Motivated by the need to predict how the Arctic atmosphere will
change in a warming world, this article summarizes recent advances made by
the research consortium NETCARE (Network on Climate and Aerosols: Addressing
Key Uncertainties in Remote Canadian Environments) that contribute to our
fundamental understanding of Arctic aerosol particles as they relate to
climate forcing. The overall goal of NETCARE research has been to use an
interdisciplinary approach encompassing extensive field observations and a
range of chemical transport, earth system, and biogeochemical models. Several
major findings and advances have emerged from NETCARE since its formation in
2013. (1) Unexpectedly high summertime dimethyl sulfide (DMS) levels were
identified in ocean water (up to 75 nM) and the overlying atmosphere (up to
1 ppbv) in the Canadian Arctic Archipelago (CAA). Furthermore, melt ponds,
which are widely prevalent, were identified as an important DMS source (with
DMS concentrations of up to 6 nM and a potential contribution to atmospheric
DMS of 20 % in the study area). (2) Evidence of widespread particle
nucleation and growth in the marine boundary layer was found in the CAA in
the summertime, with these events observed on 41 % of days in a 2016
cruise. As well, at Alert, Nunavut, particles that are newly formed and grown
under conditions of minimal anthropogenic influence during the months of July
and August are estimated to contribute 20 % to 80 % of the 30–50 nm
particle number density. DMS-oxidation-driven nucleation is facilitated by
the presence of atmospheric ammonia arising from seabird-colony emissions,
and potentially also from coastal regions, tundra, and biomass burning. Via
accumulation of secondary organic aerosol (SOA), a significant fraction of the new
particles grow to sizes that are active in cloud droplet formation. Although
the gaseous precursors to Arctic marine SOA remain poorly defined, the
measured levels of common continental SOA precursors (isoprene and
monoterpenes) were low, whereas elevated mixing ratios of oxygenated volatile
organic compounds (OVOCs) were inferred to arise via processes involving the
sea surface microlayer. (3) The variability in the vertical distribution of
black carbon (BC) under both springtime Arctic haze and more pristine
summertime aerosol conditions was observed. Measured particle size
distributions and mixing states were used to constrain, for the first time,
calculations of aerosol–climate interactions under Arctic conditions.
Aircraft- and ground-based measurements were used to better establish the BC
source regions that supply the Arctic via long-range transport mechanisms,
with evidence for a dominant springtime contribution from eastern and
southern Asia to the middle troposphere, and a major contribution from
northern Asia to the surface. (4) Measurements of ice nucleating particles
(INPs) in the Arctic indicate that a major source of these particles is
mineral dust, likely derived from local sources in the summer and long-range
transport in the spring. In addition, INPs are abundant in the sea surface
microlayer in the Arctic, and possibly play a role in ice nucleation in the
atmosphere when mineral dust concentrations are low. (5) Amongst multiple
aerosol components, BC was observed to have the smallest effective deposition
velocities to high Arctic snow (0.03 cm s
Rapid changes in the Arctic environment including rising temperatures, melting sea ice, elongated warm seasons, and changing aerosol and trace gas long-range transport patterns (IPCC, 2013) are driving a growing interest in developing a better understanding of the processes that control Arctic climate. Furthermore, because high-latitude climate change is a bellwether for change on a global scale, it is particularly important to understand the feedbacks that lead to amplification of Arctic warming (Serreze and Barry, 2011).
NETCARE Arctic field campaigns.
This article discusses key discoveries that have been made in climate-related
Arctic aerosol research by the NETCARE (Network on Climate and Aerosols:
Addressing Key Uncertainties in Remote Canadian Environments) research
network. Formed in 2013, NETCARE consists of Canadian academic and government
researchers along with international collaborators. Given the highly diverse
nature of inter-related earth system processes that couple within the Arctic
environment, the network is necessarily interdisciplinary, consisting of
climate and air quality modellers, atmospheric chemists, aerosol and cloud
physicists, biological and chemical oceanographers, biogeochemists, and
remote sensing experts. Over the past 6 years, the network has conducted a
set of field campaigns and modelling projects focused on the sources and loss
mechanisms of atmospheric particles, their chemical and optical
characteristics, and their role in climate. The field studies were conducted
using a variety of platforms including the Alfred Wegener Institute's
Map of the Arctic indicating NETCARE field work locations, including the
ground station (Alert), CCGS
The network's output is documented through a special issue across three
journals:
Written for a scientist interested in the fields of Arctic climate, atmospheric chemistry, and biogeochemistry, this article starts with a background on Arctic aerosol that is not focused on NETCARE results (Sect. 2). For additional background information, the reader is referred to Quinn et al. (2006, 2008), Law and Stohl (2007), and Willis et al. (2018). The article then presents new insights into the three topics around which NETCARE was structured: marine processes and the Arctic atmosphere (Sect. 3); the sources, sinks, and properties of Arctic aerosol (Sect. 4); and ice nucleating particles (INPs; Sect. 5). Each of these sections stands alone, so that the interested reader can focus their attention on a specific subject. However, there are clear connections between the different topics. For example, Sect. 3 (Marine processes and the summertime Arctic atmosphere) is motivated by the increasing marine impact that is arising as sea ice melts and focuses on new NETCARE Arctic measurements of dimethyl sulfide (DMS), ammonia, and oxygenated volatile organic carbon species. The oceans are an important source of such reactive gases to the atmosphere, leading to direct impacts on aerosol particles and ultimately on climate. Those connections are made in Sect. 4 (Arctic aerosol: sources, sinks, and properties), which presents insights gleaned for the summertime environment, when these marine emissions can lead to new particle formation and growth, and discusses the impacts of this aerosol on clouds. Section 4 also presents results from the Arctic haze springtime period, where the emphasis is on the sources of particles, their optical properties, and the potential for direct radiative forcing. Section 5 (Ice nucleating particles) addresses the select fraction of atmospheric particles that nucleate ice crystals. Section 6 concludes the article by discussing remaining research uncertainties and future priorities.
Over the last half century, our knowledge of Arctic aerosol and its role in climate has advanced from almost nothing to a clear understanding of its importance, although important questions remain regarding mechanistic details. This short section of the paper presents a comprehensive description of the field, leaving the recent NETCARE results for later sections.
Following early observations of visibility-reducing haze particles in the spring Arctic atmosphere (Greenaway, 1950), study of Arctic haze began in earnest in the 1970s (Holmgren et al., 1974; Rahn and Heidam, 1981). Investigations intensified through the 1980s, with observations (ground-based and airborne) and meteorological analyses indicating that haze particles were transported from mid-latitude pollution sources, often in layers that reached up to the tropopause, and that their concentrations increased in winter and spring due to efficient meridional transport and low rates of wet deposition (Barrie, 1986; Barrie and Hoff, 1985; Brock et al., 1989; Leaitch et al., 1989; Radke et al., 1984; Schnell and Raatz, 1984; Shaw, 1982).
