At the CaMI Field Research Station we recorded several weeks of continuous surface and borehole seismic data to study the feasibility of using ambient noise correlation (also called interferometry) to monitor CO2 injection. We focus here on the October 2017 dataset (prior to injection), composed of 14 days of continuously recorded data at the 98 stations of a 3C-3D permanent array (receiver grid of 10 m x 10 m). We use a standard processing (mean and trend removal, 1bit, spectral whitening) and compute the 14 daily ZZ-correlations for the 4753 pairs of stations to reconstruct the Green’s function between them. Daily correlations show stable waveform for the baseline dataset with a good correlation coefficient between the reference and the daily correlations. Variations in the elastic parameters of the subsurface due to CO2 injection will directly affect the reconstructed Green’s function, and passive recording should allow us to detect the induced change of the medium. Interferometry can also be used as a tomographic tool through the analysis of the dispersion curve of the reconstructed Green’s functions. We compute the dispersion curves for few couples of stations. A detailed analysis will be undertaken to determine why some periods show outlier values, but the group velocities obtained are similar to those found in literature.
View full article as PDF (33.58 Mb)