Abstract

Tropospheric ozone is seen as as one of the target molecules for atmospheric chemistry modelling given its direct importance for air pollution and climate change, but also because we can readily measure its atmospheric concentration and its central role in the chemistry of the atmosphere. We use a chemistry transport model (GEOS-Chem) run in its regional configuration (0.25×0.3125 degree) over Beijing during the summer-time explore our ability to simulate O3 during the summer of 2017 taking advantage of observations made from the Atmospheric Pollution & Human Health in a Chinese Megacity programme. We constrain the emissions of the O3 precursor in the model to the observations and show that this gives a good simulation of O3 during the campaign, potentially giving us confidence to use the model to explore policy options to control the concentration of O3 in the city. However, when we evaluate the model using other observations (HONO, OH, HO2 , RO2, OH Reactivity, Aerosol surface area etc) our faith in the model getting the right O3 for the right reasons drops away. We highlight a number of science questions this comparison raises and suggest that there should be a degree of humility when models are used for evaluating policy options, at least in urban environments.