Application of the Refraction and Reflection Joint Tomography in the Pre-stack Depth Migration Velocity Modeling Under the Complex Near-Surface Condition

The near-surface situation of Sichuan Basin is very complicated, and the elevation difference is large. Even in the flat area of central Sichuan, the elevation difference reaches hundreds of meters, and the velocity and thickness of the low deceleration zone lateral change rapidly. Complex topography seriously affect the traveltime of seismic waves, which brings great challenges for subsequent seismic migration imaging. How to establish high-precision near-surface velocity model is the key step of ensuring imaging quality. Therefore, a optimization of actual surface velocity model of refraction and reflection joint tomography from surface to subsurface is proposed. Based on the simple model, high-precision first arrival is picked, and the refraction tomography inversion is used to optimize the near-surface velocity model. By using refraction-reflection joint tomography of the variable grid based on sparse matrix, a united tomographic matrix was built, which was solved as a whole, and a high precision velocity model from shallow to deep was obtained. This velocity modeling technology solves the problem that the conventional reflection tomography inversion cannot optimize near-surface velocity model under complex topographic conditions, and highlights the integrity, high efficiency and high precision of velocity modeling. The method have been applied in the actual surface pre-stack depth migration velocity modeling in central Sichuan. The shallow narrow channel features are clear in the migration results, and the boundary is clearly delineated. At the same time, the well-seismic matching error caused by the near-surface velocity model is reduced, and the imaging accuracy is significantly improved.