Coupling hydrological and geotechnical models for enhanced flood–landslide cascading disaster modelling

Flash floods, fluvial floods and shallow landslides triggered by intense rainfall present substantial threats to both human lives and infrastructure. Furthermore, floods and landslides often manifest in a cascading sequence, where an initially lower-consequence event like heavy rainfall can lead to more severe floods and/or landslides, intensifying the impact to affected communities. Losses resulting from these combined hazards may be significantly greater than the sum of losses from individual hazards. Therefore, there is a crucial need to integrate hydrological and geotechnical modelling into an integrated flood–landslide cascading preparedness and hazard management. This research introduces a coupled flood and landslide initiation modelling system, integrating a temperature index-based snowmelt model (SNOW-17), the Coupled Routing and Excess STorage (CREST) model, and the Fast-Shallow Landslide Assessment Model (FSLAM). The proposed approach is evaluated in the Val d’Aran region that experienced multiple landslides and important flooding due to a combination of heavy rainfall and snowmelt in June 2013. The coupled-model involves three main steps: i) The SNOW-17 model is applied to quantify the snow melting process which is further included in ii) the hydrological model CREST in order to estimate soil water content conditions, discharge and flood extent. Later, iii) the FSLAM model generates landslide susceptibility maps based on the hydrological model outputs, and finally iv) a random walk runout model will determine the landslides trajectories and the amount of sediment that may reach the river network. Preliminary results, related to snow, hydrological and landslide model calibration, have shown good statistical performance when comparing modelled daily soil water equivalent and daily hydrographs with observations from 2012-2020. Landslide predictions also showed a good accuracy (72%). Further steps will try to include the cascading effect of sediments being delivered to drainage network during landslides episodes. This study highlights the importance of the physical connection among snow melting, hydrological processes and slope stability, and aims to provide a prototype model system for operational forecasting of floods and landslides.