Change in forest cover of the northwest region of Amazon in Mato Grosso state

Abstract

In the past few years, the intensification of anthropic activities that modify the soil-vegetation cover in Brazil's land has been occurring on a large scale. To monitor the forest cover changes, the techniques of Remote Sensing of vegetation are essential tools, especially in large areas and with difficult access, as is the case of the Brazilian Amazon. The aim of this work was to identify the changes in land use and land cover, over the past 20 years, in the municipalities of Aripuana and Rondolandia, Northwest of Mato Grosso State, in order to quantify the effective altered areas. Landsat 5 TM and Landsat 8 OLI digital classification images techniques were used in three different dates (1995, 2005 and 2015) and, later, the detection to the land use and land cover changes. The digital classification showed excellent results, with kappa index above 0.80 for the generated maps, indicating the digital classification as a potential tool for land use and land cover. Results reflect the conversion of forest areas mainly for agricultural activities, in the order of 472 km(2), representing a loss of 1.3% of Amazon forest surface in the study region.