Through the 1990s and beyond, concentrations of Arctic haze components declined at the northernmost observatories: Alert, Nunavut; Barrow, Alaska; Mount Zeppelin, Svalbard; and Station Nord, Greenland (Heidam et al., 1999; Hirdman et al., 2010; Quinn et al., 2009; Sharma et al., 2004, 2006; Sinha et al., 2017; Sirois and Barrie, 1999). Recent measurements (Fisher et al., 2011; Frossard et al., 2011; Leaitch et al., 2018; Massling et al., 2015; Sharma et al., 2017; Sinha et al., 2017) have found surface mass concentrations of sulfate, organic material and black carbon (BC) 3–10 times lower than those estimated from studies conducted prior to 1981 (Rahn and Heidam, 1981), but the total Arctic column burden of BC may have increased (Koch and Hansen, 2005; Sharma et al., 2013) with implications for climate forcing efficiency (Breider et al., 2017). The turn of the century saw renewed interest in Arctic haze with concern for the role of BC in Arctic warming (Flanner et al., 2007; Hansen and Nazarenko, 2004; Law and Stohl, 2007; McConnell et al., 2007; Quinn et al., 2008; Shindell and Faluvegi, 2009).
From the early studies of Arctic haze arose the concept of the Arctic atmosphere as a dome of cold air that regulates transport of polluted air from southerly latitudes (Barrie, 1986). The polar front extends in the winter to include more southerly industrial emissions that can be transported into the high Arctic, and the front retreats in the summer to inhibit transport from mid-latitude sources. Figure 2 shows an example of identification of the polar dome in spring 2015 through measurements conducted during the NETCARE aircraft campaign. Pollution transport into the Arctic may also be influenced by the North Atlantic Oscillation (Duncan and Bey, 2004; Eckhardt et al., 2003). Arctic haze originates from Eurasia, Siberia, southeast Asia and North America, with Eurasia as the dominant source region at lower altitudes and contributions from south and central Asian sources dominating at higher altitudes (Fisher et al., 2011; Qi et al., 2017; Sharma et al., 2013; Stohl, 2006). Sea salt contributes to the haze due to the combination of stronger winds and reduced wet deposition in the winter and spring (Huang and Jaeglé, 2017; Leaitch et al., 2018) and frost flowers may contribute some marine salt (Shaw et al., 2010). Snowpack exchange is a potential springtime source of organic precursors (McNeill et al., 2012), while stratospheric contributions appear to be small (Leaitch and Isaac, 1991; Stohl, 2006).
The potential temperature (
Arctic haze warms the Arctic in several ways. BC from anthropogenic sources and forest fires deposits to snow and ice, lowering the surface albedo (Clarke and Noone, 1985; Doherty et al., 2010; Flanner et al., 2007; Forsström et al., 2013; Hegg and Baker, 2009; Keegan et al., 2014; McConnell et al., 2007). Atmospheric haze layers containing BC are warmed while the underlying surface is cooled, which acts to increase atmospheric stability (Blanchet and List, 1983; Brock et al., 2011; Koch and Del Genio, 2010; Leighton, 1983; Pueschel and Kinne, 1995; Valero et al., 1984). Meridional temperature gradients are enhanced by BC outside the Arctic, which warms the air during transport to the Arctic, hence increasing heat transport into the Arctic (Sand et al., 2013). Dust, when present in layers over high-albedo surfaces and/or deposited to the snow, will warm the atmosphere (Bond et al., 2013; Dumont et al., 2014; Lohmann and Feichter, 2005). Arctic haze can also increase longwave radiative forcing by forming thin Arctic low-level liquid clouds (Garrett et al., 2009; Garrett and Zhao, 2006; Lubin and Vogelmann, 2006; Mauritsen et al., 2011).
However, many components of Arctic haze (e.g. sulfate; organic matter, OM; sea salt) help to cool the Arctic by scattering light back to space (Schmeisser et al., 2018) and by modifying the microphysics of liquid clouds to enhance shortwave cooling (Garrett and Zhao, 2006; Lubin and Vogelmann, 2006; Zamora et al., 2017; Zhao and Garrett, 2015). During winter and spring, sulfuric acid in Arctic haze particles may reduce their effectiveness as INPs, leading to larger crystals that precipitate more easily. As a result, there may be an increase in the dehydration rate of the atmosphere and a corresponding reduction in longwave forcing (Blanchet and Girard, 1994; Curry and Herman, 1985). At cirrus temperatures, dust, ammonium sulfate, and sea salt may also increase cloud albedo by increasing ice crystal concentrations (Abbatt et al., 2006; Sassen et al., 2003; Wagner et al., 2018).
Observed and simulated seasonal cycles of BC and sulfate typically show a maximum in near-surface concentrations in March or April (Barrie and Hoff, 1985; Eckhardt et al., 2015; Garrett et al., 2010; Sharma et al., 2006) and clean conditions in the summertime. Natural emissions of BC from vegetation fires are considerable in late spring to early summer in the Arctic and at mid-latitudes (Mahmood et al., 2016). Production of sulfate aerosol is more efficient in the warm than the cold seasons (Mahmood et al., 2018; Tesdal et al., 2015). The decline in Arctic haze after its peak in early spring and the approach to the summertime pristine conditions are largely related to changes in transport as the polar front moves northward and aerosol scavenging rather than a reduction in aerosol production. Wet deposition associated with transport across the retracted polar front, frequent low-intensity precipitation, and longer residence times within the polar dome keep the summertime near-surface Arctic nearly free of anthropogenic aerosol (Barrie, 1986; Stohl, 2006; Garrett et al., 2010; Browse et al., 2012). However, at higher altitudes up to 8 km, long-range transport from mid-latitude pollution into the Arctic was also observed in summer (Schmale et al., 2011). Marine sources have a strong influence on the Arctic summer aerosol near the surface and possibly aloft (Dall'Osto et al., 2017; Korhonen et al., 2008b; Stohl, 2006).
Summer sources of sulfate appear to be the oxidation of DMS from the Arctic Ocean as well
as connected waters to the south, volcanism, residual Arctic haze sulfate, and some
anthropogenic sulfate or
Characterized by a unimodal diameter distribution centred between 200 and 300 nm (Bigg, 1980; Heintzenberg, 1980; Leaitch and Isaac, 1991; Radke et al., 1984; Staebler et al., 1994), Arctic haze particles are effective at both scattering light (Andrews et al., 2011; Schmeisser et al., 2018) and acting as nuclei for cloud droplets (Earle et al., 2011; Komppula et al., 2005). In contrast, the summertime number distribution is dominated by smaller Aitken particles resulting from newly formed particles that have experienced modest growth in the near-pristine summer Arctic. Their small sizes render Aitken particles relatively ineffective at scattering light, but they may be able to influence cloud microphysics in the clean summertime Arctic (Korhonen et al., 2008b).
Overall, the net effect of anthropogenic aerosols has been to cool the Arctic (Fyfe et al., 2013; Najafi et al., 2015), and Navarro et al. (2016) showed that reductions in Arctic haze have contributed to the sharp increase in the rate of Arctic warming since 1990. Mitigation of BC emissions may help to slow Arctic warming so long as cooling components are not simultaneously mitigated (Kopp and Mauzerall, 2010; Sand et al., 2013; Shindell and Faluvegi, 2009).
As seen from this brief overview, understanding natural aerosol processes in addition to anthropogenic aerosol sources is vital for climate studies, as anthropogenic aerosol forcing is measured against the natural component (Carslaw et al., 2013; Megaw and Flyger, 1973). For example, in the winter and spring, sea salt aerosol may play an important climate role (Kirpes et al., 2018). At the start of NETCARE, detailed knowledge of natural particle sources and their impacts on clouds in the nearly pristine summer was incomplete, and it became a major focus of the network's research activities.
