Seasonal variation of water-soluble inorganic components in aerosol size-segregated at the puy de Dôme station (1,465 m a.s.l.), France

Seasonal variation of water-soluble inorganic componentsin aerosol size-segregated at the puy de Dôme station(1,465 ma.s.l.), France

L. Bourcier & K. Sellegri & P. Chausse & J. M. Pichon &

P. Laj

Received: 13 September 2011 /Accepted: 26 April 2012 /Published online: 6 June 2012# Springer Science+Business Media B.V. 2012

Abstract The size-segregated chemical composition of aerosol particles was investigatedduring 1 year at the puy de Dôme (1,465 ma.s.l.), France. These measurements aimed to abetter understanding of the influence of the air mass origin on the size-segregated chemicalcomposition of the aerosol at an altitude site. Mountain site measurements are importantbecause they are representative of long range transport and useful for model validation. PM1

mass concentration exhibits a seasonal variability with a summer maximum. The composi-tion of PM1 did not change significantly in terms of relative contribution of water solubleinorganic ions but is rather variable in term of total mass concentrations. For the PM10-1, adifferent seasonal behaviour was found with maxima concentrations in autumn-winter.Aerosols were classified into four different categories according to their air mass origin:marine, marine modified, continental and Mediterranean. The PM10 aerosol mass at 50 %relative humidity was close to 2.5 !g m!3 in the marine, 4.3 !g m!3 in the marine modified,10.3 !g m!3 in the continental and 7.7 !g m!3 in the Mediterranean sectors. We noted thatthe influence of the air mass origin (on the chemical properties) could be seen especially onthe PM10-1. A significant PM10-1 mode was found in marine, modified marine, and

J Atmos Chem (2012) 69:47–66DOI 10.1007/s10874-012-9229-2

L. Bourcier :K. Sellegri : P. Chausse : J. M. Pichon : P. LajLaboratoire de Météorologie Physique, CNRS, Université Blaise Pascal, 63 177 Aubière, France

K. Sellegrie-mail: [email protected]

Present Address:L. Bourcier (*)Isotope Measurements Unit, Institute for Reference Materials and Measurements, Joint Research Center,European Commission, 2440 Geel, Belgiume-mail: [email protected]

Present Address:P. ChausseLAPSCO/BAPS, Université Blaise Pascal, 63037 Clermont-Ferrand, France

Present Address:P. LajLaboratoire de Glaciologie et Géophysique de l’Environnement,Université Joseph Fourier – Grenoble 1/CNRS (UMR 5183), 38 200 St Martin d’Hères, France

Mediterranean air masses, and PM1 dominated in the continental air masses samples. As aresult, the aerosol chemical composition variability at the puy de Dôme is a function of boththe season and air mass type and we provide a chemical composition of the aerosol as afunction of each of these environmental factors.

Keywords Aerosol particle . WSII . Chemical composition . Size-segregated

1 Introduction

Information on the atmospheric composition, from the local to the global scale, is of strategicvalue for climate and air quality related studies. Despite important improvements of theEarth observing strategy in the last decade, there is still a lack of knowledge. Observations ofthe chemical composition of the atmosphere with temporal, horizontal and vertical resolu-tion are crucially needed to 1) satisfy and verify current legislation, 2) to validate and help toimprove our understanding of atmospheric processes, and 3) to allow accurate predictions offuture atmospheric states by providing inputs to forecasting models.

A process-level understanding of the atmospheric cycle of key variables requires continuousobservations with hourly to daily resolution. At present, this is only provided by ground-basedinstrumentation. Clearly, the ground-based component is an essential element of the observingsystem because long-term and high-quality observations are required for improving our under-standing of atmospheric processes. The ground-based component, however, suffers from unevenglobal and regional coverage of atmospheric observations. In particular, while observations inthe mixed layer are to some extent sufficient, at least in OECD countries (Organisation forEconomic Co-operation and Development), there is a clear lack of information from the uppertroposphere, i.e. derived from stations located at high altitude (Laj et al. 2009). Temporalvariations in the upper troposphere of aerosol properties, especially of non-regulated advancedparameters such as chemical composition, is required for investigating climate, environmental,and health effects in the context of future global changes (Andrews et al. 2011).

Monitoring of the chemical composition of aerosol particles has been performed at manylocations within the mixed layer (e.g. Mészáros et al. 1997; Heintzenberg et al. 1998; Lestari etal. 2003; Jaffe et al. 2005; Saliba et al. 2007; Putaud et al. 2010 andmany others). However, longtermmeasurement series are scarce at high altitude sites. Kasper and Puxbaum (1998) monitoredsulphate, nitrate and ammonium at Sonnblick (3,106 ma.s.l.) during 2 y and showed markedseasonal cycles with low concentrations in winter and high concentrations in summer. Cozic etal. (2008) studied organic and inorganic compounds for 7 y at the Jungfraujoch station (3,580 ma.s.l.) and showed that the maximum of total suspended particles (TSP) mass corresponds toperiods with highest temperatures, but they did not find any year to year trend. The EU-CARBOSOL project (Present and Retrospective State of Organic versus Inorganic Aerosol overEurope: Implications for Climate; Legrand and Puxbaum 2007) provided very important resultsconcerning the variability of aerosol chemistry at various high altitude sites across Europe. Theyobserved that organic matter concentrations were as high as the ones of total inorganic solublecompounds and that both had a seasonal behaviour with a summer maximum.

The present paper aims at presenting results by specifically focusing i) on the variabilityof aerosol composition, ii) on the size-dependant chemistry, and iii) on the link betweenmeteorological factors (air mass origin) and chemical composition at a mountain siterepresentative of a large footprint area (Asmi et al. 2011). Therefore, this paper reportsaerosol size-segregated chemistry and investigates the source origin of the aerosolcomponents.

48 J Atmos Chem (2012) 69:47–66

2 Experimental

2.1 Sampling site and method

Measurements were conducted at the altitude site puy de Dôme (pdD) summit(1,465 ma.s.l., 45° 46! 20" N, 2° 57! 57" E) . A detailed description of the samplingsite can be found in a number of previous studies (Wobrock et al. 2001; Venzac et al.2009). The station is located in a natural background area, restricted during winter,authorized to cars during the spring and autumn periods, and only to organized busesduring the summer period. Previous analysis showed limited impact of emissions fromboth the city of Clermont-Ferrand and local traffic on chemical measurements at thesite (Sellegri et al. 2003). Therefore, this site can be classified as a rural backgroundsite according to Putaud et al. (2004) and high altitude site (z>1,000 ma.s.l.)according to Asmi et al. (2011). According to modelling results (Fig. 1) and LIDARresults (Boulon et al. 2011), the pdD often lies in the mixed layer during the day andin the residual layer during the night during summer. During the winter, the station islocated above the mixed layer.

Meteorological parameters (temperature, wind speed and direction, and relative humidity- RH) are monitored permanently since October 1995, showing temperatures varyingbetween !20 and 10°C during winter and between 5 and 20°C during summer. The sitewas often used for investigations of cloud chemistry and aerosol-cloud interactions (Sellegriet al. 2003; Marinoni et al. 2004), due to the frequent formation of clouds (more than 50 % ofthe time between November and March).

Aerosol sampling inlets (1 m above the station roof, i.e. >2.5 m a.g.l.) were designed tosample during clear sky conditions as well as during cloudy conditions using a whole airinlet (WAI; Sellegri 2002) which ensured efficient sampling of both cloud droplets residuesand interstitial aerosol in the presence of clouds. Wind speed was decreased in the vicinity ofthe WAI by a series of metallic fences ensuring efficient sampling even at elevated windspeed. Based upon theoretical considerations, the WAI is capable of efficiently (50 %efficiency) sampling droplets <35 !m for wind speed <10 ms!1. During cloudy conditions,interstitial aerosols and cloud droplets residues were sampled simultaneously at a RH closeto 50 % by mean of heating coil.

