Toward real-time microearthquake event detection and location in anisotropic media using a multiscale approach for EGS collab experiments

Real-time, automatic microearthquake detection and location provides crucial information on fracture growth during stimulation and is planned for the second phase of the EGS Collab stimulation experiments. We study the feasibility of real-time detection and location of microearthquake events for the sensor distribution in the EGS Collab Experiment I, using eighteen 3C accelerometers. We pre-compute a microearthquake waveform database for a 43 x 43 x 43 grid with a grid interval of 1 m using an anisotropic elastic-wave modeling tool. We then compute correlation coefficients between recorded seismograms and those in the waveform database using a moving time window and search for the highest correction coefficients to detect and locate microearthquake events simultaneously. Toward real-time monitoring, we employ a multiscale event scanning method that requires only 0.03% of the computation cost of the global search for 43 x 43 x 43 grid points and every scanning time step, making real-time detection and location feasible. We verify the feasibility of our multiscale microearthquake event detection and location for real-time monitoring using noise-free and noisy synthetic microearthquake data for the sensor distribution in the EGS Collab Experiment I.