Application of geospatial tools for landslide hazard assessment for Uganda

Abstract

The need to predict possible occurrence of landslides is increasingly becoming a concern of governments and humanitarian bodies in developing countries. The occurrence of landslides and other related disasters in Uganda requires a shift from the current disaster management approach where interventions come into play following the disaster, towards a better approach where mitigation measures are planned before the disaster. In Uganda, studies that utilize spatial technologies to develop tools to support landslide disaster prediction have been limited by the problem of spatial data suitability and availability. This paper explores the suitability of the available spatial datasets as inputs into GIS-based landslide risk assessment in Uganda. The datasets used in this study included digital elevation model, soils, precipitation, vegetation cover and population. The relative importance of datasets was established by a combination of literature review, expert opinion and pair wise comparison technique. Through GIS tools, a prediction map was generated that showed risk levels of various areas in Uganda. The results from GIS analysis showed that the areas of highest risk included mountain slopes associated with high rainfall and clayey soils. The predicted high risk areas coincided with areas where landslides have recently occurred. On the basis of this, it was concluded that the current spatial datasets in Uganda, combined with data in the public domain could still be useful in providing reasonable predictions about landslides at a national level. The results of this research demonstrate the need to incorporate the use of geospatial tools in Uganda’s disaster management strategies.