Regional aerosol optical depth characteristics from satellite observations: ACE‐1, TARFOX and ACE‐2 results

Analysis of the aerosol properties during 3 recent international field campaigns (ACE‐1, TARFOX and ACE‐2) are described using satellite retrievals from NOAA AVHRR data. Validation of the satellite retrieval procedure is performed with airborne, shipboard, and land‐based sunphotometry during ACE‐2. The intercomparison between satellite and surface optical depths has a correlation coefficient of 0.93 for 630 nm wavelength and 0.92 for 860 nm wavelength. The standard error of estimate is 0.025 for 630 nm wavelength and 0.023 for 860 nm wavelength. Regional aerosol properties are examined in composite analysis of aerosol optical properties from the ACE‐1, TARFOX and ACE‐2 regions. ACE‐1 and ACE‐2 regions have strong modes in the distribution of optical depth around 0.1, but the ACE‐2 tails toward higher values yielding an average of 0.16 consistent with pollution and dust aerosol intrusions. The TARFOX region has a noticeable mode of 0.2, but has significant spread of aerosol optical depth values consistent with the varied continental aerosol constituents off the eastern North American Coast.     

ACE‐2 multiple angle micro‐pulse lidar observations from Las Galletas, Tenerife, Canary Islands

Multiple‐angle micro‐pulse lidar (MPL) observations were made at Las Galletas on Tenerife, Canary Islands during the Aerosol Characterization Experiment‐2 (ACE‐2) conducted June–July, 1997. A principal objective of the MPL observations was to characterize the temporal/spatial distributions of aerosols in the region, particularly to identify and profile elevated Saharan dust layers which occur intermittently during the June–July time period. Vertical and slant angle measurements taken 16 and 17 July characterize such an occurrence, providing aerosol backscatter, extinction, and optical depth profiles of the dust layer between 1 and 5 km above mean sea level (MSL). Additionally, horizontal measurements taken in Las Galletas throughout the 6‐week period provide a time profile of the varying aerosol extinction at the surface. This profile exhibits the alternating periods of clean maritime air and pollution outbreaks that typified the region. Horizontal measurements also provide some evidence suggesting the possible influx of Saharan dust from the free troposphere to the surface. This paper presents estimates of aerosol optical properties retrieved from the multi‐angle MPL measurements in addition to an outline of the methodologies employed to obtain these results.     

Ground‐based lidar measurements of aerosols during ACE‐2: instrument description, results, and comparisons with other ground‐based and airborne measurements

A micro‐pulse lidar system (MPL) was used to measure the vertical and horizontal distribution of aerosols during the Aerosol Characterization Experiment 2 (ACE‐2) in June and July of 1997. The MPL measurements were made at the Izaña observatory (IZO), a weather station located on a mountain ridge (28°18' N, 16°30' W, 2367 m asl) near the center of the island of Tenerife, Canary Islands. The MPL was used to acquire aerosol backscatter, extinction, and optical depth profiles for normal background periods and periods influenced by Saharan dust from North Africa. System tests and calibration procedures are discussed, and an analysis of aerosol optical profiles acquired during ACE‐2 is presented. MPL data taken during normal IZO conditions (no dust) showed that upslope aerosols appeared during the day and dissipated at night and that the layers were mostly confined to altitudes a few hundred meters above IZO. MPL data taken during a Saharan dust episode on 17 July showed that peak aerosol extinction values were an order of magnitude greater than molecular scattering over IZO, and that the dust layers extended to 5 km asl. The value of the dust backscatter–extinction ratio was determined to be 0.027±0.007 sr⁻¹. Comparisons of the MPL data with data from other co‐located instruments showed good agreement during the dust episode.     

Measurements of aerosol optical depth above 3570 m asl in the North Atlantic free troposphere: results from ACE‐2

As part of the 2nd A erosol C haracterisation E xperiment (ACE‐2), conducted during summer 1997 in the North Atlantic region between the Canary Islands and Portugal, we measured aerosol optical depths (AOD) at a mid‐tropospheric site, near the top of the volcanic mountain "El Teide"(28°16'N, 16°36' W, 3570 m asl). Our instrument was located at the highest altitude in a network of sunphotometers that extended down to sea level. Clear conditions dominated the ACE‐2 period, and, although suggested by back‐trajectories at 300 hPa, no evidence of anthropogenic pollution was found in our data. Three distinct dust episodes were observed. Vertical soundings and back trajectories suggested mineral dust from the Sahel region as a source. During these episodes, AOD increased an order of magnitude with respect to background conditions (from 0.017 up to 0.19 at λ=500 nm). A shift towards neutrality of the extinction spectral dependence (Ångstrom exponent α down to 0.13), indicated that the coarse mode (particle diameter >2 μm) dominated the aerosol size distribution. For 6 days during the episodes of mineral dust, a monomodal size distribution between 2 and 20 μm diameter was obtained from Mie based size distribution calculations. Estimates, at 500 nm, of the single scattering albedo ω₀(0.87–0.96), and the aerosol asymmetry parameter g (0.72–0.73) suggest that the dust layer causes a net cooling forcing at the top of the atmosphere.     

