Effect of cloud representation errors on simulated global atmospheric oxidant levels

Solar radiation drives reactive atmospheric chemistry through photolysis reactions. The rates of these photolysis are strongly influenced by clouds and their optical properties, which are notoriously challenging to accurately represent in global atmospheric models. To address the potential errors that simulated clouds might cause for photolysis rates and atmospheric oxidants, we recently developed SatJ, a global 3-D dataset of photolysis rates constructed from satellite observations at cloud-resolving scales. In this work, we use SatJ to evaluate the photolysis rates simulated in a global atmospheric chemistry model: GEOS-Chem driven by MERRA-2 meteorology. We find modest and variable biases in monthly mean photolysis rates. Using regional and monthly scale factors to adjust the simulated photolysis rates to match SatJ, we explore how atmospheric oxidant levels, principally hydroxyl and ozone, change in the improved model.