JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 107, NO. D24, 4766, doi:10.1029/2001JD000827, 2002
Marine and anthropogenic aerosols at Punta Del Hidalgo, Tenerife, and the aerosol nitrate number paradox F. M. McGovern,1 M. J. Nunes,2 F. Raes,3 and H. Gonzales-Jorge4 Received 14 May 2001; revised 25 February 2002; accepted 28 February 2002; published 20 December 2002.
[1] Results from analysis of aerosol ion composition and condensation nuclei (CN)
concentration measurements carried out at the second Aerosol Characterization Experiment (ACE-2) Tenerife site, in the eastern North Atlantic subtropical region between July 1995 and May 1997, are described. Sea-salt derived Na+ dominated the samples having an average concentration of 2.4 mg m3. Extensive anthropogenic influences are evident with average non-sea-salt sulfate (nssSO42) and nitrate (NO3) concentration values of 2.1 mg m3 and 1.3 mg m3 respectively. High levels of these species are linked to air mass transport from Europe: observed as short-term (2–3 days) pollution peaks. The pollution peaks overlie a summertime maximum in background levels. From comparison with Southern Hemisphere data, it is estimated that 80% of the observed nssSO42 originated from anthropogenic sources. Regression analysis shows a high degree of linearity between the NO3 and CN concentration, which is stronger than found with between NssSO42 and CN concentration. This is considered paradoxical as in the maritime atmosphere NO3 has primarily been found in supermicron size range linked to sea-salt, while nssSO42 and CN are primarily considered to be submicron aerosol. The CN:NO3 linearity is considered to arise from NO3 formation and transport INDEX TERMS: 0305 Atmospheric Composition and Structure: processes over ocean regions. Aerosols and particles (0345, 4801); 0368 Atmospheric Composition and Structure: Troposphere—constituent transport and chemistry; 0312 Atmospheric Composition and Structure: Air/sea constituent fluxes (3339, 4504); 0315 Atmospheric Composition and Structure: Biosphere/atmosphere interactions; KEYWORDS: Marine, aerosol, sulfate, nitrate, condensation nuclei Citation: McGovern, F. M., M. J. Nunes, F. Raes, and H. Gonzales-Jorge, Marine and anthropogenic aerosols at Punta Del Hidalgo, Tenerife, and the aerosol nitrate number paradox, J. Geophys. Res., 107(D24), 4766, doi:10.1029/2001JD000827, 2002.
1. Introduction [2] Fossil fuel combustion and biomass burning are major sources for a range of short and long-lived atmospheric pollutants. Understanding and quantification of the influences of diverse combustion related species, such as carbon dioxide and aerosol sulfate, on the global radiation balance is pivotal to understanding anthropogenic impacts on climate. The Third IPCC report (2001) shows that estimated radiative forcing by the main greenhouse gases is relatively well understood. However, large levels of uncertainty exist in estimates of radiative forcing caused by direct and particularly indirect aerosol radiative influences. Andrea et al. [1997], in a review of scientific work since the publication of the acronymic CLAW hypothesis by Charlson et al. [1987], point to the lack of progress in this area: they cite the absence 1
Environmental Protection Agency, Richview, Dublin, Wexford, Ireland. CECUL, DQB, Faculty of Sciences, University of Lisbon, Lisbon, Portugal. 3 Environment Institute, EC Joint Research Centre, Ispra, Italy. 4 Department of Physics, University of LaLaguna, Tenerife, Canary Islands, Spain. 2
Copyright 2002 by the American Geophysical Union. 0148-0227/02/2001JD000827
AAC
of sufficient long-term measurements of key atmospheric species as hampering progress. Unfortunately, the range of aerosol data available for large regions of the globe has decreased in recent years with the closure of a number of measurement sites (J. M. Prospero, private communication, 1998). Here we present an analysis of aerosol measurements carried out at a relatively new site, Punta del Hidalgo (PDH), on the Spanish island of Tenerife (16W, 28.5N) over a twoyear period. This site was used extensively during June and July 1997 for studies undertaken as part of the second Aerosol Characterization Experiment (ACE-2) [Raes et al., 2000] and been used for a number of short-term campaigns. The data presented here were obtained as part of the ACE-2 Longterm project. The Longterm project was designed to provide a temporal and spatial context for the intensive ACE-2 measurements. A study of regional impacts of European pollution at Tenerife and other Longterm sites; Madeira Island, Azores Islands and Cabo Sao Vicente, (south west Portugal) has been carried out by Nunes [2002]. This study shows Cabo Sao Vicente to be the most highly polluted of the Longterm Sites with pollutants from Europe having least influence at the Azores of the Longterm Sites. [3] The Azores High-Pressure system, which gives rise to the Trade Wind circulation, has a dominant role on air mass movement in the subtropical North Atlantic region. The
9-1
AAC
9-2
MCGOVERN ET AL.: AEROSOLS AT PUNTA DEL HIDALGO
Azores High frequently extends into Europe, typically following the passage of midlatitude low-pressure systems. On such occasions pollutants from a range of European source regions accumulate in the circulating air and are transported to the North Atlantic region. Long-range transport of European pollutants in the Trade Winds has been observed off the east coast of the USA [Savoie et al., 1989a]. The PDH site, in the East North Atlantic region, is located in the path of the Trade Wind circulation and provides an excellent observation point for examination of the characteristic these pollutants at an earlier stage. These data are analyzed in terms of the influences of marine and continental/anthropogenic sources. The main source components are described through analysis of variations in marine aerosol species i.e. sea-salt, methanosulfonic acid and variations in aerosol sulfate, nitrate levels, which have large regional sources. Relationships between these species and the condensation nuclei (CN) level are examined in detail. The importance of sulfate aerosols in climate studies is linked to their roles in light scattering and as cloud condensation nuclei (CCN) [e.g., Charlson et al., 1992]. This arises as in athropogenically influenced air masses sulfate aerosols are primarily, but not exclusively, found in the submicron size range where then can efficiently scatter solar radiation and under suitable thermodynamic conditions can act as CCN. Consequently considerable attention has been given to understanding relationships between sulfate and submircon aerosol number and volume [e.g., Hegg, 1994; Van Dingenen et al., 2000]. However, the data presented here suggest that in polluted air masses transported over Ocean areas for a number of days the relationship between total aerosol number, measured as CN, and the NO 3 concentration is stronger than that found between CN and nssSO2 4 concentration. To an extent this observation contradicts the generally held view that the CN concenlevels while tration would be primarily related to SO2 4 NO 3 is primarily found in the super-micron size range linked to coarse mode sea-salt, i.e. the titular aerosol number, nitrate, paradox. These observations are considered here in terms of the evolution of polluted air masses in the marine atmosphere.
