Design And Development of A Flood-Prone Area Mapping System Using The K-Means Clustering Method Based On Web (Case Study: Lhoksukon District)

Abstract

Natural disasters are a form of natural events that result in a major impact on human populations. The natural disaster that frequently occurs in Lhoksukon District is flooding. Flooding is caused by continuous rainfall. Based on data from BPS (Central Statistics Agency) of North Aceh, the rainfall height in 2021 averaged 152.19 mm/month. High rainfall has caused many areas to be affected by floods and experience many losses, including disrupted road access, submerged houses, a paralyzed economy, and even loss of life. Based on data from BPBD (Regional Disaster Management Agency), in 2018 there were 19 villages affected by floods out of 75 existing villages, while in 2022 there were 54 villages affected by floods out of 75 villages in Lhoksukon District. In this research, a flood-prone area mapping system was created using the K-Means Clustering method, where the K-Means Clustering method is used to cluster villages affected by floods using 5 variables: duration of water inundation, water height, watershed (DAS - Daerah Aliran Sungai), elevation, and land cover. Based on the test results that have been conducted, there are 2 villages in the green cluster, 52 villages in the yellow cluster, and 21 villages in the red cluster. The results of this clustering are digitalized into a Geographic Information System using the Mapbox API. The digital map displayed shows the area in Lhoksukon District divided by village with green zone identification for non-vulnerable level, yellow zone for vulnerable level, and red zone for highly vulnerable level.