Resposta de índices de vegetação às precipitações em diferentes regiões do Estado da Bahia.

Abstract

The vegetation cover of a given location or region is closely related to the climatic conditions and predominant soil, which are influenced by a set of meteorological factors acting in the region. Researches have been developed from data extracted from satellite images to estimate biophysical parameters of the vegetation, aiming to monitor it. In this context, this study aims to analyze vegetation dynamics from the NVDI (Normalized Difference Vegetation Index) and EVI (Enhanced Vegetation Index) vegetation indexes, as well as to evaluate the response time of vegetation in relation to rainfall in sub- regions of the state of Bahia in the period 2001-2015. For this, the technique of the multivariate analysis ACP (Principal Component Analysis) was used to obtain the patterns of spatial variability and the seasonal correlations of the variables. The homogeneous regions of the vegetation indexes and precipitation were determined from the hierarchical cluster analysis (AA) proposed by Ward. In general, both indices had similar homogeneous regions. The discrepancy between them lies in the fact that EVI takes into account the atmospheric correction that optimizes the sign of vegetation, especially in areas of dense cover. The results showed a strong dependence on the rainfall indexes in the subregions, that is, as rainfall decreases during the months of the year, the values of EVI and NVDI decrease as a function of the coverage response to water stress. The correlations obtained between EVI and precipitation indicate that in the homogeneous region 1 (RH1) the highest correlations occurred in the rainy and less rainy period, with a 1 month lag. In RH4 and RH6, which present a more regular rainfall regime and forest cover, the lowest correlations are found. In RH2, RH3 and RH5, where the Caatinga, Cerrado and Caducifolia predominate, the correlations are higher in both the dry and rainy periods, with a lag of 1 to 2 months. For the correlations between NDVI and precipitation, it is observed that the highest correlations in RH1, RH3 and RH5 occurred in the rainy and less rainy period, With lag of 1 month. In RH2 and RH4 the highest correlations were found in the rainy season and in RH6 the lowest correlations were found, presenting a lag of 1 to 2 months.