2. Measurement Site, Instrumentation [4] The location of the PDH site is shown in Figure 1. The measurements were carried out at the top of newly constructed lighthouse at a height of approximately 65 m above sea level. The site has open access to the Atlantic Ocean between 270– 0 – 90 (Ocean Sector). Airflows at the site are predominately from a North Easterly direction associated with the Trade Wind circulation. Under Ocean Sector conditions, the site is free from any local influences, including those of sea-salt generated at the shoreline. Aerosol filter samples were collected on a daily basis and using a Riemer High Volume sampler with a sample collection flow rate of 500 l min1. Whatman 41 filters were used as the collection substrate. Sample collection was wind sector controlled so that samples were only collected from Ocean Sector air masses. Filter samples were collected on a 24 h basis with filter samples being changed automatically at midnight UTC. Sample blanks were obtained on a
Figure 1. The location of the ACE-2 Longterm site at Punta del Hidalgo (PDH), Tenerife, is shown in this map of the eastern North Atlantic regions. The locations of the other Longterm sites are also shown.
three-day basis. After sampling the filters were placed in sealed plastic bags and kept in refrigerated conditions until analysis. [5] Samples with low sample volumes, < 150 m3, were not used. Quarter sections of the filters from the High Volume system were extracted with 20 cm3 of deionized water (18.2 M cm1) by centrifuge extraction (3 times with 7 cm3 water during 151 min at 1000 r.p.m.). The centrifuge tube filters (Costar ) were weighed before and after the extraction to determine the exact water extraction volume. The extracts were analyzed for major water soluble 2 + + ionic species: anions (Cl, NO 3 , SO4 ), cations (Na , NH4 , + 2+ 2+ K , Mg , Ca ) and organic acids including methanosulfonic acid (MSA) by ion chromatography (IC). The analysis was performed simultaneously by three Dionex systems DX-500. Helium C was used to sparge and degas ionic eluents and Nitrogen C was used for the automatic injection valve. The chromatographic conditions were similar except for the following. The injection volumes of cations and anions were 25 ml and 100 ml respectively. An injection volume of 400 ml was used for organic analysis. For suppression the autosuppression recycle mode was used with an electrode current of 100 mA. The detection limits for all the samples were conditioned by the blank samples. [6] The CN concentration was measured using a TSI 3010 instrument, which can detect particles from approximately 10 nm diameter upwards at standard operation i.e. manufacturers calibration. Data from this instrument were logged at 1 minute intervals along with the meteorological parameters of wind speed, wind direction (Theise), temperature, pressure and relative humidify (Lasterm). A standard personal computer (PC) fitted with a Keithly DAS-802 interface board was used for data logging. Software control of filter sample collection and filter changes was via a serial port RS232 connection to the High Volume sampler. Aerosol black carbon levels were
MCGOVERN ET AL.: AEROSOLS AT PUNTA DEL HIDALGO
Figure 2. Time series of equivalent of hourly average aerosol black carbon (EBC) and condensation nuclei (CN) concentration data recorded at the PDH site between September and December 1995. Breaks in the data are due to instrument problems.
determined on a 20 minute basis using an Aethalometer [Hansen et al., 1984].
3. Results of Measurements Carried Out at PDH [7] The long-term measurements show that the PDH site experiences a variety of air conditions, which vary from relatively clean background North Atlantic marine to strongly polluted. These influences are reflected in the aerosol black carbon, referred to here as Equivalent Black Carbon (EBC), and CN levels measured at the site. Figure 2 shows time series of EBC and CN data measured between 1st September 1995 and 1st December 1995 which exemplify the variations observed. A high degree of similarity is apparent in the EBC and CN levels indicating a similar source for these species. As EBC principally arises from combustion it is an excellent tracer for anthropogenic influences. It is evident from Figure 2 that on occasions the site experiences relatively high pollution levels. Putaud et al. [2000] provide a detailed description of the characteristics of such pollution events observed during ACE-2. McGovern et al. [1999] also show the occurrence of similar concurrent peaks in EBC and aerosol light scattering (550 nm) levels in measurements carried out during July 1994 and July 1995 at PDH. This feature is also apparent in the Longterm data, a separate manuscript on the analysis of these data is under preparation. The continuous record may also be influenced by local sources. However, filter sample collected for ion composition analysis was sector controlled, as described in Section 2, to avoid such local influences. The main ionic species, which reflect anthropogenic influ ences, i.e., SO2 4 and NO3 , and relationships between these species and the CN concentration, are examined in the following section. As sea-salt is an important and typically predominant component of the aerosol at the PDH site, variations in the sea salt levels are considered first. 3.1. Sea-Salt Aerosols at PDH [8] The important role of sea-salt has been discussed elsewhere, for example, Keene et al. [1998]. McGovern et al. [2000] show that, under background conditions, sea-salt can dominate light scattering at PDH while under polluted related aerosols dominated scattering. condition nssSO2 4
AAC
9-3
Sea-salt, as determined by analysis of Na+ levels, generally dominates the aerosol mass loading at PDH. The average Na+ concentration was 2.4 ± 1.3 mg m3, higher than found for any of the other ionic species. Generation of sea-salt aerosols involves a variety of wind related mechanisms which, as discussed by Gong et al. [1997] and references therein, vary under different sea-conditions. The ambient sea-salt load is therefore expected to display a general dependence on wind speed. A scatterplot of Na+ levels against wind speed is shown in Figure 3. A positive correlation between the Na+ concentration and wind speed, measured at 65 m asl, is found in a widely dispersed data set, for which the following regression equation is obtained. Naþ ¼ 0:32 0:3 Wind Speed ms1 þ 0:55 0:25 R2 ¼ 0:35 ð1Þ
No significant seasonal differences are found in the sea-salt loading at PDH. This reflects the fact that high wind speeds conditions occur throughout the year at the site. No significant relationship was found between Na+ levels and the CN concentration. 3.2. European Pollution Impacts and Sulfate and Nitrate Levels [9] Time series of nitrate (NO 3 ) and non sea-salt sulfate (nssSO2 4 ) levels at PDH between July 1995 and June 1997 are shown in Figure 4. The NO 3 data are considered to represent total NO 3 , i.e., combined gaseous and aerosol phase [Savoie et al., 1989a]. The average ocean sector and NO nssSO2 4 3 concentration found at PDH over the measurement period are 2.14 ± 1.96 mg m3 and 1.26 ± 0.89 mg m3 respectively. These values are significantly higher than values reported for long-term measurements at other North Atlantic sites, i.e., at Barbados [Savoie et al., 1989a] and Iceland [Prospero et al., 1995] and for Southern Hemisphere sites [Savoie et al., 1989b]. The large standard deviation values for these data arise from the episodic nature of the pollution events. The pollution events seen
Figure 3. Scatterplot of sea-salt derived sodium (Na+) against wind speed. The data show increasing sea-salt aerosol load with wind speed. The wide scatter found arises from the range of sea-salt production mechanisms, which exist under different ocean conditions.
