A philosophical case for process-based modelling of land use change

Abstract

Modelling is used to describe, explore and predict changes in land use and other human systems. ‘Top-down’ and ‘bottom-up’ modelling approaches are both popular, and each has important philosophical implications that correspond closely to major debates in social science. We outline some key contributions to these debates and argue that social processes such as those underlying land use decisions are fundamentally determined by individual intentionality, interacting with social norms of language, culture and institutions, rather than by general and predictable ‘laws’. Therefore, bottom-up models that reflect these processes offer far more informative accounts of system development. However, prediction remains outside the scope of either approach, and methods of validation based on tests of historical predictive ability risk over-fitting to trends and underestimating uncertainties. We explore the implications of these arguments for model design and use, and for general understanding of such fundamentally complex, spontaneously evolving systems.