Physically based shallow translational landslide susceptibility analysis in Tibo catchment, NW of Portugal

Abstract

The Northwest of Portugal is a steep slope mountain area where high amounts of rainfall (3,000 mm/year) occur especially on winter, promoting high natural propensity to slope failure, mainly shallow landslides. This article aims to assess shallow landslide susceptibility, in Tibo catchment, Serra da Peneda, North of Portugal, using two physically based models—SHAllow Landslide STABility (SHALSTAB) and Safety Factor (SF)—applying a set of mechanical and hydrological parameters, assessed in situ and laboratory testing of soil samples collected in the field, calibrated by back analysis of landslides inventoried in the study area, as well as accurate topographic information derived from a high-resolution digital elevation model (DEM). The validation of results was made using shallow landslide scars, directly inventoried in the field. Both susceptibility model results were validated by scar concentration (SC) and landslide potential (LP). SHALSTAB model was also validated by minimum log q/T. SHALSTAB predicts 50 % of the area to be on unstable classes (log q/T < −2.5), 77 % of the SC on unstable classes and a LP index of 7 and 4.7 % for the two most unstable classes. By minimum log q/T, SHALSTAB predicts 91 % of the scars to occur on unstable classes. Safety factor predicts 47.99 % of the area as unstable, 79.9 % of the SC for unstable classes, and a LP index on unstable classes of 4.63 and 2.77 % on partially unstable class. For the most unstable classes of both models, the greatest values of LP were between 3.5 and 7 %. The simple physically based models used in this study (SHALSTAB and SF) proved to be effective as shallow landslide susceptibility predictors, being in consequence useful tools for municipal planning on landslide hazards, but their application requires, beyond detailed topographical information, good estimates of the mechanical and hydrological soil properties.