AAC
9-4
MCGOVERN ET AL.: AEROSOLS AT PUNTA DEL HIDALGO
Europe dominated during the summer months. Equivalent analysis of the nssSO4: NO 3 relationship for the PDH data shows that this ratio remains relatively constant throughout the year. This is expected as only transport from European is considered to have had a significant influence on the PDH measurements. Linear regression analysis given the following equation 2 2 NO 3 ¼ 0:36 0:02 nssSO4 þ 0:3 0:04 R ¼ 0:70
Figure 4. Time series of non sea-salt sulfate (nssSO2 4 ) and nitrate (NO 3 ) measured at PDH. The pollution peaks in this plot are linked to air mass transport from Europe. as short term peaks in Figure 4 can last for a number of days and occur throughout the year. Back trajectory analyses principally carried out by Merrill [1994] shows that the peak values are typically linked to air mass transport from Europe. North African sources may also influence the measurements. Such events are clearly evident from the meteorological record as periods of anomalous high temperature and low relative humidity and are generally associated with high levels of crustal aerosol species. These types of events were not frequent during the period under consideration with only one significant outbreak from North Africa being recorded. The pollution peaks seen in Figure 4 are therefore attributed to transport of pollution from European source regions. The role of the Azores High in determining the nature of this transport has been discussed by McGovern et al. [1999] and in greater detail by Verver et al. [2000]. Figure 5a shows a typical air mass trajectory for periods when the Azores High extends into Europe. During such events the air masses generally pass along the western coast of Spain and Portugal and may be influenced by pollution sources in these areas. During the winter months, when the Azores High is less dominant, more complex meteorological patterns can also result in transport of pollutants from other European regions, as shown in Figure 5b. In general, the trajectories indicate that the air masses have been over ocean regions for 2 – 5 days before reaching the measurement site. Over the relatively short time period under consideration here considerable inter-annual variability is observed. For example, February 1997 was highly polluted while February 1996 was relatively clean. July 1996 was highly polluted while the July 1997 ACE-2 measurement period was by comparison relatively clean. [10] Savoie et al. [1989a] show that nssSO2 4 and NO3 levels at Barabados were influenced by long range transport of continental aerosols from Africa and Europe. They 2 reported finding a seasonal difference in the NO 3 : nssSO4 ratio. This was linked to the continental source region that influenced the measurements, i.e. transport from Africa dominated during the winter months while transport from
ð2Þ
The slope value found is surprising similar to that found by Savoie et al. [1989a], i.e. 0.36 ± 0.04, for European influenced Barbados data. 3.2.1. Variation in Background NssSO2 4 and NO3 at PDH [11 ] As described by Charlson et al. [1989] DMS released by phytoplankton may be oxidised to form sulfate or MSA. Ayers and Gras [1991] and Andrea et al. [1999] show that this is an important source of sulfate aerosol in Southern Hemisphere. MSA levels have been used as a tracer of biogenic activity and multiple linear regression between nssSO2 4 and concurrent NO3 and MSA levels has been used to apportion marine and non-marine sources for the observed sulfate levels [Savoie et al., 1993]. The regression equation for this type of analysis for the PDH data is given below. The high R2 value obtained for these data would indicate that biogenic sources influenced the relationship between these species. 2 nssSO2 4 ¼ 1:75 0:1 NO3 þ 19:7 3:2 MSA R ¼ 0:76
ð3Þ
However, there is evidence that MSA levels may be enhanced in anthropogenicaly influenced air masses, e.g.,
Figure 5. Two examples of isentropic back-trajectory analysis for PDH. (a) For a period when the Azores high enters Europe and pollutants from Northern Europe are transported to the subtropical region by the Trade Winds. (b) More transport of pollutants from other European areas.
MCGOVERN ET AL.: AEROSOLS AT PUNTA DEL HIDALGO
Figure 6. Scatterplots of Monthly NssSO2 4 , NO3 and methanosulfonate (MSA) values for PDH. Monthly average values for these species at the Southern Hemisphere Norfolk Island site are also shown. The Norfolk Island data have been time shifted by 6 months so that the seasonal cycle is in phase with Northern Hemisphere biogenic activities.