Lagrangian air mass tracking with smart balloons during ACE‐2

A 3rd‐generation smart balloon designed at National Oceanic and Atmospheric Administration, Air Resources Laboratory Field Research Division, in collaboration with the University of Hawaii, was released from ship‐board during the recent Second Aerosol Characterization Experiment (ACE‐2) to provide Lagrangian air‐mass tracking data. ACE‐2 is the 3rd in a series of field experiments designed to study the chemical, physical, and radiative properties and processes of atmospheric aerosols and their role in climate and is organized by the international global atmospheric chemistry (IGAC) program. The adjective smart in the title of this paper refers to the fact that the buoyancy of the balloon automatically adjusts through the act of pumping air into or releasing air from the ballast portion of the balloon when it travels vertically outside a barometric pressure range set prior to release. The smart balloon design provides GPS location, barometric pressure, temperature, relative humidity, and other data via a transponder to a C130 research aircraft flying in the vicinity of the balloon. The addition of 2‐way communication allows interactive control of the balloon operating parameters by an observer. A total of 3 cloudy Lagrangian experiments were conducted during the ACE‐2 field program which lasted from 16 June through 26 July 1997. This paper reviews the design and capability of the smart balloons and their performance during the ACE‐2 Lagrangian experiments. Future development and applications of the technology are discussed.     

Clear‐sky closure studies of lower tropospheric aerosol and water vapor during ACE‐2 using airborne sunphotometer, airborne in‐situ, space‐borne, and ground‐based measurements

We report on clear‐sky column closure experiments (CLEARCOLUMN) performed in the Canary Islands during the second Aerosol Characterization Experiment (ACE‐2) in June/July 1997. We present CLEARCOLUMN results obtained by combining airborne sunphotometer and in‐situ (optical particle counter, nephelometer, and absorption photometer) measurements taken aboard the Pelican aircraft, space‐borne NOAA/AVHRR data and ground‐based lidar and sunphotometer measurements. During both days discussed here, vertical profiles flown in cloud‐free air masses revealed 3 distinctly different layers: a marine boundary layer (MBL) with varying pollution levels, an elevated dust layer, and a very clean layer between the MBL and the dust layer. A key result of this study is the achievement of closure between extinction or layer aerosol optical depth (AOD) computed from continuous in‐situ aerosol size‐distributions and composition and those measured with the airborne sunphotometer. In the dust, the agreement in layer AOD (λ=380–1060 nm) is 3–8%. In the MBL there is a tendency for the in‐situ results to be slightly lower than the sunphotometer measurements (10–17% at λ=525 nm), but these differences are within the combined error bars of the measurements and computations.     

Boundary layer and aerosol evolution during the 3rd Lagrangian experiment of ACE‐2

Aircraft measurements are presented of the Lagrangian evolution of a marine boundary layer over a 30‐h period during the ACE‐2 field campaign. At the start of the observational period, a 500‐m deep polluted marine internal boundary layer (MIBL) was overlain by the remnants of a polluted continental boundary layer extending to around 2 km below a clean, dry free troposphere. The MIBL grew rapidly to a thickness of 900–1000 m in response to increasing sea surface temperatures. No significant aerosol spectral evolution was observed in the boundary layer. Low concentrations of SO₂ were observed in the MIBL suggesting that the air mass contained relatively aged aerosol. Aerosol spectra show a broad mode with a modal diameter of around 0.1μm. The polluted layer between the MIBL and the unpolluted free troposphere was only weakly and intermittently turbulent which prevented significant entrainment of clean air into the polluted layer from aloft. The polluted layer depth was thus controlled mainly by subsidence which as a result becomes shallower, decreasing from over 2000 m to around 1200 m during the observational period. The aerosol characteristics of the polluted layer were similar to those in the MIBL and so although the MIBL entrained considerable amounts of air from above the MIBL the aerosol characteristics underwent no significant change. This has important implications for the rate at which a polluted continental air mass is converted to a clean marine one. The dataset should prove useful in the validation of the modelling of continental pollution outbreaks.     

