ANALYSIS OF GROUNDWATER POLLUTION LEVELS IN THE BANDUNG REGENCY AREA THROUGH CLUSTER AND DISCRIMINANT METHODS

• Main Author: Vira Nandhita, Nabila
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/82368
• Institution: Institut Teknologi Bandung

Water is an essential component of life. In Indonesia, the availability of clean water is crucial as the demand increases with the growing population. Bandung Regency, which has the third highest population in West Java (West Java BPS 2020), has 53.07% of households relying on bore wells or pumps as their main water and sanitation source. This high percentage necessitates an analysis of well water contamination in Bandung Regency for better management. Cluster methods can be used for this analysis, employing both hierarchical and non-hierarchical approaches for various numbers of clusters. The most appropriate method is chosen using icdrate and pseudo-f values. For the metal content data in well water in Bandung Regency, the most suitable clustering method is k-means, which divides the samples into five clusters: 6 samples in Cluster 1, 36 samples in Cluster 2, 19 samples in Cluster 3, 119 samples in Cluster 4, and 3 samples in Cluster 5. The clustering results are then compared with the quality standards from the Minister of Health Regulation No. 492 of 2010 and Government Regulation No. 22 of 2021 to determine the water condition status based on pollution levels. The results show that Clusters 1, 3, and 5 have the highest percentages of samples classified as contaminated compared to the other clusters. Wells located in residential areas near industries have three out of five average metal contents higher than those in other locations.