Gas/Aerosol partitioning: a simplified method for global modeling

The main focus of this thesis is the development of a simplified method to routinely calculate gas/aerosol partitioning of multicomponent aerosols and aerosol associated water within global atmospheric chemistry and climate models. Atmospheric aerosols are usually multicomponent mixtures, partly composed of acids (e.g. H2SO4, HNO3), their salts (e.g. (NH4)2SO4, NH4NO3, respectively), and water. Because these acids and salts are highly hygroscopic, water, that is associated with aerosols in humid environments, often exceeds the total dry aerosol mass. Both the total dry aerosol mass and the aerosol associated water are important for the role of atmospheric aerosols in climate change simulations. Still, multicomponent aerosols are not yet routinely calculated within global atmospheric chemistry or climate models. The reason is that these particles, especially volatile aerosol compounds, require a complex and computationally expensive thermodynamical treatment. For instance, the aerosol associated water depends on the composition of the aerosol, which is determined by the gas/liquid/solid partitioning, in turn strongly dependent on temperature, relative humidity, and the presence of pre-existing aerosol particles. Based on thermodynamical relations such a simplified method has been derived. This method is based on the assumptions generally made by the modeling of multicomponent aerosols, but uses an alternative approach for the calculation of the aerosol activity and activity coefficients. This alternative approach relates activity coefficients to the ambient relative humidity, according to the vapor pressure reduction and the generalization of Raoult s law. This relationship, or simplification, is a consequence of the assumption that the aerosol composition and the aerosol associated water are in thermodynamic equilibrium with the ambient relative humidity, which determines the solute activity and, hence, activity coefficients of a multicomponent aerosol mixture. Thus, the necessary equilibrium equations can be solved analytically, so that numerical and therefore expensive iterative calculations are avoided. Subsequently, a new thermodynamic gas/aerosol partitioning model has been developed, called EQSAM (Equilibrium Simplified Aerosol Model). EQSAM has been compared with various other thermodynamical models presently in use, which shows that the results of EQSAM are well within the range produced by these more complex models. The application to global modeling further shows that EQSAM is indeed sufficiently fast and accurate. Especially the results of the global gas/aerosol partitioning calculations show that differences resulting from the thermodynamical treatment affect much less the aerosol composition compared to other, non-thermodynamical parameters, such as the model resolution or the boundary layer mixing scheme used. This indicates that the gas/aerosol partitioning calculations in atmospheric chemistry models are largely governed by transport processes, including meteorology, emission sources, as well as wet and dry deposition processes. Modeling results further indicate that gas/aerosol partitioning, especially at lower temperatures (during winter and nights), is of great importance for both the gas phase concentrations and the aerosol composition, including aerosol associated water. For instance, the mean surface gaseous nitric acid concentration is predicted to partition almost completely into the aerosol phase during winter and summer nights. This considerably increases the predicted aerosol load, compared to model calculations excluding gas/aerosol partitioning. This consequently affects the aerosol associated water (because the aerosol water is proportional to the amount of dissolved matter). Additionally, aerosol mass from gas/aerosol partitioning, such as ammonium nitrate, has a longer residence time than the precursor gases (NH3 and HNO3) and might, therefore, be subject to long-range transport from the sources. This is, for example, the case for ammonium nitrate originating from gas-to-particle conversion over northern India. Our model results indicate that these particles, through convective redistribution, can be transported at altitudes of 200-300 hPa as far as Europe during the Indian summer monsoon. Verification of these results, however, would require aircraft measurements, which are presently not available. Comparison with ground-based measurements indicates that the simplified aerosol module coupled to a global atmospheric chemistry model (TM3), for the considered ammonium/sulfate/nitrate/water system, yields realistic results at locations where ammonium nitrate is important. For remote locations, the comparison also indicates that it is important to account for other aerosol species such as sea salt and mineral dust. Although these compounds have not (yet) been included in the global gas/aerosol partitioning calculations with TM3, it seems to be possible to consider them with our simplified approach, as indicated by the results of box-model calculations.