Assessment of Landslide susceptibility and risk implication to road network in Mt Elgon, Uganda

Abstract Globally landslides occurrence is reportedly frequent particularly in the mountainous regions causing both direct and indirect effects to various sectors including the road transport. Existing literature reveals limited assessment of road vulnerability to landslides in the mountain regions in Africa. The objective of this study was to investigate the risk to different segments of the road network in the Mt Elgon region. A Fuzzy logic model was used to assess and map the landslide susceptibility of the study area. A total of 478 landslide sites were used in the model development. Ten conditional factors were applied for generating the dataset for training and validation of the model. The results reveal that mid to high altitude steep and rugged areas are more susceptible to landslides. The model performance was good as revealed by high Area Under the Curve (AUC) of 83% and thus can be relied upon in landslide susceptibility mapping. The hotspot segments, which are high risk sections of the road network need to be prioritized for monitoring so as to initiate and strength existing risk mitigation strategies.