Velocity structure of The Geysers geothermal area (California) from ambient noise cross-correlation.

Abstract:

Ambient seismic noise recorded in 2012 - 2013 with a dense array of broadband seismometers located at The Geysers, California, is used to investigate the spatiotemporal distribution of seismic velocities in the reservoir which is extensively used for geothermal power production. We processed the raw seismic data to obtain reliable empirical Green's functions from cross-correlating data streams from station pairs. We retrieve the dispersion characteristics of surface waves by applying Time-Frequency Analysis of the empirical Green's functions to. We show that an analysis method involving the Wigner-Ville distribution has higher resolution in time and frequency compared to conventional filter techniques used for surface wave analysis. Group travel times of the Rayleigh waves at specified frequencies are measured and represent the input for the seismic tomography. A 2D model of group velocity distributions is derived and tested for robustness by subjecting it to a checkerboard test. In addition, temporal variations of group velocities are investigated for selected station pairs.