Abstract

When spectral ultraviolet (UV) irradiance measurements are carried out, the signal rendered by the spectroradiometer should be proportional to the cosine of the angle between the incident radiation and the normal to the entrance optics. The deviation from this ideal response is referred to as cosine error. The cosine error effect on solar irradiance measurements performed by spectroradiometer systems can be accounted for by using a correcting factor B. If the cosine response is reasonably good, the factor B can be considered to be 1, such that the irradiance estimates are not affected. However, the uncertainty of B is not zero and it can be under certain conditions the greatest uncertainty source affecting solar spectral UV irradiance measurements. In this paper, we evaluated the B uncertainty of a state-of-the-art UV diffuser under different radiance conditions. We applied a Monte Carlo-based uncertainty propagation technique that allowed us to estimate an upper limit for the cosine error influence. We found that irrespective of the wavelength and the solar zenith angle, the standard uncertainty of B is about 2.5% in the case of high aerosol load. Under low aerosol and cloudless conditions, the standard uncertainty increased with the solar zenith angle, reaching a maximum of 3.5% at about 55° and decreasing thereafter. We conclude that the cosine response of even modern entrance optics is still a significant uncertainty source affecting ground-based spectral UV irradiance measurements. © 2008 BIPM and IOP Publishing Ltd.