The Arctic Research of the Composition of the Troposphere from Aircraft and Satellites (ARCTAS) Mission: Design, Execution, and First Results

. The NASA Arctic Research of the Composition of the Troposphere from Aircraft and Satellites (ARC-TAS) mission was conducted in two 3-week deployments based in Alaska (April 2008) and western Canada (June–July 2008). Its goal was to better understand the factors driving current changes in Arctic atmospheric composition and climate, including (1) inﬂux of mid-latitude pollution, (2) boreal forest ﬁres, (3) aerosol radiative forcing, and (4) chemical processes. The June–July deployment was preceded by one week of ﬂights over California (ARCTAS-CARB) focused on (1) improving state emission inventories for greenhouse gases and aerosols, (2) providing observations to test and improve models of ozone and aerosol pollution. ARCTAS involved three aircraft: a DC-8 with a detailed chemical payload, a P-3 with an extensive aerosol and radiometric payload, and a B-200 with aerosol remote sensing instrumentation. The aircraft data augmented satellite observations of Arctic atmospheric composition, in particular from the NASA A-Train. The spring phase (ARCTAS-A) revealed pervasive Asian pollution throughout the Arctic as well as signiﬁcant European pollution below 2 km. Unusu-Correspondence high of carbonaceous Satellite observations of BrO column hotspots were found suggesting the presence of inorganic Fresh ﬁre from California during the phase indicated low NO x factors no secondary and no signiﬁcant ozone enhancements except when with PILS Particle Liquid Sampler; SP2 Particle Soot Photometer; AFS Fluorescence


5192
D. J. Jacob et al.: ARCTAS mission: design, execution, and first results interpret and augment the continuous observations of Arctic atmospheric composition from satellites. The aircraft were based in Alaska in April (ARCTAS-A) and in western Canada in June-July (ARCTAS-B). ARCTAS was part of the international POLARCAT program of field observations of Arctic atmospheric composition during the  and was coordinated with several concurrent field programs. This paper provides an overview of the ARCTAS mission and a synthesis of first results.
The Arctic has been warming rapidly over the past century, and the pace has accelerated in the past decades (Trenberth et al., 2007). Sea ice cover and permafrost show rapid declines (Lemke et al., 2007). General circulation models (GCMs) project strong 21st-century warming in the Arctic, reflecting in part the sea ice feedback (Christensen et al., 2007). Even so, these models underestimate the observed warming over the past decades (Shindell and Faluvegi, 2009), suggesting that they might underestimate future warming. One possible factor missing from current climate models is proper representation of the regional radiative forcing from aerosols and tropospheric ozone, which may be particularly important in the Arctic (Law and Stohl, 2007;Shindell et al., 2007;Quinn et al., 2008). Decreasing the forcing from these short-lived agents could provide an opportunity to curb Arctic warming. The primary goal of ARCTAS was to better understand the sources of aerosols and tropospheric ozone in the Arctic, as well as the radiative properties of Arctic aerosols and of the underlying surface, in order to improve estimates of the associated radiative forcings.
Inflow of pollution from northern mid-latitudes is a wellknown source of aerosols and ozone to the Arctic. It prompts not only climatic but air quality and ecological concerns. It has been known for decades that the Arctic is heavily affected by mid-latitude pollution in winter-spring, manifested by visible "Arctic haze" aerosol layers (Radke et al., 1984;Barrie and Hoff, 1985;Brock et al., 1989;Shaw, 1995). Arctic observations have been mostly at the surface, leading to the conventional view of a strong seasonal variation in midlatitude pollution inflow with northern Eurasia as the dominant contributor (Rahn, 1981;Barrie, 1986). However, recent models and limited observations show transport from mid-latitudes taking place year-round through different processes at different altitudes and involving all three source continents Koch and Hansen, 2005;Stohl, 2006;Shindell et al., 2008). Sources at mid-latitudes are rapidly changing, with increases in Asia driven by industrialization and decreases in North America and Europe driven by air quality regulations. Better understanding of pollution transport to the Arctic and of the associated sourcereceptor relationships was a priority for ARCTAS.
Chemical processing plays a major role in the evolution of aerosols, ozone, and other pollutants in the Arctic. The chemical environment in the Arctic is characterized by remoteness, intense cold, and long periods of light and dark-ness. Photochemical reactions on snow and sea ice surfaces following the winter darkness provide a springtime source of halogen radicals that drive rapid ozone destruction and mercury oxidation (Simpson et al., 2007a). Halogen radical generation may be highly sensitive to climate change through sea ice cover (Simpson et al., 2007b). A goal of ARCTAS was to better understand the radical photochemistry of the Arctic atmosphere and its implications for the regional budgets of aerosols, ozone, and mercury.
Boreal forest fires in Siberia, Canada, and Alaska represent a major perturbation to the Arctic atmosphere in summer affecting the carbon cycle (Preston and Schmidt, 2006), climate (Amiro et al., 2001), air quality , and land ecology (Soja et al., 2007). Long-term records for Canada show an increase in fires over the past 50 years correlated with rising temperatures (Gillett et al., 2004;Girardin, 2007), and this increase is expected to continue as a result of warming (Stocks et al., 1998;Flannigan et al., 2005). Aerosols emitted by the fires could represent a significant forcing or feedback to regional climate change in the Arctic. The summer deployment of ARCTAS focused on better quantifying the fire emissions, the evolution of fire plumes, and the associated radiative effects.
