Foothills seismic imaging for deep exploration in Junggar Basin — A case study

Abstract

Seismic imaging for deep exploration in foothills areas is currently one of the most active endeavors on land. In response to intricate challenges posed by subsurface and near-surface geology, a robust imaging workflow for land seismic data has been successfully developed. First, turning-ray tomography constrained by uphole velocity or micrologging velocity is applied to estimate a reliable near-surface velocity model by using range-limited and full-offset first arrivals. Second, a robust and seamless initial depth velocity model is constructed for prestack depth migration (PSDM) velocity model building. This model integrates the near-surface model with a legacy subsurface model. Third, tilted transverse isotropy joint turning-ray and reflection tomography, constrained by well logs and geologic interpretation, is performed to iteratively update the depth velocity model. This approach achieves improved well ties and spatial positioning of depth images from shallow to deep horizons. In the joint tomography process, 5D interpolation is employed to reduce the trace interval of common-image gathers and increase the trace numbers or common-depth point fold. As a result, the reflection tomography performs better, especially at the gap zone between the near-surface and subsurface structures. Application of the proposed methodologies and workflow to the MiQuan foothills 3D seismic data (acquired from the south rim of Junggar Basin in northwestern China) has had significant success in imaging complex subsurface structures. The final PSDM velocity model aligns with geologic expectations, and the final depth migration offers improved delineation of deep reservoirs, revealing meaningful faulting structures within a regional anticline in MiQuan. This holds significance for seismic exploration, not only in MiQuan within Junggar Basin, but also in other regions worldwide that share similar complexities in near-surface and subsurface structures.