Numerical models of atmospheric composition will play an important role in POLARCAT, both in addressing the key scientific issues of the project using the measurements made, and in operational support of the field campaign periods. A hierarchy of models will be used, ranging from process-based photochemical box models capable of detailed representation of reactive radical chemistry and aerosol microphysics to 3-D Chemical Transport Models (CTMs) and General Circulation Models (GCMs) able to resolve the interaction of aerosol, chemical and dynamical processes and to determine the impacts on regional or global climate.

The Arctic presents a unique photochemical environment characterized by low UV intensities, cold temperatures, halogen radical chemistry, and ice-covered surfaces, and strongly stratified conditions lead to aging times of several weeks. The dynamical and chemical conditions here are challenging to model, and the current generation of global CTMs show widely divergent behaviour over polar regions, as seen in recent studies of surface ozone (Stevenson et al., 2005). POLARCAT observations will provide a valuable test of the ability of CTMs to simulate the chemical and microphysical evolution of air masses in the region, and in combination with more detailed process-based box model analysis will contribute to improved treatments of stratification, slow and/or novel chemistry and surface processes. Reducing uncertainties in modelling the Arctic region is an important goal of the project and will contribute to an improved understanding of the impacts on regional chemistry and climate.

Analysis of POLARCAT observations will make use of box models, trajectory models and CTMs. Photochemical box (0-D) models including detailed representation of chemical processes (Crawford et al., 1999; Evans et al., 2003) will be used to interpret aircraft observations in terms of radical chemistry in the Arctic with particular attention to processes involving halogen radicals and heterogeneous reactions. Another class of box models including detailed representation of aerosol microphysics will be needed to describe the unique Arctic environment for nucleation and growth of particles. 3-D particle dispersion models will be used to derive flow climatologies for the Arctic region, and to determine source-receptor relationships to aid in interpretation of aircraft observations (Stohl et al., 2003). Photochemical box models following air mass trajectories will be used to trace the chemical evolution of air masses entering and leaving the region (e.g., Methven et al., 2003). Finally, 3-D chemical transport models (CTMs) will integrate the information from the surface, aircraft, and satellite platforms in terms of the constraints that they provide on source regions affecting the Arctic atmosphere, transport between mid-latitudes and the Arctic , large-scale vertical motions, and the chemical and aerosol evolution coupled with these dynamical processes. The CTMs will need to be at least hemispheric in scale to describe the range of motions affecting Arctic atmospheric composition.

Beyond their value for post-mission data analysis, the CTMs will be of critical importance for the planning and execution phases of the POLARCAT field missions. Model simulations conducted before the mission using hindcast meteorological fields, and evaluated with pre-existing surface and satellite observations in the Arctic , will provide critical input for selecting optimal mission time windows, bases of operations, and flight regions. The hindcast simulations will be used to develop a menu of flights to guide mission execution. During the execution phase of the missions, the same CTMs driven by meteorological forecasts will provide chemical forecasts to guide the aircraft on a day-to-day basis. These forecasting activities will involve a number of CTMs to provide different perspectives and to address the broad range of mission objectives. This hindcast-forecast methodology has been applied very successfully in a number of recent aircraft measurement campaigns including TRACE-P (Jacob et al., 2003; Kiley et al., 2003), ITCT-2K2 (Parrish et al., 2004; Forster et al., 2004) and the recent ICARTT/INTEX campaign.

In the 20th century, the anthropogenic emissions resulted in a general increase of the aerosol load (sulfate as well as carbonaceous) in North America , Europe , and East Asia . In the first part of the 21st century, according to the IPCC SRES scenario A1B, the most polluted regions are found at lower latitudes (Brazil, Africa, the Arabian Peninsula, India and China) whilst sulphate and carbonaceous aerosols have both decreased in North America and Europe. To study scenarios of possible future climate conditions and emission distributions, fully coupled aerosol-chemistry-climate models are needed. Transient as well as time-slice simulations using these models will be performed to study how Arctic Haze will develop in the future. Furthermore, these simulations will be used to identify possibly important feedback processes in the climate system involving aerosol and pollution transport at high latitudes.