On the use of SRTM and altimetry data for flood modeling in data-sparse regions

The growing availability of remotely sensed data has fostered the implementation of hydraulic modeling in poorly gauged regions. However, these applications suffer the lack of knowledge of river bathymetry, which cannot be directly inferred from satellite instruments. This study explores the possibility to set up, calibrate, and validate a hydrodynamic model which geometry is based on global and freely available satellite data. First, the study tests two different procedures for inferring the river bathymetry under the water surface level. Second, focusing on a Po River stretch of ∼140 km (Northern Italy), the study further assesses the suitability of spaceborne topographic and remotely sensed altimetry data (i.e.; ERS-2 and ENVISAT) for implementing and calibrating hydrodynamic models. Referring to 90 m SRTM (Shuttle Radar Topography Mission) digital elevation model for the representation of the riverbed morphology, the work analyzes the performances of different 1-D numerical models which cross sections are modified according to two approaches: (1) Channel Bankfull depth (CB) and (2) Slope-Break (SB) approach. The calibration and validation processes are performed by referring to extended altimetry time series (∼16 years of data), while the accuracy and trustworthiness of 1-D models are tested with reference to a quasi-2-D model based on detailed geometry data. Results show that both CB and SB approaches enhance the performance of SRTM-based models. In particular, the SB approach is completely based on satelliteborne data and shows Nash-Sutcliffe efficiency, MAE, and RMSE values similar to those obtained with the benchmark model.