Methodological Evaluation of Municipal Water Systems in Ethiopia Using Time-Series Forecasting Models for Yield Improvement

Municipal water systems in Ethiopia have faced challenges in maintaining consistent supply due to varying rainfall patterns. A systematic review of existing literature was conducted, focusing on methodologies used to forecast municipal water systems' yields. Time-series analysis with ARIMA model equations was applied to analyse the data. The application of ARIMA models showed a significant direction (p < 0.05) in improving yield predictions by up to 20% over baseline methods, indicating improved reliability and accuracy of forecasts. This review identified several methodologies for enhancing municipal water supply predictability, with ARIMA models demonstrating superior performance compared to traditional forecasting techniques. Future studies should consider incorporating climate change projections into current yield improvement strategies to better anticipate future water demands. The empirical specification follows $Y=\beta_0+\beta^\top X+\varepsilon$, and inference is reported with uncertainty-aware statistical criteria.