A central motivation for ARCTAS was to enhance the ability of satellite observations of atmospheric composition to address the above issues. Polar-orbiting satellites have unique potential. They observe the long-range transport of mid-latitude pollution to the Arctic. They detect fires and the associated smoke plumes. They measure aerosol optical depths and other aerosol properties, and the concentrations of gas-phase species including ozone, CO, NO 2 , and BrO that are important drivers of Arctic chemistry. Table 1 lists the most important satellite measurements of atmospheric composition during ARCTAS. Observation of the Arctic from space poses unique challenges including bright and cold surfaces, low sun angles, seasonal darkness, and extensive cloud cover. ARCTAS aimed to use aircraft observations to provide the validation, retrieval constraints, correlative data, and process information needed to develop the potential of satellites for Arctic research.
The summer ARCTAS-B deployment was preceded by one week of flights over California sponsored by the California Air Resources Board (CARB) to address regional issues of air quality and climate forcing. Many areas of California exceed the state ozone and fine particulate matter (PM 2.5 ) air quality standards, and little improvement in ozone has been observed over the last decade despite continued emission reductions (Cox et al., 2008). Controlling emissions from heavy-duty diesel trucks has been far more difficult than from passenger cars. Shipping emissions have increased. Long-range transport of pollution from Asia or Mexico may be offsetting benefits from local pollution control measures. In addition to these air quality concerns, CARB is embarking on strategies to reduce state greenhouse gas emissions. The ARCTAS-CARB flights focused on collecting data to MOPITT (Terra, 1999) NIR/TIR CO Deeter et al. (2009) MISR (Terra, 1999 Vis/NIR Aerosol optical depth, injection heights Kahn et al. (2005Kahn et al. ( , 2007  aid California improve its emission inventories and air quality models.

ARCTAS scientific themes and mission design
ARCTAS was organized around four main research themes: (1) long-range transport of pollution to the Arctic, (2) boreal forest fires, (3) aerosol radiative forcing, and (4) ozone budget and chemical processes. We elaborate here on each of these themes and how they defined the objectives and design of the mission. We also discuss the separate California air quality objectives of the ARCTAS-CARB component.

Long-range transport of pollution to the Arctic
Transport of pollution from northern mid-latitudes to the Arctic takes place throughout the year but the processes and source regions vary with season . Intense cooling in winter produces a "polar dome" of cold, stable air that shields the Arctic lower troposphere from isentropic transport from lower latitudes . Low-altitude transport from cold northern Eurasia can still take place and provides then the dominant source of pollution measured at the surface (Stohl, 2006). Transport from North America and East Asia takes place mainly at higher altitudes, in particular through warm conveyor belts (WCBs) associated with storm tracks on the east coast of the continents (Stohl, 2001). This cyclonic activity increases from winter to spring, resulting in a springtime peak of sulfate concentrations in the free troposphere (Scheuer et al., 2003). Midlatitude pollution influence over the Arctic still takes place in summer and fall but is diluted by the weaker general circulation and by more efficient photochemical loss and aerosol scavenging (Barrie, 1986;Brock et al., 1989;Klonecki et al., 2003). Pollution transport to the Arctic in winter-spring is highly layered (Radke et al., 1984;Brock et al., 1989). This layering reflects the stratification of the Arctic atmosphere, the lack of precipitation, and also possibly the local heating by lightabsorbing aerosol (Quinn et al., 2008;Stone et al., 2008). Scavenging of aerosols and other water-soluble species takes place during WCB lifting of air masses, but this scavenging is not complete Miyazaki et al., 2003) and preferentially removes sulfate relative to black carbon (BC) (Park et al., 2005).
The long-range transport theme of ARCTAS targeted the following questions: -What are the transport pathways for different pollutants to the Arctic?
-What are the contributions from different source regions and what are the source-receptor relationships?
Satellites have obvious potential for observing the longrange transport of pollution to the Arctic. This potential is least in the winter when conditions are cold and dark, and when most of the transport takes place at low altitudes under a prevailingly cloudy atmosphere. We chose therefore to focus on spring as holding more promise for satellite observation, and as offering more opportunities to observe the long-range transport from Asia which is likely to grow in the future. Spring is also of more relevance than winter for snowmelt and aerosol radiative forcing. Previous springtime aircraft missions to the Arctic included the Arctic Gas and Aerosol Sampling Program (AGASP) in 1983 (Schnell, 1984) and the Tropospheric Ozone Production about the Spring Equinox (TOPSE) experiment in 2000 . Further context was provided by previous NASA aircraft campaigns along the Asian Pacific Rim in spring 1994 (PEM-West B; Hoell et al., 1997) and spring 2001 (TRACE-P; Jacob et al., 2003), and over the Northeast Pacific and southern Alaska in spring 2006 (INTEX-B;Singh et al., 2009). PEM-West B and TRACE-P observed frequent lifting of Asian pollution in WCBs followed by northeastward transport toward the Arctic in the general direction of Alaska (Fuelberg et al., 2003). This led to the selection of Fairbanks, Alaska as the main operational base for the ARCTAS-A spring deployment, and an overnight base in Thule, Greenland to achieve broader spatial coverage.

