Performance Comparison of PhaseNet with Conventional Event Identification of Microearthquake Data in Geothermal Sites
Putu Raditya Ambara Putra, Muhamad Firdaus Al Hakim*, Fanata Yudha

1) Geophysical Engineering, Faculty of Mineral Technology, UPN Veteran Yogyakarta, Indonesia
2) Petroleum Engineering, Faculty of Mineral Technology, UPN Veteran Yogyakarta, Indonesia


Abstract

To evaluate the reservoir^s permeability and understanding the potential seismic hazard in geothermal fields, microearhtquake (MEQ) monitoring is mandatory. However, the process to obtain the information from seismic data is often tedious and time-consuming. Therefore, in this study, we used the PhaseNet to identify the arrival times of P and S to obtain the microseismic events catalogue. We compared the results of PhaseNet identification with existing catalogue to see the quality of the data. In this study we use one month data from geothermal field. PhaseNet succeeded in detecting a greater number of phases and events compared to catalog data, where the identification match rate was 85%. Furthermore, the time required for automatic detection of PhaseNet is relatively short, thus this method is good for initial step for MEQ analysis. We also compared the hypocenter location of PhaseNet catalogue and existing catalogue to determine the stability of the method.

Keywords: PhaseNet, Machine Learning, Microearthquake

Topic: Engineering

ICARSESS 2024 Conference | Conference Management System