Scaling methods in ecological modelling

Abstract

Modelling is often confronted with scaling problems, because modelling—directly or indirectly—always implies scaling. This is because models simplify. Simplification usually means aggregation, and aggregation is a scaling process. As scaling cannot be avoided in modelling, it should carefully be addressed and resolved, at least to the degree possible. In this paper, we give an overview of scaling approaches in ecological modelling. We propose to classify scaling approaches into pre-model scaling, in-model scaling and post-model scaling depending on the timing of the scaling relative to the main modelling process. We show general approaches, examples and potential application problems for each category. We suggest that scaling problems might be more widespread than previously thought. These scaling problems are matched with a range of solutions, but often these solutions will have to be adapted and tailored to the specific scaling case. Thus, we recommend that ecologists be aware of scaling challenges especially where models do not explicitly aim at scaling. Developing general test systems for scaling methods may help to broaden and enhance the application of scaling methods in ecology.