NOAA Office of Satellite and Product Operations

Correction for GOES Imager Midnight Blackbody Calibration Error

Michael Weinreb

Dejiang Han
Integral Systems, Inc.
October 2003

Description of Problem

  • At certain times of the year, the magnitudes of the computed calibration slopes for the GOES Imagers' infrared channels exhibit anomalous dips during the approximately six hours centered on satellite midnight.
  • The amplitudes of the dips is greatest for Imager channels at the shortest wavelengths.
  • For GOES-8, the anomalous dips occurred during the months between April and October. However, GOES-10 experiences this phenomenon all year round. GOES-12, which became operational April 1, 2003, has so far exhibited the effect all the time, but we do not know whether it will continue all year.
  • Assuming these slope dips are errors, they will cause erroneous decreases in measured scene temperatures. Such decreases appear to have been observed.
  • The cause of the slope dips is believed to be interference by radiation from solar-heated structural components that reaches the Imagers' detectors during the blackbody look—see below.

Calibration Correction

Normal Calibration      Calibration Error from Extraneous Radiation
Error and Normal Calibration Diagram
Calibration error occurs when formula on left is used for case on right. (“m” is calibration slope.)

Correction Algorithm

  • The Midnight Blackbody Calibration Correction (MBCC) was developed to overcome this problem.
  • The algorithm is based on the observation that when this problem is absent, there is a very high correlation between the calibration slope and the temperature of several optics components, particularly the telescope's primary mirror.
  • So, when the problem occurs, we replace the bad slope values by estimates computed by regression on the primary-mirror temperature.
  • Slope values in “spikes” are erroneous.
  • Corrected values, based on regression, lie on curve.

Original and Estimated Slopes - CH2 Det 1


Scene Temperature Errors Predicted from Slope Errors vs Time

Temperature Errors Predicted from Slope Errors vs Time Chart


Observed Scene Temperature Errors vs Time

Observed Scene Temperature Errors vs Time Chart


Observed vs Predicted Errors

Observed vs Predicted Errors Chart

These images present two figures above aligned in time. The point is that the scene temperature errors predicted from the slope errors are well correlated with the observed 1 K depression, which is believed to be erroneous.


Slopes vs. Time

The next two images show all the slopes (computed once every half hour) for channel 2 (3.8 :m), detector 1, of the GOES-8 Imager plotted over seven months as a function of time.

The first image has the original slopes, including those affected by the midnight problem. The erroneous slopes constitute the "fuzz" above the main sequence.

The second shows all the slopes with the MBCC activated. The points that were in the fuzz are now in the main sequence.

Uncorrected Slopes vs Time

Uncorrected Slopes vs Time Chart


Corrected Slopes vs Time

Corrected Slopes vs Time Chart


Parallel Test of Midnight Blackbody Correction (MBCC)

  • Parallel (to operations) test of MBCC occurred Jan. 13-16, 2003, with data from GOES-10 Imager.
  • Parallel (MBCC) GOES-10 GVAR stream was distributed via the GOES-9 transponder.
  • Operational and parallel data were collected and compared at SSEC Data Center (U. Wisconsin.)
  • The following two figures show their results for Jan. 14 and 15, 2003:

Average Difference in Brightness Temperature Chart


Average Difference in Brightness Temperature Chart


  • The MBCC was effective from 0600 to 1000 UTC in Imager channels 2 (SW—3.8 :m), 3 (WV—6.7 :m), and 4 (LW—10.7 :m). (Local midnight for GOES-10 is at 0900 UTC.) At those times, the MBCC increased the scene temperatures, as expected.
  • An unexpected "correction," having opposite sign, was also made between 1000 and 1300 UTC in channels 3 and 4. This "correction" is not valid and occurred because of an algorithm error. The error has since been corrected. In the future, this invalid "correction" will not be made. (This catch illustrates the benefits of parallel testing. We are greatly indebted to our colleagues at U. Wisconsin.)



  • GOES Imager midnight blackbody calibration errors, caused by effects of solar heating in the Imager's calibration cycle, artificially depress scene temperature observations by as much as 1 K in channel 2 (the 3.8 :m channel). The errors are smaller in the channels at longer wavelengths.
  • We have developed a correction algorithm, the Midnight Blackbody Calibration Correction (MBCC), and have tested it in a parallel run with GOES-10 data. The MBCC will produce significant (up to 1 K in channel 2) increases in scene temperature measurements over a period extending from 3-4 hours before local midnight to 1-2 hours after local midnight.
  • We plan to make the MBCC operational in the fall of 2003.
  • Read more details: Correction For GOES Imager Midnight Calibration Errors