Modeling of transition dynamics among Brazilian savannah (Cerrado) physiognomies

Abstract

Models are fundamental tools in studies of complex systems, helping, for example, in understanding and forecasting of landscape processes, also as on complexity of its patterns. The objective of this project is to model the dynamics and to evaluate the complexity related to transition among Cerrado (Brazilian savannah) phytophysiognomies. For this, will be generate a cellular automaton whose cells can be occupied by three types of Cerrado's plant formations: grassland, savannah, and woodland. Automaton dynamics will follow probabilistic rules and each step of system's temporal evolution will corresponds to a model configuration related to occupation of each cell by a particular plant formation. Heterogeneity of spatial pattern formed on each step of temporal dynamic of cellular automaton will be used as indicative of system complexity. To quantify such complexity, will be utilized measures derived from fractal geometry and informational entropy. Joining the analysis of measures of patterns complexity and of the model behavior, may be identified phase transitions (e.g., from ordered to self-organized state) and critical points associated with them. This way, the results obtained will be confronted with the hypothesis elaborated by Mattos (2010) that certain physiognomies of Cerrado self-organize near critical points that can lead to phase transitions. (AU)