Size segregated sampling was performed for a full year (April 2006–April 2007) with a30 lpm 13-stages low pressure cascade impactor (Dekati type) with cut-off aerodynamicdiameters of 0.03, 0.06, 0.10, 0.17, 0.26, 0.40, 0.65, 1.02, 1.65, 2.51, 4.07, 6.73 and10.42 !m. The different modes have been calculated by summing the impactor’s stages.

0

200

400

600

800

1000

1200

1400

1600

1800

0 3 6 9 12 15 18 21 24

hours

hmix

ed la

yer (

m a

.s.l.

)

DJFMAMJJASONpdD height

Fig. 1 Mixed layer height at pdDcalculated with data fromECMWF for the year 2007

J Atmos Chem (2012) 69:47–66 49

The PM0.1 mode is constituted by particles with a diameter comprised between 30 nm and105 nm, the PM1-0.2 mode is composed by particles with a diameter comprised between 105 nmand 1.02 !m and the PM10-1 mode by particles which diameter is comprised between 1.02 !mand 10.42 !m. Sampling time integration varied between 3 and 4 d and was performed on aweekly basis for a full year. The collection plates were custom-made out of aluminium foil andanalysed for ionic species. Field blank were collected on a weekly basis.

2.2 Chemical analysis

Collection plates from the impactor were handled and extracted under a class 100 laminarflow hood. Before sampling, the aluminium foil were washed with MeOH and rinsed withMilli-Q water. Then, the collection plates were dried in a class 1,000 clean room (T022°C±1; "P0+ 60 Pa; RH050 %±10; Clermont-Ferrand, France). Before and after sampling, thecollection plates were weighed, using a microbalance UMT2 Mettler Toledo, after 24 h inthe clean room in order to reach the equilibrium temperature and RH. After weighing, thecollection plates were analyzed with the chromatographic method described below.

The water soluble inorganic ions (WSII) were analyzed by ion chromatography using themethodology described by Ricard et al. (2002). Impaction plates were extracted in theirstorage bottle for about 10 min using 10 ml of Milli-Q water and manual agitation. Extractwere filtrated with the filter caps which have 20 !m porosity. Cations (sodium, ammonium,potassium, magnesium, and calcium) were analysed with a Dionex ICS-1500 chromato-graph, using a CS16 column, a CG16 guard column and chemical regeneration was madewith a CSRS ULTRA II autosuppressor and a 0.20 % MSA eluent. Concentrations of majorwater soluble anions (chloride, nitrate, sulphate and oxalate) were determined with a DionexIC25 chromatograph, using an AS11 column, an AG11 guard column and an ASRS ULTRAII autosuppressor. Injection was performed using a KOH gradient and an EG40 eluentgenerator. Anions and cations were injected in parallel with an AS40 automated samplerwith an injection loop of 750 !L.

The measurement of NO3- is problematic and may be affected by large uncertainties

(Chang et al. 2001). On the one hand, there is the possibility of a positive artefact by theadsorption of gaseous HNO3 onto the filter material, which will be then accounted as NO3

-

(Ten Brink et al. 2009). However, Weber et al. (2003) found that this may not be a significantartefact. On the other hand, NO3

- can be evaporated in the form of semi volatile NH4NO3. Itis worthy to note that there is some NO3

- losses during impactor sampling due to lowpressures in the instrument, but they are generally lower than those observed with filtersamplers (Zhang and McMurry 1987, 1992; Wang and John 1988). For impactor sampling,Wang and John (1988) determined wall losses of NO3

- of 1.1 % for particles larger than2 !m, of 3.0 % for particles between 0.5 and 2 !m and of 8.1 % for particles smaller than0.5 !m. They also estimated evaporation losses up to 7 %, in agreement with Wall et al.(1988). Concerning the storage before weighing, Dougle and Ten Brink (1996) found nosignificant loss of NO3

- at 0°C and 20°C over a period of several days and a small loss ofweigh (within 10 %) at room temperature during 2 months. In the following, NO3

- data willbe corrected only from the field blank.

The arithmetic averages of concentrations for the field blanks were subtracted fromsample concentrations. The contribution of field blanks varied according to the chemicalspecies, the highest contribution was observed for the compounds which had the smallestconcentrations such as Cl- and Na+ was 3.6 and 4.6 % respectively. When concentrations arelower than the detection limit (see Table 1), the level was set at the detection limit for thecalculation of the means, which are therefore upper limits. Atmospheric concentrations

50 J Atmos Chem (2012) 69:47–66

sampled at the atmospheric pressure of the pdD (around 850 hPa) were normalized to thestandard conditions of pressure (1,013 hPa).

3 Results

3.1 Time variation of the weighed mass

The average monthly mass (Fig. 2) is calculated by averaging the 3 or 4 weeklyvalues obtained with the sum of the weighed masses of all impactor stages of a givenrun. The PM10 mass is calculated with the sum of the weighed mass from stage 1 tostage 12 and the PM1 is the sum of the weighed masses of stages 1 to 7, the concentrationvalues are given in Table 2.

We observe a maximum during summer and a minimum during winter, which is slightlydifferent from what has been observed in the mixed layer in Southern Europe. Bergametti etal. (1989) and Querol et al. (1998) showed a maximum in spring and a minimum in autumn.This behaviour was not observed everywhere in mixing layer stations. In some places, amaximum was observed during winter and a minimum during summer, presumably due to alarger impact of combustion sources during winter and the “trapping” of these particles in athinner mixed layer (Mészáros et al. 1997; Röösli et al. 2001; Calvo et al. 2008; Vecchi et al.2009). At high altitude stations such as the Jungfraujoch (3,454 ma.s.l.), Switzerland(Baltensperger et al. 1991) or in Manali (2,050 ma.s.l.), India (Gajananda et al. 2005), asummer maximum in aerosol mass was usually observed.

The PM1 concentration is higher than observed at Jungfraujoch (1.7 !g m!3 and2.5 !g m!3 in winter and summer respectively, Cozic et al. 2008). We observe a maximumin July and a minimum in January for the sub-micron particles, as for the PM10 mass. ThePM10-1 concentration is of the same order of magnitude than at the Jungfraujoch (2.4 !g m!3

and 2.0 !g m!3 in winter and summer respectively, Cozic et al. 2008) with little seasonality.Hence, the seasonal behaviour of the PM10 concentrations is mainly driven by the PM1

concentration seasonal behaviour.

Table 1 Detection limit of the ion chromatography (!g)

Cl- NO3- SO4

2- C2O42- Na+ NH4

+ K+ Mg2+ Ca2+

0.02 0.08 0.3 0.2 0.01 0.007 0.1 0.01 0.02

0

5

10

15

20

Apr

-06

may jun jul

aug

sep

oct

nov

dec

jan

feb

mar

Apr

-07

mas

s co

nc. (

µg m

-3)

0

25

50

75

100

mas

s fr

actio

n (%

)

mass_sub mass_supersub_fraction

Fig. 2 Monthly mean concentra-tion of the weighed mass of thesub-micron and the super-micronaerosol particles at the puy deDôme (histogram) on the left axis.The black dotted line representsthe submicron mass fraction, onthe right axis

J Atmos Chem (2012) 69:47–66 51

In fact, the PM1 fraction represents the dominant fraction of the PM10 at pdD (64±12 %),as opposed to sites closer to natural primary emissions sources (Van Dingenen et al. 2004).This is particularly the case in spring-summer, when the PM1/PM10 reaches about 80 %.This seasonal behaviour is similar to those observed on particle number concentrations, withsummer particle number concentrations being 4 times higher than winter particle numberconcentrations (Venzac et al. 2009). Venzac et al. (2009) noticed the importance of theatmospheric vertical dynamics on the seasonal variability of the aerosol number concentration.This feature could also be partly attributed to higher photochemical reactions (Lee et al. 2001)and higher particles formation caused by higher solar radiation (Birmili and Wiedensohler2000) in summer, as the observed actinic flux experiences a strong maximum in summer.

