Volcanic regions with stratovolcanoes and monogenetic fields are very common and have the potential to trigger along their stream systems landslides and debris flows that damage human settlements, industrial development, cattle raising, forestry, and agricultural activities. However, a practical and standardized landslide mapping methodology using GIS for landslides that occur continually along the stream systems in volcanic terrains has not been applied in Mexico. As a result, landslide inventory maps and related geo-databases that support the prediction of future slope instability in volcanic terrains are lacking. Also, little work has been done on the systematic comparison of different models to predict landslide susceptibility. In this study, a landslide inventory map is derived from a representative sample of landslides using multi-temporal aerial-photo-interpretation and field investigations. The stability is modeled using LOGISNET (Multiple Logistic Regression, Geographic Information System, and Neural Network) as a tool to compare and contrasts the advantages and limitations of two landslides susceptibility models: Stability Index MAPping (SINMAP) and Multiple Logistic Regression (MLR). Both models are embedded in LOGISNET to predict landslides and facilitate the analysis. The validations of the resulting susceptibility maps were performed by comparing them with the inventory map in a contingency table. This research uses the stream system of the Rio Chiquito-Barranca del Muerto watershed as a case-study area. The study area is located in the SW flank of Pico de Orizaba, Volcano, Veracruz-Puebla.
Read full abstract