Errors in determining broadband ultraviolet irradiances from spectral measurements

The errors in determining the biologically effective and broadband UV irradiances from spectroradiometer measurements depend critically on procedures employed in the calibration process. A series of intercomparisons of two calibration standards and a spectroradiometer were undertaken in two laboratories. The near UV (320-400 nm) and shorter wavelength (280-320 nm) waveband irradiances of one standard obtained employing the other lamp for calibration agreed to within 3% and the erythemal irradiances agreed to within 5 to 6%. This was also the case if the lamps were swapped. There was no significant drift with time over five weeks in the measured standard lamp output. The variation of the spectroradiometer calibration during this period was less than 6%. It was observed that the calibration of the spectroradiometer varied by about 10% per degree of the temperature change in the temperature range between 22 and 25 degrees Celcius.