New Harvard Forest Publication: Spatial Variation Essential In Predicting Landscape Change

Postdoctoral Fellow Robert McDonald examined deforestation patterns in the Raleigh-Durham metropolitan region over the last decade, building a set of models that enable prediction of patterns of deforestation into the near-term future. Different functional forms of the land-use change model are tested, and it is shown that having a model that allows for spatial variation in the rules of landscape change is crucial for accurate model estimation and inference. Failure to allow for spatial variation in the rules of landscape change leads to chronic underestimation of deforestation rates in hotspots of deforestation, and overestimation for the rest of the landscape. This result has wide implications for efforts to model urban development elsewhere. 

McDonald, R.I. and Urban, D.L., 2006. Spatially Varying Rules of Landscape Change: Lessons from a Case Study. Landscape and Urban Planning 74:7-20.