Using the CLUE framework to model changes in land use on multiple scales

Abstract

The analysis of changes in land use and land cover has recently received ample attention in the scientific literature. So far, dynamic and integrated modeling approaches which are essential for the modeling of complex systems are relatively few. CLUE (Conversion of Land Use and its Effects), an empirically based framework for modeling changes in land use on various scales was developed to fill part of this gap. CLUE combines a statistical description of land use with scenarios of changes in the demand for regional commodities to model the possible pathways of future land use development. In the Atlantic Zone of Costa Rica CLUE was applied for the first time on sub-national scale. Multiple regression analyses produced equations with coefficients of determination between 0.58 and 0.91 for the major kinds of land use in the area (forest, pasture and banana). The statistical analysis demonstrated the importance of considering both biophysical and socio-economic variables as the driving forces of land use changes. Predicted changes in geographic patterns between 1984 and 2005 under different scenarios could be related to processes that are already known to take place in the Atlantic Zone. Forest was predicted to be largely replaced by pasture and to become limited to areas unfavorable for agriculture. Model validation yielded highly significant results with correlation coefficients ranging from 0.87 to 0.95 for the major kinds of land use. The study demonstrates that dynamic, multi-scale empirical modeling is a suitable tool to model land use changes on sub-national levels.KeywordsEuropean UnionGross Domestic ProductionMultiple Regression ModelNational ParkBase ScenarioThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.