The regional climate model RegCM4 extended with the land surface model CLM4.5 was coupled to the chemistry transport model CAMx to analyze the impact of urban meteorological forcing on surface fine aerosol (PM2.5) concentrations for summer conditions over the 2001-2005 period, focusing on the area of Europe. Starting with the analysis of the meteorological modifications caused by urban canopy forcing, we found a significant increase in urban surface temperatures (up to 2-3 K), a decrease of specific humidity (by up to 0.4-0.6 gkg(-1)), a reduction of wind speed (up to -1 ms(-1)) and an enhancement of vertical turbulent diffusion coefficient (up to 60-70 m(2)s(-1)).
These modifications translated into significant changes in surface aerosol concentrations that were calculated by a "cascading" experimental approach. First, none of the urban meteorological effects were considered.
Then, the temperature effect was added, then the humidity and the wind, and finally, the enhanced turbulence was considered in the chemical runs. This facilitated the understanding of the underlying processes acting to modify urban aerosol concentrations.
Moreover, we looked at the impact of the individual aerosol components as well. The urbanization-induced temperature changes resulted in a decrease of PM2.5 by -1.5 to 2 mu gm(-3), while decreased urban winds resulted in increases by 1-2 mu gm(-3).
The enhanced turbulence over urban areas resulted in decreases of PM2.5 by 2 mu gm(-3). The combined effect of all individual impact depends on the competition between the partial impacts and can reach up to 3 mu/gm3 for some cities, especially when the temperature impact was stronger in magnitude than the wind impact.
The effect of changed humidity was found to be minor. The main contributor to the temperature impact is the modification of secondary inorganic aerosols, mainly nitrates, while the wind and turbulence impact is most pronounced in the case of primary aerosol (primary black and organic carbon and other fine particle matter).
The overall as well as individual impacts on secondary organic aerosol are very small, with the increased turbulence acting as the main driver. The analysis of the vertical extent of the aerosol changes showed that the perturbations caused by urban canopy forcing, besides being large near the surface, have a secondary maximum for turbulence and wind impact over higher model levels, which is attributed to the vertical extent of the changes in turbulence over urban areas.
The validation of model data with measurements showed good agreement, and we could detect a clear model improvement in some areas when including the urban canopy meteorological effects in our chemistry simulations.