In remote marine atmospheres such as the summertime Arctic, assessing the impact of natural marine biogenic aerosol (MBA) sources on cloud formation is pivotal to accurately estimating climate forcing (Carslaw et al., 2013; Charlson et al., 1987). While a variety of organic compounds, such as marine microgels, may be relevant primary MBA sources in the Arctic (Leck and Bigg, 2005; Orellana et al., 2011), DMS-derived sulfate is thought to be a key precursor to secondary marine aerosol mass over biologically productive regions (Ghahremaninezhad et al., 2016; Leaitch et al., 2013; McCoy et al., 2015; Park et al., 2017). The production of DMS and other organic compounds in polar regions is linked to the productivity of microalgae, as well as to the dynamics and the structure of pelagic (oceanic) and sympagic (ice-associated) microbial food webs (Gabric et al., 2017; Levasseur, 2013; Simó, 2001; Stefels et al., 2007). Peaks in the DMS proxy MSA have been observed in association with bursts of phytoplankton productivity in the high Arctic (Becagli et al., 2016). As well, atmospheric DMS mixing ratios in the marine boundary layer have been shown to transiently peak during the phytoplankton growth period from May to September (Park et al., 2013, 2018). Particle nucleation and growth events have been observed even at moderate levels of atmospheric and oceanic DMS in the high Canadian Arctic (Chang et al., 2011b; Rempillo et al., 2011).
Despite these compelling indications of the key role played by marine biogenic DMS in contributing to sulfate aerosols (Rempillo et al., 2011), measurements of seawater and sea-ice DMS during the biologically productive summer months (June to August) that coincide with clean aerosol time periods are still scarce (Jarníková et al., 2018; Levasseur, 2013). The paucity of DMS measurements in ice-associated habitats, such as under the sea ice, in melt ponds atop the ice, or directly at the Arctic sea-ice margin, is even greater (Levasseur, 2013). Sea ice not only acts to modulate gaseous exchange but also hosts active microorganisms (Gradinger, 2009), making it a fundamental driver of various MBA precursors, including DMS (Arrigo, 2014; Gabric et al., 2017; Korhonen et al., 2008b). Our understanding of the processes that control other key gases that can lead to aerosol formation in marine environments, including ammonia and volatile organic compounds (VOCs), is particularly weak. There have been very few measurements of their Arctic abundance in the past and we have a poor understanding of their sources. In this context, NETCARE targeted the spatio-temporal variability in DMS and the underlying ecosystemic mechanisms controlling its abundance in the eastern Canadian Arctic (Canadian Arctic Archipelago, henceforth CAA, and northern Baffin Bay), along with the atmospheric abundances and sources of other key gases.
The two NETCARE summer campaigns (July–August 2014 and 2016; see Fig. 1 and Table 1)
revealed high open-water concentrations of DMS (interquartile range of
5.1–10.9 nmol L
Novel measurements made during NETCARE also substantiated the potentially
important role played by melt ponds. An in-depth study of nine melt ponds
revealed that brackish melt ponds over first-year sea ice (FYI) may have DMS
concentrations ranging from 3 to 6 nmol L
While marginal ice zones (MIZs) and various ice-edge systems have long been recognized
for their teeming biological activity (Mundy et al., 2009; Perrette et al., 2011) and
potential for heightened DMS production (Galí and Simó, 2010; Levasseur, 2013;
Matrai and Vernet, 1997), they remain surprisingly under-documented for their specific
role in MBA production in the eastern Canadian Arctic during summer. Two distinct MIZs
explored during the summer of 2014 revealed highly contrasting DMS dynamics, suggesting
that whether the sea ice is FYI or multi-year ice (MYI) is of paramount importance in
shaping marine food webs and the net production of DMS in the water exiting the ice pack.
Contrasting DMS dynamics between FYI and MYI systems were likely linked to differences in
light penetration through the ice pack and its availability to primary producers in the
waters just below the ice. At the MYI edge in Kennedy Channel (ca. 81
High levels of DMS have previously been associated with aerosol formation and growth in
the CAA (Chang et al., 2011b; Park et al., 2017; Rempillo et al., 2011). As part of
NETCARE, new atmospheric measurements of DMS were performed from both the
VOCs were measured in the marine atmosphere during the 2014 CCGS
A large suite of oxygenated VOCs (OVOCs) were measured on the CCGS
NETCARE provided the opportunity to make some of the first observations of ammonia in the
Arctic atmosphere. Previous measurements of atmospheric ammonia over the Norwegian Sea
and Arctic Ocean during the summer ranged between the detection limit (35 ppt) and
400 ppt (Johnson et al., 2008). Simultaneous measurements of sea surface ammonium
(
Prior to the NETCARE field campaigns, the existing un-extrapolated DMS climatology, averaged over the most productive time of the year (months of July and August), clearly demonstrated the scarcity of surface ocean DMS measurements in the Arctic (Lana et al., 2011). The updated Lana DMS climatology and its precursor (Kettle et al., 1999) have long represented useful tools for oceanic model validation (e.g. Le Clainche et al., 2010; Tesdal et al., 2016; Kim et al., 2017) and lack of data over the Canadian Polar shelf and the Baffin Bay area challenged the representativeness of the standard (extrapolated) version of this climatology for these specific regions (Fig. 5c). Observations gathered through NETCARE field campaigns (Fig. 5b) significantly enhanced coverage in these regions.
As part of NETCARE, a new process-based sea-ice–ocean biogeochemical model representing ecosystems in both the sea ice and water column of the marine Arctic was developed. The model was initially developed in a one-dimensional (1-D) configuration (Mortenson et al., 2017). Subsequently, sulfur and inorganic carbon cycling were developed and implemented into the model (Hayashida et al., 2017; Mortenson et al., 2018). The simulated Arctic sea-ice ecosystem and sulfur cycle were next incorporated into a three-dimensional (3-D) regional configuration (Hayashida, 2018; Hayashida et al., 2018). This model advances previous Arctic-focused DMS model studies (Elliott et al., 2012; Jodwalis et al., 2000) in that many of the parameters concerning the DMS production are derived from recent field observations in the Arctic, enabling quantification of the relative contributions of ice algae and phytoplankton to DMS production and emissions. The 1-D simulations indicated a notable contribution of ice algae: an 18 % enhancement of DMS concentrations under the ice and a 20 %–26 % enhancement of sea–air DMS fluxes during the melt period for Resolute Passage (Hayashida et al., 2017). Also in the vicinity of ice margins, simulated spikes in sea–air fluxes of DMS originating from bottom and under-ice production by algae were comparable to some of the local maxima in the summertime flux estimated for ice-free waters in the Arctic.
Pan-Arctic distribution of July–August concentrations of surface ocean DMS.