3.2 WSII time series

The concentrations are given in Table 3. The PM1 WSII presents a minimum in winter and amaximum in summer. As for the PM1, the PM10 WSII mass is higher in summer and lower inwinter. Again, the seasonal variation of WSII follows an opposite seasonal trend at mixedlayer stations in Hungary (Mészáros et al. 1997), but a similar trend at high altitude sitessuch as the Sonnblick (3,106 ma.s.l.), Austria (Kasper and Puxbaum 1998), at the Jung-fraujoch (3,580 ma.s.l.), Switzerland (Cozic et al. 2008) and at the Schauinsland (1,205 ma.s.l.), Germany (Salvador et al. 2010). The PM10 WSII winter concentration is lower thanthose found by Sellegri et al. (2003) at pdD during the winters 2000 and 2001. Within theCARBOSOL project, Pio et al. (2007) measured PM10 WSII average concentrations atpdD station of 3.7 !g m!3 in summer and 2.2 !g m!3 in winter, comparable to theWSII that wereport.

Table 2 Mean concentration of the weighed mass (!g m!3)

year DJF MAM JJA SON

PM10 5.6±4.6 2.5 7.0 7.8 5.8

PM1 3.9±3.4 1.3 5.3 6.0 3.2

PM10-1 1.7±1.9 1.2 1.7 1.9 2.5

Table 3 Mean concentration of PM1 and PM10-1 compounds for the whole year and each season (ng m!3)

year DJF MAM JJA SON

PM10 WSII 2,400±1,900 1,100 2,800 3,500 2,000

PM1 WSII 1,800±1,600 700 2,200 2,700 1,300

PM1 NO3- 260±710 60 820 90 40

PM1 SO42- 1,200±920 420 940 2,200 960

PM1 NH4+ 280±210 120 360 390 230

PM1 und. mass 2,200±2,000 670 3,100 3,300 2,000

PM10-1 NO3- 250±270 160 290 220 320

PM10-1 SO42- 180±240 60 120 40 10

PM10-1 NH4+ 17±33 1 40 20 5

PM10-1 Na++Cl- 140±180 190 120 80 190

PM10-1 und. mass 1,100±1,500 720 1,000 1,100 1,800

52 J Atmos Chem (2012) 69:47–66

In the following section, we will analyze the aerosol composition in more details in orderto further understand what drives the seasonal variability of the chemical composition(Fig. 3).

3.3 Detailed chemical compositions

3.3.1 PM1

Nitrate concentrations show a minimum in autumn and a maximum in spring. Nitrate isderived from precursors (e.g. HNO3, N2O5) emitted by anthropogenic sources and cancondense on the surface of particles (Tolocka et al. 2004; Anlauf et al. 2006). In the finefraction, it is linked with ammonia (Putaud et al. 2004). Such spring maximum has beenattributed to mountain breeze at Montseny (Pey et al. 2010), although the configuration ofthe pdD mountain chain does not favour the formation of mountain breezes. It could also beattributed to regional pollution episodes (Pérez et al. 2008). Differently, for SO4

2- and NH4+,

the minimum is observed in winter and the maximum is observed in summer, which is inagreement with the PM10 mass seasonal variation, linked to the seasonal variation of themixed layer height. The high levels of SO4

2- during summer have been described byRodriguez et al. (2002), they pointed out the importance of photochemical processes.

We calculated the ratio between analysed cations and anions following the season. Asshown by Fig. 4, the cation-to-anion ratio is smaller for particles with an aerodynamicdiameter comprised between 0.2 !m and 1 !m and increasing when going to either larger orsmaller particle sizes. This feature is the same of the one observed at remote and rural sites inNorthern Europe (Kerminen et al. 2001). Except in summer, the averaged cation-to-anion

Winter

0%

25%

50%

75%

100%

0.03

0.06 0.1

0.17

0.26 0.4

0.65

1.02

1.65

2.51

4.07

6.73

mas

sfr

actio

nm

ass

frac

tion

mas

sfr

actio

nm

ass

frac

tion

0

1

2

3

mas

sco

nc.

(dM

/dlo

gDp)

Spring

0%

25%

50%

75%

100%

0.03

0.06 0.1

0.17

0.26 0.4

0.65

1.02

1.65

2.51

4.07

6.73

0

3

6

9

12

mas

sco

nc.

(dM

/dlo

gDp)

Summer

0%

25%

50%

75%

100%

0.03

0.06 0.1

0.17

0.26 0.4

0.65

1.02

1.65

2.51

4.07

6.73

Dp (µm) Dp (µm)

Dp (µm) Dp (µm)

0

3

6

9

12

mas

sco

nc.

(dM

/dlo

gDp)

Autumn

0%

25%

50%

75%

100%

0.03

0.06 0.1

0.17

0.26 0.4

0.65

1.02

1.65

2.51

4.07

6.73

0

3

6

9

12

mas

sco

nc.

(dM

/dlo

gDp)

NO3 SO4 Na+Cl NH4 Ca

other OM OM+min WSII w eighted

Fig. 3 Size distributions of the aerosol particles at the puy de Dôme sampled according to the season. The massfraction is represented by histogram (left axis). TheWSII analyzed (black dotted line) and the weighedmass (blackplain line) are on the right axis. The compounds presented here are nitrate, sulphate, sea salt (Na+ + Cl-),ammonium and calcium. The mention “other” refers to all compounds analyzed with the chromatographic methodbut not detailed in the study, OM is the difference betweenweighed and analysedmass in the sub-micron sizes andOM+ min is the difference between weighed and analysed mass in the super-micron sizes

J Atmos Chem (2012) 69:47–66 53

ratio is close to unity. During summer, there was a large amount of sulphate, making ourionic balance not satisfied and these aerosol particles quite acidic.

The mean NO3-/SO4

2- ratio is 0.3, which is similar to what was observed at MonteCimone, high altitude site (2,165 ma.s.l.; Marenco et al. 2006). We observe a high seasonalvariability of the mass ratio between NO3

