Impact of atmospheric correction in AVHRR/ NOAA-17 images for calculation of NDVI vegetation index

Abstract

In order to obtain true surface reflectance in bands 1 and 2 of the AVHRR sensor, atmospheric correction was performed based on input provided by the atmospheric parameters from MODIS sensor. The use of MODIS data is directly related to obtaining the needed information for atmospheric correction,
considering the variability of parameters in time and space. Using the SCORADIS algorithm, which is based on the 5S radioactive transfer model, it was proposed an adjustment viewing to allow the entry of atmospheric images corresponding to the plans through the use of different methodologies for atmospheric correction. These analyses indicated that the performed corrections had results consistent with the elimination of the effects of atmospheric scattering and absorption. We tested the magnitude of these effects on the NDVI vegetation index that is widely used in agrometeorological studies. The percentage difference between the images with and without correction lasted up to approximately 60 to 80% for the analyzed dates, emphasizing the importance of atmospheric correction of images.