Semi-empirical method for the conversion of spectral UV global irradiance data into actinic flux

A large set of synchronous actinic flux and global irradiance data recorded in Thessaloniki (Greece) is used to determine the characteristics of the ratio of actinic flux to global irradiance as a function of global irradiance. The shape of this function is dominated by the values for clear and overcast sky conditions. Hence, model calculations carried out with the radiative transfer model UVSPEC for these two types of atmospheric conditions serve as boundary conditions. Then the ratio of actinic flux to global irradiance as a function of global irradiance is reproduced by an analytic function. This function depends on the boundary conditions and on one wavelength-dependent parameter. With the analytic function, the boundary conditions and the empirically determined parameter it is possible to convert any global irradiance data into actinic fluxes within reasonable uncertainties. This is demonstrated with data recorded in Thessaloniki (Greece), at the high mountain station Jungfraujoch (Switzerland), in Innsbruck (Austria) and in Weybourne (Great Britain). In the UVA and UVB for solar zenith angles up to about 60° the uncertainty in calculated actinic fluxes is 6% (twofold standard deviation, k=2) for the data taken from all four measurement sites. For larger solar zenith angles the uncertainties are on average 8% (k=2) in the UVB and about 12% (k=2) in the UVA. Afterwards photolysis rates for ozone and nitrogen dioxide derived from measured and calculated actinic fluxes are compared. The uncertainty in photolysis rates obtained from calculated actinic fluxes for ozone is about 7% (k=2) and for nitrogen dioxide about 12% (k=2).