revised

Abstract

Spectral measurements of global UV irradiance recorded by Brewer spectrophotometers can be significantly affected by instrument-specific optical and mechanical features. Thus, proper corrections are needed in order to reduce the associated uncertainties to within acceptable levels. The present study aims to contribute to the reduction of uncertainties originating from changes in the Brewer internal temperature, which affect the performance of the optical and electronic parts, and subsequently the response of the instrument. Until now, measurements of the irradiance from various types of lamps at different temperatures have been used to characterize the instruments' temperature dependence. The use of 50 W lamps was found to induce errors in the characterization due to changes in the transmissivity of the Teflon diffuser as it warms up by the heat of the lamp. In contrast, the use of 200 or 1000 W lamps is considered more appropriate because they are positioned at longer distances from the diffuser so that warming is negligible. Temperature gradients inside the instrument can cause mechanical stresses which can affect the instrument's optical characteristics. Therefore, during the temperature-dependence characterization procedure warming or cooling must be slow enough to minimize these effects. In this study, results of the temperature characterization of eight different Brewer spec-trophotometers operating in Greece, Finland, Germany and Spain are presented. It was found that the instruments' response changes differently in different temperature regions due to different responses of the diffusers' transmittance. The temperature correction factors derived for the Brewer spectrophotometers operating at Thessaloniki, Greece, and Sodankylä, Finland, were evaluated and were found to remove the temperature dependence of the instruments' sensitivity.