Cloud optical thickness and albedo retrievals from bidirectional reflectance measurements of POLDER instruments during ACE‐2

The POLDER instrument is devoted to global observations of the solar radiation reflected by the Earth–atmosphere system. The airborne version of the instrument was operated during the ACE‐2 experiment, more particularly as a component of the CLOUDYCOLUMN project of ACE‐2 that was conducted in summer 1997 over the subtropical northeastern Atlantic ocean. CLOUDYCOLUMN is a coordinated project specifically dedicated to the study of the indirect effect of aerosols. In this context, the airborne POLDER was assigned to remote measurements of the cloud optical and radiative properties, namely the cloud optical thickness and the cloud albedo. This paper presents the retrievals of those 2 cloud parameters for 2 golden days of the campaign 26 June and 9 July 1997. Coincident spaceborne ADEOS‐POLDER data from 2 orbits over the ACE‐2 area on 26 June are also analyzed. 26 June corresponds to a pure air marine case and 9 July is a polluted air case. The multidirectional viewing capability of airborne POLDER is here demonstrated to be very useful to estimate the effective radius of cloud droplet that characterizes the observed stratocumulus clouds. A 12 μm cloud droplet size distribution appears to be a suitable cloud droplet model in the pure marine cloud case study. For the polluted case the mean retrieved effective droplet radius is of the order of 6–10 μm. This only preliminary result can be interpreted as a confirmation of the indirect effect of aerosols. It is consistent with the significant increase in droplet concentration measured in polluted marine clouds compared to clean marine ones. Further investigations and comparisons to in‐situ microphysical measurements are now needed.     

Cloud condensation nuclei and cloud droplet measurements during ACE‐2

During the 2nd Aerosol Characterization Experiment (ACE‐2), relationships between stratocumulus cloud properties and aerosols were examined. Here, the relevant measurements including the cloud condensation nuclei (CCN) activation spectrum, updraft velocity, cloud microphysical and aerosol properties are presented. It is shown that calculations of droplet concentration based on updraft velocity and the CCN activation spectrum are consistent with direct observations. Also discussed is an apparent disparity among measurements of the CCN activation spectrum, the accumulation mode size distribution, and the composition of the submicrometric aerosol. The observed consistency between CCN, updraft and cloud droplets is a necessary refinement; however, extended analyses of the ACE‐2 data set are needed to guide improvements in model simulations of the interaction between aerosols and cloud microphysics. In particular, there is need for an examination of aerosol size spectra and chemical composition measurements with a view towards validating droplet activation schemes which relate the aerosol and cloud dynamical properties to cloud albedo.     

An overview of the ACE‐2 CLOUDYCOLUMN closure experiment

CLOUDYCOLUMN is one of the 6 ACE‐2 projects which took place in June‐July 1997, between Portugal and the Canary Islands. It was specifically dedicated to the study of changes of cloud radiative properties resulting from changes in the properties of those aerosols which act as cloud condensation nuclei. This process is also refered to as the aerosol indirect effect on climate. CLOUDYCOLUMN is focused on the contribution of stratocumulus clouds to that process. In addition to the basic aerosol measurements performed at the ground stations of the ACE‐2 project, 5 instrumented aircraft carried out in situ characterization of aerosol physical, chemical and nucleation properties and cloud dynamical and microphysical properties. Cloud radiative properties were also measured remotely with radiometers and a lidar. 11 case studies have been documented, from pure marine to significantly polluted air masses. The simultaneity of the measurements with the multi‐aircraft approach provides a unique data set for closure experiments on the aerosol indirect effect. In particular CLOUDYCOLUMN provided the 1st experimental evidence of the existence of the indirect effect in boundary layer clouds forming in polluted continental outbreacks. This paper describes the objectives of the project, the instrumental setup and the sampling strategy. Preliminary results published in additional papers are briefly summarized.     

