GEOthermal SEISmic AI Platform (GEOSEIS-AI): Shear-wave Splitting Analysis Module and A Case Study of Geothermal Site in Miaoli, Taiwan

Shear-wave splitting analysis module is part of the GEOSEIS-AI platform, primarily utilized to characterize stress states and subsurface fracture distributions in geothermal sites. However, microseismic data in geothermal sites often face on inherent limitations, including low signal-to-noise ratios (SNR), cycle skipping, fast/slow wave misidentification, and null measurements, all of which compromise the accuracy of automated processing.

To solve these limitations, this study optimizes the pre-processing stage by utilizing adaptive time-window selection to maximize SNR. Furthermore, an automated quality-controlling workflow was developed, based on three diagnostic metrics: (1) peak-picking determination of fast and slow waves; (2) cross-correlation (CC) coefficients; and (3) the energy variation rate between the principal S-wave component and perpendicular component. These tests facilitate the robust identification and remove low-quality seismic events.

This methodology was validated using microseismic monitoring data from the geothermal site in Miaoli, Taiwan. The results reveal two predominant fracture sets oriented NW-SE and N-S. The NW-SE orientations align with the regional focal mechanism solutions, reflecting stress states, while the N-S trends correspond to surface-mapped fault orientations. This workflow was integrated into the GEOSEIS-AI Platform—alongside AI catalogs, focal mechanisms, and seismic tomography—to establish a reliable microseismic monitoring system for geothermal exploration.

Keywords: GEOSEIS-AI; Geothermal Energy; Microseismic Monitoring; Shear-Wave Splitting; fracture distribution.