Xiaohui Qi, Joao Mendes, Rupert Bainbridge , Stuart Dunning, Mike Winter, Michael Lim

Shallow landslides are a major natural hazard in the UK, causing more than £10M of economic losses annually and posing clear threats to life. Rainfall is an essential control on shallow landslide risk, but network operators currently use generic ‘on-off’ warnings based on localised rainfall intensity and duration, giving no strategic information about where the hazard is highest or evolving the quickest for any particular rainfall event. A new procedure has been developed to automatically evaluate the stability of slopes across regional scales in response to spatially and temporally variable rainfall events, quantified by rainfall radar data updated every 5 minutes. A physically based programme, TRIGRS, was used to assess the factor of safety (FS) of slopes in response to spatially variable rainfall radar data. The dynamic FS maps produced provide early warning and informed decision-making for the management of regional to national-scale infrastructures. Sensitivity analyses were performed to investigate the effect of the soil thickness, shear strength parameters, hydraulic parameters, and antecedent rainfall on the FS during a known landslide-causing rainfall event in Glen Croe, western Scotland. The results specifically highlight the slope where the failure occurred as a high-hazard area 1.0 hour before the event. The dynamic FS maps were particularly sensitive to soil thickness, the saturated hydraulic diffusivity and Gardner’s unsaturated conductivity, suggesting that more effort should be put into improving network-scale datasets of these parameters. Ultimately, the ability developed here to account for near-real-time spatial variabilities in rainfall data and slope responses measured through the relative change in FS values provides a potentially transformative new tool for network operators to proactively mitigate the impacts of specific storms as they develop.