Polarising properties of the aerosols in the north‐eastern tropical Atlantic Ocean, with emphasis on the ACE‐2 period

Measurements of the polarisation state of the atmosphere were performed at Tenerife in June–July 1997, in the framework of ACE‐2 (second Aerosol Characterization Experiment), by 2 ground‐based instruments: RefPol (a LOA prototype) which took measurements at 445, 665, 870, 1610 nm in the solar principal plane; and an automatic CIMEL (CE 318) sun/sky‐photometer which measured polarised radiation at 870 nm in the same observational geometry. Measurements acquired during the campaign, as well as AERONET (AErosol RObotic NETwork) measurements acquired at the sites of Cape Verde and M'Bour, are processed with an algorithm determining the polarised single‐scattering sky‐radiance due to aerosols, directly proportional to the aerosol polarised phase function (representing the probability to scatter polarised radiation in the direction of the scattering angle). A good correlation between the Ångström exponent α, representing the spectral dependence of the extinction measurements, and the polarised phase function is observed on each set of data. The uncertainty of retrievals at 445 nm makes the determination of the spectral dependence of polarisation inconclusive but does not prevent confirming the dependence of the aerosol polarised phase function on α, at all wavelengths. An Ångström exponent of 1 corresponds to a polarised phase function of around 0.1 (±0.04), at 870 nm and at a scattering angle of 60°. For α between 0 and 0.4, the average value of the polarised phase function is 0.05. The correlation shows that polarisation is more sensitive to small particles than to large particles. The discrepancy between retrievals and Mie calculations from an AERONET size distribution, inverted from Izaña measurements acquired during a dust event, suggests the presence of small particles, not detected by total sky‐radiance measurements.     

Evolution of boundary‐layer aerosol particles due to in‐cloud chemical reactions during the 2nd Lagrangian experiment of ACE‐2

The second Aerosol Characterisation Experiment (ACE‐2) was aimed at investigating the physical, chemical and radiative properties of aerosol and their evolution in the North Atlantic region. In the 2nd "Lagrangian" experiment, an air mass was tracked over a 30‐h period during conditions of extensive stratocumulus cover. Boundary‐layer measurements of the aerosol size distribution obtained with a passive cavity aerosol spectrometer probe (PCASP) during the experiment show a gradual growth in size of particles in the 0.1–0.2 μm diameter mode. Simultaneously, SO₂ concentrations were found to decrease sharply from 800 to 20 ppt. The fraction of sulphate in aerosol ionic mass increased from 0.68±0.07 to 0.82±0.09 for small particles (diameter below 1.7 μm) and from 0.21±0.04 to 0.34±0.03 for large particles (diameter above 1.7 μm). The measurements were compared with a multicyclic parcel model of gas phase diffusion into cloud droplets and aqueous phase chemical reactions. The model was able to broadly reproduce the observed transformation in the aerosol spectra and the timescale for the transformation of SO₂ to sulphate aerosol. The modelled SO₂ concentration in the boundary layer fell to below half its initial value over a 6.5‐h time period due to a combination of the entrainment of cleaner tropospheric air and cloud chemical reactions. NH₃ and HCl gas were also found to play an important rôle in cloud processing in the model.     

Microphysical properties of stratocumulus clouds during ACE‐2

Microphysical measurements performed during 8 flights of the CLOUDYCOLUMN component of ACE‐2, with the Meteo‐France Merlin‐IV, are analyzed in terms of droplet number concentration and size. The droplet concentration is dependent upon the aerosol properties within the boundary layer. Its mean value over a flight varies from 55 cm⁻³, for the cleanest conditions, to 244 cm⁻³, for the most polluted one. For each flight, the variability of the concentration, in selected cloud regions that are not affected by mixing with dry air or drizzle scavenging, ranges from 0.5 to 1.5 of the mean value. The mean volume diameter increases with altitude above cloud base according to the adiabatic cloud model. The frequency distribution of mean droplet volume normalized by the adiabatic value, for the selected regions, shows the same dispersion as the distribution of normalized concentration. The values of droplet concentration versus mean volume diameter are then examined in sub‐adiabatic samples to characterize the effects of mixing and drizzle scavenging. Finally, the ratio of mean volume diameter to effective diameter is analyzed and a simple relationship between these 2 crucial parameters is proposed.     

