Little work has been done on the effect of the pixel neighborhood information when modeling landslide susceptibility using Multiple Logistic Regression (MLR). This research uses in situ and neighborhood cartographic information to evaluate how pixel distance of sampling sites affects the precision and accuracy of the MLR landslide susceptibility model. Two landslide susceptibility models are used: MLR-in situ, calibrated and validated by using variables that are collected at the site of the sampling point; and MLR in combination with Continuous Neighborhood Spatial Analysis (CNSA) to incorporate a search radius to extract pixel values for each cartographic variable based on a distance ratio. La Ciénega watershed on the eastern flank of the volcano Nevado de Toluca is selected as a study area. Its climate, topography, geomorphology, and geology predispose it to episodic landslides. The resulting susceptibility maps are validated in terms of the area under the curve (AUC) of the receiver operating characteristic (ROC), and they are compared with an inventory map in a contingency table; the MLR-CNSA model yields the better spatial prediction and representation of landslide susceptibility. The AUC evaluation indicates a predictive capability for the MLR-CNSA model of 0.969.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.