Dual-Purpose and Low-Cost IoT-Based Sensor Network for Real-Time Rainfall and Urban Flood Monitoring

This work aims to develop an integrated framework based on measurements from a dual-function, low-cost Internet of Things (IoT) based sensor network capable of measuring rainfall and surface water level simultaneously to enable real-time data for actionable decision-making during extreme events and address observational gaps in flood-prone urban areas. This study presents a novel urban flood monitoring network using observations from co-located water level and rainfall sensors. The high-density IoT sensor network measures rainfall using a Hydreon RG-15 Optical sensor and water level using a Maxbotix MB7389 ultrasonic distance sensor at 1-minute intervals. This sensor network allows for real-time monitoring of local flood and rainfall events, which is an improvement over the traditional high-water mark observations. The proposed network is operational in two cities in New Jersey, USA, and constitutes the first step towards building NJFloodNet, a state-wide network for flood monitoring in real time. The analysis of the collected data shows a sensitivity of water level observations to air temperature and higher measurement error with height. A method was proposed to correct the temperature-introduced bias. The observation from the rainfall sensor was in good agreement with an adjacent reference rain gauge station, as the comparison led to an RMS of 2.49 mm. The analysis of both observations from the water level and rainfall sensors revealed an added value in the understanding of the hydrologic processes, as rainfall peaks preceded the peak of inundation, laying the groundwork for the potential of using the network in an urban flood alerting system.