<![CDATA[Indonesia is situated at the intersection of three major tectonic plates, resulting in high seismic activity and significant earthquake vulnerability.1 Rapidly determining initial earthquake parameters—including origin time, epicenter location, depth, and magnitude—is critical for effective early warning systems. This study evaluates the reliability of automated versus fast manual picking (<3 minutes, S-wave-based) by comparing their performance against final validated results. Utilizing data from the BMKG SeisComP system for the period of May 18, 2024, to May 17, 2025, the study analyzed 2,790 seismic events across Indonesia, including low-seismicity regions such as Kalimantan. Performance was assessed across six key parameters (depth, origin time, RMS, azimuth gap, magnitude, and epicenter) using a numerical scoring system (0–100) based on deviation from validated data. The results indicate that while automated picking processed a significantly higher volume of events (1,857 events; 66.6%) compared to manual picking (327 events; 11.7%) within the target timeframe, manual picking achieved a superior 'good' quality rating (score 75–100) at 96.9%, compared to 88.5% for automated methods. Nevertheless, automated picking remains the preferred method for rapid dissemination (<3 minutes) due to its operational speed. Furthermore, the study establishes regional thresholds for the minimum seismic phases required for reliable automated picking, ranging from 8 to 16 phases depending on the region, with a national average of 15 phases.]]>
Copyrights © 2025