Boreal forest fires
Wildfires in North America and Eurasia are important regional and hemispheric sources of aerosols, tropospheric ozone, and possibly mercury (Lavoue et al., 2000;Turetsky et al., 2006;Generoso et al., 2007;Leung et al., 2007). Fires are particularly active in the boreal forest zone in summer. Buoyancy generated by the fires can result in injection above the boundary layer (Kahn et al., 2008;McMillan et al., 2008) and in extreme cases in the stratosphere (Fromm and Servanckx, 2003). This lifting of emissions has important implications for the transport, evolution, and impacts of the fire plumes (Morris et al., 2006;Turquety et al., 2007;Hyer et al., 2007;Leung et al., 2007).
Emissions from boreal fires depend on burn conditions (Goode et al., 2000;Reid et al., 2005;Preston and Schmidt, 2006). Rapid aerosol evolution takes place in the fresh plumes through particle coagulation and condensation of organic gases, ammonia, and water (Radke et al., 1995;Fiebig et al., 2002;Colarco et al., 2004). This near-field evolution of the fire plumes determines their regional and global radiative impact (Westphal and Toon, 1991;Fiebig et al., 2003). Ozone production takes place in the fresh plumes from the precursor nitrogen oxide (NO x ) radicals, though this is limited by conversion to peroxyacetylnitrate (PAN) (Jacob et al., 1992). Subsequent decomposition of PAN as the plumes age and sink yields additional ozone production on the hemispheric scale (Real et al., 2007).
The boreal forest fires theme of ARCTAS targeted the following questions: -What are the emissions from boreal forest fires, and what is the near-field chemical evolution of the fire plumes?
-What are the optical properties of the aerosols emitted by boreal forest fires, and how do they evolve during aging of the fire plume?
-What are the injection heights of fire emissions and what processes determine these heights?
-What are the implications of boreal forest fires for regional and global atmospheric composition?
Addressing these questions was the principal objective of the ARCTAS-B summer deployment. July was selected to maximize the probability of sampling large fires in boreal North America. The DC-8 and P-3 aircraft were based in Cold Lake, Alberta (54 N, 110 W), and the B-200 in Yellowknife, Northwest Territories (62 N, 114 W), on the basis of climatological fire data and available facilities. An overnight base in Thule for the DC-8 was included in the summer deployment to increase spatial coverage and the opportunities for sampling aged biomass burning plumes.

Aerosol radiative forcing
Aerosol radiative forcing, both direct (clear-sky scattering and absorption) and indirect (aerosol effects on clouds), could be an important regional driver of Arctic climate change (Quinn et al., 2008). Major aerosol types include anthropogenic particles associated with Arctic haze (Barrie and Hoff, 1985;Bodhaine et al., 1989;Clarke, 1989), biomass burning particles from both boreal and lower-latitude fires, and mineral dust particles. These different origins can be separated on the basis of chemical, microphysical, and optical aerosol properties (Bergstrom et al., 2007;. Radiative effects of aerosols vary not only with aerosol composition but also with solar irradiance, surface albedo, and altitude (Pfister et al., 2008;Stone et al., 2008). BC aerosol deposited on snow has a warming effect and may promote snow and sea ice melting (Clarke and Noone, 1985;Hansen and Nazarenko, 2004;Flanner et al., 2007).
Aerosol optical properties and the resulting radiative forcing evolve with atmospheric aging (Fiebig et al., 2003;Preston and Schmidt, 2006). This has been observed in particular for boreal forest fire plumes in Canada (O'Neill et al., 2002). Anthropogenic Arctic haze has a wide range of ages and scavenging histories, and this is expected in turn to yield a range of optical properties and radiative forcing .
The aerosol radiative forcing theme of ARCTAS targeted the following questions: -What is the regional radiative forcing from Arctic haze and fire plumes?
-How does this forcing evolve during plume aging?
-What are the major sources of BC aerosol to the Arctic?
-How does BC aerosol deposition affect ice albedo?
Characterizing the forcing associated with aerosol layers was the central task of the P-3 aircraft equipped with an ensemble of in situ aerosol and radiative observations. Validation and interpretation of the CALIOP satellite observations of altitude-resolved aerosol extinction was a primary charge of the B-200 aircraft equipped with a high spectral resolution lidar (HSRL). During the spring deployment, the B-200 operated out of Barrow, Alaska to coordinate with the Indirect and Semi-Direct Aerosol Campaign (ISDAC) aircraft mission of the US Department of Energy (DOE). ISDAC focused on aerosol-cloud interactions and was centered over the long-term Atmospheric Radiation Measurement (ARM) North Slope of Alaska (NSA) Climate Research Facility (CRF) at Barrow.