Upper panels show the comparison between
The data obtained during the two NETCARE ship campaigns, together with data
previously available in the PMEL sea surface database
(
Under future global warming conditions, sea-ice extent is expected to decline
significantly, affecting the temporal and spatial evolution of ice algae and under-ice
and open-water phytoplankton blooms. This may lead to changes in oceanic DMS emissions,
although the sign and magnitude of the change is highly uncertain. Using the satellite
approach mentioned above, Galí et al. (2019) showed that DMS emission has increased
at a rate of about 30 % decade
To estimate the sensitivity of Arctic aerosols and radiative forcing to surface seawater concentrations of DMS in the Arctic, simulations with different specified surface seawater DMS concentrations and spatial distributions in the Arctic were performed for future sea-ice conditions using the Canadian Atmospheric Global Climate Model (CanAM4.3). For all of the specified surface seawater DMS conditions in the model, simulated Arctic sulfate aerosol amounts respond only weakly to a reduction in sea-ice extent owing to increases in precipitation and aerosol wet deposition associated with the receding ice and increased open water (Mahmood et al., 2018). However, nucleation rates for sulfate aerosol respond significantly to reductions in sea-ice extent, which leads to a strengthening of cloud radiative forcing in the future. Furthermore, the simulated response of the mean cloud radiative forcing in the Arctic is approximately proportional to the mean surface seawater DMS concentration in the Arctic. Thus potential future changes in sea-ice extent may result in a negative climate feedback of DMS on radiative forcing in the Arctic, as suggested by Charlson et al. (1987).
The overall motivation of Arctic summertime research is to determine how the atmosphere will respond to melting sea ice, as an ocean that was largely covered by sea ice through much of the summer will potentially be ice free in summer by mid-century (AMAP, 2017; Comiso, 2011; Gregory et al., 2002; Holland et al., 2006). Given the evolution of the summertime Arctic Ocean from a bright ice cap to a dark ocean that can readily absorb solar radiation, it is of particular importance to understand factors controlling the overhead aerosol and cloud that could mediate the positive radiative feedback of declining sea ice. Precipitation associated with low clouds and fogs is common in the summertime (Browse et al., 2014). Wet deposition is a highly efficient aerosol removal mechanism, giving rise to a clean boundary layer in which new particles may be formed or into which they may be input. In these clean boundary layers, increases in the numbers of particles acting as cloud condensation nuclei (CCN) may increase longwave warming by clouds if the absolute concentrations of CCN are sufficiently low (Mauritsen et al., 2011); otherwise, increases in CCN concentrations lead to enhanced shortwave cooling. In this context, it is important to better understand the processes that give rise to new particle formation and growth to CCN sizes, and the associated impacts on clouds. For example, how do the emissions of biogenic gases described in Sect. 3 affect new particle formation and growth in such environments, and what is their importance relative to anthropogenic inputs from local shipping and long-range transport?
In contrast, the springtime atmosphere, with its associated Arctic haze, has
been better studied than the summertime atmosphere. The results from high
profile campaigns such as ISDAC (Indirect and Semi-Direct Aerosol Campaign,
Aerosol size distributions from Alert and Zeppelin Arctic field stations. The pronounced accumulation mode in the winter and spring is characteristic of Arctic haze. The mode of Aitken particles is a common feature of the Arctic summertime atmosphere. Figure from Croft et al. (2016b).
Lastly, the deposition rates of aerosol constituents need to be measured to better constrain models. Ideally, both wet and dry deposition rates would be individually evaluated throughout the year, to map out the transition from a system dominated by the relatively slow loss with ice cloud scavenging versus the more efficient removal via warm clouds and fogs.
As described in Sect. 2, a pronounced Aitken mode in the aerosol size distribution is a common feature during the Arctic summertime, as demonstrated by Croft et al. (2016b), who identified this feature in long-term monitoring data sets from both the Alert and Zeppelin ground stations (Fig. 6). One of the major findings from NETCARE is the widespread prevalence of 5–50 nm ultrafine particles in the summertime Canadian Arctic (Burkart et al., 2017b; Collins et al., 2017; Willis et al., 2016, 2017) and their ability to activate as CCN (Burkart et al., 2017a; Chaubey et al., 2019). While previous ship-based measurements in similar regions in late summer and fall had demonstrated new particle formation and growth events, their frequency was low. For example, in the fall period of late August to the end of September 2008, only three such events were observed over a 5-week observation period, whereas no events were observed at all in October 2007 (Chang et al., 2011b). By comparison, NETCARE measurements in mid-July to mid-August 2016 observed enhancements in 5–50 nm particles 41 % of the time in a spatially heterogeneous manner (Collins et al., 2017). Characterization of the summertime increase in particles is provided in Fig. 7, wherein the number of particles between 15 and 30 nm (N15–N30) is highly enhanced at Alert in July and August, before rapidly declining in September (see the Supplement for details). As discussed in the Supplement, natural sources are estimated to contribute 20 %–80 % of the 30–50 nm particles during July and August. NETCARE aircraft measurements in July 2014 also demonstrated the spatial heterogeneity of 5–50 nm particle numbers in the inversion layer, with the highest concentrations observed over marine and cloudy regions and little detectable enhancement over ice-covered areas (Burkart et al., 2017b). These aircraft measurements also indicate that the numbers of these tiny particles in the free troposphere are spatially homogeneous and considerably lower than those measured in the inversion layer, indicative of a boundary layer source.
The changing composition and size distributions of aerosol in the
high Arctic, see the Supplement for details.
Panel
Significant growth of 5–50 nm particles to CCN sizes was clear from each observational platform. At Alert (Fig. 7), the summertime enhancement in particles between 15 and 30 nm (N15–N30) coincides with the increase in particles in the 50 to 100 nm size range (N50–N100), which is also the size of particle activation diameters observed in the field (see Sect. 4.3). Interestingly, using Fourier Transform Infrared (FTIR) absorption of particulates collected on filters, the ratio of aerosol organic material to sulfate was also observed to increase during this time period, and the region of amide functional groups indicates a contribution of organic components from breakdown of seabird urea in guano (Leaitch et al., 2018). Likewise, a particle growth episode was clearly observed over the ice-free Lancaster Sound, in which the numbers of 5–50 nm particles and CCN increased in concert with the measured organic content of the PM1 aerosol (Willis et al., 2016). Across the entire aircraft campaign, the numbers of CCN were most strongly enhanced above background levels when the air had recently been at low altitude over open water (Fig. 8a), when the wind speeds were low, and when the organic-to-sulfate ratio of the particles was high (Willis et al., 2017). This marine influence is consistent with summertime single-particle mass spectrometric measurements of trimethylamine-containing particles in the marine boundary layer that were largely externally mixed with sea-salt-containing particles (Fig. 9; Köllner et al., 2017).
The lack of a wind speed dependence and the observations of externally mixed particulate
trimethylamine suggests that secondary sources are important. Similarly, microphysical
models of growing particle size distributions could only be reconciled with observations
from the CCGS
Single-particle mass spectrometry results from the NETCARE 2014 summer campaign,
where the detected particle fraction is plotted against the aerodynamic diameter of the
particle. The total number of particles detected in a specific size bin is plotted in
red. The classifications of particle types containing different species are: Na
Natural emissions of ammonia are also important to new particle formation and growth.
Wentworth et al. (2016) used GEOS-Chem model simulations to interpret NETCARE ammonia
measurements (see Sect. 3.3) and found that migratory seabird colonies (emitting 36 Gg
Time series of measured and modelled numbers of particles 10 nm and larger at Alert during 2011. Seabird ammonia is included in the blue curve simulation but not in the red curve simulation. Measurements are in black. Figure from Croft et al. (2016a). GCT represents GEOS-Chem-TOMAS.
