An examination of historical and future land use changes in Uganda using change detection methods and agent-based modelling

Abstract

We analyse historical land use/land cover changes beginning from 1996 to 2013 and apply an agent-based model to simulate potential agricultural land use change scenarios in Uganda. The model incorporates farmers’ decision-making processes into biophysical and socioeconomic factors and uses these to analyse the effect of farmers’ decisions on agricultural land use changes. Geographic information system tools are employed to build spatial relations between farmers and land cover systems. Satellite images are used to represent the initial land cover conditions and serve as observed land cover datasets to calibrate the simulated results. Significant agricultural and grassland cover and urban land uses are experienced in 72 and 36% of the regions, respectively, while wetland land uses increased significantly in 82% of the regions. On the other hand, 91% of the regions experience reductions in forest cover except for Teso region which reports gains of 62% in forest cover. Acholi is the only region (extreme outlier) that reports dramatic gains in wetlands of over 880%. The results of the simulation model are promising, and the model was successful at representing historical and future scenarios of agricultural land use patterns at a national level.

Keywords:

• agent-based model
• agricultural land use
• change detection
• land cover change
• simulation

Acknowledgments

The authors would like to thank the editor and anonymous reviewers for their constructive comments and suggestions. Initial results were published in a conference proceeding in Geo-Informatics in Resource Management and Sustainable Ecosystem, Second International Conference, GRMSE 2014, Ypsilanti, MI, USA, October 3, 2014 to October 5, 2014, by Springer Berlin Heidelberg. Special thanks to staff and colleagues in our respective institutions, Department of Geography, University of Cincinnati; Department of Preventive Medicine, University of Tennessee; and Department of Geography, Geo-informatics & Climatic Sciences, Makerere University.