Cell-based methodologies for rainfall-induced landslide and debris-flow risk assessment and early warning

The 12 May 2008 Wenchuan earthquake triggered numerous landslides. A large number of additional slope failures and debris flows were triggered by storms in the last several years after the earthquake. Since a large amount of loose materials is still retained on the steep terrains, widespread slope failures and debris flows are likely to occur in the future. Hence, appropriate methods are needed for hazard analysis, risk assessment, and early warning for regional rainfall-induced slope failures and debris flows. The risks to travellers posed by slope failures and debris flows along an 18 km section of Provincial Road 303 (PR303) near the epicentre of the Wenchuan earthquake are considered in this thesis research.;Numerous hazardous soil deposits are distributed in the study area at high elevations up to 2000 m above the road. A real-time distributed risk-based early warning system based on monitored rainfall by rain gauges is developed to safeguard the travellers on the road. The system accommodates several types of landslide hazards, particularly uncertain and inaccessible hazards, and is robust and cost-effective. The early warning system comprises six components; namely, a digital terrain module, a spatial rainfall distribution module, a slope failure prediction module, a debris flow simulation module, a multi-hazard quantitative risk assessment module, and a warning message issuing module. Given monitored rainfall data, the temporal and spatial response of the digital terrain to rainfall can be analysed by the slope prediction module and the debris flow simulation module. The multi-hazard quantitative risk assessment module evaluates the temporal and spatial risks posed by the slope failures and debris flows. The 18 km road is divided into 18 sections, one kilometre each, and each section is a warning entity. The warning level for each road section is determined based on the sectional average individual risk.;A physically-based distributed cell model for predicting regional rainfall-induced slope failures and runout on three-dimensional digital terrain is developed in this thesis research, which forms the slope failure prediction module for the early warning system. The model can predict the locations, volumes, runout traces, and deposition locations of the slope failures efficiently. Since soils are layered at some locations, an infiltration analysis module for slopes with two-layer soils under arbitrary rainfall conditions is developed. Empirical equations that are developed based on a slope failure inventory in the study area are adopted as runout cessation criteria. The size effect of slope failures and uncertainties in shear strength parameters and the runout process are considered in the runout prediction.;A new depth-integrated numerical model for simulating debris flow (EDDA) is developed, which forms the debris flow simulation module for the early warning system. The model considers the changes in debris flow density, yield stress, and dynamic viscosity, as well as the influences of such changes on the runout characteristics of the debris flow. The yield stress of the debris flow mixture estimated using the Mohr-Coulomb equation is suitable for any of clear water flow, hyper-concentrated flow, and fully developed debris flow. A variable time step algorithm is developed to assure both numerical stability and computation efficiency. The model is able to simulate two-dimensional dam-break water flows, confined debris flows in a channel, unconfined debris flows in a flat area and catchment-scale debris flows considering erosion, deposition and property changes.;A multi-hazard quantitative risk assessment (QRA) method for slope failures and debris flows has been developed, which forms the multi-hazard quantitative risk assessment module for the early warning system. The method quantitatively assesses the risk posed by regional slope failures and debris flows and the influence of slope failures on debris flows. Both individual risk and societal risk are assessed. The integrated QRA for slope failures and debris flows also considers the scenario of a location being impacted by several slope failures or debris flows or both. The method is applied to the study area and verified by the slope failures and debris flows triggered by the 13 August 2010 storm. The established early warning system is able to cover the 165 km2 study area efficiently, and proves to be a valuable tool for mitigating the risks posed by regional slope failures and debris flows.