Answers are grounded in this paper and cite their sources.
Seismic acoustic impedance inversion is an important method for reservoir prediction in oil and gas exploration and development. Since seismic data are band-limited, noisy, and incomplete, seismic inversion is an ill-posed problem that requires regularization. However, commonly used L1-norm regularization tends to compress reflection amplitudes and thereby reduce amplitude fidelity, whereas conventional horizontal multi-trace constraints often introduce non-geological staircase artefacts in dipping strata. Furthermore, incorporating spatial constraints transforms the inversion into a large-scale coupled problem, substantially increasing computational and memory costs. To address these issues, we propose a fast multi-trace seismic inversion method based on the Smoothly Clipped Absolute Deviation (SCAD) penalty. In this method, the SCAD penalty enables nearly unbiased estimation of large coefficients, thereby effectively preserving reflection amplitudes. To ensure lateral continuity, a structure-oriented difference operator guided by a dip field is employed to smooth the inversion results along local structural directions. Moreover, we adopt a Red-Black alternating iteration strategy that reorders the coupled spatial system into conditionally independent single-trace subproblems, thereby greatly improving computational efficiency. Experiments on both synthetic and field data demonstrate that horizontal discontinuities and staircase artefacts in the inversion results are significantly reduced. Structural features in faulted and folded areas are clearly resolved, both weak and strong reflection amplitudes are accurately recovered, and computational efficiency is substantially improved while memory requirements are reduced.
No takes yet. Share an insight, caveat, or question.
Liu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: