13 May 2026
Nikita Razguliaev recently defended his dissertation in Urban Water Engineering.
Sensors provide a clearer picture of urban stormwater
Rainwater running off roads and urban areas carries pollutants that affect water quality and aquatic ecosystems. Researchers at Luleå University of Technology have shown how sensors and data-driven approaches can provide a more detailed picture of stormwater quality, capturing variations that would otherwise be missed.
"To understand how pollution gets distributed in urban environments, we need high-resolution data, not just occasional samples", says Nikita Razguliaev, PhD in Urban Water Engineering at Luleå University of Technology.
Continuous monitoring instead of spot sampling
In the thesis “Sensor-based monitoring and modelling of urban stormwater quality”, he investigates how well sensors can measure how cloudy the water is, its acidity and how much dissolved substances it contains, directly in the field.
The results show that sensors are influenced by particle composition and weather conditions – in cold climates, snow and ice add further complexity by affecting sensor performance through introducing different patterns in data during cold periods.
"Sensors work well, but they need to be adapted to real stormwater conditions", he says.
The research also highlights the role of missing and erroneous data and how they can be handled using interpolation methods and machine learning, making it possible to build more consistent and representative time series.
New ways of working with stormwater
With continuous monitoring, it becomes possible to follow how pollutant levels change throughout an entire rainfall event, rather than relying on isolated measurements. This provides a clearer understanding of when peak loads occur, how long they persist and how different parts of the system respond.
It also allows practitioners to see how control measures and strategies aimed to mitigate adverse effects perform, enabling more informed decision-making.
From measurement to decision-making
When dynamics become visible, decisions can be made with greater precision. Problems can be detected earlier, system behavior better understood and actions prioritized more effectively. This enables systems to be adjusted based on actual conditions rather than assumptions – whether in design, operation or maintenance.
"With better data urban water management can be optimized", says Nikita Razguliaev.
Published:
Updated: