Speaker
Description
Accurate seismic event detection is crucial for understanding seismic activity and ensuring compliance with the Comprehensive Nuclear-Test-Ban Treaty Organization (CTBTO). The National Data Center (NDC) of Israel uses data from local and regional stations, as well as the International Monitoring System (IMS) of the CTBTO. With increasing seismic stations and available data, administrators have many options for selecting stations. However, stations vary in terms of proximity to noise sources, instrumentation, maintenance and the environment, all of which affect data quality and detection capabilities. To optimize station selection, it is essential to understand the strengths and weaknesses of individual stations and networks. This study aims to characterize detection capabilities and compare thresholds across different stations (e.g. deployment depth, instrumentation), regions (e.g. geology, noise proximity) and networks. Using decades of IDC bulletins, which reflect IMS detection and location algorithms and analyst expertise, we apply supervised learning to explore detection thresholds for IMS stations. Preliminary results suggest that event magnitude and station-event distance are the most influential features for detection likelihood. Station-event azimuth also plays a significant role, indicating spatial variations in detection capabilities. These insights will help system operators select stations for monitoring and improve station maintenance where performance is lacking.
| [email protected] |
