Automatic workflow for detection, localization, and calculation of spectral parameters for induced seismicity on the Episodes Platform

Recently the automatic processing of seismological data recorded in areas with anthropogenic seismicity has become an important issue, as the number of available recordings has increased significantly over the past few years. The problem is also addressed within the DT-Geo Project WP8: Anthropogenic Geophysical Extremes, where a specific workflow is being developed for the automatic processing of induced seismicity-related waveforms. The workflow is designed as a set of independent applications implemented inside the interactive, publicly available Episodes Platform (https://episodesplatform.eu/). The applications created for the workflow include:

• A tool for picking the first arrivals of seismic waves using neural network-based solutions available within the SeisBench library,
• A phase association tool that employs the PyOcto algorithm,
• Location procedures already existing within the Episodes Platform.
• An application for calculating spectral parameters using the spectral fitting method,

Ultimately, the workflow will enable fully automated processing of raw and continuous seismic data, including event detection, localization, and spectral parameter calculation. The workflow can be used on the Episodes Platform either with data collected from various Episodes, which are geophysical datasets related to regions where induced seismicity has been observed, or with datasets uploaded by the user.

This work is supported by Horizon Europe grant DT-Geo 101058129 and a project co-financed by the Minister of Science Republic of Poland under contract no. 2024/WK/05. We gratefully acknowledge Polish high-performance computing infrastructure PLGrid (HPC Centers: ACK Cyfronet AGH) for providing computer facilities and support within computational grant no. PLG/2024/017279.