Hubert et al. [1996]. The use of this type of analysis to apportion marine and non-marine sources in polluted regions may therefore provide ambiguous results. Another approach is used here. Background levels are estimated from data for a similar latitude Southern Hemisphere site. The University of Miami, Norfolk Island (29020S, 167570E), DOE site was selected for this analysis. This
AAC
9-5
site was chosen rather than Cape Grim, used during ACE-1, as its latitude is more closely matched to PDH. A large aerosol chemical database also exists for Norfolk Island. These data are available from the World Data Centre for Aerosols (J. Wilson, private communication, 2001). [12] Monthly nssSO2 4 , MSA and NO3 data for PDH and monthly average values for the Norfolk Island data are plotted in Figure 6. The Norfolk Island data are time shifted by 6 months so that the seasonal cycle found at the site would be in phase with a Northern Hemisphere cycle. A logarithmic scale is used for the vertical axis to better display the baseline values. The seasonal cycle in the Norfolk Island nssSO2 4 data, which display a summertime maximum, is broadly similar to the baseline values seen at PDH. Figure 6b shows the variations found in the MSA levels at PDH are also broadly similar to the levels found at the Norfolk Island. Although the average Norfolk nssSO2 4 levels are marginally lower than the baseline PDH values during the later summer and autumn months, the values are considered to adequately represent the PDH baseline values. A similar summertime peak is observed in the baseline NO 3 levels at PDH. A less significant cycle is found at Norfolk Island. Apart from a small number of values, the baseline NO 3 levels measured at PDH are generally higher, i.e., approximately double, than those found at Norfolk. Using nssSO2 4 as a reference species this would suggest that baseline Northern Hemisphere, NO 3 values are significantly enhanced by comparison with the Southern Hemisphere levels. levels at [13] Anthropogenic impacts on the nssSO2 4 PDH were estimated from the difference between monthly average values at Norfolk (baseline values) and monthly average values at PDH. These data are given in Table 1 and seen at PDH is indicate that over 80% of the nssSO2 4 related to anthropogenic or continental sources. A similar percentage is found for NO 3 . This calculation is based on the assumption that anthropogenic non-marine influences can be neglected at the Southern Hemisphere site. However, there are ample data to show that continental and anthropogenic sources do influence measurements even at remote sites. Accounting for such impacts is likely to lead to an increase in the estimated anthropogenic impacts at PDH but in the context of the strong anthropogenic influences already estimated for the PDH data this calculation would not greatly change the overall results.
a Table 1. Monthly Statistics for nssSO2 4 , NO3 and MSA Levels Measured at the PDH Site
PDH Data (average ± std dev) Month January February March April May June July August September October November December
3 nssSO2 4 , mg m
0.50 0.83 0.70 1.53 2.08 3.18 3.54 1.88 2.81 2.68 1.68 0.88
± ± ± ± ± ± ± ± ± ± ± ±
0.49 0.68 0.56 1.16 1.23 2.53 2.44 1.62 2.15 2.10 1.95 0.72
NO3 , mg m3 0.61 0.66 0.44 0.82 1.62 1.36 1.12 0.89 1.48 1.52 1.19 1.09
± ± ± ± ± ± ± ± ± ± ± ±
0.62 0.69 0.29 0.51 1.35 0.88 0.60 0.37 1.01 1.03 1.06 0.95
PDH - Nor, (average,% increases) MSA, mg/m3 0.01 0.01 0.01 0.05 0.02 0.06 0.09 0.04 0.03 0.02 0.01 0.01
± ± ± ± ± ± ± ± ± ± ± ±
0.011 0.004 0.015 0.031 0.023 0.042 0.067 0.017 0.012 0.012 0.011 0.016
3 nssSO2 4 , mg/m
nssSO2 4 , %
NO3 , mg/m3
NO3 , %
0.31 0.59 0.45 1.24 1.65 2.71 3.08 1.47 2.52 2.44 1.48 0.71
0.62 0.72 0.64 0.81 0.80 0.85 0.87 0.78 0.90 0.91 0.88 0.80
0.47 0.46 0.23 0.56 1.39 1.12 0.89 0.69 1.27 1.34 1.04 0.95
0.77 0.70 0.52 0.69 0.86 0.82 0.80 0.78 0.86 0.88 0.87 0.88
a Estimates of anthropogenic enhancement of nssSO4 and NO3 at PDH based on the assumption that background levels can be reasonably estimated from Norfolk Island data.
9-6
AAC
MCGOVERN ET AL.: AEROSOLS AT PUNTA DEL HIDALGO
super-micron size range [Savoie and Prospero, 1982], while is considered to exist in polluted air masses nssSO2 4 primarily in the sub-micron size range as are the bulk of the CN population. It should be noted that the correlation between CN and nssSO2 4 is not in itself exceptionally low but is lower than the correlation between CN and NO 3. Similar analyses of relationship between CN and other measured ionic species show weak correlation. Analysis of monthly data shows that this feature is also found on shorter timescales also. Data for September 1995, shown in Figure 8, clearly show that the CN concentration displays a highly linear relationship with NO 3 whiles a more scatter relationship is found when plotted against nssSO2 4 . Similar results are found for other months. The coherence of the monthly data may be attributed to relatively homogenous sources influencing the measurements over the shorter time period. The scatter seen in the full data set can therefore be attributed to the variability of source regions influencing these measurements over the longer integration period. [15] In order to further assess this feature the variations of the CN:NO 3 ratio with respect to available parameters that may potentially influence this ratio were examined. This showed that variations in parameters such as; sea-salt levels,
Figure 7. Scatterplot of condensation nuclei (CN) conand NO centration against nssSO2 4 3 concentration. As shown by the regression equation data the CN displays a high degree of linearity with respect to the NO 3 concentration. 3.3. NssSO2 4 , NO3 and CN Concentration [14] The influence of anthropogenic pollution on the aerosol number concentration is of particular interest in climate and atmospheric chemistry studies. Scatterplots of the average CN concentration, over the sampling period, with the equivalent nssSO2 4 and NO3 data, are shown in Figure 7. These plots are based on all valid data collected during the measurement period. Problems with the CN counter during early 1997 mean that data are not available for this period. Linear regression equations are shown on the charts. Given that these species arise from similar sources the analysis shows, as expected, a positive correlation between CN and these ionic species. However, an unexpected feature of the data is the high degree of linearity 2 found between the CN and NO 3 levels. Also the R value for linear regression analysis of the CN and NO3 data is higher than that found between CN and nssSO2 4 .
