Satellite imagery, combined with the power of artificial intelligence, is becoming the new guardian of our water resources
To mark World Space Week, which aims to celebrate the contributions of space science and technology to improving the human condition, we take a look back at the results of several OiEau projects utilising satellite technology.
Against a backdrop of increased pressure on water resources, resulting from climate change, it is important to carry out detailed monitoring, in terms of both quantity and quality, in order to implement proactive and adaptive management.
To this end, having data collected in real time for close monitoring, and in sufficient quantity to be representative across a wide area, is a major asset in understanding the state of the resource.
However, in situ measurements are not always easy to obtain: some areas are difficult to access, making it challenging to install monitoring equipment. Furthermore, establishing and maintaining a large measurement network remains a costly undertaking. Thanks to satellite technology, these obstacles have now largely been overcome.
For over 10 years, OiEau has been involved in the collection and use of satellite data for water management. Since 2014, OiEau has coordinated the ‘space hydrology’ group in France, comprising French research, institutional, technical and operational stakeholders (CNES, IRD, AFD, IRSTEA, BRLi, CNR, CLS), who sought to reconcile their technical and scientific perspectives with user needs regarding satellite data relevant to water management.
This work culminated in December 2022 with the launch by the French space agency (CNES) and NASA of the SWOT satellite (Surface Water Ocean Topography), the first space mission to study the Earth’s surface waters.
In South America
M.A.R.U. (Monitoramento das Águas Residuais urbanas e do seu Impacto Ambiental)
The M.A.R.U. project aims to monitor the quality of surface water in Brazil. It focuses specifically on the automated detection of major pollution hotspots using remote sensing and artificial intelligence.
The methodology is based on images from the Sentinel-2 satellite, the optical component of the European Copernicus programme, which enable the analysis of four parameters: chlorophyll a, dissolved organic matter, suspended solids and turbidity. Data processing includes the identification of the flow direction and a pollution gradient.
This method has been successfully validated by comparing the predicted points with reference points identified manually in the field.
The results demonstrate the tool’s ability to automatically detect pollution hotspots.
Bio-Plateaux, an initiative for cross-border coordination on water and biodiversity
The Bio-Plateaux project aims to promote the sharing of data, information and expertise concerning water and biodiversity in aquatic environments between French Guiana, Brazil and Suriname. It focuses in particular on the transboundary basins of the Oyapock and Maroni rivers, which are vast and often difficult to access.
To compensate for the lack of ground-based measurement infrastructure, the project relies heavily on spatial hydrology. This approach incorporates several satellite technologies, including satellite rainfall measurement. Combined with radar data, it feeds into the MGB hydrological model to simulate flow rates, forecast water levels and anticipate the risks of floods or droughts.
The MGB model (Modelo de Grandes Bacias or MGB-IPH model) is a distributed hydrological and hydrodynamic model designed to simulate the water cycle and propagate flow rates across large basins. Suitable for large basins (generally larger than 1,000 km²), such as the Amazon basin or the Guiana Shield, it is used to forecast floods, monitor low water levels (droughts) and manage navigability or hydraulic structures.
Furthermore, space-based altimetry, using satellites such as SWOT, makes it possible to measure the depth of watercourses in the most remote areas.
The project also uses remote sensing, notably data from the Sentinel-2 satellite, and artificial intelligence to monitor surface water quality.
Analysis of water colour enables four key parameters (turbidity, chlorophyll, suspended solids, organic matter) to be mapped, thereby detecting pressures on the environment, such as gold panning or agriculture.
The methodology relies on the processing of satellite imagery to identify the location and extent of the impact of mining activities.
This includes generating flow direction masks, analysing concentration gradients, and using algorithms to detect pollution hotspots.
The results show that AI enables the automatic detection of pollution hotspots, which are then validated by comparison with manually identified reference points.
These satellite data provide a comprehensive overview and complement field measurements for water resource management.
On the African continent
The integration of remote sensing and spatial data is a strategic tool for addressing the shortage of ground-based monitoring stations and tackling climate-related challenges in major African basins, such as those of the Congo, Senegal, Niger, Volta, Lake Chad and the Nile.
In this context, the DYNOBA and SCREEN projects work together in a complementary manner to modernise operational hydrology and strengthen the technical expertise of river basin organizations.
DYNOBA: strengthening African cross-border river basin organizations to improve water resource management
Led by OiEau and funded by the AFD, in close collaboration with CNES, IRD and various specialist private sector organisations, the DYNOBA project places space-based technologies at the heart of water management at a regional level. Its activities focus primarily on space-based altimetry, drawing on satellite missions such as SWOT to remotely measure water levels in lakes and rivers in order to accurately assess flow rates. The project also utilises satellite imagery to estimate rainfall and evapotranspiration across vast, remote areas. Furthermore, the use of spatial time series makes it possible to reconstruct missing hydrological records, feed into flood forecasting models and anticipate droughts. To ensure these tools are taken up and utilised, DYNOBA is facilitating a major skills transfer by organising regional workshops on operational spatial hydrology, designed to train technicians from various African organisations such as the OMVS, the ABV, the ABN and the CBLT.
SCREEN: Monitoring and management of water resources in the Senegal River basin
Building directly on this momentum, the SCREEN project, led by the OMVS, is deploying these applications operationally across the Senegal River basin. Its second component is entirely dedicated to the use of satellite technologies to mitigate the impacts of climate change.
In practical terms, the project aims to modernise the basin’s hydrology by modelling flows to estimate discharge rates in near real time, thereby complementing or supplementing in situ measurement networks.
SCREEN also incorporates mapping and continuous monitoring of the river’s ecological quality, whilst refining the prevention of extreme risks linked to floods and droughts.
Knowledge is harnessed by centralising all data on a dedicated platform. This tool helps inform strategic decisions in key sectors such as hydropower, agriculture, navigation and drinking water supply.
OiEau also supports the training of teams from the OMVS High Commission and national hydrological services to ensure the sustainable implementation of these innovative solutions.
In conclusion
The combined use of space-based remote sensing and artificial intelligence represents a major breakthrough for water resource management, particularly in vast and hard-to-reach areas.
This synergy not only overcomes the constraints associated with deploying physical sensors – which are often costly and complex to maintain – but also provides a comprehensive, real-time overview of the condition of basins.
AI maximises the potential of satellite data by automating the processing of large volumes of information and accurately identifying sources of pollution.
By cross-referencing hydrological and qualitative parameters, this approach provides decision-makers with robust decision-support tools to better anticipate risks, target interventions and ensure the sustainable management of transboundary aquatic environments.
Find out more about space technologies for water resources management
It is a technique for remote observation of the Earth’s surface based on the use of sensors carried on board satellites. By measuring the electromagnetic radiation reflected or emitted by the Earth, it enables data to be collected on the environment (vegetation, water bodies, soil) without any direct physical contact with the environment under study.
It enables very large areas to be covered, including isolated, cross-border or hard-to-reach basins where ground-based monitoring stations are few and far between or costly to maintain. Relying largely on open-access data (such as the Sentinel satellites from the European Copernicus programme), it provides regular time-series monitoring to map water quality, track physico-chemical indicators and anticipate droughts or floods.
Artificial intelligence automates the processing of large quantities of satellite imagery to detect deterioration in water quality. By cross-referencing satellite data with flow directions and concentration gradient analyses, machine learning algorithms are able to automatically identify the exact location of sources of pollution (mining activities, urban or agricultural discharges).
Find out about our projects using data and information systems
