Abstract's details
Reservoir monitoring based on SWOT products, comparison with imaging and altimetric multi-mission approach: applications to the Grand lacs de Seine reservoirs (France)
Event: 2025 SWOT Science Team Meeting
Session: Hydrology: SWOT Lakes, Estuaries and Wetlands (SLEW)
Presentation type: Poster
Reservoirs are key tools in the management of water resources. They provide a means of reducing the effects of inter-seasonal and inter-annual flow fluctuations, thereby facilitating water supply, flood control, power generation, recreation and other water uses. For more than 20 years, satellite radar altimetry has been an effective technique for monitoring variations in the elevation of continental surface waters, such as inland seas, lakes and reservoirs, rivers and, more recently, wetlands.
This study presents the capabilities of SWOT to monitor water surfaces, heights and variations in water stocks. It also highlights the contribution of multi-sensor synergy and the combination of current imaging and altimetry satellites.
The demonstration of this potential is ongoing over the largest French reservoir, the Lac du Der. With 48 km2, the Lac du Der-Chantecoq, also known as the Marne reservoir, is the largest man-made lake in mainland France. Together with the Orient, Amance and Auzon lakes, it is part of the system of large Seine lakes intended to protect Paris from flooding. Water is drawn from the Marne from November to June, filling the reservoir. From July to October, water is released to support the flow of the rivers. As a result, the water surface area changes considerably throughout the year, from about forty square kilometers during the high-water period to less than ten square kilometers during the very low-water period. In addition to its specific dynamics, this lake is tracked by several altimetry satellites.
Regarding SWOT data, this reservoir is theoretically located in the No Acquisition diamond area, but the signal from this large water body is clearly visible in the SWOT PIXC, the Lake SP products and the SWOT NADIR track that passes over the lake. An analysis of the heights measured by SWOT sensors was carried out both on the PIXC and the Lake SP products over a period of more than one year, from July 2023 to mid-September 2024. Both types of products reproduce well the hydrological evolution of the central reservoir, with a decrease in water levels from September to December, then an increase in levels with water storage and a maximum reached during the summer. These SWOT water level values were compared with in-situ data, to validate the observed phenomenon. For this period, an RMSE of 0.09 m and at 1 sigma 0.30 m were obtained.
Lac du Der is also tracked by the Sentinel-6, since March 2021 with a ten-day frequency. These Sentinel-6 data were first processed using FFSAR to achieve fine resolution of the radargram in the along-track direction, and then a retracker specially designed for this focused data was applied to estimate the water height. Comparison of the in-situ water level data and the relative Sentinel-6 data over 14 months of data shows an impressive consistency of the data, with a RMS of 0.02 m (Median and 1 sigma at 0.03 m). A detailed analysis of the Sentinel-6 data over time also revealed a specific phenomenon, reflecting a reservoir management strategy involving frequent alternation between water release and storage, linked to the needs associated with the Paris Olympic Games.
Another part of this study focuses on evaluating the combination of Sentinel-2 surfaces and altimetry data from SWOT, Sentinel-6 and ICESAT-2 missions to generate a surface/height hypsometric curves for the lake, and to monitor volume variations. These results were also compared with in-situ volume variations provided by the lake operators.
The results obtained illustrate the strong current capabilities of altimetry satellite data such as SWOT and Sentinel-6, combined or not with optical data from Sentinel-2, to access water surfaces, heights and volumes over relatively small reservoirs.
Back to the list of abstractThis study presents the capabilities of SWOT to monitor water surfaces, heights and variations in water stocks. It also highlights the contribution of multi-sensor synergy and the combination of current imaging and altimetry satellites.
The demonstration of this potential is ongoing over the largest French reservoir, the Lac du Der. With 48 km2, the Lac du Der-Chantecoq, also known as the Marne reservoir, is the largest man-made lake in mainland France. Together with the Orient, Amance and Auzon lakes, it is part of the system of large Seine lakes intended to protect Paris from flooding. Water is drawn from the Marne from November to June, filling the reservoir. From July to October, water is released to support the flow of the rivers. As a result, the water surface area changes considerably throughout the year, from about forty square kilometers during the high-water period to less than ten square kilometers during the very low-water period. In addition to its specific dynamics, this lake is tracked by several altimetry satellites.
Regarding SWOT data, this reservoir is theoretically located in the No Acquisition diamond area, but the signal from this large water body is clearly visible in the SWOT PIXC, the Lake SP products and the SWOT NADIR track that passes over the lake. An analysis of the heights measured by SWOT sensors was carried out both on the PIXC and the Lake SP products over a period of more than one year, from July 2023 to mid-September 2024. Both types of products reproduce well the hydrological evolution of the central reservoir, with a decrease in water levels from September to December, then an increase in levels with water storage and a maximum reached during the summer. These SWOT water level values were compared with in-situ data, to validate the observed phenomenon. For this period, an RMSE of 0.09 m and at 1 sigma 0.30 m were obtained.
Lac du Der is also tracked by the Sentinel-6, since March 2021 with a ten-day frequency. These Sentinel-6 data were first processed using FFSAR to achieve fine resolution of the radargram in the along-track direction, and then a retracker specially designed for this focused data was applied to estimate the water height. Comparison of the in-situ water level data and the relative Sentinel-6 data over 14 months of data shows an impressive consistency of the data, with a RMS of 0.02 m (Median and 1 sigma at 0.03 m). A detailed analysis of the Sentinel-6 data over time also revealed a specific phenomenon, reflecting a reservoir management strategy involving frequent alternation between water release and storage, linked to the needs associated with the Paris Olympic Games.
Another part of this study focuses on evaluating the combination of Sentinel-2 surfaces and altimetry data from SWOT, Sentinel-6 and ICESAT-2 missions to generate a surface/height hypsometric curves for the lake, and to monitor volume variations. These results were also compared with in-situ volume variations provided by the lake operators.
The results obtained illustrate the strong current capabilities of altimetry satellite data such as SWOT and Sentinel-6, combined or not with optical data from Sentinel-2, to access water surfaces, heights and volumes over relatively small reservoirs.