Visualizing water demand and predicting future consumption in Malaysia

Abstract

- Access to clean water has long been a major issue in Malaysia. Urbanization and flooding have made the need for innovative and sustainable solutions even more pressing. This study aims to analyse the challenges and opportunities of clean water access across states in Malaysia. Datasets from the Department of Statistics Malaysia (DOSM) and MyWater Portal are used to examine accessibility to treated water through data visualization and identify disparities in Kelantan, Sabah, and Sarawak, where coverage remains below 90% despite the abundance of water resources. To explore the underlying factors, the comparison focuses on how dams are built and used in these states versus in Selangor, which, despite its dense population and heavy demand, still manages to provide high access to treated water. Then, time series models (Exponential Smoothing, DES, ARIMA) and machine learning algorithms (XGBoost, Random Forest) are built to forecast state-level water consumption. The result shows that Double Exponential Smoothing (DES) performs the best to capture historical water consumption levels and trends. Water demand is expected to grow rapidly due to population and industrial development in Selangor, while other states are projected to experience steady increases. However, dam capacity or water production alone does not guarantee equitable access. To close these gaps, the study this study proposes two sustainable water treatment technologies; electrocoagulation and photocatalysis. Both eco-friendly recommendations can strengthen Malaysia’s long-term water security while supporting Sustainable Development Goal 6: Clean Water and Sanitation.