next up previous
Next: CPFM-Derived Quantities Up: The CPFM Spectroradiometer Previous: The CPFM Spectroradiometer

Uncertainties

There are numerous sources of error which contribute to the total uncertainty. The largest of these is the absolute error, estimated at 7.0/4.6% (UV/vis). Errors due to noise can be estimated by examining the differences in successive spectra. These were found to be less than 1.0/0.5%. Errors associated with wavelength assignments contribute less than 1%. There is up to a 15$^{\circ }$C temperature difference between the hangar, where the absolute calibration is performed, and the in-flight temperature of the instrument (controlled at 10$^{\circ }$C). As a result, uncertainties of up to 2.5% will be introduced.

The horizontal flux has additional error sources which arise due to the cosine-correction and the required navigational information (relative zenith angle and change in azimuth between the sun and the direction of view). The cosine error is estimated at 2.0/3.0% and the navigational error at 5.0%. Similarly, the direct-viewing ports have an additional error of 1.2% from the use of a calibrated diffuser.

The final uncertainty is 9.4/8.0% for absolute irradiance and 7.7/5.6% for absolute radiance. Omitting the absolute error contribution, the uncertainty in the radiances become 3.3/3.2%. To estimate the error in degree of linear polarization only relative errors need be considered. By propagating these uncertainties through, the error in polarization is determined from,

\begin{displaymath}\Delta LP = \frac{\Delta I}{I} \left[ \frac{I_{\parallel}+I_{...
...p}\vert} + 1 \right] LP = \frac{\Delta I}{I}(1+\vert LP\vert).
\end{displaymath} (5.3)


next up previous
Next: CPFM-Derived Quantities Up: The CPFM Spectroradiometer Previous: The CPFM Spectroradiometer
Chris McLinden
1999-07-22