Application Of Artificial Intelligence In Rainy Season Water Level Analysis And Forecasting At Phu Loc Station, Vietnam

Abstract

Phu Loc is a region strongly influenced by the rainy season from May to November, causing significant fluctuations in surface water levels. This study analyzes the variations in water levels and rainfall during the 2024 rainy season at Phu Loc station, An Giang Province, and employs advanced machine learning and deep learning models, including Linear Regression, Random Forest, Support Vector Regression (SVR), and Long Short-Term Memory (LSTM), to forecast water levels for the period 2025–2029. Among these models, LSTM demonstrated superior predictive performance with the highest accuracy and ability to capture abrupt changes in water levels during heavy rainfall events. Forecast results indicate a continued trend of elevated water levels during the rainy months, with a slight increase in peak water levels attributed to climate change and tidal influences. These findings provide essential insights for proactive water resource management, flood risk mitigation, and climate adaptation planning in the region. The study also discusses the integration of additional hydrological variables and the need for continuous model updating to enhance forecast reliability.