Comparative analysis of SHALSTAB and SINMAP for landslide susceptibility mapping in the Cunha River basin, southern Brazil

Abstract

The Shallow Landsliding Stability Model (SHALSTAB) and Stability Index Mapping (SINMAP) models have been applied to various landslide management and research studies. Both models combine a hydrological model with an infinite slope stability model for predicting landslide occurrence. The objectives of the present study were to apply these two models to the Cunha River basin, Santa Catarina State, southern Brazil, where many landslides occurred in November 2008, and perform a comparative analysis of their results. Soil samples were collected to determine the input parameters. The models were calibrated with a landslide scar inventory, and rainfall data were obtained from three rain gauges. A comparison of their results obtained from the models was undertaken with the success and error index. Based on the maps of stability and instability areas for the study basin, the models performed well. Since the initial equations of both models are not particularly different, their results are similar. Locations with steep slopes, as well as areas with concave relief that tend to have larger contribution areas and moisture, have lower stability indexes. SHALSTAB classified only ~13 % of the total area of the Cunha River basin as unstable, while SINMAP classified ~30 % as unstable. The analysis of maps based on the results of the two models shows that if SHALSTAB is correctly calibrated, based on hydrological parameters, its results could be more accurate than SINMAP in the prediction of landslide areas. Although SINMAP showed better calibration of the landslide scars, its classification over the basin results in an overestimation of stability areas. The conclusion is that SHALSTAB is more suitable than SINMAP for the prediction of landslides in the Cunha River basin, Brazil.