Prelaunch calibration of the CZCS used a 76 centimeter diameter integrating sphere as a source of diffuse radiance for channels 1 through 5 and a blackbody source for calibration of channel 6. The integrating sphere was especially constructed for calibration of the CZCS and was calibrated from a standard lamp from the National Bureau of Standards utilizing a spectrometer and another integrating sphere to transfer calibration from the lamp to the sphere. In addition to the sphere and the blackbody, a collimator was used to calibrate the CZCS in vacuum testing.
The instrument manufacturer, Ball Aerospace, provided calibration coefficients, in the form of a slope and intercept, to convert measured digital counts into radiance values for each band and gain setting
It was intended that the on-orbit calibration of the CZCS would be accomplished (for the first five bands) by using a built-in incandescent light source. This on-orbit calibration source was calibrated using the instrument itself as a transfer against the referenced sphere output. The light source was redundant in the instrument so that in case of failure of one of the lights, another one can be ordered to operate on command. However, as the light from these lamps did not pass through the entire optical train, they could not be used for a true calibration. The lamps were also not of sufficient intensity to provide a meaningful signal for band 1. The lamp data was used only to verify instrument stability with time.
The thermal channel, band 6, was calibrated on-orbit by viewing the blackened housing of the instrument whose temperature is monitored. Deep space was another calibration source viewed during the 360 degrees rotation of the scan mirror. As this band failed early on, no further discussion of it's calibration is warranted.
In 1981, the Nimbus Experiment Team (NET) determined that the sensitivity of the visible bands was degrading with time. This degradation was significant for the 443nm band. As a result, Evans and Gordon (1994, EG94) derived a long-term degradation correction, which is a time-varying, band dependent multiplicative factor applied on top of the Ball calibration. The OBPG re-evaluated the sensor degradation using a clear-water radiance model. Using this model-based approach (Werdell, et. al., 2006), the OBPG was able to replicate the results of Evans and Gordon, in a gross sense, using an exponential fit to the derived gain coefficients. As the exponential fit and the EG94 piecewise linear fit produced comparable results, it was decided to continue the use of the EG94 temporal calibration for the current OBPG processing
The EG94 calibration included adjustments to the Ball Aerospace pre-launch calibration. These vicarious calibration coefficients were derived by forcing the peak of the frequency distribution of 10-day average global radiance values to match the clear-water radiance values for bands 2 and 3 (520 and 550nm respectively). The vicarious coefficient for band 1 was determined by forcing agreement with simultaneously measured in situ data collected during the NET post-launch validation cruises.
The OBPG left these EG94 derived coefficients in the calibration and derived vicarious calibration coefficients in addition to these by using a clear-water radiance model based on a climatological (in situ based) chlorophyll time-series at the Bermuda Atlantic Time Series (BATS) site. Unlike the EG94 vicarious calibration, this approach was used for band 1 as well. This additional vicarious calibration is done to account for the differences in the atmospheric correction approach employed by EG94 and the OBPG. The EG94 coefficients could have been removed prior to the OBPG vicarious calibration, but as they are simply multiplicative factors, it was not necessary. By leaving the EG94 vicarious coefficients in the calibration, end users can get back to the EG94 result by simply setting the OBPG vicarious coefficients to unity.
Ball Aerospace Division, (1979) "Development of the coastal zone color scanner for Nimbus-7, vol 2: Test and performance data, Final report"
(2) Evans and Gordon (1994) "Coastal zone color scanner 'system calibration': A retrospective examination", JGR, vol 99 no. C4, pp 7293 -- 7307.
(3) Werdell, P.J., Antoine, D., Bailey, S.W., Feldman, G.C., Hooker, S.B., McClain, C.R., and Zibordi, G., "Alternate on-orbit vicarious calibration schemes for ocean color satellites", NASA Ocean Color Research Team Meeting, Newport, RI, 11-13 April 2006.