NETCARE has been configured around four research activities that address key uncertainties in the field. The first three are focused on specific aerosol-climate connections that remain poorly characterized, and will be addressed through a variety of observational approaches. The fourth activity integrates the results from Activities 1-3, approaching the subject from a comprehensive modeling perspective so as to provide a broad assessment of aerosol climate effects.

It is widely recognized that black carbon (BC) aerosol directly warms the atmosphere, especially in snow-covered regions where the high surface albedo accentuates its effects. Unlike other climate warming agents such as the long-lived greenhouse gases, the radiative effects of BC on the atmosphere are short-lived, determined by its multi-day lifetime for deposition to the surface. Given the large uncertainties in models simulating the dynamic radiative effects of BC [Koch et al., 2009], there is considerable interest in better identifying sources, atmospheric loadings, and loss processes for such light-absorbing aerosols. Indeed, measurements of the size distribution of BC aerosol are highly limited. Major sources are industry and BB, either through wildfires or biofuels. Anthropogenic versus natural input is difficult to assess, given the increasing prevalence of forest fires arising through regional warming, forest management processes and forest die-off (such as that associated with the pine bark beetle infestation in British Columbia). Whereas there have been many measurements of BC in heavily populated regions, there is a notable paucity of data from remote environments. This issue is of particular importance in snow-covered regions where BC deposition can affect the albedo of the snowpack and regional radiative forcing [McConnell et al., 2007]. 

Key questions to be addressed:

• How do BC and OC loadings from biomass burning compare with anthropogenic BC and OC over the Arctic and Western Canada?
• What is the relative importance of the mechanisms for BC and OC deposition to Arctic snow and ice?  In particular, what is the vertical distribution of BC in the Arctic atmosphere?  Is there evidence for dry deposition in the boundary layer and/or via ice clouds?
• What are the levels and sources of BC and OC, including brown carbon, in snow? What are the implications of carbonaceous loadings in snow on radiative forcing?
• What will be the consequences of reducing or eliminating BC?

Small changes in cloud properties can cause a large change in radiative forcing, comparable in magnitude to the radiative forcing caused by an increase in anthropogenic greenhouse gases. Currently, clouds provide some of the greatest uncertainties in predictions of climate change [Lohmann and Feichter, 2005; IPCC, 2007]. This is in large part because the properties of clouds and their formation processes and responses to small changes are poorly understood, especially ice and mixed-phase clouds understood [Cantrell and Heymsfield, 2005; Hegg and Baker, 2009].  The ability of aerosol to nucleate liquid water, i.e. to act as cloud condensation nuclei (CCN), is much better understood.

Ice particles can form in the atmosphere when ice homogeneously nucleates in aqueous aerosol particles or heterogeneously nucleates on solid particles.  Only a very small fraction of atmospheric particles cause heterogeneous ice nucleation, and this subset of particles are referred to as ice nuclei (IN) [DeMott, 2002]. Although homogeneous nucleation in aqueous particles is now relatively well understood [Hegg and Baker, 2009; Koop et al., 2000], heterogeneous ice nucleation on solid particles remains highly uncertain, which translates to large uncertainties in models used to predict radiative forcing [Lohmann and Feichter, 2005]. To improve climate predictions, concentrations of IN in different environments need to be determined and the types of atmospheric particles that act as good IN need to be identified. This information can then be used to constrain and validate ice nucleation and ice cloud formation parameterizations in atmospheric models.

This team will address key uncertainties regarding the marine sources of primary and secondary aerosols in the Arctic and how these emissions may be affected by the decline of summer sea ice. Observations will directly feed current Arctic ocean-ice-atmosphere models.  In particular, DMS is the most important source of material for secondary aerosol formation in the marine boundary layer. Atmospheric DMS is oxidized to form sulfate aerosols that may act as CCN lessening radiation at the Earth’s surface [Charlson et al., 1987]. While the proposed climate regulation mechanism involving feedback between climate warming and DMS plankton production (the CLAW hypothesis; Charlson et al. 1987) is increasingly being questioned [Quinn and Bates, 2011], feedbacks between oceanic DMS emissions and regional climate do appear to be significant, especially in the Arctic, where warming is magnified by reduction in the extent of the summer sea ice [Perovich et al., 2007; Zhang et al., 2008].

It is only recently that the ocean as a source of primary organic aerosols has been addressed. In the Arctic, CCN could be formed by aggregates of marine organic materials originating from open water at the ice edge and in leads in the pack ice [Bigg and Leck, 2008; Leck and Bigg, 2005]. In particular, marine microgels may act as a source of CCN [Leck and Bigg, 2005; Orellana et al., 2011]. Such materials are ubiquitous in oceanic surface waters [Wurl et al., 2011], and similar organic compounds occur at very high concentrations in sea ice, where they are called exopolymeric substances (EPS) [Riedel et al., 2006; van der Merwe et al., 2009]. Only recently has upward transport of these gelatinous polymeric carbon materials from sea ice been observed.

