Brief communication: Introducing rainfall thresholds for landslide triggering based on artificial neural networks

Abstract

In this communication we show how the use of artificial neural networks (ANNs) can improve the performance of the rainfall thresholds for landslide early warning. Results for Sicily (Italy) show how performance of a traditional rainfall event duration and depth power law threshold, yielding a true skill statistic (TSS) of 0.50, can be improved by ANNs (TSSg Combining double low lineg 0.59). Then we show how ANNs allow other variables to be easily added, like peak rainfall intensity, with a further performance improvement (TSSg Combining double low lineg 0.66). This may stimulate more research on the use of this powerful tool for deriving landslide early warning thresholds.