Original paper

UV Index forecasting on a global scale

Staiger, Henning; Koepke, Peter

Meteorologische Zeitschrift Vol. 14 No. 2 (2005), p. 259 - 270

published: May 10, 2005

DOI: 10.1127/0941-2948/2005/0029

BibTeX file

ArtNo. ESP025011402021, Price: 29.00 €

Download preview PDF Buy as PDF


A method is presented to forecast the UV Index (UVI) on a global scale for up to 3 days, taking into account ozone, aerosol, cloudiness, altitude and albedo. Particular consideration is made for the adjustment of the UVI to altitude and for seasonal changes in aerosol optical depth. The UVI describes the level of erythemal weighted solar irradiance and consequently is an indicator of the potential for skin damage by UV radiation. Thus the UVI is an important vehicle to raise public awareness of the risks of UV radiation, and forecasted UVI may help to adapt the human behaviour. The UVI is forecasted by the German Meteorological Service (DWD) using adequate values of the relevant atmospheric and surface properties. The UVI is calculated via lookup tables valid for sea level, clear sky conditions, an aerosol optical depth at 550 nm (AOD) of 0.2, and an albedo of 3 %. The lookup tables have a high resolution in total column ozone and sun zenith angle. The ozone values are forecasted within the global numerical weather prediction model GME of DWD. The module structure of the UV model allows for a subsequent adjustment of UVI to changes in AOD, to the topography, to the albedo of predicted snow, and to forecasted cloudiness, the latter by using empirical cloud modification factors. The factor applied to account for changes in AOD depends on solar zenith angle and on aerosol single scattering albedo, and that for altitude effects additionally on AOD. Since forecasts of AOD are not yet available, regional monthly mean values of AOD have been derived from MODIS and TOMS measurements. The forecasted UVI for all effective atmospheric conditions are compared with measurements at 12 locations in Europe for the summer seasons 1999 and 2003. 80 % of the forecasts for all atmospheric conditions were within the range of ┬▒1 UVI; the predominant error can be attributed to cloudiness.