-/SO42-; it is higher during winter and spring (0.1

and 0.9, respectively) and lower during summer and autumn (0.03 for both).We find an ubiquitous significant fraction of undetermined mass. The influence of water

present on impaction foil is estimated assuming a particle hygroscopic growth factor at 50 %RH. Most of hygroscopic particles (sea salt, ammonium sulphate, ammonium nitrate) showgrowth factors equal to 1 at 50 % RH (Hansson et al. 1998; Krämer et al. 2000; Cruz andPandis 2000; Hämeri et al. 2000; Mikhailov et al. 2004). Carbonaceous particles also have ahygroscopic growth factor of 1 (Weingartner et al. 1997). In addition to that, Putaud et al.(2004) showed that gravimetric measurements at 50 % RH led to PM10 and PM2.5,respectively, 1.09 and 1.07 times as high as gravimetric measurements at 20 % RH andwater should not contribute to the weighed mass at 50 % RH. Moreover, the main metaloxides concentrations previously detected by PIXE in pdD samples in the past were Al2O3,SiO3, and FeO, and very little contribution from other metal oxides were found. Altogether,the contribution of these three metal oxides of crustal origin were always less than 4 % of thetotal mass of the aerosol, except in strong Saharan dust events in which it reached 26 % ofthe PM10-1 mode mass (Sellegri et al. 2003). Thus, based on the results from Sellegri et al.(2003) who achieved a mass balance closure within 15 % on cascade impactor samples, wetentatively attribute the undetermined mass (weighed but not analyzed) to EC + OM in thePM1. In favour of the hypothesis, we found that the oxalate concentration correlatessignificantly with the undetermined mass with R2 comprised 0.88 and 0.99. The concentra-tion of the undetermined mass is higher during the summer and lower in winter. A spring-summer maximum and the autumn-winter minimum have been observed for OM as well atSchauinsland, Germany (1,205 ma.s.l.) and at Sonnblick, Autria (3,106 ma.s.l.) (Pio et al.2007) with summer concentrations being 2 to 3 times higher than the winter concentrations.We calculate concentrations lower than those reported for CARBOSOL at pdD (Pio et al.2007), possibly due to year to year variability but also expected because of different cut-off(PM10 for CARBOSOL).

The composition of the PM0.1 range has been the focus of several studies, as it representsthe most numerous fraction of particles acting as CCN (Raga and Jonas 1993; Martin et al.1994; Khlystov et al. 1997, 1998). The PM0.1 has a chemical composition dominated byundetermined mass (Table 4). The SO4

2- is the predominant WSII compounds, the NH4+ and

NO3- contributions being low. Even if the proportions of each compound are slightly

different between the PM1 and PM0.1, the same compounds are predominant in both modes.

0

1

2

0.01 0.1 1 10

AEDca

tion-

to-a

nion

rat

io

winter

spring

summer

autumn

Fig. 4 Cation-to-anion ratioaccording to the season and im-pactor stages

54 J Atmos Chem (2012) 69:47–66

We notice that the behaviour of the PM1 is mainly driven by the undetermined mass andthen by SO4

2-, NO3- and NH4

+, in agreement with what is usually observed at altitudestations and mixed layer sites. We do not observe a strong seasonality in the relativecomposition of the PM1. This observation favours the hypothesis of the mixed layerdynamics being the main parameter driving the aerosol concentrations at the site.

3.3.2 PM10-1

Nitrate is a compound which can be found both in the PM1 and in the PM10-1 (Putaud et al.2004). A clear seasonal pattern is observed with a higher contribution of the PM10-1 NO3

- tothe PM10 NO3

- in the autumn-winter period (Fig. 5) even if the concentration of PM10-1 NO3-

is higher in autumn but lower in winter. The autumn maximum could be attributed to thehigh number of samples originating from Mediterranean region (see later, section III.5.1). Inthe PM10-1, it is bound with alkaline sea salt and/or mineral dust particles (Harrison and Kitto1990; Pakkanen et al. 1996; Hanke et al. 2003; Krueger et al. 2004; Sullivan et al. 2007). Thereaction of HNO3 to sea salt has been pointed out in several publications (Robbins et al. 1959;Martens et al. 1973; Harrison and Pio 1983; Pakkanen et al. 1996; Plate and Schulz 1997;Vignati et al. 2001; Sørensen et al. 2005), resulting in a loss of Cl- relative to the Cl-/Na+ ratio inthe sea composition. The minimum of PM10-1 NO3

- in winter is consistent with the fact thatthere is no loss of Cl- during this season (NaCl molar ratio of 1.05, which is very close to what isdescribed for sea water i.e. 1.2 (Wilson 1975)).

The PM10-1 SO42- exhibits a spring maximum and a winter minimum. The mean NO3

-/SO4

2- ratio is 2, which is higher than what was observed for rural sites in Spain, between 0.4and 0.7 (Rodriguez et al. 2004; Viana et al. 2008; Pey et al. 2009) and higher than the one weobserved for PM1.

The NH4+ maximum is in spring and the minimum in winter.

The winter sea salt (Na+ + Cl-) average concentration is twice as high as the ones foundby Sellegri et al. (2003) at the pdD during winter. The average sea salt concentrations ofevery seasons at pdD are 10 times higher than the ones measured (for PM10) by Cozic et al.(2008), which is expected as the PdD is farther west compared to the Jungfraujoch. We

Table 4 Mass fraction of mainWSII and undetermined mass (%) NO3

- SO42- NH4

+ Und. mass

PM0.1 <4 14–70 4–7 28–80

0%

25%

50%

75%

100%

april

may

june july

augu

st

sept

embe

r

octo

ber

nove

mbe

r

dece

mbe

r

janu

ary

febr

uary

mar

ch

april

2006 2007

mas

sfr

actio

n(%

)

accumulation coarseFig. 5 Monthly mean of NO3-

mass fraction for PM1 andPM10-1

J Atmos Chem (2012) 69:47–66 55

observe that sea salt has a higher concentration in winter and autumn. A sea salt massfraction maximum in winter was already observed by Jaffe et al. (2005) for sites locatedinland from the coast by 200 km, who attributed this maximum to higher wind speed duringthis season. Indeed, Yan et al. (2002) observed highest wind speed during winter and lowestwind speed during summer over the Atlantic Ocean.

As previously discussed, we find a significant fraction of undetermined mass. Since theundetermined mass correlates significantly with C2O4

2- in spring (R200.85) and in winter,summer and autumn (R200.5) and with nssCa2+ (0.3<R2<0.7), we hypothesize this fractionto be mostly composed by OM and minerals. The undetermined mass mean concentrationdoes not present specific seasonal behaviour. It is largely predominant since it represents 58to 71 % of the PM10-1.

3.3.3 Size

Size distributions (Fig. 3) globally show PM1with amodal diameter at 0.4!m,which is also thesame for each individual major water soluble inorganic compound (NO3

-, SO42- andNH4

+), andPM10-1 with a modal diameter at 1.6!m.We observe that the mass ratio between PM1/PM10-1 ishighly variable according to the season, together with the composition of the PM10-1.

Moreover, we observe that the intra-seasonal variability of the aerosol is higher than theinter-seasonal variability, for both PM1 and PM10-1. Within a single season, we still need toestablish which factor is determining aerosol concentration and composition. We will in thefollowing section focus on the effect of air mass type on the aerosol composition at pdD.

4 Average chemical composition in various air mass types

4.1 Classification of air masses

The goal of this classification is to identify air masses with a specific chemical signature. Toestablish this classification, 5 d back-trajectories were calculated using the HYSPLITtransport and dispersion model and/or READY website (http://www.arl.noaa.gov/ready.html) (Draxler and Rolph 2003; Rolph 2003) for the sampling period of the impactor(ca. 3 d on average). The back-trajectories were calculated at 00:00 UTC and 12:00 UTC andtheir end point was centered at the pdD (45°46!20"N, 2°57!57"E, 1,465 ma.s.l.). The airmasses were classified following their origin (Fig. 6) if they were homogeneous (consideringthe geographical origin within the concerned sector, not the meteorological conditionsduring transport or sampling) during the three calculation days. If a change of air mass typeoccurred, they were classified as undefined.

These sectors were selected depending on the main aerosols sources. Marine air massesoriginated from the Atlantic Ocean and arrived at pdD from the west coast of France.Marine-modified air masses also originated from the Atlantic Ocean, but crossing a popu-lated area of North-Western Europe (England, Denmark, etc.) before reaching France andthen pdD. Continental air masses originated from areas located to the east of the site.Mediterranean air masses originated from south of the site, often from North Africa or theMediterranean Sea.