Ozone budget and chemical processes
Tropospheric ozone at high northern latitudes shows a pronounced spring maximum (Logan, 1985;Monks, 2000). The TOPSE aircraft campaign over the North American Arctic in February-May 2000  found that this maximum is driven by tropospheric production rather than by stratospheric input, and that the ozone budget in the region largely reflects a balance between in situ production and loss (Browell et al., 2003;Emmons et al., 2003). However, ozonesonde data from Canadian stations show significant stratospheric impact on tropospheric ozone in both spring and summer (Tarasick et al., 2007;Thompson et al., 2007). Stratospheric influence could be important in the upper troposphere where ozone is of most concern as a greenhouse gas, and this influence could increase in the future as stratospheric ozone recovers and the Brewer-Dobson circulation increases due to greenhouse forcing (Shindell et al., 2007). The Canadian ozonesonde data show a decrease in tropospheric ozone in the 1980s followed by an increase since then, for reasons that are unclear (Tarasick et al., 2005;Kivi et al., 2007).
Tropospheric ozone chemistry in the Arctic is largely determined by the supply of hydrogen oxide radicals (HO x ) and nitrogen oxide radicals (NO x ) but the processes controlling the abundances of these radicals are poorly understood, including the possible role of photochemistry taking place in snow (Grannas et al., 2007). An additional feature of the Arctic boundary layer in spring is the presence of high concentrations of bromine and chlorine radicals produced by photochemical reactions involving sea salt particles deposited on sea ice. High concentrations of bromine radicals in the Arctic boundary layer lead to catalytic destruction of ozone (Barrie et al., 1988) and drive rapid oxidation of elemental mercury to oxidized mercury which is subsequently deposited (Goodsite et al., 2004), resulting in ozone depletion events (ODEs) and mercury depletion events (MDEs) (Simpson et al., 2007b;Steffen et al., 2008). Chlorine atoms can also lead to rapid oxidation of volatile organic compounds (VOCs) (Zeng et al., 2006). Fresh sea ice and ice leads are thought to be preferential environments for generation of halogen radicals (Bottenheim and Chan, 2006;Simpson et al., 2007a), but the ODEs can spread over 1000 km downwind Zeng et al., 2006). Satellite observations of BrO columns indicate hotspots over the Arctic in spring that have been attributed to the boundary layer (Chance, 1998;Hollwedel et al., 2004) but these observations had not been validated prior to ARCTAS.
The chemical processing theme of ARCTAS targeted the following questions: -What controls HO x and NO x chemistry in the Arctic?
-What drives halogen radical chemistry in the Arctic, and what is its regional extent?
-What are the regional implications for ozone, aerosols, and mercury?
-How does stratosphere-troposphere exchange affect tropospheric ozone in the Arctic?
The DC-8 included a detailed chemical payload to address these questions. One priority was to validate the satellite observations of high BrO events. These events are most frequent in March-April (Hollwedel et al., 2004). We chose April as more favorable in terms of photochemistry, satellite observation capabilities, and operations weather.

California air quality (ARCTAS-CARB)
The objectives of the California deployment (ARCTAS-CARB) were to: (1) improve state emission inventories for greenhouse gases and aerosols; (2) characterize offshore emissions of sulfur and other pollutants from shipping and natural sources; and (3) characterize upwind boundary conditions for modeling local surface ozone and PM 2.5 . Despite extensive abatement efforts, air quality standards are widely exceeded in southern California and the Central Valley (Cox et al., 2008). Little improvement in ozone has been observed in the past decade despite continued reduction in precursor emissions. The causes for this lack of progress are unknown but could reflect poorly understood non-linearities in the photochemical system. There is far more sulfate in southern California than can be accounted for in present emission inventories and ship emissions may be implicated (Spencer et al., 2009). Another issue is the poorly quantified import of pollution from Asia and Mexico Oltmans et al., 2008;. California is now embarking on synergistic strategies to control greenhouse gas emissions together with air pollutants. A first step is to improve inventories for these emissions.
The ARCTAS-CARB flights targeted the following questions: -How good is our current understanding of the HO x -NO x -O 3 -aerosol photochemical system over the Los Angeles Basin as represented in air quality models?
-How should upwind boundary conditions for simulating air quality in California be specified? -What are the state emissions of VOCs and greenhouse gases from urban and industrial activities, agricultural operations, and wildfires?
The flights took advantage of the fully loaded DC-8 and P-3 aircraft based in California just before the ARCTAS-B summer deployment. In addition to the local air quality objectives, these California flights helped to address the ARC-TAS Arctic objectives by providing contrasting data, expanding the ranges of the instruments, and characterizing emissions from California wildfires that would then be sampled as aged plumes from the aircraft based in Canada. Table 2 gives the characteristics of the DC-8, P-3, and B-200 aircraft as used in ARCTAS. Tables 3a-c lists the payloads for each. The DC-8 had the largest range and payload. It was responsible for addressing the long-range transport and chemical processing themes of ARCTAS, and also contributed to the boreal fires theme. Its payload included measurements of ozone (in situ and remote); HO x , NO x , and BrO radicals; reservoirs for these radicals and related species; aerosols, carbon gases, and mercury; tracers of transport; and actinic fluxes. Measurements of HO x radicals were made by two independent techniques (LIF and CIMS), since previous comparisons between data sets and with models in past aircraft missions had been inconsistent (Ren et al., 2008).

