Avaliação do albedo nos diferentes tipos de cobertura do cerrado do Amapá - Brasil com imagens MODIS

This study focuses on comparisons of land surface albedo estimates for different types of land cover in Amapa state, Brazil. Land surface albedo is an important parameter for radiation balance estimates on land surfaces. Liang and Tasumi algorithms which are used to estimate land surface albedo are analysed over different types of land cover such as urban regions, forest, scrubland, eucalyptus agroforestry etc.. Three images from MODIS sensor which operates on board TERRA satellite from 2002, 2006 and 2012 are used to validate the algorithms' spatial estimates. The results shows that both methods present accuracy on land surface albedo estimates. Regarding mean land albedo values for all types of land cover, the linear correlation coefficient observed is 0.88, with standard deviation of 0.012 and mean percentage error of 7.8 % with significance level α=0.05.