Determining the precursors to Arctic MSOA is of crucial importance. Aerosol mass
spectrometry measurements from the aircraft campaign in summer 2014 indicate that the
organic chemical character of this aerosol is distinctly different from that which arises
from oxidation of common continental precursors, such as isoprene or the monoterpenes
(Willis et al., 2017). The mass spectral signatures indicate molecules that instead have
substantial alkyl components, such as functionalized fatty acids (Fig. 8b). Long-chain
fatty acids can sometimes be a significant component of the sea surface microlayer
(Cunliffe et al., 2013). Croft et al. (2018) have shown that a steady flux of condensable
organic material from the ocean that oxidizes with a lifetime of a day is essential for
consistency between GEOS-Chem-TOMAS modelled aerosol size distributions and those
measured at Alert and from the CCGS
Studies at mid-latitudes have routinely shown that the smallest particles that can serve
as nuclei for liquid cloud droplets are 80–120 nm in diameter (Hoppel et al., 1985;
Leaitch et al., 1986). The smaller Aitken particles, 20–80 nm in size, are commonly
considered to be too small to activate into cloud droplets. However, there are two
circumstances which enable Aitken particles to activate at cloud base: (1) rapid cooling
rates, generally associated with higher updraft speeds, increase cloud base
supersaturation; and (2) very low concentrations of larger particles (
During the NETCARE flights conducted out of Resolute Bay in July 2014, number
size distributions of cloud droplets and aerosol particles measured in and
around clouds showed that 50 nm particles were routinely activated and that
particles as small as 20–30 nm were activated on a few occasions where
updraft speeds were likely higher (Leaitch et
al., 2016). These results substantiate the prediction made by Korhonen et
al. (2008b). However, Leaitch et
al. (2016) found
no evidence for an association of cloud liquid water content with aerosol
variations when droplet concentrations are less than about 10 cm
Lastly, experiments are in progress to evaluate the Single Column Model of Arctic Boundary Layer Clouds (SCM-ABLC) and version 18 of the Canadian Climate Centre's radiative transfer model with the cloud observations conducted from Resolute Bay discussed above. The modelling work will attempt to reproduce the observations and quantify the uncertainty in modelling the radiative balance of low clouds and fog in the summertime Arctic.
As discussed in Sect. 2, Arctic haze is a prominent feature in springtime, yet its
composition and sources remain uncertain. During the NETCARE 2015 aircraft campaign,
vertically resolved observations of trace gases and aerosol composition were made in the
high Arctic springtime, with six flights north of 80
NETCARE flight observations based at Alert and Eureka revealed that within the polar
dome, submicron aerosol composition varied systematically with potential temperature. In
the lower polar dome (i.e. below 252 K), measured aerosol mass (non-refractory aerosol
and BC) was dominated by sulfate (74 %), with smaller contributions from BC
(1 %), organic aerosol (OA, 20 %), and ammonium (
Model simulations of air mass history using FLEXPART indicate differences in transport
history as a function of potential temperature in the polar dome. Air masses at lower
potential temperature (lower altitude) spent long times (
An analysis of results from simulations with four different models in NETCARE (Mahmood et al., 2016) indicates that the main source of BC in the Arctic is long-range transport from mid-latitudes. The long-range transport of BC to the Arctic is particularly efficient in midwinter and then decreases in efficiency, reaching a minimum in March and April. At the same time, dry deposition decreases, and wet deposition from clouds in the low and mid troposphere becomes more important during the transition from winter to spring. Overall, sources and sinks of BC in the Arctic are well balanced, leading to nearly steady Arctic burdens during the time period from December to May. Subsequently, during the transition from spring to summer, precipitation increases and wet deposition becomes highly efficient, which leads to substantial reductions in BC burdens in the Arctic despite increased emissions from vegetation fires. At high altitudes in the Arctic, the model results indicate that convective transport of pollution from the lower to the upper troposphere at lower latitudes and subsequent long-range transport to the Arctic represents an important source of BC.
Xu et al. (2017) interpreted a series of airborne and ground-based BC measurements made using multiple measurement techniques with the GEOS-Chem global chemical transport model and its adjoint to attribute the sources of Arctic BC (Fig. 11). This was the first comparison of BC measurements from a Single Particle Soot Photometer (SP2) at Alert with a chemical transport model. The inclusion of seasonally varying domestic heating and of gas-flaring emissions was crucial to successfully simulating ground-based measurements of BC concentrations at Alert and Barrow and airborne BC measurements across the Arctic. Sensitivity simulations suggest that anthropogenic emissions in eastern and southern Asia have the largest effect on the Arctic BC column burden in spring (56 %), with the largest contribution in the middle troposphere. At the Arctic surface, anthropogenic emissions from northern Asia (40 %–45 %) and eastern and southern Asia (20 %–40 %) are the largest BC contributors in winter and spring, followed by Europe (16 %–36 %). This dominant role of Asian sources is consistent with some recent studies (e.g. Ma et al., 2013; Wang et al., 2014; Ikeda et al., 2017) but differs from many earlier studies (e.g. Stohl, 2006; Shindell et al., 2008; Gong et al., 2010; Huang et al., 2010; Bourgeois and Bey, 2011; Sharma et al., 2013) due to decreased European emissions and increased Asian emissions. The adjoint simulations enabled identification of pronounced spatial heterogeneity in the contribution of emissions to the Arctic BC column concentrations, with noteworthy contributions from emissions in eastern China (15 %) and western Siberia (6.5 %). The Tarim oilfield in western China stood out as a specific influential source with an annual contribution of 2.6 %. Emissions from as far away as the Indo-Gangetic Plain could have a substantial influence (6.3 %) on Arctic BC as well.
GEOS-Chem adjoint modelling results for BC sources to the Arctic.
Panels
Kodros et al. (2018) combined measurements of BC mixing state in the
springtime Canadian high Arctic with simulated size-resolved aerosol mass and
number concentrations to constrain model estimates of the direct radiative
effect (DRE). Airborne measurements using an SP2 (soot particle photometer) and Ultra-High
Sensitivity Aerosol Spectrometer on board the
Some of the first vertically resolved and concurrent measurements of aerosol composition
and optical properties in the springtime high Arctic are presented in Leaitch et
al. (2019). As shown in Fig. 12a, observations from the
While in situ field campaigns provide detailed information over a short period of time, remote sensing provides annual measurements and thus information about the transitions from winter to spring and then into summer. In particular, ground-based lidar and star photometry (carried out at the PEARL observatory in Eureka, Nunavut) and satellite-based lidar (CALIOP/CALIPSO) during the latter half of two polar winters suggest the frequent Arctic-wide presence of submicron particles in the boundary layer with aerosol optical depths (AODs) significantly greater than the AOD predicted by GEOS-Chem, in which the AOD largely results from sulfate particles (O'Neill et al., 2016). Ground-based sun photometry (AEROCAN/AERONET) measurements acquired between 2009 and 2012 at five western-Arctic stations (Hesaraki et al., 2017) revealed Arctic-wide springtime peaking of both submicron and super-micron AODs that were roughly consistent with submicron and super-micron AOD estimates from GEOS-Chem (predominantly associated with Arctic haze sulfates and Asian mineral-dust aerosols, respectively). A summertime peak in submicron particles, which was determined to be smoke induced at the four westernmost AEROCAN Arctic stations, was not simulated by GEOS-Chem. Rather, GEOS-Chem indicated a continuous spring-to-fall decrease in submicron AOD that was principally associated with a decrease in sulfate contributions.