Aircraft payloads and support
The P-3 and B-200 payloads focused on aerosol characterization and radiation, targeting the aerosol radiative forcing theme of ARCTAS and addressing aerosol source issues related to Arctic haze and boreal fires. The P-3 payload included in situ measurements of aerosol properties; solar irradiance and direct-beam transmission measurements; surface reflectivity measurements; and CO as a pollution tracer. The B-200 payload included a high-spectral resolution lidar (HSRL) and the Research Scanning Polarimeter (RSP), a precursor to the Aerosol Polarimetry Sensor (APS) to be launched on the NASA Glory satellite.
Intercomparisons of instruments were conducted aboard the DC-8 for redundant measurements (HO x , NO 2 , HNO 3 , oxygenated VOCs, SO 2 , sulfate). Wingtip-to-wingtip legs involving the DC-8, the P-3, the NOAA WP-3D, and the DLR Falcon were conducted to intercompare instruments across aircraft. Intercomparison results for H 2 O, O 3 , CO, HNO 3 , HO x and BC are presented by Kleb et al. (2009) and a complete report is available from the Tropospheric Airborne Measurements Evaluation Panel (TAbMEP) at http: //www-air.larc.nasa.gov/TAbMEP.html.
Frequent ozonesonde launches during ARCTAS were made at the 17 sites of the Arctic Intensive Ozonesonde Network Study (ARC-IONS) ( Table 3d). The spring sites were chosen to emphasize latitudinal gradients, opportunities to sample Arctic pollution, and coordination with aircraft and ground stations. The summer sites were aligned to capture cross-continental flow. Special launches were timed to coincide with TES satellite instrument overpasses for ozone validation (Boxe et al., 2010).
ARCTAS execution was also supported by ground-based measurements (Table 3e) and various forecasts and analyses (Table 3f). The ground-based measurements included aerosol lidars, sun-sky photometers, in situ measurements from an instrumented trailer, and multi-axis differential optical absorption spectroscopy (MAX-DOAS) measurements of BrO at Barrow. Meteorological and chemical transport model (CTM) forecasts, near-real-time satellite and CTM data, and fire reports were used to guide flights on a dayto-day basis.

Deployments
The spring ARCTAS-A deployment took place from 1 to 21 April 2008. Flight tracks are shown in Fig. 1a. The DC-8 and P-3 were based in Fairbanks (65 N, 148 W) and the B-200 in Barrow (71 N, 157 W). The DC-8 made overnight flights to Thule (77 N, 69 W) and Iqaluit, Nunavut (64 N, 69 W). Iqaluit was used as a contingency airport because of bad weather in Thule on the second planned overnight. The P-3 also made an overnight flight to Thule. Flights were coordinated with the Aerosol, Radiation, and Cloud Processes affecting Arctic Climate (ARCPAC) aircraft mission of the National Oceanic and Atmospheric Administration (NOAA) as well as with the DOE ISDAC aircraft mission, both concurrently based in Fairbanks. ARCPAC focused on the factors determining the radiative forcing by aerosols and tropospheric ozone in the Arctic, dovetailing with the objectives of ARCTAS. It used a WP-3D aircraft with in situ instrumentation for aerosols and clouds, ozone, related species, and radiation. ISDAC focused on aerosol-cloud interactions in the Arctic, including indirect forcing and scavenging processes. It used a Convair-580 aircraft with in situ instrumentation for aerosol and cloud microphysical properties. Flight tracks were mainly to and over the ARM CRF site at Barrow and involved coordination with the B-200. Actinic fluxes and photolytic frequencies spectral radiometry R. Shetter, NCAR Shetter and Miller (1999) 1 Chemical concentrations unless otherwise indicated. NO y = total reactive nitrogen oxides, including NO x and its oxidation products; PANs = peroxyacylnitrates; NMHCs = nonmethane hydrocarbons; OVOCs = oxygenated volatile organic compounds; CCN = cloud condensation nuclei; EC = elemental carbon; OC = organic carbon; LWC = liquid water content.     (Table 3d).  Fuelberg et al. (2010) present a detailed meteorological overview for the ARCTAS periods. There were no major departures from climatology. ARCTAS-A in spring sampled air with a wide range of origins. Cyclone tracks delivering Asian air to the Arctic were shifted northward from their climatological position. Strong mid-latitude cyclonic activity persisted in the summer, transporting Asian pollution and Siberian fire plumes across the Pacific to the west coast of North America. The North Atlantic Oscillation (NAO) was in a negative phase throughout the mission. Figure 2 shows fire activity during the spring and summer deployments as indicated by MODIS fire counts. April 2008 saw typical seasonal biomass burning in Southeast Asia but unusually early fire activity in eastern Siberia due to low snow cover, and unusually large agricultural burning in Kazakhstan. April 2008 was the second highest fire month in Russia (behind May 2003) over the 2001-2008 duration of the MODIS satellite record (Fisher et al., 2010). Flights from the NOAA ARCPAC mission out of Fairbanks also observed strong and persistent influence from these Russian fires (Warneke et al., 2009).
