Microseismicity detection and seismic ambient noise correlation at the CaMI Field Research Station, Newell County, Alberta

Marie Macquet, Donald C. Lawton, J. Helen Isaac

We present the continuous seismic data recorded at the CaMI Field Research Station and present results obtained with various methods. We focus here on the microseismicity detection using the STA/LTA method, and on the ambient noise correlation method, applied to imaging and monitoring. The STA/LTA method is applied on borehole continuous seismic data and detects 10's of thousands of events. The parameters chosen (STA, LTA and threshold) change the absolute number of detected events but the temporal distribution remains the same. We observe an increase in the number of detected events from February to end of March, and the number reduces after end of March. If some days of injection show a large number of detected events, some days without injection also show a large number of events. The frequent human activity (both at the site and in the surroundings) makes difficult to draw conclusions without a better discrimination between human produced events and injection related event. The ambient noise correlation method is used to produce a near-surface VS model which is compared with the one obtained with S-wave active source refraction analysis. Ambient noise correlation is also used to try to track the velocity change due to CO2 injection through the Moving-Window Cross Spectrum analysis applied on the reconstructed Green's function. A last application using this method is the attempt to reconstruct body waves, especially using the downhole continuous seismic data. Preliminary results show a good reconstruction of the body wave when correlating a surface station with downhole stations.