Refugee settlement induces accelerated land use/cover change in Northern Uganda

Abstract

ABSTRACTThe establishment of refugee settlements has caused profound segmented impacts on the coverage of land use/cover types. Few studies have computed the spatial drivers of land use/cover changes in Refugee prone areas. This study explored how refugee settlements induced geo-changes in land use/cover in Bidi-Bidi refugee settlement in Uganda. Sentinel-2B images (2015, 2017 and 2020) were used to assess and predict (2030 and 2040) the spatial areal extent of land use/cover changes. The images were classified using Supervised algorithm (Maximum-likelihood) and used CA-Markov model to generate transitions into the future. A Binary Logistic regression was used to compute the spatial drivers of geo-changes. Our results reveal that the settlements triggered more geo-changes in built-up (0.6%), refugee settlements (4.1%), and subsistence farmlands (7.0%) at the expense of woodlands (−0.3%), wetlands (−2.9%), and grasslands (−8.3%). The same trajectory will also be expected between 2030 and 2040. The most critical spatial drivers of these changes are as follows: population growth, increased temperatures, precipitation variation, distance to water sources, distance to roads, distance to police posts, and distance to trading centres. To reduce the acceleration of changes in land use/cover types, this study agitates for equal spatial establishment of woodlots, promotion of alternative sources of energy and livelihoods, and settlement construction materials.

Keywords

• CA-Markov model
• projection
• refugees
• land use/cover changes
• Uganda