|
The angular response of entrance optics is an important parameter for solar spectral UV measurements, and ideal cosine
entrance optics is required to measure ground-based global solar spectral UV irradiance including direct and diffuse
radiation over a solid angle of 2π sr. Early international comparisons have shown that deviations from the ideal cosine
response lead to uncertainties in solar measurements of more than 10%.
A special spectroradiometer used for solar spectral UV measurements was developed at National Institute of Metrology
(NIM). Based on Polytetrafluoroethylene (PTFE) integrating sphere, seven kinds of cosine-entrance system were
designed and compared. A special cosine measurement apparatus was developed to measure the angular response of the
entrance optics. Experimental results show that, the integral cosine error <f2> is 1.41% for a novel combination entrance
optics, which is composed by a PTFE integrating sphere, a spherical ground quartz diffuser and a special correction ring,
and the cosine error is 0.08% for an incidence angle of θ=±30°, 0.84% at θ=±45°, -0.47% at θ=±60°, -0.74% at θ=±70°,
and 5.47% at θ=±80°.
With the new non-plane entrance optics, the angular response of the solar UV spectroradiometer is improved evidently,
but on the other side, the system's absolute calibration becomes more difficult owing to the curved geometry of the new
diffuser. The calibration source is a 1000W tungsten halogen lamp, but the measurement object is the global radiation of
the solar, so a small error of the calibration distance will lead to an enormous measurement error of solar spectral UV
irradiance. When the calibration distance is 500mm, for an actual diffuser with spherical radius 32.5mm and spherical
height 20mm, the calibration error will be up to 3%~10% on the assumption that the starting point was calculated just
from the acme or the bottom of the half-spherical diffuser. It was investigated that which point inside the threedimensional
entrance optics should be used as the starting point of the calibration distance in this paper. According to
the information of the geometrical shape of the diffuser, the different irradiance value on the spherical surface and the
angular response of the receiver, mathematical methods are adopted to calculate the optical reference plane of the
spherical entrance system. Furthermore, an experimental method was used to verify the feasibility of the theoretic
formula.
|
