A hybrid optimization method for full-waveform inversion

Seismic full waveform inversion (FWI) is one of the most attractive seismic imaging tools for estimating subsurface fluid and rock properties by taking most, if not all types, of seismic waves into account. By iteratively fitting both the kinematic and the dynamic characteristics of the observed seismic data, FWI is able to build high-resolution subsurface models. Due to the highly nonlinear parameter to data mapping, there are multiple local minima in the parameter space. Traditional FWI based on a local optimization method might fail to converge to a geologically meaningful model if the starting model of the inversion does not contain sufficiently accurate macro-structures. In this research, we propose a hybrid optimization method for FWI by combining the very fast simulated annealing (VFSA) algorithm with a derivative-based local search step to address this issue. We show that the proposed hybrid optimization method can generate an accurate background velocity model, and the result is not sensitive to the choice of the starting model. The output of the proposed method can then be utilized as a starting model for local optimization FWI methods to further refine the model.