Landslide Susceptibility Mapping in a Geologically Complex Terrane: A Case Study From Northwest Mindoro, Philippines
J. T. Padrones, N. T. Ramos, C. B. Dimalanta, K. L. Queaño, D. V. Faustino-Eslava, G. P. Yumul, Jr., and K. Watanabe (pp. 25-44)
Abstract
Northwest Mindoro is underlain by inherently weak materials that are the product of arccontinent collision involving the Philippine Mobile Belt and the Palawan Microcontinental Block. Consequent to having an arc-continent collision setting, the frequent exposure to precipitation and the human-induced pressure on the environment is the occurrence of landslides in the island. Because of the complicated tectonic makeup of the area and the complex controls to slope failures, a variety of approaches to landslide susceptibility assessment was used and compared. Using a geographic information system-based platform and the weighted overlay method, three different grading techniques were investigated for their reliability in quantifying landslide susceptibility in northwest Mindoro. The first set of overlay analysis involved the assignment of weights to the different factors based on subjective, expert knowledge. Application of the analytic hierarchy process (AHP) was utilized in the generation of two other susceptibility models in which different sets of factor weights were assigned. Parameters that influence slope failure such as lithology, slope gradient, elevation, distance from faults and lineaments, land cover, distance from historical earthquakes, and slope aspect were assigned susceptibility ratings using two methods: One set assumed progressive increases in landslide susceptibility equivalent to increases with typical value scales (e.g., steepest and highest slopes are most susceptible to slope failure), and the other set utilized expert knowledge based on field observations of active and inactive landslides. Generated landslide susceptibility maps generally show zones of highest susceptibility to landslides coinciding with areas that are underlain by old and highly weathered sedimentary sequences (i.e., Lasala Formation), while geomorphological constraints such as slope gradient and slope aspect appear to have minimal contribution to slope failure. Results further reveal the importance of expert opinion, especially for lithological factors, in the generation of the most appropriate landslide susceptibility map for a particular region. A comparison of three susceptibility models shows that the combination of the AHP method with factor ratings derived from expert knowledge produced a susceptibility map that closely approximates the actual distribution of landslide occurrences observed. The usefulness of these landslide susceptibility maps, particularly for science-directed land use planning, may be validated with continued monitoring and inventory of landslides in northwest Mindoro Island.