BAYESIAN NETWORK MODEL TO PREDICT AREAS FOR SUGARCANE EXPANSION IN BRAZILIAN CERRADO

Abstract

The growing demand for ethanol has powered the shift of the sugarcane frontier into the Brazilian Cerrado, mainly in the states of Goiás and Mato Grosso do Sul. Therefore, this study aims to propose Bayesian Network models for identifying potential areas for sugarcane expansion in Goiás and Mato Grosso do Sul states. The models take into account constraint factors in relation to the sugarcane expansion such as topography, soil aptitude, climate conditions, and available infrastructures. Results showed that Bayesian Network models proposed in this study were able to represent the tendency of sugarcane expansion.