NOAA Office of Satellite and Product Operations

Operational Calibration of the Imagers and Sounders
on the GOES-8 and -9 Satellites

Michael Weinreb, Michael Jamieson, Nancy Fulton, Yen Chen, Joy Xie Johnson,
James Bremer, Carl Smith, and Jeanette Baucom

4.2 Derivation of emissivities

The scan-mirror emissivity values were determined soon after launch from in-orbit observations of space between the extreme west and the extreme east. Data for a number of scans were averaged to reduce random noise. For the sounders, the instrument output varies not only with scan position, but with time, as the temperature of the instrument changes over the minute or so required for each scan. (This is not a problem for the imagers, which clamp on space every few seconds.) We removed the temperature-driven drifts from the sounder data after estimating them by regression on the optics temperatures.

Emissivities were derived separately for each detector in each channel of each instrument. In the following, we suppress reference to instrument, channel, and detector, since the procedure was the same for all of them. Since measurements on space were available at all values of q, not just at the east and west limits, the definition of the calibration slope in Eqs. (7) and (8) could be simplified to

where Xsp(45) is the instrument output in counts when it views space at a scan angle of 45o. The emissivity was given by

where RM is the radiance of the mirror computed from its temperature during the scans, and Xsp(q) is the instrument's output in counts when it views space at scan angle q. The emissivities at 45o in all imager and sounder channels were provided from laboratory measurements on witness samples10.

The entire process was carried out for data collected approximately once an hour for 24 hours. A daily-average profile of e vs q was then computed as the average over all hours.

For each channel, we fitted a quadratic, i.e.,

to the profiles of emissivity vs q. (Also, the variable q was transformed to become the east-west position of the scan mirror expressed in inductosyn increment numbers, which is uniquely related to scan angle.) This produced the coefficients a0, a1, and a2. They are stored in a database in the ground-system computer and are used in the processing to generate the emissivities vs angle.

Approximately once every three months, new emissivity-vs-angle profiles are computed from east-west scans of space, as previously described. Thus far, the profiles have changed slightly for some instruments, and not at all for others, since the launches of the GOES-8 and -9 satellites.

Figure 2 attests to the success of the generalized calibration equations. It shows the radiances of space vs scan angle derived from the original calibration equations ("before") and the generalized equations ("after") in channels 4 (10.7 µm) and 5 (12 µm) of the GOES-8 imager. The generalized equations removed apparent east-west radiance differences of approximately 1mW/(m2-sr-cm-1 ) and 2 mW/(m2-sr-cm-1) in channels 4 and 5, respectively.

Fig. 2. Radiances of space vs mechanical scan angle (degrees) in channels 4 and 5 of GOES-8 Imager.
Fig. 2. Radiances of space vs mechanical scan angle (degrees)
in channels 4 and 5 of GOES-8 Imager.

The label, "before", indicates data before correction for scan-mirror emissivity variation; "after" indicates data after correction. The uncorrected radiance values are negative because the calibration establishes the zero-radiance level at the space clamp, which in this case was on the east.

Next  |  Previous  |  Table of Contents

Contact Michael P. Weinreb at
Latest Revision: July 9, 1997