Land changes are the result of a complex web of interactions between human and biophysical factors, which act over a wide range of temporal and spatial scales. Understanding processes of change from local to global scale and their impacts on the coupled human-environmental system is a main scientific challenge. No single model or scale can fully capture such interactions and processes of land change. This thesis presents a methodology for building multiscale, multi-locality land change models that include top-down and bottom-up relations. At first, we conceptualize two types of spatial relations among geographic objects at different scales. To handle the interaction of nested spatial objects at different scales, we propose hierarchical relations. To handle the interaction between networks and spatial objects, we propose action-at-a-distance relations. Then in a second step, we propose a modular software organization to build multiscale land change models. We consider the case when single-scale models, using diferente modeling approaches, are independently built and then dynamically coupled. We introduce the concepts of Model Couplers to define the bi-directional flow of information between the scales. We implement these concepts using the TerraME modeling environment. As a proof of concept, we present a hierarchical two-scale example for the Brazilian Amazon. The conclusion of this work points out that combining hierarchical and network-based spatial relations provides a comprehensive conceptual framework to include top-down and bottom-up interactions and feedbacks in multi-scale land-change models. The modular software organization and concept of Model Couplers are general enough to be used for other types of applications, and to contribute to the creation of Integrated Environmental Models from local to global scales.
Disponível em: http://mtc-m16c.sid.inpe.br/col/sid.inpe.br/mtc-m18@80/2008/220.127.116.11/doc/publicacao.pdf