Evaluation of the simulated photolysis rates and their response to solar irradiance variability

The state of the stratospheric ozone layer and the temperature structure of the atmosphere are largely controlled by the solar spectral irradiance (SSI) through its influence on heating and photolysis rates. This study focuses on the uncertainties in the photolysis rate response to solar irradiance variability related to the choice of SSI data set and to the performance of the photolysis codes used in global chemistry-climate models. To estimate the impact of SSI uncertainties, we compared several photolysis rates calculated with the radiative transfer model libRadtran, using SSI calculated with two models and observed during the Solar Radiation and Climate Experiment (SORCE) satellite mission. The importance of the calculated differences in the photolysis rate response for ozone and temperature changes has been estimated using 1D radiative-convective-photochemical model. We demonstrate that the main photolysis reactions, responsible for the solar signal in the stratosphere, are highly sensitive to the spectral distribution of SSI variations. Accordingly, the ozone changes and related ozone-temperature feedback are shown to depend substantially on the SSI dataset being used, which highlights the necessity of obtaining accurate SSI variations. To evaluate the performance of photolysis codes, we compared the results of eight, widely used, photolysis codes against two reference schemes. We show that, in most cases, absolute values of the photolysis rates and their response to applied SSI changes agree within 30%. However, larger errors may appear in specific atmospheric regions because of differences, for instance, in the treatment of Rayleigh scattering, quantum yields or absorption cross-sections.