Analysis of microseismic event picking with applications to landslide and oil-field monitoring settings.

摘要

The methods considered for microseismic event detection consist of energy, multi-window, autoregressive-Akaike Information Criterion (AR-AIC) and S-transform techniques. Synthetic seismic data were developed to analyze each method under various conditions. The application of an onset time correction resulted in the energy and multi-window techniques selecting waves to within 1 ms and 0.01 ms, respectively. The energy and multi-window techniques were found unsuitable for detecting the P- and S-wave arrivals in low signal-to-noise ratio environments.; Having obtained sound results with the synthetic data, specific methods were applied to data acquired in an oilfield at Cold Lake, Alberta and in an unstable topographic area at Turtle Mountain, Alberta. While the AR-AIC and S-transform techniques resolved emergent events from Turtle Mt., the multi-window technique had much more accurate results on the few events it could detect. The energy and multi-window procedures had difficulties detecting the S-wave arrivals at Cold Lake as a result of the amplitudes being clipped.