Advanced Regularization Strategies and Applications in Double Complex Surveys of Qaidam Basin

Abstract

Abstract The processing of land surface-seismic survey data in the Qaidam Basin has for many decades been recognized as facing the dual challenges of both surface and subsurface complexity. To obtain improved seismic imaging of the complex subsurface structures, near-surface modelling constrained by near-surface survey information and a series of multi domain noise suppression algorithms targeting specific noise types have been applied achieving good results. However, in response to the ongoing issues of low signal-to-noise ratio and high uncertainty in imaging caused by complex surface and subsurface structures, we consider it important to evaluate approaches to data regularization and interpolation. In this paper we compare and discuss the effectiveness of several different strategies for seismic data regularization and interpolation based on matching pursuit Fourier interpolation (MPFI) and time domain radon interpolation (TDRI). Furthermore, we conclude that in areas such as the Qaidam basin which exhibit both surface and subsurface complexity, the radial domain 5D MPFI can effectively improve signal-to-noise ratio, enhance the continuity of subsurface reflection events, and provide regular sampling at each azimuth direction, which is beneficial for subsequent structural interpretation and azimuth anisotropy analysis.