Development of a Sensor-Based Flood Early Warning System with Rainfall and Water Level Monitoring

Floods are one of the most frequent and destructive natural hazards, with their intensity and frequency expected to increase under climate change and rapid urbanization. Accurate and timely detection of rainfall and rising water levels is crucial for reducing disaster risk at the community level. This study aims to develop a low-cost and sensor-based Flood Early Warning System (FEWS) that integrates a tipping-bucket rainfall sensor and an ultrasonic water level sensor with a web-based monitoring platform. The system was designed, implemented, and evaluated through calibration, sensitivity testing through accuracy analysis, and Root Mean Square Error (RMSE) calculation. Experimental results showed that the rainfall sensor achieved an average accuracy of 96% with an RMSE of 2.38, while the water level sensor demonstrated 100% accuracy with an RMSE of 0 under controlled testing conditions. These results confirm that both sensors can provide reliable measurements for real-time flood monitoring. The integration of rainfall and water level observations in a single system enhances the capacity for early detection, enabling rapid dissemination of alerts through online platforms. The findings highlight the feasibility of deploying affordable and flexible FEWS prototypes to strengthen disaster preparedness and resilience in flood-prone communities.