Abstract

This study demonstrates high-resolution, geologically consistent depth interval velocity model building using full waveform inversion (FWI) applied to narrow-azimuth towed-streamer seismic data from a deep-water block in the Mahanadi Basin, India. We have applied an advanced acoustics anisotropic FWI approach combining adaptive waveform inversion, reflection waveform inversion and least square FWI. Utilization of different type of inversion objective functions prevent the velocity to fall in local-minima and help to achieve viable solution. The final velocity model yields flattened common image gathers, aligns closely with well log velocity trends and demonstrates high trace-totrace correlation between modelled and observed seismic data. It effectively captures the sedimentary architecture and distinctly delineates the Mio-Pliocene Channel-Levee Complexes (CLCs) which is a proven hydrocarbon-bearing system. This integrated approach delivers a highresolution velocity model that significantly enhances seismic imaging and supports more accurate hydrocarbon prospect evaluation.