Speaker: Jorge Guizar

Date: Thursday 12th of March 2026, 1:15pm

Abstract:

Seismic data processing plays a fundamental role in determining the accuracy, resolution, and reliability of results obtained from full-waveform inversion (FWI), seismic imaging, and seismic interpretation. High-fidelity processing workflows—including noise attenuation, multiple suppression, deghosting, deconvolution, amplitude and phase corrections, and velocity model building—directly influence the quality of the wavefield used in advanced inversion and imaging techniques. Furthermore, the careful design of the processing sequence impacts the final seismic image by shaping its frequency content, reflector continuity, and structural reliability. Recent advances in AI and machine-learning approaches have demonstrated strong potential for seismic interpretation—supporting applications such as automatic fault extraction, salt boundary interpretation, gas-chimney detection, and stratigraphic classification. Nonetheless, the performance and reliability of these models remain fundamentally dependent on the quality of the input seismic data, and therefore on the robustness of the underlying processing workflow.