Velocity variation coefficient-based angle-dependent gradient conditioning scheme: a new strategy for an enhanced full waveform inversion

Abstract The development of an accurate velocity model is the significant target in Full Waveform Inversion (FWI) process where data fitting carried out based on ill-posed technique. In FWI technique optimization plays a crucial role through which objective function minimizes, related to misfit between observed and modelled data. However, influence external factors such as errors (local minima) presence noise are involved success this processing artefacts that arise during gradient computation also affect This study presents strategy mitigate these local minima other variation coefficient angle-dependent conditioning approach has been proposed. It auto-controlled primary mechanism updates from large angle scale smaller when iteration begins. At each iteration, it preserves previous result whereby does not scatter or overlap with one. covers all angles smoothly helps minimizing providing high-resolution model. proposed demonstrated by implementing Marmousi model, proves method provides much-improved inversion attained reasonable iterations. represents suitable procedure for less sensitive identified negligible time consumption. Furthermore, reduce cycle skips improve convergence any complex scenario.