Dynamic deforestation modeling in the Amazon

Abstract

Dynamic models operating on a cellular automata (CA) basis have arisen as a feasible alternative for the analysis of the ever increasing deforestation rates in the Amazon. In this sense, this work aims at simulating deforestation in the area of São Félix do Xingu, southeastern part of Pará State, employing such class of models as a tool for understanding the main driving factors of deforestation. Real data of deforestation in 1997 and 2000 were used to assess the forest suppression extent in the study area. A set of explaining variables related to deforestation together with internal parameters of a CA model were used to generate a simulation for the year 2000. The provided simulation output presented a high fuzzy similarity index, what indicates the model effectiveness. The main driving variable takes the "distance to deforestaded areas in 1997", into consideration which confirms statements of previous scientific studies for this region concerning the predominance of new deforestation patches in the surroundings of pioneer areas.

Keywords: Deforestation; Amazon Forest; Cellular automata; Dynamic modeling.