Summer 2008 saw in general low fire activity in Canada (1.6 Mha burned, as compared to an average of 2.8 Mha for the 1990s, Flannigan et al., 2005) but there was unusually large lightning-generated fire activity in northern Saskatchewan within 1-h flight time from Cold Lake, enabling sampling of fresh fire plumes and their evolution. Burning in northern Saskatchewan was four times the 20year average (Soja et al., 2008). Summer 2008 also saw considerable fire activity in northern California and Siberia, and plumes from these fires were repeatedly sampled by the aircraft. Sessions et al. (2010) simulated the transport of boreal   fire plumes during ARCTAS-B with the WRF-Chem model (Grell et al., 2005) and found that the plume rise parameterization in that model (Freitas et al., 2007) could successfully simulate the injection heights observed from CALIOP and the subsequent long-range transport of smoke and CO observed from MISR and AIRS.  April and remain anomalously high in July. Fires in California and Saskatchewan also stand out in July.

Flights
Tables 4 and 5 lists the research flights of the three aircraft during the spring and summer deployments. ARCTAS-CARB flights are included in Table 5. The tables describe some specific features of each flight including coordination with other platforms. Coordination between aircraft involved vertically stacked flights to provide complementarity between in situ and remote measurements, and wingtipto-wingtip legs at different altitudes to intercompare instruments. Vertical spirals to complement other measurements were also conducted over several surface sites including Barrow (ARM site, tropospheric BrO), Summit (surface and sondes), and Eureka (vertical column measurements of trace gases). Coordination with satellites involved underpasses and vertical spirals along the satellite orbit tracks. Top priorities were validation of the CALIOP aerosol observations by the B-200 and P-3 aircraft, and of the OMI tropospheric BrO observations by the DC-8 aircraft.
All flights of the spring ARCTAS-A deployment targeted mid-latitude pollution and fire plumes transported to the Arctic. The P-3 and B-200 focused on aerosol layers. The P-3 also devoted flight time to measuring surface albedo and reflectance, including for sea ice at Elson Lagoon just north of Barrow and over the Arctic ice and clouds near Greenland. The DC-8 conducted extensive vertical profiling from its nominal floor (0.15 km over ocean, 0.3 km over land) to the stratosphere. To ensure that the boundary layer was properly sampled, the DC-8 occasionally descended to 0.1 km over land and made deliberate missed approaches of the Barrow and Prudhoe Bay airports.
All flights of the summer ARCTAS-B deployment targeted wildfire plumes. Again, the P-3 and B-200 focused on aerosol layers. The DC-8 focused on plume evolution, including sampling of fresh Canadian fire plumes as well as aged Asian and California fire plumes. A few flights sampled plumes from oil sands recovery operations near Fort McMurray in northeast Alberta and found high concentrations of heavy alkanes, aromatics, and SO 2 (Simpson et al., 2010). On the outbound flight to Cold Lake, the DC-8 conducted an intercomparison with the NSF HIAPER aircraft as it completed its pre-HIPPO (HIAPER Pole-to-Pole Observations) mission out of Colorado. Pre-HIPPO focused on testing the HIAPER aircraft payload for global observations of greenhouse gases and BC aerosol. It included several flights to the Arctic in the May-June time frame, bridging the two ARC-TAS deployments.
The ARCTAS-CARB flights focused on satellite validation, sampling of emissions, and providing upwind boundary conditions and atmospheric observations to improve air quality models. DC-8 satellite validation focused on TES (ozone, CO), OMI (NO 2 ), and MOPITT (CO), while P-3 satellite validation focused on MISR and MODIS (aerosols). Plumes from large wildfires in northern California were repeatedly sampled. These plumes were transported from northern to southern California, allowing opportunities to study their interactions with urban pollution. Extensive low-level sampling was made of emissions from ships, population centers, industrial centers, agricultural crops, feedlots, and dairy farms. Flights also investigated pollution import from Asia and Mexico.

Synthesis of first results
Interpretation of the ARCTAS data is still in its early stages but promises to successfully address all the mission objectives. Results are already providing important new information on pollution transport and chemistry in the Arctic, its radiative implications, and the utility of satellites for monitoring Arctic atmospheric composition. We summarize here a few highlights.
The ARCTAS observations provided a first validation of satellite observations of atmospheric composition and radiation in the Arctic. Boxe et al. (2010) found that TES V004 retrievals of tropospheric ozone north of 60 • N during ARC-TAS agreed with ARC-IONS ozonesonde profiles within the estimated TES error (15%), and they reported significant error reduction relative to previous versions of the TES ozone product. Fisher et al. (2010) showed that AIRS daily observations of CO could successfully track the springtime transport of polluted air masses from mid-latitudes to the Arctic in the free troposphere. Lyapustin et al. (2010) validated the airborne measurements of the bidirectional reflectance distribution function (BRDF) of snow against ground observations, and used them to construct an improved snow BRDF model for satellite retrievals. Extensive data on aerosol optical properties and correlation length scales for biomass burning plumes were obtained to assist satellite AOD retrievals (Shinozuka et al., 2010). Gatebe et al. (2010)  by the P-3 and AERONET near Barrow, Alaska to demonstrate a new method for simultaneously retrieving aerosol and surface properties from combined airborne and groundbased direct and diffuse radiometric measurements.

