Detecção de mudanças da cobertura e uso do solo a partir de índices espectrais de vegetação em ambiente semiárido: enfoque para o município de Teofilândia, Bahia

Abstract

The vegetation indices enable the evaluation of the spectral characteristics of land cover in a given area. This work has as main objective to model the different spectral index derived Normalized Difference Vegetation on land uses land cover to evaluate the effectiveness of the identification of the dynamics of vegetation in semiarid environment. To this end, we used satellite images Landsat 5 TM scene 216/68 with different time intervals and in different seasons to assess the spectral changes between them. The images were processed according to the steps of digital image processing to obtain the information necessary to achieve those goals. Among the major steps taken, it was the radiometric correction, the conversion of digital number values of apparent reflectance and supervised classification for comparison between the results, which algorithm was used Maximum Likelihood. The techniques found possible to conduct detection of major classes resulting from the dynamics of coverage and use of municipal land, showing the greatest change detection between the class areas in seasons than necessarily the end of the chosen time period.