An Open Source Landslide Model for Slope Stability Analysis Integrating Remote Sensing and Geotechnical Data

Abstract

Computer-based landslides models are employed on large regions for the assessment of slope failures. However, these models have specific operating framework strategies for their configuration and input data with long processing time. To address these concerns, a python-based slope stability model has been developed with a modular framework that integrates parametrized geotechnical data and Digital Elevation Model (DEM) to compute the Factor of Safety (FoS) and Failure Probability (Pf). The model mechanics couples the classic infinite slope equation based on Monte Carlo Simulations which allows generation of probability density functions representing all random variables and probability of failure by repeated calculations. For each node, multiple FoS and Pf values are computed and mean values for each node provide predictive susceptibility. For a deeper insight elevation, slope and aspect maps are likewise produced. The model was calibrated and validated for a test site in Italy and a comparison was made with other opensource models in terms of efficiency and accuracy. The developed model tested with the upper and lower limits of parameterized dataset showed an accuracy in the range of 79% to 88% and the results have shown that the developed model provides a proportionate level of prediction accuracy and processing efficiency. A component with the capability of spatial distribution of infiltration process would further improve its accuracy.