Assessing the accuracy of satellite radar altimetry for inland water level monitoring

Inland water monitoring is essential to quantify the available and accessible water resources and to assess the amount that is useful for meeting human needs. It plays an essential role for both society and the environment, and due to the many problems associated with ground hydrological monitoring networks, the measurement of inland waters represents a political and economic challenge. A viable alternative for deriving surface water information on a large scale is satellite Earth Observation, and in recent decades satellite altimetry has become an established method for providing water level measurements. The expansion of satellite constellations guarantees a continuous and comprehensive coverage in time and space, but often the harmonisation between different missions is difficult and mostly reduced to a bias correction. For this purpose, in the framework of the FDR4ALT (Fundamental Data Records for Altimetry) project funded by the European Space Agency (ESA), the Altimetry Fundamental Data Records are generated to produce the Inland Water Thematic Data Products based on the exploitation of the measurements acquired by the altimeters on board the ESA ERS-1, ERS-2 and ENVISAT missions. In this paper, we present the results of the FDR4ALT project for rivers, showing the analysis of the different retrackers (Ice1, Ice3, MLE3, TFMRA, Adaptive) over a one-year period. A Round Robin analysis is carried out to evaluate the performance of each retracker, with the final aim of identifying the best one, able to describe the inland water flow, to be implemented at global level. The validation of the derived altimetric water levels is carried out for three large basins in the world by comparing the altimetric data against external measurements obtained from both reliable ground-based measurements and other altimetry sources freely available on the web (Theia Hydroweb, Dahiti, HydroSat). In general, quite accurate long time series of altimetric water levels have been reconstructed from 1993 to 2010 by merging ERS-1, ERS-2 and ENVISAT data.