Seismic waveform inversion using decomposed one-way wavefields

Conventional seismic full waveform inversion (FWI) estimates a velocity model by matching the two-way predicted data with the observed data. The non-linearity of the problem and the complexity of the Earth's subsurface, which results in complex wave propagation may lead to unsatisfactory inversion results. Moreover, having inaccurate overburden models, i.e water velocity model in the ocean bottom acquisition setting, may lead to erroneous velocity models. In order to obtain better velocity models, we simplify the FWI problem by decomposing the total two-way observed and predicted wavefields into one-way wavefields at the receiver locations using acoustic wavefield decomposition. We then propose to fit the less-complex one-way wavefield to obtain a velocity model that contains valuable information about the overburden. We use this inverted velocity model as a better initial model for the inversion using the other more-complex one-way wavefield. We demonstrate our proposed algorithm on acoustic non-inverse crime synthetic data produced from the Marmousi2 model. The proposed method provides improved inversion results compared with conventional FWI due to properly accounting for separate model updates from the up- and down-going wavefields.

https://slim.gatech.edu/Publications/Public/Conferences/SEG/2018/alfaraj...