To address both DMS and primary organics, it is necessary to study the formation processes of these materials. In particular, the production cycle begins in spring with massive microalgal blooms developing at the base of the sea ice when solar and ice and snow conditions allow light penetration [Arrigo et al., 2010; Mundy et al., 2005]. Sea ice algae represent an important reservoir of dimethylsulfoniopropionate (DMSP), the non-volatile precursor of DMS [Levasseur et al. 1994]. The fraction of the ice-DMSP that is converted to DMS and eventually reaches the atmosphere is still unknown [Nomura et al., 2012].  Dimethylsulfoxide (DMSO) may also act as a source or sink of DMS either through the production of DMSO by algae and its subsequent reduction into DMS or through the formation of DMSO by photochemical or bacterial oxidation of DMS, respectively [Hatton et al., 2004]. During the melting period, the sea ice surface ice becomes punctuated by biologically active melt water ponds which could also represent a source of primary and secondary aerosols. The extent of these melt ponds in spring and summer will increase with global warming.

The sea-surface microlayer film that forms at the interface between the ocean and atmosphere, known to affect air-sea gas exchange, may also affect aerosol formation in regions where wind-driven bubble bursting occurs. The microlayer may have different DMS concentrations as compared to the subsurface [Yang et al., 2005], while it is nearly always enriched in gelatenous organic matter[Wurl et al., 2011]. High biological productivity coupled with constrained surface area in sea ice leads likely enriches the Arctic sea-surface microlayer with organic material, but this has yet to be confirmed [Knulst et al., 2003; Wurl et al., 2011]. Nonetheless, a recent study has shown that in leads, the sea-surface microlayer may represent a source of Arctic bio-aerosols [Leck and Bigg, 2005]. Thus, we are lacking a fundamental understanding of what controls microlayer enrichments and the formation of organic aerosols in the presence and absence of sea ice.

To assess ocean-atmosphere interactions, it is crucial to quantitatively connect levels of dissolved aqueous DMS and microlayer properties to the atmospheric aerosol, including new particle formation and growth events. Gas-to-particle formation occurs where gases such as DMS are at high levels, oxidants are present, and pre-existing aerosol surface area is low. After nucleation, particles then experience growth to become CCN-sized. In our past Arctic-SOLAS study, we reported nucleation events in summer 2008 [Chang et al., 2011a], tying them for the first time to oceanic DMS levels. By contrast, in autumn 2007, low solar radiation and low oceanic production resulted in the absence of nucleation. It is also important to determine whether there is a connection between primary aerosol and the composition of different ocean sources (ice edge, melt pounds, microlayer). There is an indication that the sub-micron-sized fraction of oceanic primary particles is organic rich with a microlayer component [Leck and Bigg, 2005], whereas the larger size mode is more inorganic rich. Measurements indicate that particles containing marine algae (diatoms) may also act as IN [Knopf et al., 2011).

Modeling experiments suggest that the interaction between a larger ice-free surface available for gas exchange and a potential stimulation of biological DMS production could partly offset the warming caused by the loss of ice albedo [Gabric et al., 2005]. Current models are in their infancy and suffer from a lack of data for calibration and from the absence of parameterizations for key biological processes [Gabric et al., 2005; Elliot et al., 2012]. The measurements obtained in NETCARE will directly advance models describing these processes.

Key questions to be addressed are:

• What are the relative contributions of bacteria, ice algae, under-ice phytoplankton blooms, melt ponds and open water phytoplankton blooms to DMS production at the ice edge in spring-early summer?
• Can DMS escape directly through the ice and if so what is the importance of this source for the Arctic atmosphere?
• Is the sea-surface microlayer a source of primary organic atmospheric aerosol, and what are the cloud nucleating properties of these particles?
• What oceanic and atmospheric conditions favour particle nucleation and growth arising from oceanic emissions?
• What is the vertical extent of new particle formation and growth events and do such events occur primarily in the atmospheric boundary layer, or do ventilated emissions above the boundary layer promote nucleation more efficiently
• How might warming-induced changes in the ice cover, and the resulting increased extent of seasonal ice and ice edge, affect the production and emission of oceanic aerosols and their precursors?

The development of chemical transport models (CTMs) and general circulation models (GCMs) includes a rigorous effort to represent aerosol processes of relevance for climate.  However, many uncertainties remain.  A central objective of this activity is to integrate the NETCARE measurements from the previous three activities by evaluating and improving aerosol representation in global models. This objective will then enable improved estimates of the role of radiative forcing of aerosols in present-day climate, and improved predictions of future climate impacts.

Key questions to be addressed are:

• How well do models represent aerosol processes in remote areas of Canada?
• How do aerosol processes drive uncertainties in model simulations and how can aerosol simulations be improved?
• How do aerosols contribute to changes in climate at mid-high latitudes?

The NETCARE field observations will be expanded upon by a suite of ground-based and satellite observations that extend our analysis of aerosol processes in space and time.  The additional observations available include:

1. long-term in situ measurements from Barrow, Ny-Alesund (in addition to Alert, Whistler)
2. indicators of aerosol optical depth (AOD) and particle size from more than 19 ground-based sun-photometers as part of a ground based network (AEROCAN / AERONET)
3. AOD measurements from the MODIS, MISR, and VIIRS satellite instruments
4. vertical profiles of aerosol backscatter across Canada from ground-based lidars (collocated with AEROCAN), and space-based lidars (CALIOP and ATLID)
5. cloud linkages through the CloudSat radar and the EarthCARE Cloud Profiling Radar.

Modelling activities outlined below will serve to integrate these with the other observational data sets in the network for carbonaceous aerosols, ocean-atmosphere interactions, and ice cloud formation.  Modelling will also be used to help identify specific measurement needs. In addition, implications of model improvements for multi-decadal changes in climate will be analyzed.