technical field [0001] The present disclosure relates to petroleum seismic exploration velocity modeling technology, in particular to a high-precision VSP (Vertical Seismic Profilling, vertical seismic profile) data full waveform inversion modeling method based on ground seismic structure constraints. Background technique [0002] With the increasing refinement and complexity of oil and gas exploration, high-precision velocity modeling plays an increasingly important role in geophysical processing. Seismic wave velocity not only determines the quality of migration imaging, but is also closely related to rock properties, which can reflect rock types and fluid-rich conditions, and ultimately affect the credibility of seismic interpretation results. Therefore, seismic wave velocity is a very important stratigraphic physical property. parameter. [0003] In recent years, the full waveform inversion method based on seismic data amplitude, phase, and waveform matching is based on...

AI Technical Summary

Problems solved by technology

Due to the limited angular coverage of VSP data, and the ray tomographic velocity modeling schemes are often based on high-frequency approximation assumptions, the tomographic modeling of VSP seismic data can only obtain low-wavenumber background velocities within a limited angular range, but cannot be refined. Characterize the exact position and structural shape of the abnormal velocity body

Although the full waveform inversion can effectively describe the structural details of the model, when the initial model is inaccurate or the data is incomplete, the gradient will generate noise, causing the inversion result to fall into a local minimum problem

Images