Estimativa do fator C da Equação Universal de Perdas de Solos (EUPS) utilizando Indice de Vegetação pela Diferença Normalizada (NDVI) na bacia hidrográfica do córrego Monjolo Grande/Ipeúna/SP
Edvania Aparecida Correa1
Sergio dos Anjos Ferreira Pinto1
1 Universidade Estadual Paulista, UNESP/Rio Claro/SP edvaniacorrea@ig.com.br
sanjos@rc.unesp.br
Anais XV Simpósio Brasileiro de Sensoriamento Remoto - SBSR, Curitiba, PR, Brasil, 30 de abril a 05 de maio de 2011, INPE p.5708
ABSTRACT
Considering of the problems of the traditional method of classification the C Factor of USLE, which does not represent the changes in vegetation cover along of classified area, the aim of this paper was to estimate the values of Factor C obtained through the NDVI using regression analyses in the scenarios of 1988 and 2006 in the basin of brook Monjolo Grande/Ipeuna/SP. With the satellite images from TM sensor Landsat 5 of scenarios 1988 and 2006, was applied the atmospheric correction in the bands 3 and 4 (1988) and subsequently performed the Radiometric Normalization of bands 3 and 4 of 2006 using, to this, as reference adjusted images of 1988 . After pre-processing of images, we obtained the NDVI of the scenarios analyzed and applied a regression equation where they obtained the maps of Factor C from 1988 to 2006. The presence of El Nino in 1988 resulted in the differences found between the mean NDVI of the scenarios analyzed. Likewise, the prevalence of high levels of factor C (0.4 to 0.6) denotes the presence of areas of agriculture and pasture, which, combined with areas of high slope, may have high values of factor C. The proposed methodology is as quick alternative to obtain the C factor for large areas of study. The Remote Sensing was effective in obtaining the results proposed in this study, making the extraction of data related to Factor C more quickly and accurately.
Keywords: soil erosion modeling, remote sensing, linear regression, land cover, modelagem de erosao do solo,sensoriamento semoto, regressao linear, cobertura do solo.
|