Deposition fluxes in the Arctic are very poorly characterized, in large part because of the logistical challenges of collecting continuous data series. To address this, NETCARE scientists collected temporally resolved data for the chemical composition of snow (common metals, BC, soluble ions, and small organics) that fell throughout the cold season at Alert (Macdonald et al., 2017). In particular, new snow was collected after each appreciable period of precipitation, resulting in samples every 4 days, on average, from September 2014 to May 2015.
Using measurements of the amount of snow that had fallen in a given area, the chemical compositions were converted to fluxes for comparison with modelled values. In combination with ambient air concentrations of the equivalent chemical species, the measured fluxes were then expressed as an effective deposition velocity, which encompasses both wet and dry deposition processes (Fig. 13; Macdonald et al., 2017). Interestingly, effective deposition velocities are higher for the warmest months (September, October, May) than for the cold months, arising from the switchover from liquid water to a combination of dry deposition and ice cloud scavenging. The effective deposition velocities for BC were the smallest of all species characterized, consistent with its low hygroscopicity and poor ice-nucleating abilities.
To take advantage of the high temporal resolution of the samples, the data were also used to assess potential sources contributing to chemical species in snow using a combination of positive matrix factorization and FLEXPART potential emission sensitivity analysis (Macdonald et al., 2018). The best positive matrix factorization solution consisted of seven source factors (sea salt, crustal metals, BC, carboxylic acids, nitrate, non-crustal metals, and sulfate), reflecting a balance between natural and anthropogenic sources. Notable findings include identification of anthropogenic sources (but not biomass burning) as dominant for BC during this study period, and a potential source of volcanic sulfur in the fall of 2014.
The monthly average (circles) effective deposition velocity of different chemical species to snow at Alert during 2014–2015. Median values (bars) are also shown. The effective deposition velocity encompasses both wet and dry deposition processes. In general, the warmer months have higher deposition velocities than the colder months, likely due to enhanced wet deposition in the former. Figure from Macdonald et al. (2017).
A simple parameterization of BC in snow was developed and tested in the Canadian Atmospheric Global Climate Model (CanAM). According to the parameterization, the temporal evolution of the concentration of BC near the top of the snowpack is determined by changes in dry and wet deposition of BC, the snowfall flux and scavenging by snow meltwater. Comparison of model results with a multi-year climatology of BC concentrations in snow produces good agreement for locations in the Canadian Arctic and sub-Arctic (Doherty et al., 2010, 2014; X. Wang et al., 2013) as well as for other regions in the Northern Hemisphere (Namazi et al., 2015). Simulated changes in BC loading in snow in the second half of the 20th century had much smaller cryospheric impacts on surface air temperatures than other aerosol and greenhouse gas radiative forcings.
Understanding the impacts of ship emissions on climate and air quality of the Arctic environment is challenging but important, given the likelihood of future increases in Arctic shipping (Corbett et al., 2010; Pizzolato et al., 2014; Winther et al., 2014). The Arctic atmospheric boundary layer exhibits different dynamics from mid-latitudes, being characterized by thermally stable conditions with reduced turbulent mixing (Aliabadi et al., 2016a). Ships navigating northern latitudes operate under partial engine load and ice-breaking conditions as opposed to full speed cruising. Uncertainties are compounded by the lack of accurate predictions for increased ship traffic patterns in the Arctic as the ice cover retreats, as well as the lack of a robust regulatory framework to control emissions via emission control areas set by the International Maritime Organization (Aliabadi et al., 2015).
The NETCARE campaign near Resolute Bay in July 2014 characterized typical ship emissions
and plume evolution by mapping the plume of the CCGS
Currently, clouds are responsible for some of the greatest uncertainties in climate
change predictions. This is in large part because the properties of clouds and their
formation processes are poorly understood, especially in the case of ice and mixed-phase
clouds (Cantrell and Heymsfield, 2005; Hegg and Baker, 2009; Murray et al., 2012).
Particles that can initiate ice formation in the atmosphere at temperatures and relative
humidities lower than those required for homogeneous freezing of solution droplets are
referred to as ice nucleating particles (INPs; Vali et al., 2015). Only a very small
fraction of atmospheric particles (1 in 10
The sea surface microlayer is the interface between the atmosphere and the ocean and is a source of particles to the atmosphere via wave-breaking and bubble-bursting. INPs have previously been detected in bulk seawater (Schnell, 1977; Schnell and Vali, 1975, 1976); however, concentrations and properties of INPs in the microlayer have not been confirmed prior to the start of NETCARE. This lack of information led to large uncertainties in quantifying the importance of the microlayer as a source of INPs to the atmosphere (Burrows et al., 2013). In initial experiments, the concentration of INPs in the microlayer collected off the west coast of Canada were measured (Wilson et al., 2015); while in parallel, researchers from the University of Leeds measured the properties and concentrations of INPs in the microlayer collected off the east coast of the United States and Greenland (Wilson et al., 2015). We built on this work by measuring the concentrations and properties of INPs in the microlayer collected in the Canadian Arctic (Irish et al., 2017, 2018).
Microlayer samples were collected in the Canadian Arctic during the summers
of 2014 and 2016 from the CCGS
The size of INPs collected from the atmosphere at Alert in late spring and early summer
2014 were also measured (Mason et al., 2016). The size of atmospheric INPs can help
distinguish which types of atmospheric particles are important as INPs. During this
campaign, the average concentrations of atmospheric INPs were 0.05, 0.22, and
0.99 L
Results from measurements in the Arctic marine boundary layer during summer 2014
(Irish et al., 2019).
During March 2016, INP measurements at Alert were made daily (Si et al.,
2018). In these high-frequency data, INP concentrations were strongly
correlated with tracers of mineral dust at a freezing temperature of
During the summer of 2014, we measured atmospheric concentrations of INPs in the Canadian
Arctic marine boundary layer on board the CCGS
Sulfuric acid coatings strongly affect INPs and thus their effect on clouds
and precipitation. This is particularly important during Arctic haze events.
Laboratory studies (Eastwood et al., 2009), in situ measurements (Jouan et
al., 2014), and large-scale observations from the CloudSat and CALIPSO
satellites (Grenier et al., 2009; Grenier and Blanchet, 2010) support the
hypothesis of a dehydration–greenhouse feedback (Blanchet and Girard, 1994)
linking acidified aerosols to the favoured formation of larger ice crystals
and light precipitation through a reduction of INP activity. In cold Arctic
conditions, thin ice clouds (TIC), like cirrus, are ubiquitous in the coldest
free troposphere (Grenier et al., 2009). Two types have been identified:
TIC-1, which has many small crystals (smaller than
A far IR radiometer (FIRR) developed with Canadian Space Agency support and especially
designed to measure TIC properties and water vapour was flown on board the
The NETCARE research outlined above has provided novel insights into (1) the biogenic sources of gases that can impact the size and composition of Arctic aerosol; (2) new particle formation and growth into sizes that were demonstrated to be activating to form cloud droplets, with growth occurring largely via formation of Arctic marine secondary aerosol; (3) the sources and properties of Arctic haze aerosol, in particular BC-containing particles; (4) Arctic INPs in the air, ocean water, and the sea surface microlayer; and (5) deposition rates of pollutants to snow. Many of these advances arose as a result of the highly interdisciplinary approach taken within NETCARE. Nevertheless, despite these advances, many research questions remain, as outlined in this section.