We report the number of samples performed in each air mass type according to the season(DJF for winter, MAM for spring, JJA for summer and SON for autumn) (Fig. 7). We cannotice that air mass types are not evenly spread over the seasons, which might haveinfluenced the seasonal variability observed in the previous sections. The Mediterranean

56 J Atmos Chem (2012) 69:47–66

sector comprises most of the air mass types sampled during autumn, which might have animpact on the aerosol mean composition during this season in our data set. The massconcentrations are given in Table 5.

4.2 PM1

The highest weighed mass on cascade impactor samples is measured for the continental samples,then for the mediterranean samples and last for the marine modified and marine samples.

Concerning concentrations, the major WSII and the undetermined mass concentrationspresent a maximum for the continental origin. In these air masses, SO4

2- is the predominantWSII compound, then NO3

- and NH4+. In the other sectors, the SO4

2- and NH4+ concentrations

are decreased by a factor comprised from 1.5 to 4.9. For NO3-, the decrease is much more

important, the factor is comprised between 5 and 22. The concentration of Ca2+ is the same forthe marine, continental and mediterranean samples whereas it is lower for the marine modifiedair mass. Because dust source is more important in the Mediterranean region, higher calciumconcentration was expected. It has to be kept in mind that our Ca2+ is describing only the water

Mediterranean

Marine

Marine modified

Continental

Fig. 6 Sectors of backwardtrajectories

0

1

2

3

4

5

6

7

8

winter spring summer autumn

Num

ber

of s

ampl

es

marine

marine modif ied

continental

mediterranean

undefined

Fig. 7 Number of samplesfor each air mass origin accordingto the season

J Atmos Chem (2012) 69:47–66 57

soluble part (about 80 % of the total Ca according to Koçak et al. 2007). It is mainly producedby the heterogeneous reactions between insoluble Ca2+ compound in mineral dust and acidicspecies in the atmosphere, which are affected by many factors (Clarke and Karani 1992).

Table 5 Mass concentrations (ng m!3) according to air mass origin

marine Marine mod. continental mediterranean

PM1 weighed 1,700 3,200 8,200 3,900

PM1 WSII 650 1,400 3,500 1,700

PM1 NO3- 50 180 890 40

PM1 SO42- 450 910 2,000 1,300

PM1 NH4+ 110 250 540 300

PM1 Na++Cl- 30 20 20 20

PM1 Ca2+ 4 1 4 4

PM1 und. mass 1,000 1,800 4,600 2,200

PM10-1 weighed 820 1,100 2,100 3,800

PM10-1 WSII 370 610 690 980

PM10-1 NO3- 90 300 230 420

PM10-1 SO42- 50 120 300 180

PM10-1 NH4+ 2 8 50 20

PM10-1 Na++Cl- 190 140 30 200

PM10-1 Ca2+ 10 20 60 120

PM10-1 und. mass 450 450 1,400 2,800

Marine

0%

25%

50%

75%

100%

0.03

0.06 0.1

0.17

0.26 0.4

0.65

1.02

1.65

2.51

4.07

6.73

mas

sfr

actio

nm

ass

frac

tion

mas

sfr

actio

nm

ass

frac

tion

0

1

2

3

mas

sco

nc.

(dM

/dlo

gDp)

Marine modified

0%

25%

50%

75%

100%

0.03

0.06 0.1

0.17

0.26 0.4

0.65

1.02

1.65

2.51

4.07

6.73

0

2.5

5

7.5

10

mas

sco

nc.

(dM

/dlo

gDp)

Continental

0%

25%

50%

75%

100%

0.03

0.06 0.1

0.17

0.26 0.4

0.65

1.02

1.65

2.51

4.07

6.73

Dp (µm) Dp (µm)

Dp (µm) Dp (µm)

0

5

10

15

20

mas

sco

nc.

(dM

/dlo

gDp)

Mediterranean

0%

25%

50%

75%

100%

0.03

0.06 0.1

0.17

0.26 0.4

0.65

1.02

1.65

2.51

4.07

6.73

0

2.5

5

7.5

10

mas

sco

nc.

(dM

/dlo

gDp)

NO3 SO4 Na+Cl NH4 Ca

other OM OM+min WSII w eighted

Fig. 8 Size distributions of the aerosol particles at the puy de Dôme sampled in air masses of different origins.The mass fraction is represented by histogram (left axis). The WSII analyzed (black dotted line) and the weighedmass (black plain line) are on the right axis. The compounds presented here are nitrate, sulphate, sea salt (Na+ + Cl-),ammonium and calcium. The mention “other” refers to all compounds analyzed with the chromatographic methodbut not detailed in the study, OM is the difference between weighed and analysed mass in the sub-micron sizes andOM+ min is the difference between weighed and analysed mass in the super-micron sizes

58 J Atmos Chem (2012) 69:47–66

Consequently, those values are not representative of the total Ca2+. The sea salt has a higherconcentration in the marine air mass and is constant in the other air masses.

The PM1, mostly representative of the anthropogenic contribution (Kittelson et al. 2004;Tunved et al. 2006) shows a variable concentration but relatively homogeneous composition(WSII vs undetermined mass) (Fig. 8). In all air masses, the undetermined mass representsthe biggest contribution of the mass fraction (Table 6). Within the WSII, despite the variationof the NO3

-/SO42- ratio (between 0.03 and 0.4), SO4

2- is still the predominant compound andNO3

- is the lowest anthropogenic compounds.Even if the aerosol sources are expected to differ greatly from one sector to another, we

found little difference in the chemical composition of the PM1 fractions in the various airmass types, as observed during seasonal study. At pdD, the PM1 is likely aged, and hencemore homogeneous whatever the air mass type.

4.3 PM10-1

Regarding the PM10-1 mass concentration, the maximum is found for the Mediterranean airmass, then the continental air mass and last, as for the PM1, the marine modified and themarine air masses.

The WSII and the undetermined mass concentrations are highest for the mediterraneanorigin. In detail, SO4

2- and NH4+ have highest concentrations in continental air masses, as

observed for the PM1 whereas the NO3- has maximum concentration for the mediterranean air

masses. The concentration of Ca2+ is the highest in mediterranean samples and then in thecontinental samples. It is the lowest in marine and marine modified samples. The highest seasalt concentrations are found in the marine and mediterranean air masses. Those concentrationsare slightly lower in the modified marine air mass and are minor in the continental air mass.

The PM10-1 concentrations and compositions differ significantly when sampled in one airmass type or the other (Fig. 8). As already mentioned, the presence of NO3

- on the PM10-1

sea salt can be attributed to reaction of NO3- precursors (e.g., HNO3 and N2O5) with sea salt

aerosol Cl- (Harrison and Kitto 1990; Mamane and Gottlieb 1992; Anlauf et al. 2006). Infact, contrarily to what we observed for PM1, PM10-1 NO3

- is the predominant WSII in theparticles sampled in the marine - marine modified air masses sampled at pdD. The pdD issituated 320 km from the oceanic coast; hence this result indicates the level of modificationof sea salt after it has aged in polluted plumes. Indeed, shorter is the distance between oceanand pdD, smaller is the NO3

-/Na+ ratio (Table 7), in agreement with the observations madeby Anlauf et al. (2006). Cavalli et al. (2004) sampled at the coastal western Irish site of MaceHead, and found that NO3

- mass contributes to 1 % on the PM10-1 during spring and autumn,

Table 6 Mass fraction (%) according the air mass origin

WSII NO3- SO4

2- NH4+ Na++Cl- Ca2+ Und. mass

PM1 39–44 1–11 25–34 7–8 <2 <1 56–61

PM10-1 26–57 11–28 5–14 <2 2–23 1–3 43–74

Table 7 Nitrate-to-sodium ratioin the PM10-1 mode marine Marine mod. continental mediterranean

