Full Waveform Inversion Using Passive Seismic Data Based on Seismic Interferometry and Source Independent Method

Abstract

Summary Passive seismic signals correspond to real seismic waves that transmit through the subsurface media, thus they contain sufficient information of the media. Using seismic interferometry, the passive seismic data recorded on surface can be converted into virtual source records. These virtual source records can be processed using conventional methods for active source data. The main problems when conducting full waveform inversion (FWI) using virtual source records are that the virtual source wavelets are unknown and the shallow spurious events are serious. In this paper, we first simulate the noise source observed data by finite difference method, and then reconstruct virtual records by corsscorrelation. We introduce the convolution type source independent FWI algorithm to overcome the effect of wavelet uncertainty. After that, we conduct large scale smoothing to retain the long wavelength components of the result and use it as the initial model of active source FWI. The numerical example illustrates that the low frequency information contained in passive seismic data can be used to invert the long wavelength components of the media. Using it as initial model can help to suppress cycle-skipping of the conventional active source FWI.