Integrating rainfall severity and soil saturation indices to define hydro-meteorological thresholds for landslides

Abstract

Rainfall thresholds identifying the meteorological conditions critical for landslides triggering are widely used within operational territorial landslide early warning systems (Te-LEWS). Recent studies demonstrated that hydro-meteorological thresholds, combining soil hydrological information and rainfall, can improve the prediction of landslide occurrences at territorial scale. Soil moisture is predominantly used to characterize the soil wetness conditions that predispose slopes to failure. In this study, we develop hydro-meteorological thresholds by investigating the potential use of both antecedent (i.e., before the beginning of the rainfall event) and triggering (i.e., during the rainfall event) saturation variables derived from a reanalysis product. A procedure based on a Bayesian probabilistic analysis of rainfall severity and soil saturation indices is designed and tested in an area of Campania region, southern Italy. Different hydro-meteorological thresholds demonstrate a good predictive capability, with those considering maximum saturation at the uppermost soil layer performing best. Overall, this study proves that hydro-meteorological thresholds employing antecedent and triggering saturation variables derived by time series analysis can improve the prediction of landslide occurrences at territorial scale.