An excess of nitrate and phosphate loads in rivers causes a disbalance in the river ecosystem leading to extensive (toxic) algae bloom and loss of clean water used for irrigation and drinking water purposes. The Vlaamse Milieu Maatschappij (VMM) monitors nitrate levels in Flanders to better understand the pressures and impacts on Flemish rivers.
- Challenge: Continuous monitoring of nitrate levels in rivers in agricultural regions.
- Solution: Utilization of high-end UV/VIS sensors, Internet of Things technologie (IoT), data analysis and machine learning to identify peak events.
- Results: The comparison between results from in-situ UV/VIS sensors and laboratory-analyzed samples shows a very good performance of the UV/VIS sensors. Periodic maintenance, advanced data processing techniques,and semi-automatic quality control are necessary to achieve this good performance. Process knowledge and expertise are essential components to acquire a highly quality data set.
Challenge
Understanding river systems impacted by human activity requires continuous monitoring of nitrate levels. Real-time data are crucial for optimizing water intake, especially for drinking water purposes. Traditional methods involve manually collecting water samples and analyzing them in a laboratory. Monitoring during peak events is essential for good management, yet this is unfeasible by using these in-situ sampling methods.
Solution
Fluves HYDRO addresses this with an autonomous monitoring system using UV/VIS sensing technology. This system continuously collects data on nitrate by emitting light at specific frequencies into the river water and measuring the light absorption.
- Autonomous monitoring with IoT technology: Provides updates every 15 minutes.
- UV/VIS sensor setup: Ensures accurate data collection.
- Regular maintenance: Periodic maintenance prevents sensor drift and failure.
- Advanced data processing: Uses data algorithms and machine learning for quality assurance.
- Interoperable: integrates smoothly in the VMM data pipeline.
Results
Comparing lab samples with UV/VIS sensor data shows high performance and reliability of the continuous monitoring system. The uncertainties related to lab samples are similar to those related to UV/VIS monitoring, validating the sensor's accuracy. Lab samples can be used to calibrate the sensor, and in addition, quality control can be improved by using processing the maintenance log.
Conclusion
Continuous nitrate monitoring is essential for detecting peak events in river systems affected by human activity. The Fluves Hydro system operates autonomously by using IoT, providing real-time data with updates every 15 minutes, and is fully interoperable with the current VMM data pipeline. Although the system requires periodic maintenance to ensure optimal performance, the advanced dataprocessing and machine learning algorithms significantly enhance the quality of the collected data. The UV/VIS sensing approach has demonstrated a high level of performance comparable to traditional lab samples, making it a reliable and efficient solution for continuous nitrate monitoring. This system not only optimizes water intake processes but also offers valuable insights into the health of critical water sources.
This project was funded by and executed with the Vlaamse Milieumaatschappij (VMM). The report of this study can be found here.
.png)