used measurements
Total column BrO was measured by both OMI and GOME-2 during ARCTAS. The two satellite instruments showed similar patterns and magnitudes of BrO hotspots over the Arctic in spring (enhancements of 2 to 3×10 13 cm −2 relative to the zonal mean). Validation by the aircraft showed that previous attribution of these hotspots to the boundary layer is incorrect. The aircraft revealed both false positives, where hotspots seen from space were not associated with measurable boundary layer BrO, and false negatives, where aircraft observations of elevated boundary layer BrO (and associated ODEs and MDEs) were not associated with elevated column BrO. False negatives could be due to cloud cover shielding boundary layer BrO from detection by satellite. As to the false positives, Salawitch et al. (2010) showed that numerous BrO hotspots observed from space correspond to tropopause depressions, and hypothesized that the elevated BrO actually resides in the lower stratosphere. Free tropospheric BrO enhancements above the boundary layer large  Cold Lake local 1 29 June Siberian fire plume, pyroconvective outflow, Saskatchewan fires 18 Cold Lake local 2 1 July Detailed characterization of Saskatchewan fire plumes over aging times ranging from minutes to hours 19 Cold Lake local 3 4 July Near-field evolution of Saskatchewan fire plumes, California and Siberian fire plumes, deep convective injection of smoke and biogenic VOCs, TES validation 20 Cold Lake local 4 5 July Saskatchewan fire plumes, biogenic emissions 21 Cold Lake-Thule 8 July Siberian fires and Asian pollution 22 Thule local 9 July MOPITT validation over Summit, intercomparison with DLR Falcon, Siberian fires and Asian pollution 23 Thule-Cold Lake 10 July P-3 intercomparison, oil sands plumes 24 Cold Lake-Los Angeles 14 July California fire plumes 1 Flights 12-15 were part of the ARCTAS-CARB deployment. Flight durations were typically 8 h.
were also observed during ARCTAS-A and ARCPAC (Neuman et al., 2010). Closure of the budget for total column BrO observed from space could be achieved by assuming that short-lived biogenic bromocarbons supply 5 to 10 ppt of inorganic bromine (Br y ) to the stratosphere (Salawitch et al., 2010). This supply of Br y is higher than simulated in current models and may have important consequences for the photochemical budget of ozone in the upper troposphere and lower stratosphere.
The April ARCTAS flights did find extensive ODEs and MDEs in the Arctic boundary layer associated with high concentrations of halogen radicals (Koo et al., 2009;   30 June CALIPSO validation and coordination with P-3 to study Saskatchewan fire plumes Yellowknife local 4 2 July Two flights; coordination with P-3 over fresh and aged fire plumes; CALIPSO validation on second flight Yellowknife local 5 3 July CALIPSO validation and P-3 coordination along CALIPSO track Yellowknife local 6 6 July Two flights; second flight was coordination with P-3 for fires and plumes Yellowknife-Cold Lake-Edmonton 7 July Two flights: Asian/Siberian plume over Alberta Edmonton local and Edmonton-Yellowknife 8 July Two flights: Asian/Siberian plume over western Alberta/eastern British Columbia; CALIPSO validation over northern Alberta Yellowknife local 7 9 July CALIPSO validation, coordination with P-3 along CALIPSO track and in fresh fire plumes Yellowknife local 8 10 July CALIPSO validation, coordination with P-3 and DC-8 near Yellowknife al., 2010). These events were mainly confined below 1 km and did not extend to the free troposphere. Major ODE regions were identified in the Canadian archipelago and north of Alaska and Siberia (Koo et al., 2009). Mercury depletion in these events was consistent with oxidation of elemental Hg by Br atoms (Kim et al., 2010). The ARCTAS observations in April revealed a pervasive background of Asian pollution through the depth of the Arctic tropospheric column, apparently reflecting large-scale accumulation during the dark winter months. Asian pollution influence was strongest in the middle troposphere (2-6 km). Pollution plumes from all three mid-latitude continents as well as from Siberian fires were superimposed on this elevated background (Table 4 and Fisher et al., 2010). The plumes were relatively dilute: the maximum CO concentration observed from the DC-8 during ARCTAS-A north of 55 • N was 296 ppbv. From a model analysis of the aircraft CO data, Fisher et al. (2010) found a general dominance of Asian pollution over the Arctic in spring, except below 2 km where European pollution influence was of comparable magnitude. North American pollution influence was relatively small. Siberian fire plumes were often transported to the Arctic together with Asian pollution in WCBs but made relatively little contribution to the budget of CO because of the dominance of accumulated pollution during the winter months. Fisher et al. (2010) found from the multiyear AIRS record that CO in the North American Arctic was lower than average in spring 2008, despite the large Siberian fires, due to a weak Aleutian Low hindering WCB transport from Asia.
The Siberian fires in spring 2008 had a large effect on carbonaceous aerosols in the Arctic. They were the main source of organic aerosol during ARCTAS-A (Warneke et al., 2009;) and affected the entire Arctic (Saha et al., 2010). Fuel combustion and Siberian fires made contributions of comparable magnitude to the BC aerosol concentrations over the Arctic during ARCTAS-A . Combined measurements of multi-wavelength aerosol extinction, aerosol mass spectrometry, and BC incandescence on the P-3 in ARCTAS-A indicate that aerosols below 2 km altitude were mainly anthropogenic, whereas those at 2-4 km altitude were mainly from biomass burning (Shinozuka et al., 2010).
