Materials and Methods
Fourteen lamps of various makes and models were obtained for testing purposes (see Table 1). A Zoo Med ReptiSun 5.0 unit which had been used for 10 months was also included because it could provide general information about the declining of UV radiation of a high-quality lamp over a long period of time. Some of the lamps included (e.g. True-Light) are so-called full spectrum lamps and because of this, their irradiance in the UVB range will not be at the same level as actual UV lamps. Naturally this has to be accounted for when evaluating the results.
As a reference, the spectrum of the sun was also measured. The measurement was made when the sun was at its highest point of elevation in the southern sky of Finland and natural UV radiation is strongest. Measurement was made in Raisio, 8 kilometres northwest from Turku (60o29’19N 22o10’4E).
The lamps to be measured were borrowed from retailers or purchased for personal use. Anja Kairisalo kindly donated the 10 months old ReptiSun 5.0 unit. Rolf C. Hagen Inc. provided their own models: Exo Terra Repti Glo models 2.0, 5.0 and 8.0, as well as Reptile Super UV Daylight and Reptile Desert 7 % UVB, the latter two being manufactured by Energy Savers Unlimited.
As there was only one unit of each lamp, the statistical reliability of the measurements is weak (sample=1). Nevertheless, it represents quite well the situation of a consumer purchasing a new lamp. When buying only one unit, the chance of consumer getting a faulty unit is the same as in this test. To eliminate obvious errors, all exceptional, inconsequent or otherwise suspicious measurement data was confirmed with the manufacturers.
All new lamps were pre-conditioned by using them in a normal terrarium lighting fixture for 100 hours, 12 hours a day. Prior to actual measurements, each lamp was preheated for 30 minutes to allow it to reach its nominal working temperature and to ensure lamp stability.
Measurements were made by Suomen Aurinkosimulaattori Oy/Solar Simulator Finland Ltd. The spectroradiometer used for all measurements was IL700A Research Radiometer, manufactured by International Light Inc., Newburyport, Massachusetts, USA. The measuring head used was an S-20 photomultiplier PM271D. The spectroradiometer is being calibrated on a yearly basis by wavelength and sensitivity. In order to control wavelength calibration, the device is quick-calibrated by measuring a HeNe laser’s known wavelength of 633 nm before each session. Wavelength measuring accuracy is ±3 nm.
Measurements of spectral irradiance were made in a free field, at a distance of 30 centimetres from the surface of the lamp. Fluorescent lamps were measured at their centre point, perpendicular to the longitudinal axis of the lamp. In case the lamp had a fixed reflector or a particular aperture, the measurement was made from the main direction of radiation. Lamps with a European E27 socket were measured from the direction of base longitudinal axis at a distance of 30 centimetres from the face of the lamp.
Numerical analysis and spectrum diagrams of measured data were made with Microsoft Excel 2000. Measurement data shows measured irradiance (mW/m2) at one nanometre resolution between 250-800 nm, 551 data points in all. These were imported to an Excel spreadsheet, converted to µW/cm2, and irradiances on different wavelength ranges were integrated. Because of the lack of a fixed standard determining the boundaries of UVA, UVB and UVC ranges, the ranges commonly used in current literature were adopted. Due to quantized nature of measurement data, the following limits were used in analyses: UVA 320-399 nm, UVB 280-319 nm and UVC 250-279 nm. Visible light is taken to be between 400-749 nm, and near infrared to be 750-800 nm.
The basis for the calculation of the D3 Yield Index was the action spectrum of 7-DHC to preD3 conversion in human skin. The index was calculated in the 252-313 nm range by weighing the irradiance at each wavelength in accordance with the action spectrum. The effective irradiances thus obtained were summed over this range, and this value was finally adjusted to the reference by a suitable multiplication factor, resulting in the final D3 Yield Index.