NO3-/Na+ 0.5 1.0 3.7 1.4

J Atmos Chem (2012) 69:47–66 59

which is more than 10 times lower than our measurements. Regarding sea salt, we find that itcan represent up to 23 % of the PM10-1 mass (the highest mass fraction is observed for themarine samples). At Mace Head, it represents 94 % in spring and autumn (Cavalli et al.2004). In the Mediterranean aerosol, a large fraction of the mass is not WSII, and we suspectthat mineral dust is largely contributing to the unaccounted mass, as witnessed by the largestconcentrations (but not mass fraction which is not significantly higher compared to other airmass types) of Ca2+ found in these air masses. This is in agreement with the observations ofPey et al. (2009); they observed high concentrations of crustal material and low concen-trations of OM + EC and NH4NO3 in the Mediterranean aerosol composition. The SO4

2- andNH4

+ concentrations, which are predominant together with the NO3- in the PM1, are of

minor importance in the PM10-1. In the past, SO42- in this mode has been associated

Table 8 Chemical composition of the marine aerosol, calculated from impactor data (mean value + standarddeviation). The values are given in mass concentrations (ng m!3)

NO3- SO4

2- NH4+ Na++Cl- Ca2+ Und. mass Weighed mass

PM0.1 Winter 1±1 5±1 1±1 1±0 2±2 56±11 66±10

Spring 2±2 11±7 2±2 1±1 0±0 89±52 110±53

Summer 3 26 7 3 0 120 160

Autumn 1±0 15±9 2±1 3±1 0±0 110±24 130±14

PM1-0.1 Winter 12±10 160±97 51±35 45±59 9±12 190±200 470±300

Spring 96±150 300±160 87±37 19±16 1±1 1,300±1,100 1,800±1,300

Summer 53 2,400 460 28 1 3,400 6,300

Autumn 12±5 290±75 80±9 35±28 1±0 430±140 860±120

PM10-1 Winter 68±45 37±25 1±1 250±290 14±15 430±350 820±560

Spring 62±37 37±18 0±0 120±150 3±2 380±160 620±370

Summer 240 190 21 110 20 890 1,500

Autumn 140±38 41±27 0±0 270±240 12±10 280±120 770±330

Table 9 Chemical composition of the marine modified aerosol, calculated from impactor data (mean value +standard deviation). The values are given in mass concentrations (ng m!3)

NO3- SO4

2- NH4+ Na++Cl- Ca2+ Und. mass Weighed mass

PM0.1 Winter 6±6 19±10 7±4 2±1 0±0 112±70 150±80

Spring 12±9 46±30 23±6 4±2 0±0 204±45 290±69

Summer 3±1 180±80 18±13 3±1 0±0 0 190±150

Autumn 4±2 25±16 9±6 1±0 0±0 187±19 230±14

PM1-0.1 Winter 130±150 570±290 180±80 12±7 1±0 910±560 1,800±890

Spring 430±420 880±640 300±60 34±13 2±1 2,900±1,600 4,500±2,600

Summer 28±14 1,500±470 290±90 19±2 1±0 1,300±460 3,100±1,500

Autumn 45±40 700±610 180±130 14±12 1±0 1,500±770 2,400±1,500

PM10-1 Winter 190±80 44±18 3±4 110±50 14±13 250±130 630±240

Spring 480±450 100±80 25±22 240±220 20±12 680±400 1,600±1,000

Summer 400±30 360±260 3±3 150±8 15±2 180±41 1,100±190

Autumn 160±120 55±46 1±1 66±42 18±17 560±420 870±640

60 J Atmos Chem (2012) 69:47–66

with sea salt aerosols (Harrison and Pio 1983; Anlauf et al. 2006); it can also comefrom dust particles (Wall et al. 1988; Matsuki et al. 2005). It has been shown by modellingstudy that the sulphate-deposition rate onto the PM10-1 particles is slow in comparisonwith that onto the PM1 particles (Song and Carmichael 1999). This would explain whySO4

2- predominates in the PM1 mode and not in the PM10-1.We observed that the size-segregated chemical composition of the aerosol is varying with

the origin of the air mass as one would expect from the sources associated with each air masstrajectory but still, the variability within an air mass is as important as within a season andwe can conclude that both the season and air mass type should be taken into account in orderto reduce our uncertainty in attributing an aerosol chemical composition to given environ-mental conditions.

As a result, we calculate the chemical composition of aerosol as a function of the air masstype and the season (Tables 8, 9, 10 and 11) (refer to Fig. 7 for the number of samples).These tables sum up our previous observations.

Overall, we find that the seasonal variation of major WSII compounds showing amaximum during spring-summer is still observed within a given air mass type. This is alsothe case for the undetermined mass. For sea salt and Ca2+ concentrations, the seasonalbehaviour is not as clear for a given air mass as for the whole data set. This implies that theseasonal variation observed on the whole data set might be driven by the seasonal variationof the air mass types, which shows a higher proportion of mediterranean air masses duringautumn and a higher proportion of continental air masses in spring-summer.

Table 10 Chemical composition of the continental aerosol, calculated from impactor data (mean value +standard deviation). The values are given in mass concentrations (ng m!3)

NO3- SO4

2- NH4+ Na++Cl- Ca2+ Und. mass Weighed

mass

PM0.1 Spring 8±6 27±14 18±5 1±1 2±2 190±100 250±110

Summer 4±1 190±300 19±14 3±3 1±1 95±120 310±180

PM1-

0.1

Spring 1,800±1,500 1,600±1,300 640±300 12±9 4±4 4,500±2,100 8,600±4,800

Summer 130±150 2,200±740 430±90 25±6 3±4 4,500±2,300 7,300±3,000

PM10-1 Spring 390±440 210±200 80±74 26±17 96±120 1,800±1,600 2,600±2,500

Summer 110±80 370±400 23±16 35±18 32±31 1,100±780 1,700±1,000

Table 11 Chemical composition of the Mediterranean aerosol, calculated from impactor data (mean value +standard deviation). The values are given in mass concentrations (ng m!3)

NO3- SO4

2- NH4+ Na++Cl- Ca2+ Und. mass Weighed mass

PM0.1 Winter 2 32 13 1 0 101 150

Summer 5±2 130±100 13±9 2±0 0±0

Autumn 3±1 27±21 9±9 1±1 0±0 110±62 160±83

PM1-0.1 Winter 17 900 200 16 7 1,200 2,400

Summer 68±22 2,300±650 420±71 36±11 15±10

Autumn 37±40 1,200±470 270±96 19±12 4±4 2,400±1,200 4,000±1,500

PM10-1 Winter 390 160 1 220 200 3,200 4,200

Summer 300±210 720±300 41±28 85±37 120±78

Autumn 440±350 140±84 10±12 210±230 120±170 2,800±2,300 3,700±3,000

J Atmos Chem (2012) 69:47–66 61

5 Conclusion

The measurements conducted at pdD (1,465 ma.s.l.), France, aimed to a better understand-ing of the influence of the season and air mass origin on the size-segregated chemicalcomposition of the aerosol. Aerosols were sampled using low pressure cascade impactor.The samples were weighed and analysed for water soluble inorganic compounds.

The seasonal variability of the PM1 at pdD showed a maximum in the summer. This canbe attributed to 1) a seasonal variability in the mixed layer/free troposphere stratification, 2)a seasonal variation of the aerosol sources. In Venzac et al. (2009), it was shown that theatmospheric vertical dynamics was playing an important role in the seasonal variability ofthe aerosol number concentration. The absence of seasonal variation of the compoundsproportions means that the seasonal variation of sources is not contributing as much to andthis confirms that the atmospheric vertical dynamic is a dominant factor influencing theaerosol concentrations at the site.

