Finite Frequency Sensitivity Kernel for the Correlation of Ambient Noise Correlations: Theory and Numerical Tests

Xin Liu, & Gregory C. Beroza

Submitted August 15, 2018, SCEC Contribution #8727, 2018 SCEC Annual Meeting Poster #108

Full waveform adjoint tomography has achieved great success in applications from global structure using earthquakes to exploration seismology using active sources. When combined with ambient seismic noise interferometry, however, the shape of the sensitivity kernel can be distorted with significant sensitivity outside the inter-station region due to non-isotropic distribution of far field noise sources. We compute the sensitivity kernel for ambient noise cross-correlation differently from existing approaches by introducing an additional station to the classic two-station setting. We cross-correlate the two noise cross-correlations from two station pairs and use the resulting cross-correlation (C2) to compute synthetic and data waveform misfit function. This differential sensitivity kernel based on C2 shows prospect for canceling the overlapping part of the original kernels for two pairs of stations in interferometry, thus reducing the effect of non-isotropically distributed noise sources. We derive analytically and calculate numerically the differential sensitivity kernel for seismic interferometry and show examples based on 2D membrane waves.

Citation
Liu, X., & Beroza, G. C. (2018, 08). Finite Frequency Sensitivity Kernel for the Correlation of Ambient Noise Correlations: Theory and Numerical Tests. Poster Presentation at 2018 SCEC Annual Meeting.


Related Projects & Working Groups
Seismology