Shipboard measurements of concentrations and properties of carbonaceous aerosols during ACE‐2

Mass concentrations of total, organic and black carbon were derived by analyzing the supermicron and submicron aerosol fractions of shipboard collected samples in the easternAtlantic Ocean as part of the second Aerosol Characterization Experiment (ACE‐2). These analyses were complemented by experiments intended to estimate the water‐soluble fraction of the submicron carbonaceous material. Our results can be summarized as follows. Depending on the sample, between 35% and 80% of total aerosol carbon is associated with the submicron fraction. Total submicron carbon was well correlated with black carbon, a unique tracer for incomplete combustion. These correlations and the approximately constant total to black carbon ratios, suggest that the majority of submicron total carbon is of primary combustion derived origin. No systematic relationship between total submicron aerosol carbon and sulfate concentrations was found. Sulfate concentrations were, with a few exceptions, significantly higher than total carbon. Our experiments have demonstrated that water exposure removed between 36% and 72% of total carbon from the front filter, suggesting that a substantial fraction of the total submicron aerosol organic carbon is water‐soluble. An unexpected result of this study is that water exposure of filter samples caused substantial removal of, nominally insoluble, submicron black carbon. Possible reasons for this observation are discussed.     

Size‐segregated chemical, gravimetric and number distribution‐derived mass closure of the aerosol in Sagres, Portugal during ACE‐2

During the ACE‐2 field campaign in the summer of 1997 an intensive, ground‐based physical and chemical characterisation of the clean marine and continentally polluted aerosol was performed at Sagres, Portugal. Number size distributions of the dry aerosol in the size range 3–10 000 nm were continuously measured using DMPS and APS systems. Impactor samples were regularly taken at 60% relative humidity (RH) to obtain mass size distributions by weighing the impactor foils, and to derive a chemical mass balance by ion and carbon analysis. Hygroscopic growth factors of the metastable aerosol at 60% RH were determined to estimate the number size distribution at a relative humidity of 60%. A size segregated 3‐way mass closure study was performed in this investigation for the first time. Mass size distributions at 60% RH derived from number size distribution measurements and impactors samples (weighing and chemical analysis) are compared. A good agreement was found for the comparison of total gravimetrically‐determined mass with both number distribution‐derived (slope=1.23/1.09; R²>0.97; depending on the parameters humidity growth and density) and chemical mass concentration (slope=1.02; R²= 0.79) for particles smaller than 3 μm in diameter. Except for the smallest impactor size range relatively good correlations (slope=0.86–1.42) with small deviations (R²=0.76–0.98) for the different size fractions were found. Since uncertainties in each of the 3 methods are about 20% the observed differences in the size‐segregated mass fractions can be explained by the measurement uncertainties. However, the number distribution‐derived mass is mostly higher than the chemically and gravimetrically determined mass, which can be explained by sampling losses of the impactor, but as well with measurement uncertainties as, e.g., the sizing of the DMPS/APS.     

北京市持续重污染天气分析

通过对2004-2008年北京市空气持续重污染过程的统计及其对非沙尘型持续重污染天气形势的特征分析,得到以下结论：持续重污染过程具有明显的季节分布特征,主要包括春季沙尘型污染和秋冬季节非沙尘型污染;非沙尘型持续重污染过程期间多对应着大雾、轻雾、霾、烟等低能见度天气,过程后期对应的天气现象多为大风或降水天气。其中大雾天气更易引发长时间持续的空气重污染事件。非沙尘型持续重污染的天气形势特点为：高空多为纬向环流,850 hPa多为暖脊控制,地面多处于弱气压场,鞍形场型污染尤为严重。北京持续重污染多对应区域性污染。     

Characteristics of Elemental Composition of PM2.5 in the Spring Period at Tongyu in the Semi-arid Region of Northeast China

Continuous observations of mass concentration and elemental composition of aerosol particles （PM2.5） were conducted at Tongyu, a semi-arid site in Northeast China in the spring of 2006. The average mass concentration of PM2.5 at Tongyu station was 260.9±274.4 μg m^-3 during the observation period. Nine dust events were monitored with a mean concentration of 528.0±302.7 μgm^-3. The PM2.5 level during non- dust storm （NDS） period was 111.65±63.37 μg m^-3. High mass concentration shows that fine-size particles pollution was very serious in the semi-arid area in Northeast China. The enrichment factor values for crust elements during the dust storm （DS） period are close to those in the NDS period, while the enrichment factor values for pollution elements during the NDS period are much higher than those in the DS period, showing these elements were from anthropogenic sources. The ratios of dust elements to Fe were relative constant during the DS period. The Ca/Fe ratio in dust aerosols at Tongyu is remarkably different from that observed in other source regions and downwind regions. Meteorological analysis shows that dust events at Tongyu are usually associated with dry, low pressure and high wind speed weather conditions. Air mass back-trajectory analysis identified three kinds of general pathways were associated with the aerosol particle transport to Tongyu, and the northwest direction pathway was the main transport route.     

