Modelling probability of rainfall-induced shallow landslides in a changing climate, Otta, Central Norway

We analyse slope stability conditions for shallow landslides under an extreme precipitation regime with regard to present and future scenarios, in order to first study the effect of changes in precipitation on stability conditions, considering uncertainty in the model parameters, and second to evaluate which factors contribute the most to model output and uncertainty. We used a coupled hydrological-stability model to study the hydrological control on shallow landslides in different precipitation regimes, with reference to the case study of Otta, located in central east Norway. We included a wide range of climatic settings, taking intensity, duration of the extreme events and two different antecedent precipitation conditions into account. Eleven future scenarios were determined using results of down-scaled meteorological models. Considering the uncertainty in the soil parameters, we used the Monte Carlo approach and probability of failure resulting from 5,000 trials was calculated for each precipitation scenario. In unstable areas the probabilities of failure at present and future conditions were compared using a bootstrapping method. Sensitivity analysis was carried out to understand how variations in input parameters influence the output of the selected model. The results show changes in the modelled stability conditions only if the effect of antecedent precipitation is not taken into account. The uncertainties in the predicted extreme precipitation events, soil parameters, and antecedent precipitation conditions do not allow any accurate estimation of changes in stability conditions for shallow landslides.