A machine learning-based prediction system for rainfall-induced landslides in Benguet First Engineering District is proposed to address the landslide risk due to the climate and topography of Benguet province. It is intended to improve the decision support system for road management with regards to landslides, as implemented by the Department of Public Works and Highways Benguet First District Engineering Office. Supervised classification was applied to daily rainfall and landslide data for the Benguet First Engineering District covering the years 2014 to 2018 using scikit-learn. Various forms of cumulative rainfall values were used to predict landslide occurrence for a given day. Following typical machine learning workflows, rainfall-landslide data set was divided into training and testing data sets. Machine learning algorithms such as K-Nearest Neighbors, Gaussian Naïve Bayes, Support Vector Machine, Logistic Regression, Random Forest, Decision Tree, and AdaBoost were trained using the training data sets, and the trained models were used to make predictions based on the testing data sets. Predictive performance of the models vis-a-vis the testing data sets were compared using true positive rates, false positive rates, and the area under the Receiver Operating Characteristic Curve. Predictive performance of these models were then compared to 1-day cumulative rainfall thresholds commonly used for landslide predictions. Among the machine learning models evaluated, Gaussian Naïve Bayes has the best performance, with mean false positive rate, true positive rate and area under the curve scores of 7%, 76%, and 84% respectively. It also performs better than the 1-day cumulative rainfall thresholds. This research demonstrates the potential of machine learning for identifying temporal patterns in rainfall-induced landslides using minimal data input -- daily rainfall from a single synoptic station, and highway maintenance records. Such an approach may be tested and applied to similar problems in the field of disaster risk reduction and management.

Rock mass strength is used to characterize the influence of geology on the topography of the Can-abag mountain range in Southern Leyte, Philippines. A geologic model comprised of Jurassic to Cretaceous ultramafic rocks (UM), pre-Miocene fine clastic rocks (FC1), Middle Miocene limestones (LS1), Late Miocene to Late Pliocene coarse clastic rocks (CC), Late Miocene to Early Pliocene fine clastic rocks (FC2), and Late Pliocene limestones (LS2) is proposed for the study area. Five litho-structural units are derived from this model for use in the succeeding discussion: coarse clastic rocks west and east of the Can-abag ridge (CCw, CCe), fine clastic rocks (FC), limestones (LS) and ultramafic rocks (UM). FC and UM, both having weak rock mass strengths, are prone to isotropic slope failures, and are characterized by low-gradient, low curvature terrain. LS has high rock mass strengths, but because of limited topographic relief development, exhibits low-gradient, low curvature terrain similar to the weak rock masses. CCw and CCe both have high rock mass strengths and are characterized by high gradient, high curvature terrain. The former has slightly lower gradient and higher curvature than the latter due to the dominance of channelized transport processes. The latter, on the other hand, is influenced by steeply-dipping planar discontinuities which facilitate episodic massive landslides, which in turn reduce the curvature of the underlying unit. The weak rock mass strength of FC is mainly due to the inherent weakness of the fine-grained intact rock. In contrast, UM, which has high intact rock strengths due to the crystalline texture, is reduced to a weak rock mass by numerous discontinuities attributed to the long history of tectonic deformation. CCw and CCe, while having low intact rock strengths just above that of FC, have higher rock mass strengths because of good discontinuity conditions. LS has high rock mass strengths because of high intact rock strengths and good discontinuity conditions. Regional geologic structures such as bedding and fault planes provide potential conditions for planar slope failures within the rock masses. However, failures are spatially limited by the geometric relationships between the geologic structures and the existing topography. Steeply-dipping structures have direct significant contributions to massive landslides in terrains underlain by materials with strong isotropic rock mass strengths, such as in CCe. The geologic environment – lithology, structures, stratigraphy, tectonic history – provides the material base on which surface processes act on to create the present landform. Rock mass strength, as a geomechanical manifestation of the geologic environment and as a physically-based and quantifiable concept, provided a good framework for understanding and explaining the spatial variability of topographic characteristics in relation to the geologic materials of the Can-abag mountain range in Southern Leyte, Philippines. Some remarks regarding the use of gradient distributions in rock mass strength characterization, the use of slope gradient in landslide susceptibility analysis, and frequency-magnitude of landslides are put forward.