TENDENTIAL MODELING OF DEFORESTATION IN CAATINGA BIOME IN PIAUÍ STATE

Abstract

The work aims to simulate the dynamics of deforestation in the area that covers the caatinga biome in the state of Piauí for the next five decades. In the simulation, the program Dinamica EGO was used. Were acquired deforestation data for 2002 and 2008 and spatial data of the explanatory variables of deforestation in the area. Five steps were performed: calculating the transition matrices, determining the weights of evidence, adjusting, validating and projecting the model in the trend scenario. Transition rates for total and annual deforestation resulted in 4.6% and 0.8%, respectively. The weights of evidence revealed that distances up to 600 m from deforested areas influence in the process of transition to deforestation. The variables with the highest weights were altitude, up to 100 meters; areas closer to urban patch and sustainable use units, in addition to land use and cover classes: urban influence, agriculture, livestock. With the simulated images it was observed a reduction of the forest remnants of caatinga area in the state, from 69% in 2008 to 43% in 2070. These results serve as a warning to the public authorities and the population. The proposed methodology can be applied to other Brazilian biomes with different approaches.