A Critical Assessment of Two Types of Personal UV Dosimeters Gunther Seckmeyer* 1 , Marcus Klingebiel 1 , Stefan Riechelmann 1 , Insa Lohse 1 , Richard L. McKenzie 2 , J. Ben Liley 2 , Martin W. Allen 3 , Anna-Maria Siani 4 and Giuseppe R. Casale 4 1 Institute of Meteorology and Climatology, Leibniz University of Hannover, Hannover, Germany 2 National Institute of Water & Atmospheric Research (NIWA), Lauder, Central Otago, New Zealand 3 Department of Electrical and Computer Engineering, MacDiarmid Institute for Advanced Materials and Nanotechnology, University of Canterbury, Christchurch, New Zealand 4 Sapienza University of Rome, Rome, Italy Received 10 June 2011, accepted 6 October 2011, DOI: 10.1111/j.1751-1097.2011.01018.x ABSTRACT Doses of erythemally weighted irradiances derived from poly- sulphone (PS) and electronic ultraviolet (EUV) dosimeters have been compared with measurements obtained using a reference spectroradiometer. PS dosimeters showed mean absolute devia- tions of 26% with a maximum deviation of 44%, the calibrated EUV dosimeters showed mean absolute deviations of 15% (maximum 33%) around noon during several test days in the northern hemisphere autumn. In the case of EUV dosimeters, measurements with various cut-off filters showed that part of the deviation from the CIE erythema action spectrum was due to a small, but significant sensitivity to visible radiation that varies between devices and which may be avoided by careful preselec- tion. Usually the method of calibrating UV sensors by direct comparison to a reference instrument leads to reliable results. However, in some circumstances the quality of measurements made with simple sensors may be over-estimated. In the extreme case, a simple pyranometer can be used as a UV instrument, providing acceptable results for cloudless skies, but very poor results under cloudy conditions. It is concluded that while UV dosimeters are useful for their design purpose, namely to estimate personal UV exposures, they should not be regarded as an inexpensive replacement for meteorological grade instru- ments. INTRODUCTION Ultraviolet radiation (UV) from the sun causes a considerable global disease burden, including acute and chronic health effects on the skin, eyes and immune system. Worldwide, an estimated 56 000 deaths a year are attributable to ultraviolet radiation, most of which are due to malignant melanoma (1). Much of the UV-related illness and death can be avoided through a series of simple prevention measures. It is crucial for humans to prevent over-exposure to UV radiation that leads to sunburn in the short term and increased risks of skin cancers, accelerated ageing of the skin and various eye diseases in the longer term (2). Moreover, it has been recently (3) shown that erythema does not provide a complete information of cutan- eous damage, which could occur at suberythemal doses. On the other hand, UV is essential for the production of vitamin D in the human body. Emerging evidence suggests an association between vitamin D levels and risk indicators relating to some cancers, cardiovascular disease and multiple sclerosis among others, along with the established link with musculoskeletal health (3–5). Therefore, there is a need for studies into the required UV dose for maintaining optimal levels of vitamin D, whereas avoiding the damaging effects of over-exposure. The ambient UV dose rate (incident irradiance on a horizontal, unshaded surface weighted with the action spec- trum of a determined biologic response) and ambient UV dose (integrating the dose rate over the time the surface is exposed) can be obtained by spectroradiometers, broad-band and narrow-band radiometers. The quantitation of human UV exposure is a more complex issue as it is related to the dose received by differently oriented surfaces in different settings and posture. To achieve this several types of dosimeters have been developed. Historically, the most widely used types are based on changes in the optical properties of certain materials (e.g. PS) after exposure to UV radiation. More recently, electronic ultraviolet (EUV) dosimeters using wide band-gap photodiode sensors have been developed. The aim of this study was to compare the performance of PS dosimeters, in use for over 30 years, and the recently developed EUV dosimeters, to assess their relative advantages and limitations. MATERIALS AND METHODS Reference spectroradiometer. The reference spectroradiometer (RS) used in this study has been developed and characterized by the Institute of Meteorology and Climatology in Hannover (Germany). It comprises a double monochromator (DTMc300, Bentham, Reading, UK), incorporating diffraction gratings with rulings of 2400 lines mm )1 , and three detectors with the capability to scan spectra between 200 and 2500 nm. The calibration of the RS is based on a 1000 W tungsten halogen lamp, which was calibrated by the Physikalisch- Technische Bundesanstalt (PTB). The RS is described in detail elsewhere and has been previously intercompared with other high accuracy spectroradiometers (6–8). The RS complies with the require- ments for Network for the Detection of Atmospheric Composition Change (NDACC) instruments (8) and with the requirements for type 2 instruments set up by the World Meteorological Organization *Corresponding author email: seckmeyer@muk.uni-hannover.de (Gunther Seckmeyer) Ó 2011 Wiley Periodicals, Inc. Photochemistry and Photobiology Ó 2011 The American Society of Photobiology 0031-8655/12 Photochemistry and Photobiology, 2012, 88: 215–222 215