Modelling the effects of land use/cover change and rainfall variability on landslide hazards: the case of Nyabihu district, Rwanda

Abstract

Landslides have become the environmentally recognized hazard in hilly regions of Rwanda such as Nyabihu district. They are often characterized by the downslope movement of debris or other earth materials which damage or destroy everything found in their way such as infrastructures, croplands, and even cause a number of deaths. The intense rainfall has been noticed as the main trigger of landslides in Rwanda, together with land use/cover change. Therefore, the objectives of this study were; to assess the land cover change effects on landslide occurrences, evaluate the rainfall variability and its effects on landslide occurrences, and predict the occurrence of landslides in the study area. Land use/cover maps of 2005 and 2015 were generated and overlaid with mapped landslides. Maximum likelihood classification was used to classify the Landsat satellite images, and Mann Kendall test was used to assess the rainfall trends. The results revealed a remarkable decrease of agricultural land, while all other land use/cover types have increased along the mentioned period. It was noted that most of the landslides occurred in agricultural land. Also, areas with high rainfall were noted to have experienced more landslides than those with low rainfall. Despite the relation of rainfall to landslide occurrences, the rainfall variability over a period of time did not always correspond to the variation in landslide occurrences. The study also indicated the influence of controlling factors (such as slope, soil depth, and distance to road) on landslide occurrences. The occurrence of landslides was also predicted using logistic regression model. The model showed that an increase in slope angle increases the chances to landslide occurrences, while the changes in land use/cover, and rainfall do not necessarily imply the increase in landslide occurrences, though they significantly relate to landslide occurrences. The study results are expected to be useful for alerting landslide hazard management decisions, land use planning and management regulations so as to minimize the likely landslide occurrences and their resultant impacts.