Variações do índice de vegetação por diferença normalizada (NDVI) do sensor modis associadas a variáveis climáticas para o estado do Rio Grande do Sul

Abstract

The vegetation indexes, derived from products of remote sensing, are powerful tools which indicate the vegetation development. Thus, this work aims at studying the behavior of the forest and crop vegetation through the analysis of the variation of Normalized Difference Vegetation Index (NDVI) to the state of Rio Grande do Sul, during the year of 2009, as well as to correlate such behavior to the climatic, temperature and precipitation variables. In order to accomplish it product images of MOD09 and meteorological data of the year 2009 were used, along with cartographic data in order to build up a data bank, generating the applicative SPRING, version 5.1.3. A prelimary assessment of the data visually analyses the generated images, and it demonstrates to be possible observing the changes in the phenological state of the vegetation on each considered date. The applicative R was applied to the statistical analysis and Excel 2007 to the explanatory data analysis and data correlation. To the explanatory analysis, the results demonstrate average values in NDVI of agricultural areas which go from a bit more than 0,5 to approximately 0,7, and from approximately 0,65 to a bit more than 0.8 for forest areas. The average values of NDVI describe curves with similar behavior, however in different response levels. The area of Julio de Castilhos presented the biggest amplitudes in the NDVI variation, both agricultural and forest, indicating the vegetation heterogeneity. The correlation analysis for the NDVI Forest and Temperature variables reached 0,7613 and 0, 6244 to São Borja e São Martinho da Serra, respectively. The Agriculture NDVI had a correlation of 0,5956 with the Temperature in Caxias do Sul. The other correlations for Temperature and all for Precipitation were considered low, they did not allow the association to the intended variables.