Evaluating the impact of chemical boundary conditions on near surface ozone in regional climate–air quality simulations over Europe

• Either constant or varying chemical lateral boundary conditions (LBCs) were applied.
• Evaluation of RegCM3/CAMx simulations against EMEP ozone measurements for 1996–2000
• Varying LBCs improved the model performance with respect to seasonal variability.
• Varying chemical LBCs resulted in improvement of model performance for all seasons.
• Slight improvement in regional simulation in comparison to the global simulation

A modeling system based on the air quality model CAMx driven off-line by the regional climate model RegCM3 is used for assessing the impact of chemical lateral boundary conditions (LBCs) on near surface ozone over Europe for the period 1996–2000. The RegCM3 and CAMx simulations were performed on a 50 km × 50 km grid over Europe with RegCM3 driven by the NCEP meteorological reanalysis fields and CAMx with chemical LBCs from ECHAM5/MOZART global model. The recent past period (1996–2000) was simulated in three experiments. The first simulation was forced using time and space invariant LBCs, the second was based on ECHAM5/MOZART chemical LBCs fixed for the year 1996 and the third was based on ECHAM5/MOZART chemical LBCs with interannual variability. Anthropogenic and biogenic emissions were kept identical for the three sensitivity runs.In order to evaluate the ability of the RegCM3/CAMx modeling system and assess the impact of varying chemical LBCs, simulated surface ozone concentrations are compared against measurements from the EMEP network using various statistical metrics. The evaluation indicates that implementation of time and space variant chemical LBCs of a global chemistry transport model (CTM) improves the RegCM/CAMx performance with respect to seasonal variability, especially at stations close to the borders of the model domain over north and northwestern Europe. The modeling system reproduces the seasonal variability of ozone when LBCs from ECHAM5/MOZART model are applied for the vast majority of stations. On a seasonal basis, the varying chemical LBCs resulted in improvement of both model bias and variance of simulated versus observed ozone for all seasons. Finally taking into account all the statistical measures assessed for the comparison of RegCM3/CAMx and ECHAM5/MOZART with observed ozone, a slight improvement for the regional air quality model is indicated but this finding should be perceived with caution as the regional scale simulation and the global scale simulation have a different meteorological forcing.