The PM10-1 exhibited a different seasonal behaviour compared to the PM1, with higherconcentrations in the cold season (winter and autumn). During winter, air mass backtrajectories are originating from more distant areas, implicating that the average wind speedalong the trajectories are higher and larger aerosols such as sea salt can be transported at highaltitudes over larger distances.

Aerosols were sampled in different air masses, which could be classified into four differentcategories according to their origin: marine, marine modified, continental and Mediterranean.The PM10 mass concentration at 50 % relative humidity was close to 2.5 !g m!3 in the marine,4.3 !g m!3 in the marine modified, 10.3 !g m!3 in the continental and 7.7 !g m!3 in theMediterranean sectors. We noted that the influence of the air mass origin (on the chemicalproperties of the aerosol) could be seen after several days especially on the PM10-1. The sizedistribution of the aerosol varied according to the air mass type. A significant PM10-1 mode wasfound in marine, modified marine, andMediterranean air masses, and a dominant PM1 mode incontinental air masses. The main WSII constituent of the PM1 was SO4

2- whatever the air massorigin and for the PM10-1, NO3

- dominated the aerosol composition in all air mass types exceptfor marine air masses in which sea salt dominated the PM10-1.

As a result, the aerosol chemical composition variability at the puy de Dôme is a function ofboth the season and air mass type and we provide a statistically relevant chemical compositionof the aerosol as a function of each of these environmental factors. These chemical character-istics can be used as inputs in future aerosol-cloud process modelling exercises.

Acknowledgment The authors gratefully acknowledge the NOAA Air Resources Laboratory (ARL) for theprovision of the HYSPLIT transport and dispersion model and/or READY website (http://www.arl.noaa.gov/ready.html) used in this publication. This work has been funded and supported by the Agence gouvernemen-tale De l’Environnement et de la Maîtrise de l’Energie (ADEME).

References

Andrews, E., Ogren, J.A., Bonasoni, P., Marinoni, A., Cuevas, E., Rodríguez, S., Sun, J.Y., Jaffe, D.A., Fisher,E.V., Baltensperger, U., Weingartner, E., Collaud Coen, M., Sharma, S., Macdonald, A.M., Leaitch, W.R.,Lin, N.-H., Laj, P., Arsov, T., Kalapov, I., Jefferson, A., Sheridan, P.: Climatology of aerosol radiativeproperties in the free troposphere. Atmos Res 102, 365–393 (2011)

Anlauf, K., Li, S.-M., Leaitch, R., Brook, J., Hayden, K., Toom-Sauntry, D., Wiebe, A.: Ionic composition andsize characteristics of particles in the Lower Fraser Valley: Pacific 2001 field study. Atmos Environ 40,2662–2675 (2006)

62 J Atmos Chem (2012) 69:47–66

Asmi, A., Wiedensohler, A., Laj, P., Fjaeraa, A.-M., Sellegri, K., Birmili, W., Weingartner, E., Baltensperger,U., Zdimal, V., Zikova, N., Putaud, J.-P., Marinoni, A., Tunved, P., Hansson, H.-C., Fiebig, M., Kivekäs,N., Lihavainen, H., Asmi, E., Ulevicius, V., Aalto, P.P., Swietlicki, E., Kristensson, A., Mihalopoulos, N.,Kalivitis, N., Kalapov, I., Kiss, G., de Leeuw, G., Henzing, B., Harrison, R.M., Beddows, D., O’Dowd,C., Jennings, S.G., Flentje, H., Weinhold, K., Meinhardt, F., Ries, L., Kulmala, M.: Number sizedistributions and seasonality of submicron particles in Europe 2008–2009. Atmos Chem Phys 11,5505–5538 (2011)

Baltensperger, U., Gäggeler, H.W., Jost, D.T., Emenegger, M., Nägeli, W.: Continuous background aerosolmonitoring with the Epiphaniometer. Atmos Environ 25A, 629–634 (1991)

Bergametti, G., Dutot, A.-L., Buat-Ménard, P., Losno, R., Remoudaki, E.: Seasonal variability of theelemental composition of atmospheric aerosol particles over the northwestern Mediterranean. Tellus41B, 353–361 (1989)

Birmili, W., Wiedensohler, A.: New particle formation in the continental boundary layer: Meteorological andgas phase parameter influence. Geophys Res Lett 27, 3325–3328 (2000)

Boulon, J., Sellegri, K., Hervo, M., Picard, D., Pichon, J.-M., Fréville, P., Laj, P.: Investigation of nucleationevents vertical extent: a long term study at two different altitude sites. Atmos Chem Phys 11, 5625–5639(2011)

Calvo, A.I., Pont, V., Liousse, C., Dupré, B., Mariscal, A., Zouiten, C., Gardrat, E., Castera, P., Lacaux, C.G.,Castro, A., Fraile, R.: Chemical composition of urban aerosols in Toulouse, France during CAPITOULexperiment. Meteorol Atmos Phys 102, 307–323 (2008)

Cavalli, F., Facchini, M.C., Decesari, S., Mircea, M., Emblico, L., Fuzzi, S., Ceburnis, D., Yoon, Y.J.,O’Dowd, C.D., Putaud, J.-P., Dell’Acqua, A.: Advances in characterization of size-resolved organicmatter in marine aerosol over the North Atlantic. J Geophys Res (2004). doi:10.1029/2004JD005137

Chang, M., Sioutas, C., Cassee, F.R., Fokkens, P.H.B.: Field evaluation of a mobile high-capacity particle sizeclassifier (HCPSC) for separate collection of coarse, fine and ultrafine particles. J Aerosol Sci 32, 139–156 (2001)

Clarke, A.G., Karani, G.N.: Characterisation of the carbonate content of atmospheric aerosols. J Atmos Chem14, 119–128 (1992)

Cozic, J., Verheggen, B., Weingartner, E., Crosier, J., Bower, K.N., Flynn, M., Coe, H., Henning, S.,Steinbacher, M., Henne, S., Collaud Coen, M., Petzold, A., Baltensperger, U.: Chemical compositionof free tropospheric aerosol for PM1 and coarse mode at the high alpine site Jungfraujoch. Atmos ChemPhys 8, 407–423 (2008)

Cruz, C.N., Pandis, S.N.: Deliquescence and hygroscopic growth of mixed inorganic_organic atmosphericaerosol. Environ Sci Technol 34, 4313–4316 (2000)

Dougle, P.G., Ten Brink, H.M.: Evaporative losses of ammonium nitrate in nephelometry and impactormeasurements. J Aerosol Sci 27(S1), 511–512 (1996)

Draxler, R.R., Rolph, G.D.: HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory) Modelaccess via NOAA ARL READY Website (http://www.arl.noaa.gov/ready/hysplit4.html). NOAA AirResources Laboratory, Silver Spring, MD (2003)

Gajananda, K., Kuniyal, J.C., Momin, G.A., Rao, P.S.P., Safai, P.D., Tiwari, S., Ali, K.: Trend of atmosphericaerosols over the north western Himalayan region, India. Atmos Environ 39, 4817–4825 (2005)

Hämeri, K., Väkevä, M., Hansson, H.-C., Laaksonen, A.: Hygroscopic growth of the ultrafine ammoniumsulphate aerosol measured using an ultrafine tandem differential mobility analyzer. J Geophys Res 105,22231–22242 (2000)

