Development of a Flood Early Warning System for Communities Along River Benue in Kogi State

Annual flooding along the River Benue corridor in Kogi State, North Central Nigeria, displaces thousands of households and causes extensive agricultural and infrastructure damage. This study develops a flood early warning system for communities along the Kogi State stretch of River Benue, integrating IoT water level sensors with a community alert application. The system uses ultrasonic water level sensors connected to Raspberry Pi nodes, transmitting data over LoRaWAN to a central server running a Python Flask API. A community-facing mobile app developed in React Native delivers tiered flood risk alerts based on configurable water level thresholds. The iterative development model guides three hardware-software integration cycles, evaluated at four sensor deployment points between Lokoja and Ajaokuta. Features include real-time water level dashboards, automated SMS alerts to registered community members, historical flood level logs, evacuation guidance, and a Kogi State Emergency Management Agency (SEMA) administrative interface. Testing evaluates alert lead time, sensor reliability, and community alert uptake over two simulated flood seasons. Results demonstrate an average flood alert lead time of 4.5 hours, enabling timely evacuation of trial communities. The study recommends formal integration with the Nigeria Hydrological Services Agency (NIHSA) for data enrichment. Keywords: flood early warning, River Benue, Kogi State, IoT, disaster management