Observations gathered during NETCARE provide a valuable benchmark upon which to base predictions of the changes in the source strength of secondary MBA precursors in response to alterations in Arctic climate. This is important for determining the amounts of both aerosol sulfate and organics. The thinning and loss of seasonal sea ice, which is driven by warming and polar amplification, is by far the most conspicuous of these alterations (AMAP, 2013; Comiso, 2011; Serreze and Barry, 2011).
The marine production of DMS could be particularly sensitive to both modifications in
seasonal ice extent and the intermittent presence of melt ponds above the ice in spring
and summer (Gabric et al., 2017; Gourdal et al., 2018; Lizotte et al., 2019). Modelling
and observational studies suggest that the northward shrinking of the seasonal ice extent
and the ensuing increase in open waters available for gas exchange could lead to
heightened primary productivity (Arrigo and van Dijken, 2015; Gabric et al., 2017; Ito
and Kawamiya, 2010) and production of DMS (Levasseur, 2013). In turn, this would lead to
higher atmospheric MSA and secondary sulfate (Sharma et al., 2012), and background
particle concentrations (Dall'Osto et al., 2017). Simulations with CanAM indicate that
associated increases in concentrations of CCN could potentially offset part of the
warming due to enhanced cloud albedo (Mahmood et al., 2018). Specifically, the projected
loss of sea ice between 2000 and 2050 leads to a substantial increase in Arctic DMS
emissions in CanAM, leading the cloud radiative forcing associated with Arctic DMS to
increase from
The Arctic system may also be vulnerable to other changes, notably ocean acidification,
as well as amplified warming and freshwater inputs (AMAP, 2013; Yamamoto-Kawai et al.,
2009). An experimental assessment of the impact of ocean acidification on DMS-producing
planktonic communities of Baffin Bay during NETCARE (Hussherr et al., 2017) revealed that
DMS production may decrease by 25 % under end-of-the-century scenario reductions of
pH (
NETCARE illustrated for the first time the influence of ammonia emissions from seabird colonies on not only atmospheric mixing ratios (Wentworth et al., 2016) but also new particle formation, aerosol neutralization, and associated indirect effects on climate (Croft et al., 2016a, 2018). How will these emissions evolve with climate change and potential changes in wildlife populations (Weimerskirch et al., 2018), habitat, and migratory patterns? NETCARE measurements from Alert suggest that Arctic soils may also be an ammonia source (Murphy et al., 2019), perhaps reflecting the redistribution of ammonia between different components of the Arctic land–ocean ecosystem. This highlights the importance of including bidirectional fluxes in chemical transport models for species that move readily between the land, atmosphere, and ocean.
Unlike lower-latitude marine boundary layers (Quinn and Bates, 2011) particle nucleation and growth was frequently observed during NETCARE campaigns in the boundary layer in Arctic marine and coastal regions. The Arctic may behave differently for a number of reasons: (1) the persistent temperature inversion lowers the rate of mixing of surface emissions; (2) the summertime atmosphere has 24-hour sunshine to drive photo-oxidation; (3) the condensation sink is particularly low; and (4) the low temperatures facilitate molecular cluster formation. It is crucial to assess the chemical components and particle formation mechanisms that prevail in this distinctive environment. Particularly valuable will be on-line mass spectrometric measurements of the chemical composition of the smallest clusters and particles that form at the early stages of the nucleation and growth process.
The growth of 5–50 nm particles into CCN size ranges was evident in multiple NETCARE campaigns (Burkart et al., 2017a, b; Collins et al., 2017; Willis et al., 2016, 2017). Surprisingly, much of the submicron aerosol mass associated with growth was organic in composition (Burkart et al., 2017b; Willis et al., 2016, 2017), providing additional support to the idea that secondary organic aerosol of marine origin is important (Rinaldi et al., 2010). Although we do not know the precise nature of the organic precursors, the NETCARE studies described in Sect. 4.2 illustrated that the secondary organic source is marine and potentially associated with oxidation or photochemistry of the sea surface microlayer (see Sects. 3.3 and 6.3). It is important to determine the balance between secondary aerosol formation versus primary particle formation from sea spray. In one NETCARE case study of new particle formation and growth over Lancaster Sound, there were indications of secondary processes occurring alongside formation of sea spray salt particles, suggesting that these processes may sometimes occur simultaneously, complicating analyses (Collins et al., 2017; Köllner et al., 2017; Willis et al., 2016). Key uncertainties in the radiative effects of the Arctic MSOA simulation in GEOS-Chem-TOMAS include Arctic nucleation processes, the chemical composition of Aitken particles, and the volatility of the SOA (Croft et al., 2018). Further understanding of these processes would better constrain climate feedbacks. We note that the composition and properties of Arctic MSOA are not necessarily the same as that formed in marine environments in other parts of the world.
Our understanding of how the sea surface microlayer impacts air–sea exchange of aerosols and gases is still largely circumstantial and is based mainly on conceptual models (Garbe et al., 2014; Lewis and Schwartz, 2013), laboratory experiments (Bigg and Leck, 2008; Wilson et al., 2015), and observations of similarities between particulate matter in the microlayer and the atmosphere (Leck and Bigg, 2005). Obtaining information on how these concepts play out in the real world has proven extremely challenging. That being said, one pronounced example of the potential importance of the sea surface microlayer comes from work in NETCARE that demonstrated that OVOCs in a marine Arctic setting were likely formed photochemically within the microlayer or by oxidation of gases arising from it (Mungall et al., 2017). Although laboratory studies have previously demonstrated that OVOCs can be chemically generated from microlayer surrogate materials (Rossignol et al., 2016; Zhou et al., 2014), such studies do not address the chemical complexity of the genuine environmental system.
Although additional experiments have previously quantified the impact of microlayer surfactant enrichment on gas exchange (Brockmann et al., 1982; Frew et al., 2004; Pereira et al., 2016), to date no one has directly tied natural variations in the sea surface microlayer to the exchange of aerosols or gases. The main difficulty lies in the different temporal and spatial scales of atmospheric and microlayer measurements. The composition of the microlayer is highly heterogeneous even on small horizontal scales (Cunliffe et al., 2013), and recovery of microlayer samples for chemical analysis is time consuming. Thus, tying those measurements to observations of temporally variable aerosols measured from ships or airplanes is innately subject to substantial uncertainties.
In order to confidently identify the relationship between the composition of the sea surface microlayer and atmospheric aerosol production, it will be necessary to collect coherent data sets from single platforms, such as autonomous surface craft (Ribas-Ribas et al., 2017). In addition, intelligently designed time series stations could provide sufficient data to identify clear relationships between the microlayer and the atmosphere (Cunliffe et al., 2013; Engel et al., 2017).
