H23O-2110 - A multi-mission satellite altimetry water surface elevation monitoring network in the Zambezi

Catchment-scale hydrological models are essential tools in water resources management and flood prediction. They are also data demanding, as reliable observations are key to set up, calibrate and validate useful simulations. However, the availability and accessibility of in-situ observations is declining in part due to difficulties in setting up and maintaining dense monitoring networks.

Spaceborne observations of terrestrial water bodies can bridge the gap between data availability and demand. In this study, we investigate water surface elevation (WSE) observations in the Zambezi basin from the two European Space Agency (ESA) altimetry missions Sentinel-3 and CryoSat-2 and their use in hydrological studies.

Both missions carry high-resolution Synthetic Aperture Radar (SAR) altimeters offering a unique opportunity to establish spatially dense, basin-scale monitoring networks driven by operational and high-resolution remote sensing WSE observations. The Sentinel-3 ground tracks cross the Zambezi and its tributaries at 333 individual locations, virtual stations (VS), with a return period of 27 days.

Hydrological seasonal trends are captured clearly at several VS and the root mean square deviation (RMSD) between the satellite and in-situ WSE is low compared to previous similar studies (root mean square deviation, RMSD, of 29 cm), proving the observations can be used as a reliable supplement to the in-situ network.

CryoSat-2 operates in a geodetic orbit with a 369-day repeat cycle resulting in a uniquely dense groundtrack pattern. The high spatial resolution holds valuable information about parameters such as riverbed elevation and roughness along the full river longitudinal profile, which are key to producing hydrodynamic simulations but difficult to measure otherwise.

By using water masks from Sentinel-1 SAR images, we extract thousands of observations covering the entire river network as well as a wide range of reservoirs. RMSD from observed WSE is down to 2.5 cm for reservoirs smaller than 100 km2. A combined Sentinel-3 and CryoSat-2 monitoring network can provide a reliable and data dense expansion to existing monitoring efforts in the Zambezi.

The data has numerous applications, including hydrological and hydrodynamic model calibration/validation and reservoir monitoring and identification.