Rock Physics-Driven High-Resolution Seismic Inversion for Thin-Layer Prediction

This study addresses the challenges posed by poor data resolution, strong heterogeneity, and complex reservoir stacking patterns in intermediate-to-deep formations. We begin by employing forward modeling to analyze the seismic response characteristics of various sand-body architectures. To overcome the limitations of seismic resolution at depth, we propose an innovative workflow that integrates seismic data optimization with rock physics-driven, high-resolution inversion. Specifically, we explore the use of a sparse frequency extension technique through compressed sensing (CS) algorithms, validating its effectiveness with synthetic model tests. This approach enhances the vertical resolution of seismic data, providing a robust data foundation for subsequent inversion. In our rock physics analysis, we construct a comprehensive saturated rock model based on established theories to predict logging curves. Sensitive elastic parameters, useful for lithology and fluid discrimination, are identified through cross-plot analysis. Guided by a geological sedimentary framework, we then perform high-resolution seismic inversion to predict rock properties and delineate high-quality reservoirs. The inversion results clarify sand-body architecture and characterize the distribution of these reservoirs. Ultimately, this work supports field development strategies and optimizes well placement, leading to successful production.