Uncertainties in Quantitative Measurements

Effect of Foreoptics on In-Flight Calibration

The calibration blackbody target is internal to the VAS instrument; thus, its radiation does not pass through the telescope foreoptics. As a result, the effect of the foreoptics on the calibration must be modelled. Uncertainties can only be quantified by a continual comparison of the VAS data to independent measurements. There is evidence that the coefficients used in the model of the effect of the foreoptics should be seasonally adjusted to accommodate the changing thermal conditions of the spacecraft. (Menzel, W.P. 1990. Summary of the VAS calibration. Unpublished.)

Thermal Stability of the Instrument

The original weekly infrared calibrations were not frequent enough to deal with calibration drifts of up to 5K arising from thermal gradients in the instrument by diurnal variations in heat load. Therefore, NOAA recommended that quantitative users of the data attempt to establish an external validation measurement by some comparative technique (Clark 1983). This problem can be avoided by using VAS data processed by the scan-by-scan calibration scheme. All VAS data in the archive at the Space Science and Engineering Center at the University of Wisconsin contain the information needed to do this. (Personal communication with W. P. Menzel, 1991.)

Detector Non-Linearities

The calibration scheme assumes a perfectly linear response of the detectors over the full dynamic range. This is not always the case and a correction for the non-linear response of the mercury cadmium telluride detectors is necessary. In the scan-by-scan calibration scheme, a correction is made at the numerical conversion of the detector output voltages to measured radiances.(a)

Detector Noise

The "noise equivalent radiances" for the detectors is distributed by NOAA for each channel of each VAS (Gibson 1984). The detector noise is reduced by the spin-budget of the dwell-sounding mode, and can be further reduced by spatial averaging over adjacent measurements. The noise characteristics of the channels of the GOES-5 VAS were given in Table 2 (Smith 1983).

Pixel Integrity

In the imaging mode, the infrared channel is sampled every 8 s, which corresponds to a ground resolution at nadir of ~3 km, compared to the ~7 km field of view. Thus the infrared imaged pixels are over-sampled by more than a factor of two.

The mis-registration between different VAS channels measured on the same scan line (but on different revolutions of the spacecraft) is estimated to be less than 1 km (Chesters et al. 1983), but in some cases the 14-km footprints may be mis-registered in geographical coordinates by 7.5 km (Chesters et al. 1983).