Testing the use of single- and multi-mission satellite altimetry for the calibration of hydraulic models

Satellite altimetry is increasingly considered as a valuable source of information in many hydrological and hydraulic applications. However, the accuracy of different sensors adopted for monitoring the water level from satellite and the limited temporal resolution that characterizes each sensor (i.e. revisit time most of time varying from 10 to 35 days; 369 days in case of CryoSat mission) still hamper their common use. Recently introduced multi-mission (MM) densified time series might represent a possible alternative to ensure higher spatial and temporal coverage. Though, a comparison of the potential of different altimetry products, including MM series, for hydrodynamic model calibration is still missing. This study attempts to fill this gap investigating how available altimetry series perform over a stretch of the Po River (nearly 140 km across Northern Italy) in calibrating a quasi-2D model built with detailed topographic information. Specifically, objectives are manifold: i) to provide a comparison of satellite altimetry products available to the research community and commonly used in hydraulic modelling (Envisat, Envisat extended mission, ERS-2, TOPEX/Poseidon, SARAL/AltiKa, Jason-2, Jason-3, Sentinel 3A, Sentinel 3B and CryoSat); ii) to evaluate the performance of MM satellite series in calibrating a hydraulic model relative to single-mission series; and iii) to investigate the importance of the number of observations (series length) for each mission. Results of the model calibration depict a general improvement of satellite performance over time, moving from the oldest to more recent missions, with the exception of Envisat extended series. In general, Jason-2, Sentinel 3A and Sentinel 3B outperform other series both in terms of calibration error and number of measurements required to achieve a reliable calibration. MM series provide errors larger than those obtained from original single-mission time series considered with their overall length, but they provide more reliable calibrations than altimetric time series with low sampling rate (i.e., Envisat, Envisat extended, and SARAL/AltiKa) or those that cover very short periods (e.g. altimetry series limited to 20–40 months in length). The analysis offers additional insights into the possible use of altimetry series in hydrodynamic applications, providing a comparison of different original products and showing the potential, as well as limitations, offered by MM series.