Removing Stripes in GOES (Geostationary Operational Environmental Satellites) Images by Matching Empirical Distribution Functions

摘要

The current and future geostationary operational environmental satellites (GOES) of the National Oceanic and Atmospheric Administration (NOAA) are designed to produce visible images of the earth with linear arrays of eight detectors. Because the visible channels are not calibrated radiometrically in orbit, differences among instrument gains associated with the different detectors may cause artificial stripes to appear in the images. In the data processing on the ground, the images are 'normalized' to remove the stripes. Images from future geostationary satellites, GOES I-M, will be normalized by the method of matching empirical distribution functions (EDFs). The paper reports on a study of EDF matching with data from GOES-7. The technique was used to generate a normalization look-up table from data taken on May 18, 1988, and the table was applied to image data obtained two weeks later, on June 1, 1988. This successfully removed the stripes from the image. The severity of striping was assessed qualitatively from photographs of the images and quantitatively with numerical measures of striping derived from the EDF matching technique. The technique is expected to be even more effective with data from GOES I-M because of improvements in instrumentation.