3 þ86 34 CN cm3 ¼ 725 32 NO 3 mg m CN cm
3
¼ 260 19 nssSO2 4
R2 ¼ 0:77 ð4Þ
mg m3 þ300 41 R2 ¼ 0:55 ð5Þ
This is unexpected as previous studies show that in marine air masses NO 3 is primarily found in the coarse mode, i.e.
Figure 8. Scatterplot of CN concentration against and NO nssSO2 4 3 concentration for September 1995. While both plots show a positive correlation between the species, the CN displays a higher correlation with respect to 2 the NO 3 concentration than with the nssSO4 concentration. The error bars indicate the standard deviation on the CN concentration values.
MCGOVERN ET AL.: AEROSOLS AT PUNTA DEL HIDALGO
Figure 9. Variation in CN:NO 3 ratio with respect metanosulfonic acid (MSA) and temperature. The data show that the CN:NO 3 ratio is more scattered at lower MSA and temperature values. wind speed and wind direction, were effectively neutral 2 with respect to CN:NO 3 and CN:nssSO4 ratios. However, data with respect to MSA and analysis of the CN:NO 3 ambient temperature levels, as shown in Figure 9, indicate that NO3:CN ratio was generally more scattered at lower MSA and lower ambient temperature levels. These results are considered to primarily indicate that the CN:NO 3 ratio exhibited more scatter during the winter months than during the summer months. Regression analysis of data classified into two seasonal groups, predominately ‘‘Summer’’, March to September and predominately Winter, October to April, shows this to be the case.
4. Discussion [16] A primary objective of this paper is to describe the extent of anthropogenic impact on the regional aerosol as observed at the PDH site. The data show that ambient levels arise from a combination of background nssSO2 4 sulfate, probably of marine biogenic origin, which displays a seasonal cycle, over which a dominant anthropogenic signal is superimposed. In order to determine the extent of the anthropogenic influence a reliable estimate of the background level is required. However, this issue is complicated by interactions that occur between natural and anthropogenic species. To overcome this difficulty, data for an equivalent latitude Southern Hemisphere site were used to
AAC
9-7
determine background values. While this approach is open to some challenges, the closeness of the monthly average Norfolk nssSO2 4 data to the baseline PDH levels, as well as the similarity of the MSA data for the two sites, is considered to warrant its use. Monthly summaries of anthropogenic nssSO2 4 enhancement are given in Table 1. These show that the PDH measurements were dominated by anthropogenic/continental influences. [17] A seasonal cycle is also apparent in the baseline NO 3 levels. Such seasonal cycles have been also been reported at Southern Hemisphere sites, e.g. at Cape Grim [Andrea et al., 1999] and Palmer, Antarctica [Savoie et al. 1993]. The main sources for the measured NO 3 are not as yet clearly identified. However, continental sources are considered to impact most sites. In the North Atlantic region various anthropogenic/continental sources are considered to be the main source for NO 3 seen at Barbados [Savoie et al., 1992]. European areas are considered to be primary source for NO 3 at PDH. Examination of the baseline levels at PDH shows that apart from a few values the baseline levels were approximately double the average levels found at Norfolk Island. This was particularly evident during the summer and autumn months. This may indicate that the continental/ anthropogenic influence on the baseline NO 3 levels is more pronounced than found for nssSO2 4 . Recent re-evaluation of the impacts of shipping by Lawrence and Crutzen [1999] suggests that NOx emissions from this source may impact on the levels of these species in Northern Hemisphere Ocean areas. These are estimated to be double the emissions found in Southern Hemisphere Ocean areas. Such emissions are also likely to impact on background NO 3 levels, particularly in the Trade Winds. Further analysis of this feature is beyond the scope of this paper. 2 4.1. CN, NO 3 and nssSO4 [18] The PDH data show concurrent increases in CN, 2 levels during periods influenced by NO 3 and nssSO4 continental air masses. During such events peaks were also observed in Aethalometer aerosol carbon measurements. As these species are considered to have originated from a similar source region, a positive correlation is expected between them. However, as discussed in section 3.3 the strength of the linear correlation between the CN and the NO 3 concentration is unexpected. [19] A simple linear relationship between the total aerosol particle number and the mass of a constituent chemical species can be expected if variations in the species mass are directly linked to variation in the number concentration. However, formation of aerosol SO2 4 and NO3 from the gas phase will not necessarily occur in a linear manner with respect to the aerosol number distribution. Formation of SO2 4 from SO2 by non-precipitating clouds will increase aerosol sulfate mass with respect to aerosol activated in the cloud formation process, typically aerosols larger than 100 nm [Fitzgerald, 1973], i.e. it will add mass to existing aerosol. Coalescence of cloud droplets may also reduce the aerosol number. Hopple et al. [1994] describe the effects of cloud processing on the marine aerosol size distribution which is considered to lead to an aerosol number concentration minimum around 0.1 mm: widely observed in marine aerosol size spectra. In anthropogenically influenced air masses a linear relationship has been observed between
AAC
9-8
MCGOVERN ET AL.: AEROSOLS AT PUNTA DEL HIDALGO
SO2 4 mass and accumulation mode aerosol, for example, Hegg and Kaufmann [1998], indicating that the SO2 4 mass is proportionally distributed in the accumulation mode and show a strong positive correlation between accumulation mode aerosol volume and number. However, the relationship between total aerosol number/CN and sulfate or aerosol volume is more complex. Van Dingenen et al. [1999] have proposed a second order polynomial to describe the relationship between total particle number and aerosol volume and relate aerosol volume to SO2 4 mass, while Hegg and Kaufmann [1998] relate sulfate and organic aerosol to the total number concentration. It is considered that the strong correlation between nssSO2 4 and accumulation mode aerosol reflects the fact that, although coarse mode sulfate is found [Sievering et al., 1991] in clean marine conditions the distribution of non-sea-salt sulfate may be skewed toward the larger size range. In aged anthropogenically influenced air masses sulfate aerosol is principally found in the accumulation mode