Multi-scale integration for spatio-temporal ecoregioning delineation

Within the environmental and geosciences modelling community, 'Integrated Modelling' can take different meanings about what is integrated within the model. The term usually refers to the need of encompassing various important aspects influencing the outcomes of a model, via integration of other data and processing models seemingly not of primary interest. Most often the vector of influence, implying data interaction and coupling models, is the support of the scientific workflow model: the spatio-temporal paradigm. The purpose of this article is to suggest that not only the interaction of domains, such as ecological, geophysical, economical are important in integrated modelling, but also scales of phenomena and their interactions. In some modelling situations the integration of multiple scales may be achievable without tremendous changes in the model, it is often the case for models involving multivariate clustering. This is illustrated for an ecoregioning delineation model. Results, using the iterative method described in this article, shows refinements of the zonation in a non-uniform way, therefore enforcing the scale dependence hypothesis for this type of model.