[S02-5] New directions in anthropogenic seismicity studies I

[S02-5-02] Picking vs Waveform based detection and location methods for induced seismicity monitoring

^○Francesco Grigoli¹, Maren Boese¹, Toni Kraft¹, Bernd Weber², Stefam Wiemer¹, John Clinton¹ (1.ETH-Zurich, Swiss Seismological Service, Switzerland, 2.Gempa GmbH, Potsdam, Germany)

Microseismic monitoring is common operation in different georesources related industrial activities, such as oil&gas, mining or geothermal energy exploitation. In microseismic monitoring operations we generally deal with large datasets requiring robust automated analysis procedures. Such seismic datasets are often characterized by multiple events with short inter-event times or overlapping events; in this case, correct phase identification and event association are challenging, and errors can lead to missed detections and/or reduced location resolution. In the last years, to improve the performance of the current data analysis procedure various waveform-based methods for the detection and location of microseismicity have been proposed. These methods exploit the coherence of the waveforms recorded at different stations and do not require any automated picking procedure. Although this family of methods have been applied to different induced seismicity datasets, an extensive comparison with sophisticated pick-based detection and location methods is still lacking. We aim here to perform a systematic comparison in term of performance among waveform-based methods and the pick-based detection and location methods (SCAUTOLOC and SCANLOC) implemented within SeisComP3. SCANLOC is a new detection and location method specifically designed for seismic monitoring at local scale and is based on a cluster search algorithm to associate detections to one or many potential earthquake sources. On the other hand, SCAUTOLOC is a more “conventional" method and is the basic tool for seismic event detection and location in SeisComp3. This approach was specifically designed for regional and teleseismic applications, thus its performance with microseismic data might be limited. We analyze the performance of these methodologies for a synthetic dataset with realistic noise conditions and for induced seismicity data related to geothermal energy exploitation in St. Gallen (Switzerland).