size range [e.g., O’Dowd et al., 1999]. The relationship between total aerosol number and SO2 4 mass in these situations will then be dependent on the aerosol number size distribution and particularly the relationship between particle number less than 0.1 mm and particle number greater than 0.1 mm. [20] Formation of NO 3 aerosol in continental areas is considered to principally involve gas phase reactions. Formation of NH4NO3 aerosol is frequently linked to the presence of catalytic species such as aerosol carbon/soot, for example, ten Brink et al. [1996]. Ammann et al. [1998] also show the importance of the presence of soot aerosol as a catalyst in the production of nitrous acid. Harrison et al. [1996] discuss the atmospheric formation of HNO2 and highlighted the general importance of surface reactions in atmospheric nitrogenous chemistry and describe substantial deposition of HNO 3 to surfaces. Ammonium nitrate (NH4NO3) can make up a significant percentage of fine mode aerosol in continental air masses, for example, ten Brink et al. [1997]. Submicron nitrate aerosol has also been found in continental air masses transported from continental to ocean areas Chen et al. [1997]. However, this site is very close to the continental source region. As discussed above, in more remote maritime atmospheres aerosol NO 3 is considered to be principally resident in the super micron size range. Size resolved aerosol measurements during ACE-2, also show NO 3 to be principally associated supermicron aerosols [Putaud et al., 2000]. Reactive uptake of gaseous HNO3 and N2O5, produced by reaction between ozone and NO2, by Na+ in sea-salt aerosols are important sources for aerosol nitrate in the marine atmosphere. These reactions have consequent losses of chlorine and bromine from the aerosol phase [Moldanova and Ljungstrom, 2001]. However, considerable uncertainties remain about the reactions and transport pathways, which determine atmospheric NO 3 levels in the maritime atmosphere. This is in part due to the volatile nature of this species [Warneck, 1999]. [21] The above discussion does not provide an explanation for the PDH observations, for which it is assumed that the strength of the NO 3 /CN correlation indicates more than a coincidental source relationship. However, it does indicate that the presence of existing aerosol surface is important in NO 3 formation. Size resolved aerosol chemical data are not available as part of the Longterm data set, however, it is
considered unlikely that NO 3 would be predominately found in fine mode aerosol during months other than June and July. In order to provide an explanation for the PDH data, the existence of a submicron refractory aerosol core is assumed to be pivotal to the observed relationship. The continental leaving air mass is considered to contain a range of gaseous pollutants, including SO2, nitrogen oxides (NOX) and ammonia (NH3). The particulate phases is considered to principally consist of ammonium sulfate, ammonium nitrate and refractory aerosol species e.g. carbonaceous/soot material. Fine mode ammonium nitrate linked to the refractory core is considered to form a substantial component of the initial continental-leaving aerosol. As the polluted continental air mass mixes with the maritime aerosol excess gaseous species are removed through reactions with sea-salt aerosols or are deposited to the ocean surface before reaching PDH. ACE-2 Lagrangian measurements in polluted air masses carried out by Andreae et al. [2000] suggest that under cloudy conditions SO2 was removed rapidly with lifetimes of the order of 12 h and that gaseous HNO3 was taken up by sea salt aerosol and removed by dry deposition. However, measurements on Tenerife suggest that low concentrations of HNO3 were always present and that higher HNO3 levels were present during polluted conditions [Bower et al., 2000]. The authors suggest that the HNO3 resulted from outgassing from the aerosol phase, in which nitrate rich aerosol is transported to the region. [22] The observed NO 3 concentration at PDH is likely to have arisen from a combination of heterogeneous formation from precursor gaseous species on salt particles and transfer, via gas phase, from the fine aerosol mode. It is considered likely that a large fraction of the NO 3 aerosol derived from heterogeneous reaction of NOx species in the marine atmosphere will have been deposited to the ocean before the air mass reached PDH. The correlation found been NO 3 and the CN concentration is therefore considered to indicate that the observed NO 3 primarily originated from the fine mode aerosol. Bassett and Seinfeld [1984] provide a theoretical basis for the observed size distribution SO4 and NO 3 based on gas kinetics and thermodynamics. This study suggests that the relative volatility of NO 3 results in its preferential transfer to aerosols in the larger size range. Consequently transfer of NO 3 between different aerosol species is assumed to take place in the evolving polluted air mass. The observed CN concentration is considered to reflect a residual refractory aerosol core, pivotal to initial formation and transport of NO 3 to the region. The widespread existence of such refractory cores, even in remote regions, has been shown by thermal volatility studies, for example, Clarke [1993]. Posfai et al. [1999] show the existence of carbonaceous cores in TEM studies of Atlantic and remote Pacific aerosol. [23] The observed correlation between the CN and NO 3 concentration may then be a relatively transient feature. The temporal duration of the CN:NO 3 correlation is likely to be governed by a number of factors including; thermodynamic and ambient pH equilibrium values, the availability of aerosol surface and the range of competing reactive species. The CN:NO 3 ratio is apparently more stable during the summer months than during the winter period. It is considered that the more diverse meteorological patterns, which
MCGOVERN ET AL.: AEROSOLS AT PUNTA DEL HIDALGO
occur during the winter months, are likely to be responsible for this variation. However, seasonal analysis of trajectories provides no clear pattern to explain this. This suggests that other factors may also be involved. The stability of aerosol NO 3 is highly temperature dependent. However, as the higher summertime temperatures are linked to the more stable ratio this is not considered to be an influencing factor. The CN:NO 3 ratio is also more stable at higher MSA levels. This may indicate that chemical factors contribute to the stability of this ratio. While this is an interesting idea more detailed data are require before further consideration of this is warranted.
