How far are we from the use of satellite rainfall products in landslide forecasting

摘要: Abstract Satellite rainfall products have been available for many years (since '90) with an increasing spatial/temporal resolution and accuracy. Their global scale coverage near real-time perfectly fit the need of early warning landslide system. Notwithstanding these characteristics, number studies employing satellite estimates predicting events is quite limited. In this study, we propose a procedure that allows us to evaluate capability different forecast spatial-temporal occurrence rainfall-induced landslides using thresholds. Specifically, assessment carried out in terms skill scores, receiver operating characteristic (ROC) analysis. The applied ground observations four estimates: 1) Tropical Rainfall Measurement Mission Multi-satellite Precipitation Analysis, TMPA, real time product (3B42-RT), 2) SM2RASC obtained from application SM2RAIN algorithm Advanced SCATterometer (ASCAT) derived soil moisture (SM) data, 3) Estimation Remotely Sensed Information Artificial Neural Network (PERSIANN), 4) Climate Prediction Center (CPC) Morphing Technique (CMORPH). As case consider Italian territory which catalogue listing 1414 period 2008–2014 available. Results show underestimate respect observations. However, by adjusting thresholds, are able identify occurrence, even though less accuracy than ground-based Among products, CMORPH performing best, differences small. This result be attributed high CMORPH, good SM2RSC. Overall, believe might important additional data source developing continental or systems.