Ambient Wave Field Characterization Using Three-component Arrays above a Gas Storage Reservoir

Abstract

We present a three-component array processor based on the Multiple Signal Characterization (MUSIC) technique. The method estimates slowness, propagation azimuth, polarization parameters, and seismic power for simultaneously active sources. This work is a step towards evaluating the hypothesis that the physical property contrasts between a partially saturated gas reservoir and the surrounding host medium are sufficient to modify the ambient seismic wave field. The new estimator is compared to the Capon high-resolution method (using the vertical component) by means of synthetic tests and two hours of night-time data from an array above an underground gas storage reservoir in France. We find that the three-component MUSIC processor provides superior sensitivity to horizontally polarized waves and considerably more accurate power estimations. Performance is otherwise similar to the Capon method. Combining propagation and polarization parameters, wave types of body waves and surface waves are identified. We observe suspected oceanic P-waves in the frequency range 1.2-1.7 Hz. As a next step, these waves and their seismic power will be compared to data acquired at the same locations six months later, when significantly more gas was stored in the reservoir.