Análise espaço-temporal do desenvolvimento das pastagens na Zona da Mata de Minas Gerais com o uso de imagens NDVI do Sensor Modis/Terra

Abstract

There is great interest in the knowledge of the conditions of Brazilian grasslands. Therefore, due to the wide territorial expansion occupied by this vegetation and the socioeconomic importance for the agriculture and livestock production chain, the data and analysis procedures for remote sensing represent an immense advance in the evaluation of its vegetative growth and degradation. The objective of this work was to evaluate the vegetative development, temporal vigor and long-term scenery of the grassland in Zona da Mata, in the State of Minas Gerais, Brazil. This was done by integrating phenological metrics obtained with NDVI/MODIS hypertemporal data related to the spring growth season, by means of Multicriteria Evaluation, Analytic Hierarchy Process and Simplified Expert System by mathematical expression (SEs) in Geographic Information System (GIS). The integration between the Growth Index (GI) temporal metrics for the spring season, Slope, and the Maximum, Minimum and Mean metrics of the temporal series allowed the necessary combination to investigate the vegetation structure and edaphic conditions, as well as degradation aspects. The R/S statistics and Hurst exponent enabled the description of the occurrence of maintenance or future reversion of the vegetative conditions of the grasslands. Approximately 61.5% of the Zona da Mata grasslands were degraded or under degradation process with long-term persistence, and 27% of the grasslands presented good long-term sustainable vegetation density conditions estimated by temporal analysis methods. It was possible to elaborate a management key for grassland development, and a methodology for continuous monitoring was consolidated.