Dinâmica espaço-temporal da cobertura vegetal, das queimadas e da expansão da bovinocultura na bacia leiteira do estado de Pernambuco usando Landsat-8 e MapBiomas

Abstract

The semi-arid is responsible for about 58% of the effective herd in the Brazilian Northeast region (NEB), however, there are local and meteorological factors that limit the semi-arid region. The objective was to evaluate the space-time dynamics of vegetation cover through the Soil-Adjusted Vegetation Index (SAVI) and Pasture Plant Coefficient (CVP) via Landsat-8/OLI images processed in the cloud on Google Earth Engine (GEE), measuring the scenarios of burned areas and characterizing the expansion of cattle farming in the dairy basin in the state of Pernambuco through land use and cover (LULC) and its impacts on the Caatinga Biome. The study region comprises the 23 municipalities that produce the most milk in the state, between 2016 and 2021, divided into dry and rainy seasons. The Vegetation Indexes were processed in the GEE and the spatialization of the rain was made through the Climate Hazards Group InfraRed Precipitation with Station data (CRHIRPS) and both data were submitted to descriptive statistics. LULC thematic maps and burned areas were processed in QGIS software. Thus, it was observed that in the dry and rainy period from 2016 to 2020, there was average precipitation of 77.98 mm and 331.19 mm, respectively. In the rainy season of 2016, less precipitation was observed when compared to the other years of the study, as it was the period that was influenced by the great droughts that preceded the year in question. As for the vegetation indices in the dry period, the SAVI presented lower spectral responses throughout the study period, as a result of low precipitation in the region and the rainy season with high biomass activity with values above 0.78. The SAVI and CVP showed medium to high variability. The burned areas showed that anthropic activities in some sectors of the basin caused the change in the LULC, resulting in degradation processes, even in regions with greater precipitation. Therefore, it is concluded that the analysis of the products CHIRPS, SAVI, CVP associated with the descriptive statistics and the maps of the LULC and burned area, was efficient for the space-time characterization of the expansion of cattle farming in the dairy basin of the state of Pernambuco throughout the study period.