Chemical effect of carbonaceous aerosols on the diurnal cycle of MBL ozone at a tropical site: measurements and simulations

During late austral summer and winter 1998, black carbon (BC) aerosols were monitored with an Aethalometer at 2 sites of La Réunion Island (Indian Ocean): Saint‐Denis, the main city and Sainte‐Rose, a quite uninhabited region situated at the east coast. BC concentration data at Saint‐Denis show a marked diurnal cycle, which may be primarily attributed to traffic. The background data found at night‐time display average BC concentrations, ranging from about 80 to 250 ng/m³ whereas during the day, BC concentrations increase by a factor of at least 4. In comparison, BC concentrations vary in the range of 10 to 60 ng/m³ at Sainte‐Rose. Ozone concentration was also measured at Saint‐Denis using a Dasibi photometer and found to be at significant levels (means: 16.5–23 ppbv in April and 28.5–34 ppbv in September). A noticeable increase of ozone concentrations during the day points out the build‐up of pollutants enhancing photochemical transformations. However, during traffic pollution peaks, ozone concentration displays systematic depletion. The comparison of ozone and BC measurements at both seasons points to some possible effects of heterogeneous interaction of ozone and its precursors with BC particles. These interactions were also simulated with a 0D time‐dependent chemistry model using conditions of a polluted site. The measured ozone concentration characteristics (mean concentration and range of variation) are well simulated in the presence of BC. Our model results show that at La Réunion Island adsorption of ozone and its precursors onto BC aerosol particles could be one of the important steps determining ozone concentration characteristics, especially in absence of photochemistry during night‐time.     

Observations of the evolution of the aerosol, cloud and boundary‐layer characteristics during the 1st ACE‐2 Lagrangian experiment

During the 1st Lagrangian experiment of the North Atlantic Regional Aerosol Characterisation Experiment (ACE‐2), a parcel of air was tagged by releasing a smart, constant level balloon into it from the Research Vessel Vodyanitskiy . The Meteorological Research Flight's C‐130 aircraft then followed this parcel over a period of 30 h characterising the marine boundary layer (MBL), the cloud and the physical and chemical aerosol evolution. The air mass had originated over the northern North Atlantic and thus was clean and had low aerosol concentrations. At the beginning of the experiment the MBL was over 1500 m deep and made up of a surface mixed layer (SML) underlying a layer containing cloud beneath a subsidence inversion. Subsidence in the free troposphere caused the depth of the MBL to almost halve during the experiment and, after 26 h, the MBL became well mixed throughout its whole depth. Salt particle mass in the MBL increased as the surface wind speed increased from 8 m s⁻¹ to 16 m s⁻¹ and the accumulation mode (0.1μm to 3.0 μm) aerosol concentrations quadrupled from 50 cm⁻³ to 200 cm⁻³. However, at the same time the total condensation nuclei (>3 nm) decreased from over 1000 cm⁻³ to 750 cm⁻³. The changes in the accumulation mode aerosol concentrations had a significant effect on the observed cloud microphysics. Observational evidence suggests that the important processes in controlling the Aitken mode concentration which, dominated the total CN concentration, included, scavenging of interstitial aerosol by cloud droplets, enhanced coagulation of Aitken mode aerosol and accumulation mode aerosol due to the increased sea salt aerosol surface area, and dilution of the MBL by free tropospheric air.     

Anthropogenic and biogenic winter sources of Arctic CO2—a model study

Long‐range transport of anthropogenic and biogenic CO₂ to a remote site in the Arctic is studied. A limited area, off‐line, Eulerian atmospheric transport model is used, and the results are compared to the observed CO₂ concentration at the "Ny‐Alesund International Arctic Research and Monitoring Facility". Inventories of anthropogenic CO₂ emissions and estimates of biogenic CO₂ emissions are used to investigate the respective impact of these emissions on Arctic CO₂ variations during 4 winter months. A direct comparison of the modelled and observed concentrations reveals remarkably good timing in the modelled variations as compared to the observed variations for most of the time. The correlation of observed versus modelled CO₂ concentration is significant at the 95% confidence level. The biogenic and the anthropogenic CO₂ emissions are shown to have approximately equal influence on Arctic CO₂ variations during winter. Europe is found to be the dominant source of anthropogenic CO₂ at the monitoring station, while Siberia and Northern America have little influence on Arctic CO₂, during the months studied. These results contradict Engardt and Holmén whose results indicate that the lower‐Ob region in western Siberia has a large impact on Arctic CO₂.