Caracterização radiométrica e variabilidade espaçotemporal do algodoeiro

Abstract

The cotton crop has important phenological variations that influence the spectral behavior and can help in the dynamic management process. The objective of this work was to evaluate the applications of the multispectral bands of the Landsat-5TM sensor, and the spectemporal characterization of the phenology of the cotton crop on the cultivar 966 Bayer Fiber Max®. Vegetation index (IV) were calculated to aid in the discrimination of cotton phenology. The temporal monitoring revealed that the best IV for temporal characterization was EVI (0,939- May / 09-0,704 August / 09), which presented results consistent with the phenological phases of the culture. The Tasseled Cap (Wetness, Greenness and Brightness) transformations were associated to physical parameters of the study area and correlated statistically with the data of the proposed IV, NDWI, NDVI and SAVI indexes with higher values of Wetness components (r = 0.869 , p <0.05), Greenness (r = 0.878, p <0.05), Brightness (r = 0.761, p <0.05), respectively. By thermal band (B6), the surface temperature was estimated, validated by data from virtual laboratories and by INMET meteorological station. Thus, it was possible to identify and correlate the changes in the spectral reflectance of the cotton in space and time, making possible to interpret the vegetative stages of the culture remotely.