5. Conclusions [24] We have presented an analysis of Na+, nssSO2 4 , NO3 and CN levels recorded at PDH. The data indicate that, while natural aerosols such as sea-salt dominate the regional aerosol mass composition, anthropogenic sources strongly influence the regional aerosol. An evaluation of the extent of anthropogenic impacts at the site has been given through determining baseline levels by comparison with an equivalent latitude Southern Hemisphere site. From this analysis, it is estimated that over 80% of the measured nssSO2 4 and are of continental or anthropogenic origin. From NO 3 previous analyses, this would suggest that anthropogenic influences are likely to dominate the light scattering and potential cloud nucleating characteristics of the region. [25] The linearity found between the CN concentration and the NO 3 levels, referred to in the title as the ‘‘aerosol, number nitrate paradox’’, is considered an important feature of the data set. It is proposed that the correlation is indicative of similar anthropogenic sources for CN and NO 3 in the region and that the observed NO3 , though likely to exist in coarse mode, is principally related to formation and transport in the sub-micron aerosol. The CN:NO 3 correlation is therefore considered to be a transient feature of the evolving polluted air mass as it mixes with the maritime aerosol. However, an empirical linear relationship between NO 3 and CN as shown here does not necessary imply that NO 3 species have been transferred from small to large particles during transport over oceans. Further data are required to assess this. A fuller theoretical analysis of this feature may provide insights into the evolution and impacts of polluted air masses in the maritime environment particularly in relation to factors determining CN and CCN levels in polluted air mass over ocean areas. In addition more complete measurements of the aerosol throughout the year at the PDH site would also greatly enhance understanding of these results. In general the PDH site, which is currently shut down, is an excellent location for monitoring the impacts of European pollutants transported to the region. It should therefore be seen, and utilised, as a vital component of any network for long-term pollutant measurements in the North Atlantic region.
[26] Acknowledgments. The Longterm sites were established and operated by the Environment Institute of the European Commission, Joint Research Centre Ispra, Italy. The authors wish to thank Dennis Savoie for access to the University of Miami database and background information on these data and to John Merrill for access to his trajectory analyses for Punta Del Hidalgo. The authors also wish to acknowledge the helpful comments made by the anonymous reviewers of this manuscript.
AAC
9-9
References Ammann, M., M. Kalberer, D. T. Jost, L. Tobler, E. Ro¨ssler, D. Piguet, H. W. Ga¨ggeler, and U. Baltensperger, Heterogeneous production of nitrous acid on soot in polluted air masses, Nature, 395, 157 – 160, 1998. Andreae, M. O., and P. J. Crutzen, Atmospheric aerosols: Biogeochemical sources and role in atmospheric chemistry, Science, 276, 1052 – 1058, 1997. Andreae, M. O., W. Elbert, Y. Cai, T. W. Andreae, and J. Gras, Non-sea-salt sulfate, methanesulfonate, and nitrate aerosol concentrations and size distributions at Cape Grim, Tasmania, J. Geophys. Res., 104, 21,695 – 21,706, 1999. Andreae, M. O., W. Elbert, R. Gabriel, D. W. Johnson, S. Osborne, and R. Wood, Soluble ion chemistry of the atmospheric aerosol and SO2 concentrations over the eastern North Atlantic during ACE-2, Tellus Ser. B, 52, 1066 – 1087, 2000. Ayers, G. P., and J. L. Gras, Seasonal relationship between cloud condensation nuclei and aerosol methansulphonate in marine air, Nature, 353, 834 – 835, 1991. Bassett, M., and J. H. Seinfeld, Atmospheric equilibrium model of sulfate and nitrate aerosols-II particle size analysis, Atmos. Environ., 18, 1163 – 1170, 1984. Bower, B. K. N., et al., ACE-2 HILLCLOUD. An overview of the ACE-2 ground-based cloud experiment, Tellus Ser. B, 52, 750 – 778, 2000. Charlson, R. J., J. E. Lovelock, M. O. Andrea, and S. G. Warren, Ocean phytoplankton atmospheric sulfur, cloud albedo and climate, Nature, 326, 655 – 661, 1987. Charlson, R. J., S. E. Schwartz, J. M. Hales, R. D. Cess, J. A. Coakley, J. E. Hansen, and D. J. Hofmann, Climate forcing by anthropogenic aerosols, Science, 255, 423 – 430, 1992. Chen, L. L., et al., Influence of continental outflow events on the aerosol composition at Cheju Island, South Korea, J. Geophys. Res., 102, 28,551 – 28,574, 1997. Clarke, A. D., Atmospheric Nuclei in the Pacific mid-troposphere, their nature concentration and evolution, J. Geophys. Res., 98, 20,633 – 20,647, 1993. Fitzgerald, J. W., Dependence of the supersaturation spectrum of CCN on aerosol size distribution and composition, J. Atmos. Sci., 30, 628 – 634, 1973. Gong, S. L., L. A. Barrie, and J. P. Blanchet, Modelling sea-salt aerosols in the atmosphere, J. Geophys. Res., 102, 3805 – 3818, 1997. Hansen, A. D. A., H. Rosen, and T. Novakov, The Aethalometer-an instrument for the real-time measurement of optical absorption by aerosol particles, Sci. Total Environ., 36, 191 – 196, 1984. Harrison, R. M., J. D. Peak, and G. M. Collins, Tropospheric cycle of nitrous acid, J. Geophys. Res., 101, 14,429 – 14,439, 1996. Hegg, D. A., Cloud condensation nucleus-sulfate mass relationship and cloud albedo, J. Geophys. Res., 99, 25,903 – 25,907, 1994. Hegg, D. A., and Y. J. Kaufman, Measurements of the relationship between submicron aerosol number and volume concentration, J. Geophys. Res., 103, 5671 – 5678, 1998. Hopple, W. A., G. M. Frick, W. Fitzgerald, and R. E. Larson, Marine boundary layer measurements of new particle formation and the effects of non-precipitating clouds have on aerosol size distributions, J. Geophys. Res., 99, 14,443 – 14,458, 1994. Hubert, B. J., L. Zhaung, S. Howell, K. Nonne, and B. Noone, Sulfate, nitrate, methanesulfonate, chloride ammonium and sodium measurements from ship, island and aircraft during the Atlantic Stratocumulus Transition Experiment/Marine Aerosol Gas Exchange, J. Geophys. Res., 101, 4413 – 4423, 1996. Keene, W. C., R. Sander, A. P. Pszenny, R. Vogt, P. J. Cruzen, and J. N. Galloway, Aerosol pH in the Marine Boundary Layer: A review and model evaluation, J. Aerosol Sci., 29, 339 – 356, 1998. Lawrence, M. G., and P. J. Crutzen, Influence of NOx Emissions from ships on tropospheric photochemistry and climate, Nature, 402, 167 – 170, 1999. McGovern, F. M., F. Raes, R. VanDingenen, and H. Maring, Anthropogenic influences on the chemical and physical properties of aerosol in the Atlantic subtropical region during July 1994 and July 1995, J. Geophys. Res., 104, 14,309 – 14,320, 1999. Merrill, J. T., Isentropic air-flow probability analysis, J. Geophys. Res., 99, 25,881 – 25,889, 1994. Moldanova, J., and E. Ljungstrom, Sea-salt aerosol chemistry in coastal areas: A model study, J. Geophys. Res., 106, 1271 – 1296, 2001. Nunes, M. J., Establishment of the chemical properties of the aerosol in the north east Atlantic Ocean, Ph.D. thesis, University of Lisbon, Lisbon, Portugal, 2002. O’Dowd, C. D., J. A. Lowe, and M. H. Smith, Observations and modelling of aerosol growth in marine stratocumulus-case study, Atmos. Environ., 33, 3053 – 3062, 1999.
