Pre-Stack Depth Migration Velocity Modeling and Velocity Update Techniques for Shallow Water Marine Data

Abstract High-precision depth migration imaging has been a hot topic in petroleum seismic exploration research in recent years. To obtain accurate underground imaging, establishing an accurate velocity field is crucial. This paper addresses the characteristics of shallow water OBN (Ocean Bottom Node) data and designs a sequence of depth migration velocity modeling and update techniques tailored for such data. Furthermore, it demonstrates favorable application results in practical projects. Due to significant lateral velocity variations near the seabed in shallow water OBN data, the velocity modeling near the seabed leverages the velocity model obtained from first-arrival tomography inversion. During the velocity modeling process for the middle and deep layers, VSP (Vertical Seismic Profile) logging data provided by the contractor is first collected. Then, incorporating geological layer information provided by the interpretation team, an initial model is established using structure-guided logging curve-constrained inversion techniques. After establishing the initial model, FWI (Full Waveform Inversion) technology is employed to update the velocity field above 3000 meters, followed by multi-azimuth tomographic inversion to further update the velocity model. Utilizing the velocity field obtained from first-arrival tomography inversion for velocity modeling near the seabed effectively restores the true velocity field near the seabed, enhancing the accuracy of shallow water depth migration imaging, while ensuring the accuracy of middle and deep layer velocity updates. VSP logging information facilitates a more realistic recovery of the velocity field for middle and deep layers, while using geological layer positions to constrain the velocity model ensures that the initial velocity model aligns better with geological structural trends. FWI technology maximizes the advantages of stable oceanic data wavelets, rich low-frequency information, and large offset ranges, thereby improving the accuracy of the velocity field. Full azimuthal grid tomography fully utilizes the advantages of full-azimuth OBN data acquisition, effectively addressing the anisotropy issues of subsurface velocity fields. The velocity field establishment and update steps described above result in significantly improved quality of depth migration profiles compared to vintage results. This paper fully considers the rapid changes in velocity near the seabed in shallow water OBN data, as well as the stability of original data wavelets, richness of low-frequency information, and large offset ranges, along with the full-azimuth data acquisition characteristics. By making full use of VSP logging data provided by the contractor and geological layer information provided by the interpretation team, and adopting targeted processing techniques, this technique sequence provides valuable reference for similar shallow water OBN data depth migration processing.