MaD-OPS: Monitoring & Detection of Organic Pollution from Sewage: Implementation of an agile sensing network for informing river health

Rivers are under pressure from many different sources, including farming and rural land use, wastewater treatment, towns and transport. In England, very few rivers achieve good ecological status, and none achieve good chemical status. This comes after many years of exploiting our freshwater systems. In 2024 there were more than 450,000 combined sewer overflow discharges in England, totalling over 3.5 million hours of spills. This sewage has direct implications on ecological and human health, increasing the environmental contaminant load in rivers. In response, Section 82 of the Continuous Water Quality Monitoring Programme mandates continuous monitoring of freshwater systems, with scope for future expansion of monitored parameters.  

Current water quality monitoring relies heavily on infrequent spot sampling, often missing key impact events, with limited spatiotemporal context. The MaD-OPS project has developed a novel sensing network for continuous monitoring of biological, chemical, and physical water quality parameters. A key focus is to demonstrate the value of a new fluorescence-based optical sensor for detecting organic pollution and bacterial contamination within a demonstrator catchment, with the potential to reveal underlying biogeochemical cycling processes. 

To isolate different pollution sources, sensor nodes have been deployed at multiple points along a river. Alongside continuous sensor data, regular spot sampling is being carried out for faecal indicator organisms, BOD₅, nutrient analysis, and microbial community profiling to provide robust ground-truthing.  

The project aims to develop a user-friendly dynamic Water Quality Index (WQI) that integrates high-frequency sensor data with machine learning, for real time assessment of river health that can be used by citizen scientists, community groups, and regulators alike. Using a novel dynamic baseline approach, the WQI will assess each sensor node relative to the least impacted section of the river at any given time.  

Preliminary results demonstrate that continuous monitoring captures point source pollution and hydrological events that are not detected through spot sampling alone. Comparison between the dynamic headwater baseline and downstream sensor nodes highlights the direct impact of point source events on river health.  

We present progress in deploying the sensing network, early insights into river health derived from high-frequency data, and how these findings are informing the development of the WQI framework.