Cluster Analysis of Sumatra Island Earthquake Distribution
Dian Anggraini (a), Sapto Wahyu Indratno (b,c*), Utriweni Mukhaiyar (b)

(a) Doctoral Program of Mathematics, Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung, Indonesia
(b) Statistics Research Division, Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung, Indonesia
(c) University Center of Excellence on Artificial Intelligence for Vision, Natural Language Processing & Big Data Analytics (U-CoE AI-VLB), Institut Teknologi Bandung
*saptowi[at]itb.ac.id

Abstract

There are many sources of earthquakes on the Sumatera Island, such as five mega-trusts, faults, and volcanic activities that extend from Aceh to Lampung. Those make the island of Sumatra vulnerable to earthquakes. The International Seismological Centre (ISC) recorded 9,414 destructive earthquakes in the range of magnitude 4.0 to 9.1 on the Sumatera Island since 1907. When a sudden earthquake occurs and the preparedness in dealing with earthquake disasters are lacking, then the local governments or even the central government have difficulties in managing the earthquake impact. Therefore, the risk classification of locations is needed. Here, the grouping analysis is executed to determine areas with a high-risk vulnerability based on the number of recorded earthquake events. The cluster analysis can be implemented in determining the grouping of earthquake areas through the ^K-means cluster^ model. The results show that the earthquake area on the Sumatera Island is divided into five groups with the order category of vulnerability, with the most occurrences being groups 3 - 2 - 5 - 1 - 4 with values of 2994, 2552, 2488, 760 and 620 earthquake events. While the order based on the average magnitude from largest to smallest is group 1 - 2 - 3 - 5 - 4, the magnitude values are 5.9, 4.88, 4.85, 4.68, and 4.65. The results of this grouping will benefit the government as a primary reference for regional vulnerability data due to earthquakes. So that planning in terms of risk mitigation and earthquake prevention will be optimal, both in terms of policymakers, technical, and financing.

Keywords: K-means cluster, Earthquake

Topic: MATHEMATICS AND STATISTICS