On measurements and modeling of ultraviolet radiation with focus on the Antarctic

The aim of the work was to study, on the basis of literature and experiments, the aspects of measuring and modeling ultraviolet radiation. For measurements, both the spectral and the non-spectral approaches were included. For modeling, physical and statistical models were applied. Thereafter the satellite UV estimates, i.e. spaceborn spectrometer measurements on solar radiation combined with physical UV modeling, were shortly introduced. Case studies were carried out for each of the five cases of spectral and non-spectral measurements, physical modeling, statistical modeling, and satellite UV. These experiments included data from a Nordic Ozone Group Intercomparison Campaign (NOGIC-2000), Finnish-Spanish-Argentinian Antarctic UV-network, US National Science Foundation (NSF) network, and satellite data from the Ozone Monitoring Instrument (OMI) onboard the NASA Earth Observing System (EOS) Aura spacecraft. When using spectral data of instruments to which lower quality radiometers are validated, it was unexpectedly found that, even though corrected for known major errors, the data may suffer from diurnal discrepancies of 2 - 9 %. Unless taken into account, these differences may cause problems in the UV data intercomparisons. For filter radiometer data, it was found that under harsh climate conditions of the Antarctic the responsitivity of individual channels may drop by up to 61 % during the six years of operation, but using biweekly lamp calibrations and frequent measurements with the traveling reference instrument, the results can be corrected. Physical modeling was found as a useful tool to study and show the effect of various parameters affecting UV radiation. Statistical modeling offered means to study the variability and trends, for example. When comparing Antarctic region satellite-based OMI daily UV dose data with the measured in situ data, the root mean square (RMS) of difference was found smaller for Ushuaia than for Marambio, revealing possible problems of OMI surface UV algorithm for albedo, and cloudiness, for example. These results show that each method to detect UV radiation has their own advantages and disadvantages. For Antarctic regions, the in situ measurements and statistical analysis of data are challenged by quality assurance in the harsh climate and remote locations; and the model calculations by the availability of proper input data of total ozone, albedo, aerosols and cloudiness; and satellite data analysis by high solar zenith angles, high albedo, as well as spatial and temporal resolutions. Combining various sources of information can provide with the best knowledge and understanding when studying the UV radiation, especially in the Antarctic region