Inland radar altimetry for intermediate scale water bodies with nadir specular echoes and a constellation of small satellites

We previously reported (COASTALT Workshop 8, Lake Konstanz, 2017; 25-years..., Ponta Delgada, 2018) on the prevalence of quasi-specular echoes from intermediate scale (50m to 250 m) inland water bodies. Specularity is a boon and bane for Inland radar altimetry. The boon is that specular echoes allow very precise range measurements with coherent 'zero-Doppler' summing of a few individual echoes (IEs). (Accuracy of <1 cm was shown with Envisat IEs.) Also, land interference is virtually eliminated. The bane is that such water echoes can only be detected when the satellite nadir is directly over the water surface. The echoes SNR fall perceptibly squinting off nadir. The implication is that with current large spacecraft altimeters only a very small fraction of water bodies are encountered. Also, water level slope (discharge) cannot be measured. Radar altimetry cannot fulfill the hydrologists' data requirements. Inland altimetry requires a paradigm shift away from large spacecraft. A wide range of Earth observations and other LEO uses are already migrating to swarms of small satellites (ESA APIES, NASA RainCube , and SpaceX Starlink). The same idea is appropriate for inland altimetry where specular echoes reduce the required radar power 20-40 dB and so much smaller satellites can be used. A constellation of 1000 can provide 100% geographical coverage, water levels and discharge, at 10-day repeat cycles. We used several Sentinel 3A/B data sets to evaluate water level measurements with specular echoes and extrapolate to smaller satellites. First set is of the Arno River which is 100 m wide, just wide enough to have one full S3 burst over water. Water levels over three years (45 revisits) are compared with river gauges maintained by Centre for Meteorological and Hydrological Monitoring, Tuscany Region, Italy. A second data set is multiple passes over Salar de Uyuni compared with surface topography measured by Borsa et al 2008. And a third data set of passes over 800 m wide Silsersee Lake (Switzerland) were there is no ground truth but multiple bursts provide inter-data accuracy and further insight on application to larger water bodies. We degrade SNR on these data sets to extrapolate performance with small sat radars. Details on the algorithms used in this analysis will be provided and illustrated in a companion poster presentation.