Journal cover Journal topic
Atmospheric Chemistry and Physics An interactive open-access journal of the European Geosciences Union
Atmos. Chem. Phys., 8, 3107-3118, 2008
© Author(s) 2008. This work is distributed
under the Creative Commons Attribution 3.0 License.
23 Jun 2008
Long-term solar UV radiation reconstructed by ANN modelling with emphasis on spatial characteristics of input data
U. Feister1, J. Junk2, M. Woldt3, A. Bais4, A. Helbig5, M. Janouch6, W. Josefsson7, A. Kazantzidis4, A. Lindfors8, P. N. den Outer9, and H. Slaper9 1Deutscher Wetterdienst, Richard Aßmann Observatorium Lindenberg, Am Observatorium 12, 15848 Lindenberg, Germany
2Public Research Centre – Gabriel Lippmann, Department of Environment and Agro-Biotechnologies (EVA), 41, rue du Bril, 4422 Belvaux Grand-Duchy of Luxembourg, Luxembourg
3Brandenburgische Techn. Univ. Cottbus, Lehrstuhl für Umweltmeteorologie, Burger Chaussee 2, 03044 Cottbus, Germany
4Aristotle University of Thessaloniki, Laboratory of Atmospheric Physics, 54124, Thessaloniki, Greece
5Universität Trier, Fachbereich VI, Fach Umweltmeteorologie, 54286 Trier, Germany
6Solar and Ozone Observatory, Czech Hydrometeorological Inst., Hvezdarna 456, 500 08 Hradec Kralove 8, Czech Republic
7Swedish Meteorological and Hydrological Institute, 60176, Norrköping, Sweden
8Finnish Meteorological Institute, P.O. BOX 503 (Erik Palmenin aukio 1), 00101 Helsinki, Finland
9National Institute for Public Health and the Environment, A. van Leeuwenhoeklaan 9, P.O. Box 1, 3720 BA Bilthoven, The Netherlands
Abstract. Artificial Neural Networks (ANN) are efficient tools to derive solar UV radiation from measured meteorological parameters such as global radiation, aerosol optical depths and atmospheric column ozone. The ANN model has been tested with different combinations of data from the two sites Potsdam and Lindenberg, and used to reconstruct solar UV radiation at eight European sites by more than 100 years into the past. Special emphasis will be given to the discussion of small-scale characteristics of input data to the ANN model.

Annual totals of UV radiation derived from reconstructed daily UV values reflect interannual variations and long-term patterns that are compatible with variabilities and changes of measured input data, in particular global dimming by about 1980/1990, subsequent global brightening, volcanic eruption effects such as that of Mt. Pinatubo, and the long-term ozone decline since the 1970s. Patterns of annual erythemal UV radiation are very similar at sites located at latitudes close to each other, but different patterns occur between UV radiation at sites in different latitude regions.

Citation: Feister, U., Junk, J., Woldt, M., Bais, A., Helbig, A., Janouch, M., Josefsson, W., Kazantzidis, A., Lindfors, A., den Outer, P. N., and Slaper, H.: Long-term solar UV radiation reconstructed by ANN modelling with emphasis on spatial characteristics of input data, Atmos. Chem. Phys., 8, 3107-3118, doi:10.5194/acp-8-3107-2008, 2008.
Publications Copernicus