During the last five years, water resources have been heavily impacted by unusual climate situations, especially in the south of Europe, and in many other places around the World.

This leads to unprecedent tense situation, or even competition, between water users, urging heart-breaking prioritization between human drinking water supply, preservation of ecosystems, agriculture irrigation needs and industrial use.

In this situation, and through the coming years during which this tense will probably repeat and grow, the need for automated estimation and monitoring of water resources becomes compelling. The monitoring solution shall answer to the concerns of any kind of stakeholders, including resource administrators, regulatory institutions, environmental NGOs and users.

Moreover, this monitoring shall be systematic, unbiased, impartial, and dispassionate, in order to bring to local or country level decision makers the proper information to make their decisions and the levers to justify their prioritization decisions.


Earth Observation provides unprecedent data to this matter. For water bodies of scale starting from 1 ha, being either lakes, dams or reservoirs, either natural or man-made, Earth Observation is the perfect tool to monitor water resources at large scale, region wide or even country wide.

The combination of the results of different research, published during the last years, allows now to overcome the challenge of monitoring water bodies, for both surface and volume, only based on satellite data.


Inspired by those scientific papers, Atos has designed and developed on online service that automates water bodies monitoring based on the combination of different satellite sources, without the need of onsite data. The project is led in the scope of the agricultural and environmental services based on Atos’ Copernicus DIAS, Mundi Webservices https://mundiwebservices.com/

For each water body to be monitored, a first theoretical model is built from DEM data (for example Copernicus Digital Elevation Model) and geophysical local parameters. In a second step, the theoretical model is tuned by shore evolution timeseries based on optical and SAR data (for example Copernicus Sentinel 1 & 2). A third step is applied to fine tune the model, with water level altitude calibration thanks to specialised altimetry missions dedicated to water bodies (for example CNES SWOT).

Once a collection of water bodies has been modelled, the proposed service is ready to automatically monitor this collection. This systematic monitoring of water bodies is achieved through optical and SAR data (including Sentinel 1 &2), with weekly revisit, monitoring both surface and volume. The service also provides past years analysis, and real-time monitoring and alerting.

This first service level addresses mainly water management institutions and all kinds of water resources users.


Despite not being in all the regions the main user of water, agriculture is usually the main consumer if it, meaning that water used by agriculture is not reinjected in the cycle of available water. In the scope of a region wide monitoring, the water bodies monitoring service can be combined with Earth Observation based irrigation detection. It is eased by region providing crop parcels inventory, and if not, this requires a parcel detection process, as a preparation phase. This service allows the correlation of water bodies volume variation, previous years rainfall in the area (for example based on ERA5 data) and evolution of irrigated surfaces among cultivated areas.

In a further step, Atos is planning to add a forecast level to the service, making use of Destination Earth digital twin data. DestinE will provide an unprecedent data source of temperature and rainfalls evolution forecast, and the ability to calculate their impact on available water volumes and irrigation needs.

This second service level addresses more specifically agricultural regulation institutions and farmers’ crop strategy.


The goal of this poster is to sketch the proposed water resources monitoring process, detail the expected results and advantages, and highlight the challenges still to overcome.