Description

Band 21 calibration is imprecise at high temperatures.

Impact to L1B Data Set

This effect causes increased uncertainty in the brightness temperature due to the indirect nature of the calibration. Detector to detector striping exists, but is within specifications.

Estimated Impact to L2 Science Products

Science research will be seriously impacted until the calibration can be extended.

Cause of Effect

The saturation brightness temperature for this fire detection band is specified at 500 K. Pre-launch test targets had a maximum brightness temperature of 340 K, which equals about 3.5% of full radiance scale. Band 21 is expected exhibit the non-linear characteristics seen in other MWIR bands, but the calibrated temperature range is not adequate for determining non-linear calibration characteristics for this band.

Mitigation Approach/Added Testing to Characterize Effect

The linear gain coefficient for this band is computed approximately quarterly using an elevated temperature cycle of the OBC-BB. In Collection 5, the linear gain coefficient for this band was derived from the BB warm-up cycle. In Collection 6, the linear gain coefficient is derived from the BB cool-down cycle. New LUTs enabling mirror-side dependent b1 calibration (memo M1046) were added in code version 5.0.7 corresponding to July 4, 2005. LUTs are updated as needed.

Early studies indicated that the calibration might be tested with observations of the moon. Literature values suggest the moon has a brightness temperature of about 400 K at 4 µm. To date, the variability in the MWIR observations of lunar radiance is too great to allow for this calibration (memo M0207).

Estimated Schedule to Complete

Band 21 gains are updated as needed.

History of Characteristic

Historically, the linear gain coefficient for this band is updated once or twice per year with more frequent updates during the first year of the mission.