The fourth phase of the Air Quality Model Evaluation International Initiative (AQMEII) focuses on deposition processes. Since 2008, AQMEII has analysed a number of critical aspects of regional-scale air quality models that require international collaboration and evaluation. Each of the previous three phases have resulted in multi-paper “special issues” in the peer-reviewed literature (two in Atmospheric Environment and one in ACP). AQMEII leverages the long-term experience in model development and application of the two largest communities of such models, namely the North American community and the European one. These two communities have been engaged in AQMEII model evaluation activities that tackled issues such as operational and probabilistic evaluations (phase 1), online model coupling between air quality and meteorology (phase 2), intercontinental transport of air pollutants (phase 3, ACP SI “Global and regional assessment of intercontinental transport of air pollution: results from HTAP, AQMEII and MICS”, 41 papers), and now dry and wet deposition processes (phase 4).

The subject tackled in phase 4 is central to air quality modelling at any scale, and it has never been examined in the systematic and the detailed manner currently underway in AQMEII4. Given the complexity of deposition processes and the variety of approaches adopted by different groups, we opted for an analysis that comprises the individual deposition modules and the full regional-scale models that used those modules as part of a more complex computational system. As customary, for AQMEII two regional modelling domains are considered, a European and a North American one, and simulations are underway for 2 representative years for the processes of interest in the two continents. The process-oriented character of the activity is defined already in the selection of the simulation year. For the North American domain, the air quality in 2010 and 2016 will be simulated to allow comparisons across emissions changes, while for Europe 2009 and 2010 are the designated years to allow comparisons across meteorologically different years. The community presently participating is composed of 15 regional modelling teams and six deposition modules. Subgroups of regional-scale models may be using the same modelling system with a variety of parameterization implementations, or use original models developed in-house. A key feature of the AQMEII multi-model exercise is the submission of all model output on a common grid and at specified observation stations, to a common database at JRC in Ispra, Italy, where a common set of analysis tools which participants can use is available. The variety of the tools used makes possible the comparison of all models with measurements and of specific models with largely used community systems. It also makes it possible to work within the sub-community to exercise different features or modelling options of the community models. Whilst for regional-scale models the evaluation will be based on data on a wide range of pollutants in gas and particle form gathered from standard monitoring networks and soundings in the two continents, in the case of point models, ozone will be the focus, and data have been gathered from sites where high-quality instrumentation has been deployed. Analysis of the ozone dry deposition modules driven by observed site-specific meteorological and biophysical data (i.e., the point models) will focus on process-oriented model evaluation and intercomparison.

As mentioned the central topic of the issue is the deposition processes, but differently from the past a deeper level of analyses has been added in AQMEII4. That pertains to the addition of diagnostic analysis to each of the participating models. These diagnostics are allowing participants to compare the parameterizations for deposition pathways across all participating systems in an unprecedented level of detail, with comparisons between individual conductance and resistances being carried out for the first time. The diagnostics have been mapped to a common set of land-use types, in order to allow a common analysis of the different deposition parameterizations for a given land-use type.

The AQMEII4 activity goes beyond the comparison of the deposited quantities as has been customary so far and will allow the gathering of insight into the reason for model differences and the efficacy of some schemes compared to others through the comparison with measurements. For the first time the diversity with which the participating models (both point models and the full regional-scale models) describe the underlying surface on which the deposition occurs will be evaluated. This is therefore not a mere model evaluation exercise but one that aims to tackle in a multi-model fashion the most difficult of the four pillars of model evaluation (Dennis et al., 2010), i.e. diagnostic evaluation. The special issue therefore includes papers on process-oriented point model comparisons against deposition supersite data, detailed process-oriented model inter-comparisons of the regional-scale output, and the potential application of model results for ecosystem assessments.