A cautious note advocating the use of ensembles of models and driving data in modeling of regional ozone burdens

We investigate the performance of two widely used chemistry-transport models (CTMs) with different chemical mechanisms in reproducing the ambient maximum daily 8-h average ozone (MDA8 O _3 ) burden over Central Europe. We explore a base case setup with boundary conditions (BC) for meteorology from the ERA-Interim reanalysis and chemical BC from CAM-Chem as well as effects of alterations in these BC based on global model fields. Our results show that changes in meteorological BC strongly affect the correlation with observations but only marginally affect the model biases, while changes in chemical BC increase model biases while correlation patterns remain largely unchanged. Furthermore, our study highlights that CTM choice (and choice of chemical mechanism) has a similar or even larger impact on MDA8 O _3 levels as the impact of altered BC. In summary, our study calls for a multi-model strategy combining different CTM and BC combinations to explore the bandwidth of MDA8 O _3 distributions and thus uncertainty in hindcasts and future projections, in analogy to climate studies considering ensemble simulations under the same anthropogenic emissions but with slightly different initial conditions.