The 2014 NETCARE aircraft campaign illustrated that the low CCN numbers prevalent in the
summer boundary layer can lead to large cloud droplet diameters, in some cases
approaching 30
As described in Sect. 2, there is a transition in scavenging regimes between the efficient processes that occur with liquid clouds and the comparatively inefficient processes associated with pure ice clouds. However, the community struggles to accurately include such scavenging processes in models (Mahmood et al., 2016). This is important for long-range transport of pollutants from more industrialized southerly locations, and for the input of biomass burning aerosol that is likely to become more prevalent with the warming climate (Marelle et al., 2015; Shindell et al., 2008; H. Wang et al., 2013).
The degree of aerosol scavenging that occurs outside the Arctic relative to
that which occurs within it must be better established. For example,
transport associated with warm conveyor belt systems is one mechanism that
supplies pollutants to the Arctic (Ancellet et al., 2014; Roiger et al.,
2011), while cloud formation associated with synoptic uplift in mid-latitudes
cleans the air. This has been nicely demonstrated by a close inverse
relationship between the accumulated precipitation along back trajectories, a
measure of wet scavenging, and BC levels arriving in the Arctic (Matsui et
al., 2011). Deciphering the efficiency of such extra-Arctic processes is one
focus of the proposed IMPAACT project
(
A related question is the degree to which oxidation processes modify the overall aerosol
composition as a function of residence time in the Arctic. For the measurements described
in Sect. 4.4, the increase in the sulfate-to-organic ratio with decreasing altitude in
springtime aerosol may in part arise from formation of sulfate as the air mass ages.
Validation of this mechanism awaits better
While aerosol particle removal is exceedingly efficient under summer
conditions, the ice nucleation processes in the colder months are much more
selective and less well understood, as described in Sect. 5. We sought to
understand which aerosol types contain the best ice-nucleating particles.
Initial indications from NETCARE measurements are that dust is an important
contributor to the INP population (Fig. 14), but that does not rule out the
role of primary sea spray particles acting as INPs. A second important
question was to what degree coatings of secondary materials, such as
sulfates or organics, modify the ice nucleation properties of primary INPs,
such as mineral dust. A major challenge is the development of better
parameterizations of INPs for use in atmospheric modelling. To that end,
work in NETCARE improved ice nucleation parameterizations in the Global
Multiscale Environmental Model (GEM-LAM) to determine the effect of
pollution on clouds in the Arctic (Keita and Girard, 2016).
To simulate pristine clouds, a parameterization of ice nucleation by mineral
dust was included, whereas to simulate ice clouds influenced by pollution, a
parameterization of ice nucleation by mineral dust coated with sulfuric acid
was used. A parameterization was developed as well to test against the 2014
and 2016 CCGS
Although NETCARE measurements of aerosol deposition fluxes to the snow at Alert were made across a full cold season (see Sect. 4.7), the degree to which this flux occurred via dry or wet deposition was not precisely determined. In particular, it remains to be determined how important ice cloud scavenging and settling is as a particle removal process. A long-term, high time resolution aerosol deposition network that separates wet and dry deposition across the Arctic would be highly beneficial in this regard.
Mixing state refers to the uniformity of the distribution of the aerosol chemical components across an array of particles; i.e. are all the particles of the same chemical composition or is their chemical distribution heterogeneous? NETCARE measurements have highlighted how this information is crucial to our understanding of aerosol sources and impacts. In particular, the springtime measurements of BC aerosol described in Sects. 4.4 and 4.5 showed that within the Arctic haze sampled in spring 2015, only 3 %–16 % of the particles contained BC inclusions and that BC-containing particles had coatings 40–45 nm in thickness on average (Kodros et al., 2018). The direct radiative forcing that is modelled using these results as constraints is distinctly different from that where it was assumed that the chemical mixing state of the aerosol was uniform (see Sect. 4.5). Likewise, in the summertime measurements from 2014, the single-particle mass spectrometry measurements at low altitudes over open water illustrate that primary and secondary marine aerosol components were externally mixed, thus indicating different formation processes (Fig. 9; Köllner et al., 2017; Willis et al., 2016). More measurements of this type, down to as small a particle size as possible, are crucial for further determining the balance between primary and secondary marine aerosols, to establishing the degrees of coating that exist on mineral dust aerosol that contain INPs, and to assessing the efficiency of cloud scavenging that occurs across different particle types. For example, does the relatively hydrophobic character of BC inhibit the rate at which it is wet cloud scavenged, and if so, how much hygroscopic coating material must be present to make the particles CCN active?
The Arctic springtime has been much more extensively studied than other seasons. This is
understandable given the importance of the Arctic haze phenomenon. However, the fall and
winter seasons are poorly characterized using intensive campaign approaches, largely
because of the operational difficulties in working under cold, dark conditions. Although
remote sensing can be used to study transitions between seasons (see Sect. 4.6), it is
still important to better understand how transport patterns of pollutants and their
deposition rates change seasonally. As well, we know very little about the polar night
and the associated formation of ice clouds. The radiative effects of these clouds and
their ability to dehydrate the atmosphere on a large scale through extensive light
precipitation are important to assess. An exciting movement in this direction is the
development of a far infrared radiometer (FIRR) that was flown on the
The vertical profiles measured as part of NETCARE in both the springtime and summertime
provide essential information for comparison to model outputs and provide a necessary
complement to the much more extensive sets of measurements from ground-based field
campaigns and stations. Additional aircraft campaigns that provide such vertically
resolved features are a top priority for future studies. Nevertheless, the continuous
measurements at Arctic ground stations remain our most valuable data set to assess
long-term trends. There is a significant need to enhance the instrumental capabilities at
these stations, for example with key continuous measurements of
The NETCARE atmospheric measurements are publicly available
through the Government of Canada open data portal
(
The supplement related to this article is available online at:
JPDA coordinated and wrote substantial portions of the paper. All other co-authors contributed text and/or reviewed the paper. All co-authors either wrote a first-author paper as part of the NETCARE project or else contributed in a substantive manner to the research conducted in the project or presented in the paper. EG contributed to the INP and ice cloud research of NETCARE but he died before submission. We regard his approval for inclusion of his name on this paper as implicit.
The authors declare that they have no conflict of interest.
This article is part of the special issue “NETCARE (Network on Aerosols and Climate: Addressing Key Uncertainties in Remote Canadian Environments) (ACP/AMT/BG inter-journal SI)”. It is not associated with a conference.
NETCARE was funded by the Natural Sciences and Engineering Research Council (NSERC) of
Canada under its Climate Change and Atmospheric Research program, with additional
financial and in-kind support from Environment and Climate Change Canada, Fisheries and
Oceans Canada, the Alfred Wegener Institute, and the Major Research Project Management
Fund at the University of Toronto. Colorado State University researchers were supported
by the US Department of Energy's Atmospheric System Research, an Office of Science,
Office of Biological and Environmental Research program, under grant no. DE-SC0011780,
the US National Science Foundation, Atmospheric Chemistry program, under grant
no. AGS-1559607, and by the US National Oceanic and Atmospheric Administration, an Office
of Science, Office of Atmospheric Chemistry, Carbon Cycle, and Climate Program, under the
cooperative agreement award no. NA17OAR430001. All authors would like to strongly thank:
(i) the editors for the NETCARE special issue in
This paper is dedicated to Eric Girard, a NETCARE scientist who died 10 July 2018. Eric Girard contributed greatly to the field of Arctic cloud and aerosol microphysics during his research career. Edited by: Ken Carslaw Reviewed by: two anonymous referees