Uncertainties in radiance calibrations of backscatter ultraviolet (BUV) instruments

Abstract
Traditional radiance calibrations of solar backscatter ultraviolet (SBUV) instruments require a measurement of the bidirectional-reflectance distribution function (BRDF) of a flat-plate diffuser. To date the BRDF of the diffusers used for calibrating the Shuttle-borne SBUV instrument (SSBUV), has relied on an initial measurement of a sample plate coated with BaSO4 made by the National Institutes of Standards and Technology (NIST). Changes in the reflectance of the calibration diffusers were tracked over time with measurements of their total hemispherical reflectance. Preliminary comparisons of direct measurements of the BRDF using a new facility at the NASA/Goddard Space Flight Center, inferred BRDF values using an integrating sphere, and the NIST-based values, show significant inconsistencies. In general, direct measurements of BRDF values compare to within 1% of those inferred from integrating sphere measurements. However, these values differ from the NIST-based values by about 6%. These differences and their uncertainties are discussed. The results imply that SSBUV-measured vertical ozone profile values have an altitude-dependent error of between 3% and -12%, but there should be little effect on the measured total ozone values.