Operational Calibration of the Imagers and Sounders
on the GOES-8 and -9 Satellites, Page 11
Michael Weinreb, Michael Jamieson, Nancy Fulton, Yen Chen, Joy Xie Johnson,
James Bremer, Carl Smith, and Jeanette Baucom
6.1 Normalization (destriping)
Without in-orbit calibration, the uncompensated responsivity differences among the elements of the visible-channel detector array can result in the appearance of artificial east-west stripes in the images. This striping can be reduced by the normalization of each detector's output to the output of a stable reference detector. The normalization is applied in real time at the Command and Data Acquisition (CDA) station at Wallops, VA, through application of normalization look-up tables (NLUTs) to the raw scene data.
The technique for generating NLUTs is carried out off-line and involves matching the
empirical distribution functions (EDFs) of the raw data from each detector to
that of the reference detector.12 The reference detectors, listed in Table 5
(for the imagers only), were chosen from considerations of long-term stability
and maximal usage of the count range of the data system without clipping at the
upper and lower limits.
Image striping is monitored daily at NOAA. Striping in
images from the GOES-8 imager has been minimal, even in the absence of
normalization. Since March 29, 1995, the data have been normalized to remove a
small amount of striping, but the detectors have been so stable that there had
been no need to update the NLUT until May 1996, when relativization (see below),
was introduced. At that time a new NLUT was made operational, and it is expected
that it will remain valid for many months. A similar statement can be made for
the visible channel of the GOES-9 imager. The first NLUT became operational on
July 31, 1995, and it remained valid until April 1996, when relativization was
introduced, whereupon a new NLUT was implemented.
Visible data from the sounders are not being normalized as of this writing. For the GOES-8 sounder,
normalization had been applied in the period between March 29, 1995, and June
19, 1996. On June 19, 1996, when relativization was introduced, it was found
that normalization was not needed, and it was discontinued. Visible data from
the GOES-9 sounder were never normalized.
Next | Previous | Table of Contents
Contact Michael P. Weinreb at michael.weinreb@noaa.gov
Latest Revision: July 9, 1997
