The above discussions illustrate the strong interconnection between analysis, "truth," and calibration issues and show some typical adjustments that are made to the on-board calibration as a result of data analysis for temperature retrievals. Work is currently under way to use combinations of HIRS and data from other instruments in an attempt to narrow calibration uncertainties. Several examples of such work are found in the Seventh Conference on Atmospheric Radiation (1990), including Baum et al. (1990), Ackerman and Wylie (1990), Ackerman and Smith (1990), and Lavanant and Brunel (1990). This work is complicated by the fact that the instruments have different fields of view, wavelength bands and observation times over a given location. Even if a perfect comparison of two satellite instruments were possible, and even if it showed a systematic difference, one would still be unsure of where the problem(s) lies.
The fact that the HIRS/2 calibration errors are not fully understood points strongly to the need for relatively frequent, independent cross-calibrations for satellite instruments.