The earthquake events and impacts mapping in Bali and Nusa Tenggara using a clustering method

This article proposes the mapping of earthquake potential zones in Regional Center 3 (Bali and Nusa Tenggara, Indonesia) using the clustering method. A suitable clustering method to group spatial and non-convex data, such as earthquake data, is the Density-Based Spatial Clustering of Application with Noise (DBSCAN). We modify the epsilon parameter with the magnitude because it reflects the impact of the earthquake happens so that the epsilon value will be different for each point. Before the clustering process, magnitude is converted to kilometers (km) to determine the earthquake impact region, which will be used as epsilon. The earthquake impact region is obtained by computing the radius of the earthquake using the McCue Earthquake Perception Radius equation. This approach gives a good result for mapping the potential earthquake zones indicated by the cluster validity index that gives an average higher than 0 using the Silhouette index