Assessing Train - Elephant Collision Risk and Identifying Optimal Mitigation Areas Using GIS Approach

Train-elephant collisions have become to be a major threat to both wildlife conservation and railway safety in Sri Lanka specially in major railway lines that are located around the elephant habitats. This study aims to map and analyze the risk areas of train-elephant collisions along the Maho-Trincomalee railway line using Geographic Information System techniques. Multiple spatial factors including land use/land cover, elevation, rainfall, proximity to water bodies, proximity to identified elephant habitats and train speed were integrated into the analysis. Analytical Hierarchy Process (AHP) based weighted overlay and fuzzy overlay methods were used to generate elephant prone area maps. These maps were overlaid with the train speed data to produce final collision risk zone maps. The results of both approaches were spatially similar with most high-risk zones found in the locations which are favorable habitat conditions areas that overlap with high train speed. Assessment of accuracy based on historical collision records as well as the ROC analysis showed that the fuzzy overlay method was slightly better than the AHP method in recognizing high risk zones. The results indicate key hotspots areas where mitigation strategies such as speed limit, early warning devices and implementation of safe crossing corridors can be given priority to minimize the occurrence of collisions. This study proves that a multi criteria GIS analysis integrated with habitat and railway data can be used to aid in evidence-based decision making in elephant conservation and railway management in Sri Lanka.