Evaluation of satellite rainfall estimates for hydrologic and geo-hydrologic hazards prediction

Satellite rainfall data are becoming available at high temporal and spatial resolution. The use of such data in ad-hoc developed modelling can extend hydrological applications to poorly instrumented areas. The use of remotely sensed information can provide also valuable insights about rainfall spatial variability, improving ground rainfall estimation. Despite the great availability of data, their use in hydrological practice is still very rare, with just few studies that have used such data for landslide hazard risk mitigation. In order to test the feasibility of using this kind of data for landslide prediction, a coupled hydrological-slope stability model has been developed and forced with remotely sensed information. Before that, several satellite rainfall products have been assessed through direct comparison with modelled and ground –based benchmarks and through hydrological validations, in order to highlight advantages and drawbacks of this source of information. Preliminary results showed that satellite rainfall data can be used with benefit in natural hazards mitigation, although the inherent errors related to the indirect nature of measurement have to be deeply assessed and corrected.