AAC
9 - 10
MCGOVERN ET AL.: AEROSOLS AT PUNTA DEL HIDALGO
Posfai, M., J. R. Anderson, P. R. Buseck, and H. Sievering, Soot and sulfate aerosol particles in the remote marine troposphere, J. Geophys Res., 104, 21,685 – 21,693, 1999. Prospero, J. M., D. L. Savoie, R. Arimoto, H. Olafsson, and H. Hjartarson, Sources of aerosol Nitrate and non-sea-salt sulfate in the Iceland Region, Sci. Total Environ., 160/161, 181 – 191, 1995. Putaud, J. P., et al., Chemical mass closure and assessment of the origin of submicron aerosol in the marine boundary layer during ACE-2, Tellus Ser. B, 52, 141 – 168, 2000. Raes, F., T. Baters, F. McGovern, and M. Van Liderkerke, The 2nd Aerosol Characterisation Experiment (ACE-2): General overview and main results, Tellus Ser. B, 52B, 109 – 111, 2000. Savoie, D. L., and J. M. Prospero, Particle size distribution of nitrate and sulfate in the marine atmosphere, Geophys. Res. Lett., 9, 1207 – 1210, 1982. Savoie, D. L., J. M. Prospero, and E. S. Saltzman, Non-sea-salt Sulfate and Nitrate in Trade Wind aerosols at Barbados, J. Geophys. Res., 94, 5069 – 5080, 1989a. Savoie, D. L., J. M. Prospero, J. T. Merrill, and M. Uematsu, Nitrate in the Atmospheric boundary layer of the Tropical South Pacific: Implications regarding sources and transport, J. Atmos. Chem., 8, 391 – 415, 1989b. Savoie, D. L., J. M. Prospero, S. J. Oltmans, W. C. Graustein, K. K. Turekein, J. T. Merrill, and H. Levy II, Sources of nitrate and ozone in the marine boundary layer of the tropical North Atlantic, J. Geophys. Res., 97, 11,575 – 11,589, 1992. Savoie, D. L., J. M. Prospero, R. J. Larsen, F. Huang, M. A. Izaguirre, T. Huang, T. H. Snowdon, L. Custals, and C. G. Sanderson, Nitrogen and sulfur species in Antarctic aerosols at Mawson, Palmer and Marsh (King George Island), J. Atmos. Chem., 17, 95 – 122, 1993. Sievering, H., J. Boatman, J. Gallowway, W. Keene, Y. Kim, M. Luria, and J. Ray, Heterogeneous sulfur conversion in sea-salt aerosol particles, Atmos. Environ., 25A, 1479 – 1487, 1991.
ten Brink, H. M., J. P. Veefkind, A. Waijers-Iipelaan, and J. C. van der Hage, Aerosol light scattering in the Netherlands, Atmos. Environ., 30, 4251 – 4261, 1996. ten Brink, H. M., C. Kruisz, G. P. A. Kos, and A. Berner, Composition/size of the light scattering aerosol in the Netherlands, Atmos. Environ., 30, 3955 – 3962, 1997. Van Dingenen, R., F. Raes, J. P. Putaud, A. Virkukula, and M. Mangoni, Processes determining the relationship between aerosol number and npnsea-salt sulfate mass concentrations in clean and perturbed marine boundary layer, J. Geophys. Res., 104, 8027 – 8038, 1999. Van Dingenen, R., A. O. Virkkula, F. Raes, T. S. Bates, and A. Widensohler, A simple non-linear analytical relationship between aerosol accumulation number and sub-micron volume, explaining their observed ratio in the clean and polluted marine boundary layer, Tellus, 52, 439 – 451, 2000. Verver, G., F. Raes, D. Vogelezang, and D. Johnson, The 2nd Aerosol Characterisation Experiment (ACE-2) meteorology and chemical context, Tellus, 52, 126 – 140, 2000. Warneck, P., Chemistry of the Natural Atmosphere, 2nd Ed, Academic, San Diego, 1999.
H. Gonzales-Jorge, Department of Physics, University of LaLaguna, Tenerife, Canary Islands, Spain. F. M. McGovern, Environmental Protection Agency, Richview, Clonskeagh Road, Dublin 14, Ireland. (
[email protected]) M. J. Nunes, CECUL - DQB - Faculty of Sciences, University of Lisbon, Edifı´cio C8, Campo Grande, 1749-016 Lisbon, Portugal. F. Raes, TP 460, Environment Institute, EC Joint Research Centre, Ispra I-21020, Italy.
