Design of an input optic for solar UV-measurements

J.G. Schreder, M. Blumthaler, M. Huber

Institute of Medical Physics, University of Innsbruck, Austria



This article outlines a new input optic for measuring spectral UV irradiance. The system provides a considerably improved measurement accuracy in comparison to traditional optics. The input optic has a weatherproof housing and is designed for use with a quartz fibre. The cosine error is less than ± 3% for incidence angles between 0° and 75° and therefore the integral cosine error for isotropic radiation is less than 2.5%.



For measuring global irradiance the incoming radiation is weighted by the cosine of the incidence angle q . This is the angle between the direction of the incidence radiation and the surface normal. Generally all standard UV input optics deviate significantly from this ideal behaviour.

The deviation of the input optics response from the ideal cosine is described by the cosine error f2(q, j) (1):


where S(q, j) is the radiometer signal and q,j are respectively the incidence angle and the azimuth angle. The integral cosine error á f2 ñ (1) describes the quality of the diffuser by simply one number. This quantity is calculated according to equation (2)


International intercomparison campaigns (2,3) have shown that discrepancies of up to 10% in the measurement data may arise as a result of different cosine responses. Nevertheless, the measurement data are comparable if the experimental results are corrected by applying a cosine correction procedure described in (4). Depending on aerosol amount and wavelength the cosine correction factor can reach 7% ± 2% (5,6). Since these routines are restricted to cloudless sky conditions and known aerosol concentrations their use is very restricted. These routines also require the assumption of an isotropic distribution of the UV radiation in the sky. However, measurements of the spatial distribution of the sky radiance show this not to be true (7).

Due to these limitations, the use of improved input optics with a very small cosine error is advised.


Material and Methods

The new input optic was designed to provide a nearly perfect cosine response at all incidence angles as well as to be used in any weather conditions. The input optic consists of a weather proof housing, a shaped teflon diffuser, a quartz glass dome, an easyly to replace silica cartridge and a spirit level. The diffuser is optimised to work under the quartz dome.

Figure 1:Diagrammatic illustration of the new input optic with the shaped diffuser D, the quartz dome Q, the flange H of the diffuser, the light guide L and the axis A of the system


The cosine response characteristic of a flat diffuser can be modified by changing it into a nonplanar form (8). To allow a specific and reproducible shaping of the diffuser it is manufactured by the use of a computer driven milling machine. A diagrammatic illustration of the input optic is shown in Figure 1 where the flange H is simultaneously the shadow ring of the diffuser.

A 1000 W Xe UV lamp was used to determine the cosine response in the laboratory. The lamp is placed behind a wall with a hole to prevent stray light from reaching the input optic. The input optic is connected by a light guide with the spectroradiometer DM150 from Bentham. The input optic is mounted on a stepping motor driven rotary table, which is aligned with a laser system. This arrangement allows the variation of the incidence angle of the radiation on the input optic. The precision of a stepping motor is desirable because at large incidence angles, a small misalignment will lead to considerable errors in the cosine response. A 1° error in the incidence angle will produce an error of 5% in the cosine response (9). With this setup, errors due to misalignments are below 2% at 85°.



Figure 2 presents the cosine error of the newly designed global input optic in comparison to the previous input optic (flat teflon diffuser) of the Institute of Medical Physics in Innsbruck and the input optic (flat diffuser) of the Brewer MKIII #119 spectral radiometer.

Figure 2: Deviation from the ideal cosine response of the new input optic (shaped teflon diffuser) ATI-NEW of the Institute of Medical Physics, the old input optic (flat teflon diffuser) ATI-OLD and the input optic (flat diffuser) of the Brewer MKIII #119

Deviation from ideal cosine response

It is obvious that the two flat diffusers show an increasing cosine error for increasing incidence angle. The better cosine response of the new input optic is distinctly seen. For q >50° the dome shaped part of the diffuser increases the incoming radiation. The shadow ring reduces the incoming radiation for incidence angles q >80°. The interaction of the dome and the shadow ring leads to a cosine error less than ± 3% for incidence angles between 0° and 75°. Several input optics have been constructed following the methodology described here, and their integral cosine errors are all less than 2.5%.



Conventional input optics with flat diffusers produce systematic errors in the measurement of global irradiance of up to 10% (2,3). To reduce these errors, a new improved input optic was designed.
For incidence angles between 0° and 75° the cosine error is smaller than ± 3% and the integral cosine error is lower 2.5%.



1. Institute of Standardization (1994) DIN 5032, part 6, Beuth Verlag, Berlin

2. , B. G., Kirsch P. J. (1995) Setting standards for European ultraviolet spectroradiometers, report to the Commission of the European Communities, Contract STEP CT90007, Office for the official publications of the European Communities, Luxembourg.

3. The Nordic intercomparison of ultraviolet and Total Ozone instrument at Izana October 1996, Finnish Meteorological Institute, Helsinki 1997

4. Seckmeyer, G., Bernhard, G.(1993) Cosine error correction of spectral UV irradiances. Atmospheric Radiation, Proc. SPIE, 2049, 140-151

5. Gröbner, J., Blumthaler,M., Ambach,W. (1996) Experimental investigation of spectral global irradiance measurement errors due to non ideal cosine response, Geophys. Res. Lett., Vol. 23, 2493-2496.

6. Alkiviadis, F., B., Kazadzis, S., Balis, D., Zerefos, C., S, Blumthaler, M. (1998) Correcting global solar UV spectra recorded by a Brewer spectroradiometer for its angular response error. Applied Optics, in press

7. Blumthaler, M., Gröbner, J., Huber, M., Ambach, W. (1996) Measuring spectral and spatial variations of UVA and UVB sky radiance, Geophys. Res. Lett., 23, 547-550

8. Bernhard, G., Seckmeyer, G. (1997) New Entrance Optics for Spectral UV Measurements. Photochem. Photobiol. 65, 923-930

9. Gardiner, B., G. (1997) Spectroradiometer calibration methods and techniques, NATO ASI Series 1, Global Environmental Change, Vol. 52, 119-132