EGU General Assembly 2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Monitoring floating riverine pollution by advanced technology

Irene Ruiz1, Oihane Barurko1, Irati Epelde1, Pedro Liria1, Anna Rubio1, Julien Mader1, and Matthias Delpey2
Irene Ruiz et al.
  • 1AZTI. Herrera Kaia – Portualdea z/g. 20110 Pasaia, Spain
  • 2Center Rivages ProTech, SUEZ Eau France, Bidart, France

Rivers act as pathways to the ocean of significant but unquantified amounts of plastic pollution. Measuring with precision the quantities of riverine plastic inputs is crucial to support and ensure the effectiveness of prevention and mitigation waste management actions. However, there is a lack of technological tools capable of monitoring and, consequently, assessing accurately plastic abundances and its temporal variability through river water surfaces. Within the LIFE LEMA project, two videometry systems were installed at the river mouths of two European rivers  (Oria in Spain and Adour in France)  and a detection algorithm was developed  to monitor litter inputs in near real time . The objective of these developments was to detect riverine pollution at water surface, with the goal of quantifying the number and providing data on the travel speed and size of the floating items. Between 2018 and 2020, the system was tested under different environmental conditions. These tests have led to develop a second version of the algorithm that improves the results reducing false positives. After these improvements, a new validation has been carried out consisting in detailed analysis of more than 300 short videos of 5 minutes duration recorded in Orio’s station under different river flows, weather conditions and plastic loads. The validation results highlighted the operational reliability of the system. In a scale of 1 to 5 scoring (being 1 very bad and 5 very good) over 70% of the recordings scored 4 to 5. This also demonstrated the great potential of the videometry system in harmonizing visual observations of floating riverine litter. The data provided by the systems is currently being used in the LEMA TOOL, a tool designed to guide local authorities on managing, monitoring and forecasting marine litter presence and abundances in coastal waters of the SE Bay of Biscay. Furthermore, the data provided is key to evaluate the sources of the pollution and the efficiency of waste management measures within the river basins, towards a successful reduction of plastic inputs into the ocean.

How to cite: Ruiz, I., Barurko, O., Epelde, I., Liria, P., Rubio, A., Mader, J., and Delpey, M.: Monitoring floating riverine pollution by advanced technology, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-22613,, 2020

Comments on the presentation

AC: Author Comment | CC: Community Comment | Report abuse

Presentation version 1 – uploaded on 04 May 2020
  • AC1: Frequently asked question about the videometry system, Irene Ruiz, 22 May 2020

    Dear EGU members and colleagues,

    The purpose of this post is to briefly describe some aspects of the system not included in the presentation which we believe may serve to better understanding the system performance. Below you can find a short list of frequently asked questions about the technical features of our videometry system. Some of the FAQs addressed in the post have come up in the past few days during the EGU General Assembly 2020. If you cannot find the answer to your question, please feel free to reach out via our email!


    Why Near NIR (Near InfraRed) cameras?

    NIR cameras enhance the contrast between the floating objects and the water, an essential factor to warrant an automatic detection.

    What does “automatically detection” means?

    Once an item is detected, the algorithm records its position and size. The following step is tracking it in all the frames (images) provided by the camera. For each of the detected items in an image (if there was any), the algorithm will try to match them in the following image. For the item detection and tracking, the algorithm simplifies the shape of the object into a circular area. Image rectifications can improve the quality of the monitoring, but they are not essential for a good performance of the system.


    Is the videometry system suitable to monitor items in the water column or even in the seabed?

    The videometry system is only applicable for items visible from outside the water, it means floating on the water surface or very close and moving downstream. It has been designed to monitor the fluxes of riverine litter to the sea, but it has been also installed in an artificial river channel to assess its effectiveness in new spots.


    Can the system handle large amounts of litter loads?

    One of the key features to bear in mind for a reliable and efficient monitoring is the computing capacity of the system. Our experience so far shows that a suitable threshold of 500 litter items detected per second is appropriate to avoid the over saturation of the system for the standard performance of a modern pc. However, this upper limit can be adjustable depending on the hardware selected.

    Which sizes of items can capture the videometry system? What about the organic fraction?

    The system verifies items within a size range of 10 cm to 2 meters. The minimum value is constrained by the camera resolution but can be also adjustable. Although sub-centimeter sized items can be present, the system has been designed to monitor the larger items floating in the water surface. For the moment, the system does not difference between organic items and floating litter.

    How large is the area that can be monitored?

    The area is flexible and can be adapted to the monitoring needs. The river width or the installation site can be crucial factors for selecting the most convenient camera to ensure a successful monitoring.