Direct Arctic aerosol effects

Anthropogenic aerosols cause a brownish Arctic haze during late winter / early spring (AMAP, 1998, Ch. 9) with sulphate and soot aerosols also containing lightabsorbing BC. In Arctic haze, BC is frequently internally mixed with scattering particles in a sizerange which efficiently influence solar radiation (e.g. Rosen et al. 1981; Heintzenberg, 1982; Valero et al, 1983; Rosen and Hansen, 1984). Absorption is important relative to scattering in the Arctic due to a high surface albedo, with measured heating rates of 0.1-0.2 K/day during Arctic haze (Valero et al., 1989). Sunlight absorbed by soot deposited on snow may furthermore speed up melting of Arctic ice and snow, with a risk of positive feed-back. Clarke and Noone’s (1985) chemically and optically anaysed Arctic snow-samples indicated considerable effects. Grennfell et al. 9 (2002), however, estimated small albedo effects during SHEBA, but recent model-calculations give large climate effects (Hansen and Nazarenko; 2004; Koch and Hansen; 2005). The latter paper emphasizes soot-sources in south-east Asia – a conclusion which is challenged by others (Stohl, 2006).

Indirect Arctic aerosol effects

The indirect aerosol effect on global climate is regarded to only be exceeded in absolute value by greenhouse gas forcing (Lohmann and Feichter, 2005). Interactions between aerosols and clouds are very uncaertain. Ice clouds are particularly uncertain and ice-nuclei poorly understood. Observations analyzed by Garrett et al. (2004) indicate that Arctic clouds are more sensitive to anthropogenic pollution than clouds in other regions, and there are strong indications that clouds in the Arctic have different microphysical and radiative properties than clouds elsewhere. Longwave radiative effects of clouds tend to dominate, and result in a positive radiative forcing at the surface (Intrieri et al., 2002a). On the other hand, Girard et al. (2005) presented evidence of dehydration caused by sulfate aerosols in Arctic air masses. Ice nuclei and humidity are reduced from the air and leads to a negative radiative forcing. In the SHEBA experiment (e.g., Curry et al., 1996), it was found that supercooled cloud droplets are common in the Arctic, even at temperatures lower than -20°C. They often occur in between layers containing ice crystals (Intrieri et al., 2002b). These remarkable conditions are poorly understood, but are probably related to the very stable troposphere in the Arctic winter. It is known from Arctic haze studies that this pollution from different remote source regions (North America, Europe, Siberia, East-Asia) can be vertically separated by intrutions of cleaner air of maritime origins (North Atlantic). Soot and mineral dust can be effective ice nuclei, but it is not known how the ice nuclei concentrations vary depending on pathways to the Arctic. It is of great interest to explain the observed cloud microphysical conditions through a careful characterization of the aerosols in the different layers, as well as their origin.

Model tools

The main model tool for studying climate effects of Arctic aerosols is the Osloversion of the Community Atmosphere Model (CAM-Oslo) of the National Center for Atmospheric Research in U.S.A. Also its 1-dimensional single column version (SCAM) is available. CAM-Oslo calculates aerosol lifecycles and links aerosol number, size, and composition to radiation and clouds. Iversen and Seland (2003), Kirkevåg and Iversen (2002), and Kristjansson (2002) describe earlier versions Updated descriptions are underway (Kirkevåg et al, 2005). The model presently calculates sea-salt, mineral dust, sulphate, BC, and organic carbon aerosols, and parameterizes optical properties and water activity by tabulations. CAM-Oslo will be used stand-alone for detailed Arctic aerosol-cloud-radiation studies, and for a small selection of climate equilibria response calculations when coupled to a “slab-ocean” model. CAM-Oslo calculations can also be controled by observed meteorological data through “nudging”.

