Full Waveform Ambient Noise Inversion for Finite Domains

Abstract:

Ambient noise tomography is a well-established tomographic technique that has been applied successfully in the last decades. Nevertheless, new developments are necessary to advance its resolution capabilities. Recently, full waveform ambient noise inversion has emerged as an alternative approach. Accounting for arbitrary noise source distributions in both space and frequency, and modeling the seismic wavefield in a heterogeneous and attenuating 3D Earth has the potential to overcome known limitations when invoking the principle of Green’s function retrieval for the interpretation of noise correlations. Full waveform ambient noise inversion, however, is computationally challenging, since even distant noise sources can have an influence on the inter-station correlation functions and therefore require a prohibitively large numerical domain, beyond that of the tomographic region of interest. We investigate a new strategy that allows us to reduce the simulation domain while still being able to account for distant contributions.

In order to decrease computational costs, current developments assume spatially uncorrelated noise sources. With this assumption, only those sources that are located within a potentially limited domain can be considered appropriately. Simply moving distant contributions into a small numerical domain leads to contributions to the correlation wavefield with an incorrect curvature. Inspired by Huygens principle, we introduce correlated sources and generate a time-dependent effective source distribution at the boundary of a small region of interest that excites the correlation wavefield of a larger domain. We experiment in a 2D framework with correlated sources surrounding an array and incorporate beamforming results as prior information regarding the azimuthal variation of the ambient noise sources.

Understanding how the effect of heterogeneous noise source distributions can be included at reduced computational costs is a key ingredient for the evolution of full waveform ambient noise inversion into a viable tomographic technique. In addition, the presented experiments pave the way for reducing a source-induced bias in monitoring applications.