TITLE:
Development of a Low-Cost IoT Based River Discharge Measurement System
AUTHORS:
Martin Wainaina Chege, Bartholomew T. Kuria, Arthur W. Sichangi, Moses M. Ngigi, Jason N. Kabi, Andreas Rienow
KEYWORDS:
Water, Internet of Things, Telemetry, Discharge, Geo-Enabled Monitoring
JOURNAL NAME:
Journal of Geoscience and Environment Protection,
Vol.14 No.1,
January
22,
2026
ABSTRACT: Water scarcity is increasingly becoming a major concern globally. Impacts of climate change and human interference through abstraction have greatly contributed to the declining discharge within rivers, affecting sustainability. The impacts of declining river discharge are dire, impacting both upstream and downstream users and marine ecosystems. The impact of low discharge results in dried riverbeds and marine life death, among other critical environmental and human challenges. In Kenya, river discharge monitoring is mandated to the Water Resources Authority (WRA), which deploys river gauges and acoustic Doppler velocimeters, which are manual and lack telemetry. This has resulted in inadequate real-time river discharge data, endangering timely scientific investigations and policy-making decisions. To bridge this gap, an innovative real-time river discharge monitoring system that utilizes the Internet of Things was developed. The device encompasses three sensors: water level, flow sensor, and temperature. The sensors are interconnected in a 6 cm by 5 cm printed circuit board assembled with other auxiliary components. Arduino Nano lies at the heart of the interconnection of the sensors, and the Lora Module transmits the collected data to a cloud database. The PCB is housed in a 50 mm by 70 mm waterproof IP47 enclosure. The River Discharge Monitoring System operates by collecting data from the deployed location and transmitting it to the cloud spatial database, where it is geo-tagged. Key factors include water level, flow velocity, and temperature, which are continuously collected by the sensors and transmitted using LoRa. Field deployment demonstrated water level variations between 22 cm and 55 cm, temperature fluctuations from 19.50˚C to 20.23˚C with higher downstream temperatures, and discharge flows around 0.3 m3/s, indicating below-average river flows. Validation against manual in-situ methods showed strong performance with a coefficient of determination (R2) of 0.995 for discharge and 0.866 for water level, confirming the system’s accuracy is comparable to traditional methods. Low root mean square error and mean absolute error values further support sensor reliability and low bias.