Speaker: Marine Denolle
“The inaccurate description of the complex elastic and anelastic crust limits the wave propagation simulations. Predicting ground motion requires precise estimations of the earthquake source, the site effects and the path. We propose to focus on the paths and improve those simulations by using a newly developed analysis of the ambient seismic field.

First, we study the ground motion by calculating the transfer functions (impulse responses) between station pairs. Those can be compared to seismic records at long-period (4-10s) for moderate earthquakes. If more realistic source parameters should be applied, at a first order, we validate this method by testing the transfer functions against waveforms of moderate earthquakes in South California.

Second, we generate ambient noise Green’s functions to recover the spatially varying Rayleigh phase velocity and attenuation coefficient in southern California. The non-uniformity of the ambient noise sources does not bias the first order coherence (spectrum of the Green’s function) and we use its properties to estimate the Rayleigh waves phase velocity and attenuation information. We validated our wavespeed tomography maps with other studies and provide new insight on the attenuation information.

Ground Motion simulations can be improved by including the ambient seismic field analysis in different forms: directly, by calculating the transfer functions or indirectly, by improving the attenuation structure of the crust.”