Aircraft measurements of total NO y showed good closure with measurements of individual NO y components, with PAN the dominant contributor to NO y in both spring and summer (Singh et al., 2010). Independent measurements of OH by LIF and CIMS aboard the DC-8 aircraft showed a median ratio [OH] LIF /[OH] CIMS =0.73, within the stated accuracy of the instruments, and a correlation coefficient r =0.51 (Chen et al., manuscript in preparation, 2010).  found that the measured HO x and H 2 O 2 concentrations in spring were significantly lower than predicted from a standard gas-phase mechanism, and hypothesized that this could reflect HO 2 radical uptake by aerosol leading to non-radical products other than H 2 O 2 . This heterogeneous chemistry would be particularly important in Arctic spring because of the cold temperatures, relatively high aerosol concentrations, and weak radiation.
Boreal and California forest fire plumes sampled during the summer phase showed similar emission factors and plume optical properties (Singh et al., 2010). NO x emission ratios were much lower than literature values for boreal forest fires (Nance et al., 1993;Goode et al., 2000), and rapid conversion of NO x to PAN in the fresh plumes further suppressed ozone production (Alvarado et al., 2010;Hecobian et al., 2010). No significant ozone or secondary aerosol production was observed in the fire plumes sampled in summer (Alvarado et al., 2010;Hecobian et al., 2010;Singh et al., 2010), although California fire plumes mixed with urban pollution produced some very high ozone levels (Singh et al., 2010). By contrast, significant ozone enhancements were observed in the aged Siberian fire plumes sampled in spring (Singh et al., 2010). Oltmans et al. (2010) found that tropospheric ozone levels were generally elevated over the west coast of North America in spring 2008, with Denali National Park in Alaska recording an 8-h average maximum value of 75 ppbv. They attributed this ozone enhancement to Siberian fire influence.
Methane observations during ARCTAS-A showed little variability and no indication of significant April emissions from Arctic ecosystems. The July observations in ARCTAS-B showed large emissions from the Hudson Bay Lowlands (HBL), one of the largest boreal wetland ecosystems in the world. Pickett-Heaps et al. (2009) combined the aircraft vertical profiles over the HBL from ARCTAS-B and pre-HIPPO, together with surface data, to infer a methane emission of 2.7 Tg a −1 from the HBL, much higher than previously estimated values of 0.2-0.5 Tg a −1 (Roulet et al., 1994;Worthy et al., 2000). Singh et al. (2010) found from analysis of ARCTAS-CARB data that methane emissions in California are substantially underestimated in current inventories.

Conclusions
The ARCTAS mission offers a rich data set for better understanding of Arctic pollution including long-range transport, boreal forest fires influences, aerosol radiative forcing, and chemical processes. It should also be of value as a baseline against which to measure future change. All ARC-TAS data are publicly available from http://www-air.larc. nasa.gov/missions/arctas/arctas.html.
Although the ARCTAS deployments covered only two 3week periods in April and June-July 2008, the conditions encountered were highly favorable for addressing mission objectives. The aircraft observed springtime pollution transport events from Asia, North America, and Europe, springtime fire plumes from Siberia, and summertime fire plumes from Canada, California, and Siberia. They encountered a number of boundary layer ozone depletion events in the spring.
Unusually large fire activity in northern Saskatchewan during the summer deployment allowed detailed study of fire emissions and the near-field evolution of fire plumes.
A central objective of ARCTAS was to increase the value of satellite observations of Arctic atmospheric composition. Validation from the aircraft showed that thermal IR satellite measurements of tropospheric ozone from TES and CO from AIRS are of sufficient quality to track pollution plumes and construct regional budgets. Satellite observations of BrO column hotspots in spring were shown not to be due to Arctic boundary layer events but instead to tropopause depressions, implying higher inorganic bromine levels in the lower stratosphere (5-10 pptv) than previously thought.
The ARCTAS-A observations showed pervasive Asian pollution influence throughout the Arctic troposphere in spring. There was also significant European pollution influence below 2 km. North American pollution influence was relatively small. Siberian fires were unusually active during April 2008 and greatly enhanced organic aerosol concentrations over the scale of the Arctic. BC aerosol had significant contributions from both fuel combustion and Siberian fires. The Siberian fires also appeared to be responsible for elevated tropospheric ozone over the west coast of North America. HO x measurements from the aircraft by two different methods showed good consistency and lower values in spring than predicted from a gas-phase photochemical model. A possible explanation is uptake of HO 2 by aerosols, which would be particularly favored under Arctic springtime conditions.
The ARCTAS-B and ARCTAS-CARB observations showed no significant ozone or secondary aerosol production in plumes from California and boreal fires. However, fast ozone production was observed in California when the fire plumes mixed with urban pollution. Flights over the Hudson Bay Lowlands revealed higher wetland emissions of methane than previously recognized, and flights over California indicated an underestimate of methane emissions in state inventories.