Spherical K-Means method to determine earthquake clusters

Abstract

Spherical clustering is a grouping technique for spherical data. A vector data set is grouped into clusters where the distance used to group the vectors is the angle between the vectors. Earthquake events are directional data that can be analyzed in three-dimensional (spherical) form because the epicenter of the earthquake spreads is very long, and the time difference between events is different. Therefore, this study aims to determine earthquake clusters in Bengkulu Province and the surrounding areas. The data used is data on earthquake events in Bengkulu Province and surrounding areas from 1970 to 2019, including latitude, longitude, magnitude, and depth, sourced from the USGS website. The analysis used in determining the earthquake clusters in Bengkulu Province is the Spherical K-means cluster analysis. The results showed that there were six earthquake clusters in Bengkulu Province and its surroundings. Most of the clustering results are in the sea.