Modelling shallow landslide susceptibility by means of a subsurface flow path connectivity index and estimates of soil depth spatial distribution

摘要

Topographic index-based hydrological models have gained wide use to describethe hydrological control on the triggering of rainfall-induced shallowlandslides at the catchment scale. A common assumption in these models isthat a spatially continuous water table occurs simultaneously across thecatchment. However, during a rainfall event isolated patches ofsubsurface saturation form above an impeding layer and theirhydrological connectivity is a necessary condition for lateral flow initiation at a point on the hillslope.Here, a new hydrological model is presented, which allows us to account for theconcept of hydrological connectivity while keeping the simplicity of thetopographic index approach. A dynamic topographic index is used to describethe transient lateral flow that is established at a hillslope element whenthe rainfall amount exceeds a threshold value allowing for (a) developmentof a perched water table above an impeding layer, and (b) hydrologicalconnectivity between the hillslope element and its own upslope contributingarea. A spatially variable soil depth is the main control of hydrologicalconnectivity in the model. The hydrological model is coupled with theinfinite slope stability model and with a scaling model for the rainfallfrequency–duration relationship to determine the return period of thecritical rainfall needed to cause instability on three catchments located inthe Italian Alps, where a survey of soil depth spatial distribution isavailable. The model is compared with a quasi-dynamic model in which thedynamic nature of the hydrological connectivity is neglected. The resultsshow a better performance of the new model in predicting observed shallowlandslides, implying that soil depth spatial variability and connectivitybear a significant control on shallow landsliding.