Landslide Susceptibility Mapping Using Machine Learning Methods: A Case Study in Colorado Front Range, USA

Due to the complex nature of landslides, statistically based landslide susceptibility mapping has been widely used to evaluate slope failure risk for landslide-prone areas. This study evaluates the capability of different machine learning (ML) methods for landslide susceptibility mapping (LSM) in mountainous regions in the Colorado Front Range. A well-documented and georeferenced landslide inventory for the Colorado Front Range area was used to construct the database for developing and testing ML models. Nine landslide contributing factors were collected for the present study based on the availability of geophysical data and the type of landslides that occurred in the study area. These landslide causative factors represent hillslope geometries, surface hydrology, and soil conditions. Five commonly used ML models were evaluated: logistic regression (LR), support vector machine (SVM), decision tree (DT), random forest (RF), and gradient boosting machine (GBM). The cross-validation technique was used to evaluate the model performance. All the trained models reflected the relationship between landslides and their causative factors in the study area based on cross-validation results. It was found that the performance varied among the ML models; the RF model exhibited the worst performance due to possible overfitting, and the RF and the GBM models achieved the highest performance. The trained models were subsequently used to predict the landslide susceptibility for the entire study area and generate a landslide susceptibility map. The landslide susceptibility map can provide situational awareness of potential landslide hazards within the Colorado Front Range area and provide guidelines for future decision-making.