Comparison of two algorithms for calculating photolysis frequencies including the effects of clouds

Abstract

Two complex models to determine photolysis frequencies for chemical transport models are used to study the effects of input data and the consideration of relevant physical processes on the derived photolysis frequencies. Within the model CTM photolysis frequencies are calculated on a coarse latitudinal grid with climatological input data (monthly mean or seasonal mean values) and are then interpolated linearly in space to derive photolysis frequencies for each grid cell of the chemical transport model. These clear sky photolysis frequencies are then corrected to account for cloud effects. The model STAR calculates photolysis frequencies for each grid cell considering the relevant physical processes on the basis of actual profiles computed with a mesoscale meteorological model and other available geophysical data. The comparison of the O3 and NO 2 photolysis frequencies shows that the approach used within the CTM model compares to STAR only under certain conditions, as climatological input data can be less suitable for episodic photolysis frequencies calculations. The ozone column content significantly alters the photolysis frequency of ozone itself and climatological Dobson data limit the quality of the calculations. The temperature dependence of the quantum yields and the absorption cross sections lead to increased uncertainties when climatological temperature profiles are used. This is especially the case for sunrise/sunset conditions. The use of one surface albedo for all landuse types and seasons within the CTM model restricts the quality of the calculations close to the surface. If clouds are present the CTM model over-/underestimates the cloud effects on the photolysis frequencies and differences up to an order of magnitude are found for below cloud values.