Stable operation of a network-based multi-algorithm earthquake early warning system: the Korea meteorological administration platform

Abstract Network-based Earthquake Early Warning (EEW) systems determine an earthquake’s hypocenter and magnitude from the earliest detected seismic waves, known as P-waves, and then issue alerts. For this process, detection by at least three seismic stations is required. Estimating these source parameters with data from only a few stations is inherently challenging. While incorporating additional stations can improve the accuracy of source estimation, it also increases the time needed to issue an alert. Therefore, achieving both rapid response and reliable parameter estimation is essential for stable warning operations. This study evaluates the operational performance of integrating three independent EEW algorithms within the real-time KMA system. By combining their outputs and applying inter-correlation checks, the system attained higher detection rates and improved the average accuracy of issued alerts compared with any single algorithm, resulting in more stable operations. This study confirms that the platform, which integrates individual algorithms, could be refined to enhance reliability.