Quality assurance of the solar UV network in the Antarctic

Measuring ultraviolet radiation in the Antarctic region, where weather conditions are extremely challenging, is a demanding task. Proper quality control of the measurements and quality assurance of the data, which are the basis of all scientific use of data, has to be especially well planned and executed. In this paper we show the importance of proper quality assurance and describe the methods used to successfully operate the NILU-UV multichannel radiometers of the Antarctic network stations at Ushuaia, 54°S, and Marambio, 64°S. According to our experience, even though multichannel instruments are supposed to be rather stable as a function of time, severe drifts can occur in the sensitivity of the channels under these harsh conditions. During 2000–2003 the biggest drifts were 35%, both at Ushuaia and Marambio, with the sensitivity of the channels dropping at different rates. Without proper corrections in the data, this would have seriously affected the calculated UV dose rates. As part of the quality assurance of the network a traveling reference NILU-UV, which was found to be stable, was used to transfer the desired irradiance scale to the site NILU-UV data. Relative lamp tests were used to monitor the stability of the instruments. Each site NILU-UV was scaled channel by channel to the traveling reference by performing solar comparisons. The method of scaling each channel separately was found to be successful, even though the differences between the raw data of the site NILU-UV and the reference instruments were, before the data correction, as much as 40%. After the correction, the mean ratios of erythemally weighted UV dose rates measured during the solar comparisons in 2000–2003 between the reference NILU-UV and the site NILU-UV were 1.007 ± 0.011 and 1.012 ± 0.012 for Ushuaia and Marambio, respectively, when the solar zenith angle varied up to 80°. These results make possible the scientific use of NILU-UV data measured simultaneously at quite different locations, e.g., the Antarctic and Arctic, and the method presented is also practicable for other multichannel radiometer networks.