Geofenced Earthquake Notification System Using Haversine-Based Distance Calculation and MMI Intensity Estimation

Abstract

Indonesia is one of the most seismically active countries in the world, making the rapid dissemination of official earthquake information essential for disaster mitigation and public situational awareness. This study presents the design and implementation of a mobile-based earthquake notification system that operates as a post-event notification platform by disseminating official earthquake information published by the Meteorology, Climatology, and Geophysical Agency (BMKG The system is designed as an event-driven client–server architecture that brings together real-time earthquake data from BMKG, a Firebase-based cloud back-end, and Firebase Cloud Messaging (FCM). The notification mechanism uses the Haversine formula to determine the distance between the earthquake source and the user, while the Fukushima–Tanaka attenuation relationship is applied to estimate the corresponding ground motion. The resulting Peak Ground Acceleration (PGA) is then converted into Modified Mercalli Intensity (MMI). The felt intensity reported by BMKG serves as the reference threshold, which is subsequently used in a binary-search procedure to estimate the radius within which notifications should be delivered. Notifications are then selectively delivered only to users located within the estimated affected area. Experimental results show that the Haversine implementation achieved a Mean Absolute Error (MAE) of 0.0845 km, the proposed filtering mechanism correctly classified users according to the estimated notification radius, and the notification pipeline achieved average end-to-end latencies of 0.83 s and 1.87 s in foreground and background states, respectively. These results demonstrate that integrating geospatial proximity analysis with shaking-intensity estimation enables efficient and spatially relevant dissemination of official earthquake information while reducing unnecessary notifications to unaffected users. The proposed framework demonstrates that integrating geospatial proximity analysis with shaking-intensity estimation can improve the spatial relevance of post-earthquake notification dissemination