CAM-Oslo tends to underestimate Arctic aerosol concentrations in the cold seasons, and potential reasons for this will be investigated in this project. Remedies have been proposed, including SO2- oxidation on ice-crystals (Rotstayn and Lohmann, 2002), and improved fog and stratus at low temperatures. Model tests (Seland, 2006, Pers. Comm.) also show sensitivity of remote anthropogenic aerosols w.r.t. processes such as below-cloud scavenging.

Modeling of the direct and indirect climate effects of aerosols with Oslo versions of NCAR GCMs, has been carried out for several years at MetOs-UiO (Kirkevåg and Iversen, 2002; Kristjánsson, 2002; Iversen et al., 2005; Kristjansson et al, 2005; Storelvmo et al., 2005). The model and several aspects of the aerosol-cloud-radiation calculations have been subject to thorugh scrutiny in the AEROCOM intercomparison project (http://nansen.ipsl.jussieu.fr/AEROCOM/data.html) (Kinne et al., 2005; Textor et al, 2005; Penner et al., 2006).

Calculations of Arctic aerosol burdens will be complemented with an off-line chemical transport model (CTM), the hemispheric Unified EMEP model (Simpson et al., 2003; Tarrason et al., 2003; Wind et al., 2003) with a higher spatiall resolution (50 km vs. mainly T42 in CAM-Oslo). It is also particularly designed for source-allocation, and is suited for studying long term calculations of aerosols.

Experiments

Regular and campaign observations during IPY (e.g. the ASTAR-campaign in April 2007 and the summer 2008 campaign) will be used for model validation, in addition to more regular data in the Arctic and other regions. The ASTAR campaign is based in Longyearbyen and Ny Ålesund, and involves two well-equipped aircrafts making simultaneous measurements of aerosol properties and cloud physical properties. The aircraft measurements will be complemented by ground-based instruments (e.g. Barrow, Ny Ålesund, Alert) including a troposphere LIDAR, a spectrophotometer, and surface chemical measurements. Calculations will also be compared with satellite data when they become available (CALIPSO), and with transport calculations in WP1 and WP4.

Source regions for aerosol layers in the Arctic troposphere will mainly be studied with the Hemispheric EMEP model, after evaluation of this model in WP4. This will allow distinguishing the influence of natural sources (biomass burning and secondary organic aerosols), and the influence of anthropogenic aerosol sources from primary sources and precursor gases. Updates of emission data of BC (from e.g. D. Lavoue, Canada) will be investigated.

We will investigate the role of uncertain parameters in CAM-Oslo w.r.t modelling Arctic haze and Arctic aerosols. The calculation of surface albedo in CAM-Oslo will take ino account soot deposition (e.g. Warren, S.G. and Wiscombe, W.J., 1980). In climate equilibria runs coupled with a slab ocean, impacts on snow- and ice-melt in the Arctic will be estimated and studied. Source-allocated soot deposited in the Arctic will mainly be made with the Hemispheric EMEP model as calculations of both anthropogenic and natural aerosols to the Arctic, with particular focus on carbonaceous particles.

In CAM-Oslo a parameterization of the indirect effect of ice clouds is being developed in collaboration with scientists at ETH in Zürich. The ASTAR campaign in March/April 2007, based in Longyearbyen and Ny Ålesund, will provide an excellent test-bed for the new parameterization. The campaign will involve two well-equipped aircraft making simultaneous measurements of aerosol properties and cloud physical properties. In particular the POLAR2 aircraft carries FSSP and 2D-C probes as well as a cloud particle imager (CPI), which together will provide detailed information on size distributions and shapes of cloud droplets and ice crystals. At the same time, the POLAR2 and Falcon aircraft will carry several instruments for characterizing the aerosols. These measurements will be complemented by ground-based instruments including a troposphere LIDAR, a spectrophotometer and surface chemical measurements. The ASTAR data set will be used to validate the model assumptions, especially concerning ice nuclei. One intriguing possibility, suggested by Bailey and Hallett (2002), is that ice crystal shape is related to the type of ice nuclei present. This will be explored, along with the implications for the cloud radiative forcing, which is strongly dependent on crystal shape (Kristjánsson et al., 2000).