Hanke, M., Umann, B., Uecker, J., Arnold, F., Bunz, H.: Atmospheric measurements of gas-phase NHO3 andSO2 using chemical ionization mass spectrometry during the MINATROC field campaign 2000 on MonteCimone. Atmos Chem Phys 3, 417–436 (2003)

Hansson, H.-C., Rood, M.J., Koloutsou-Vakakis, S., Hämeri, K., Orsini, D., Wiedensohler, A.: NaCl aerosolparticle hygroscopicity dependence on mixing with organic compounds. J Atmos Chem 31, 321–346(1998)

Harrison, R.M., Kitto, A.-M.N.: Field intercomparison of filter pack and denuder sampling methods forreactive gaseous and particulate pollutants. Atmos Environ 24A, 2633–2640 (1990)

Harrison, R.M., Pio, C.A.: Size differentiated composition of inorganic atmospheric aerosols of both marineand polluted continental origin. Atmos Environ 17, 1733–1738 (1983)

Heintzenberg, J., Müller, K., Birmili, W., Spindler, G., Wiedensohler, A.: Mass-related aerosol properties overthe Leipzig basin. J Geophys Res 103, 13125–13135 (1998)

Jaffe, D., Tamura, S., Harris, J.: Seasonal cycle and composition of background fine particles along the westcoast of the US. Atmos Environ 39, 297–306 (2005)

Kasper, A., Puxbaum, H.: Seasonal variation of SO2, HNO3, NH3 and selected aerosol components atSonnblick (3106 ma.s.l.). Atmos Environ 32, 3925–3939 (1998)

J Atmos Chem (2012) 69:47–66 63

Kerminen, V.-M., Hillamo, R., Teinilä, K., Pakkanen, T., Allegrini, I., Sparani, R.: Ion balances of size-resolvedtropospheric aerosol samples: implications for the acidity and atmopsheric processing of aerosols. AtmosEnviron 35, 5255–5265 (2001)

Khlystov, A., Kos, G.P.A., ten Brink, H.M., Kruisz, C., Berner, A.: Activation properties of ambient aerosol inThe Netherlands. Atmos Environ 30, 3281–3290 (1997)

Khlystov, A.: Cloud forming properties of ambient aerosol in the Netherlands. PhD thesis, University ofWageningen (1998)

Kittelson, D.B., Watts, W.F., Johnson, J.P.: Nanoparticle emissions on Minnesota highways. Atmos Environ38, 9–19 (2004)

Koçak, M., Mihalopoulos, N., Kubilay, N.: Chemical composition of the fine and coarse mode fraction ofaerosols in the northeastern Mediterranean. Atmos Environ 41, 7351–7368 (2007)

Krämer, L., Pöschl, U., Niessner, R.: Microstructural rearrangement of sodium chloride condensation aerosolparticles on interaction with water vapour. J Aerosol Sci 31(6), 673–685 (2000)

Krueger, B.J., Grassian, V.H., Cowin, J.P., Laskin, A.: Heterogeneous chemistry of individual mineral dustparticles from different dust source regions: the importance of particle mineralogy. Atmos Environ 38,6253–6261 (2004)

Laj, P., Klausen, J., Bilde, M., Plaß-Duelmer, C., Pappalardo, G., Clerbaux, C., Baltensperger, U., Hjorth, J.,Simpson, D., Reimann, S., Coheur, P.-F., Richter, A., De Mazière, M., Rudich, Y., McFiggans, G.,Torseth, K., Wiedensohler, A., Morin, S., Schulz, M., Allan, J.D., Attié, J.-L., Barnes, I., Birmili, W.,Cammas, J.P., Dommen, J., Dorn, H.-P., Fowler, D., Fuzzi, S., Glasius, M., Granier, C., Hermann, M.,Isaksen, I.S.A., Kinne, S., Koren, I., Madonna, F., Maione, M., Massling, A., Moehler, O., Mona, L.,Monks, P.S., Müller, D., Müller, T., Orphal, J., Peuch, V.-H., Stratmann, F., Tanré, D., Tyndall, G., AboRiziq, A., Van Roozendael, M., Villani, P., Wehner, B., Wex, H., Zardini, A.A.: Measuring atmosphericcomposition change. Atmos Environ 43, 5351–5414 (2009)

Lee, S.C., Ho, K.F., Chan, L.Y., Zielinska, B., Chow, J.C.: Polycyclic aromatic hydrocarbons (PAHs) andcarbonyl compounds in urban atmosphere of Hong Kong. Atmos Environ 35, 5949–5960 (2001)

Legrand, M., Puxbaum, H.: Introduction of the CARBOSOL project: Present and Retrospective State ofOrganic versus Inorganic Aerosol over Europe. J Geophys Res (2007). doi:10.1029/2006JD008271

Lestari, P., Oskouie, A.K., Noll, K.E.: Size distribution and dry deposition of particulate mass, sulfate andnitrate in an urban area. Atmos Environ 37, 2507–2516 (2003)

Mamane, Y., Gottlieb, J.: Nitrate formation on sea-salt and mineral particles – a single particle approach.Atmos Environ 26A, 1763–1769 (1992)

Marenco, F., Bonasoni, P., Calzolari, F., Ceriani, M., Chiari, M., Cristofanelli, P., D’Alessandro, A., Fermo, P.,Lucarelli, F., Mazzei, F., Nava, S., Piazzalunga, A., Prati, P., Valli, G., Vecchi, R.: Characterization ofatmospheric aerosols at Monte Cimone, Italy, during summer 2004: Source apportionment and transportmechanisms. J Geophys Res (2006). doi:10.1029/2006JD007145

Marinoni, A., Laj, P., Sellegri, K., Mailhot, G.: Cloud chemistry at the Puy de Dôme: variability and relationshipswith environmental factors. Atmos Chem Phys 4(3), 715–728 (2004)

Martens, C.S., Wesolowski, J.J., Harriss, R.C., Kaifer, R.: Chlorine loss from Puerto Rican and San Franciscobay area marine aerosols. J Geophys Res 78, 8778–8791 (1973)

Martin, G.M., Johnson, D.W., Spice, A.: The measurement and parameterization of effective radius of dropletsin warm stratiform clouds. J Atmos Sci 51, 1823–1842 (1994)

Matsuki, A., Iwasaka, Y., Shi, G., Zhang, D., Trochkine, D., Yamada, M., Kim, Y.-S., Chen, B., Nagatani, T.,Miyazawa, T., Nagatani, M., and Nakata, H.: Morphological and chemical modification of mineral dust:observational insight into the heterogeneous uptake of acidic gases. Geophys. Res. Lett. 32 (2005).doi:10.1029/2005GL024176

Mészáros, E., Barcza, T., Gelencsér, A., Hlavay, J., Kiss, Gy, Krivácsy, Z., Molnár, A., Polyák, K.: Sizedistributions of inorganic and organic species in the atmospheric aerosol in Hungary. J Aerosol Sci 28(7),1163–1175 (1997)

Mikhailov, E., Vlasenko, S., Niessner, R., Oöschl, U.: Interaction of aerosol particles composed of protein andsaltswith water vapor: hygroscopic growth and microstructural rearrangement. Atmos Chem Phys 4, 323–350 (2004)

Pakkanen, T.A., Kerminen, V.-M., Hillamo, R.E., Mäkinen, M., Mäkelä, T., Virkulla, A.: Distribution ofnitrate over sea-salt and soil derived particles: implications from a field study. J Atmos Chem 24, 189–205(1996)

Pérez, N., Pey, J., Castillo, S., Viana, M., Alastuey, A., Querol, X.: Interpretation of the variability of levels ofregional background aerosols in the Western Mediterranean. Sci Total Environ 407, 527–540 (2008)

Pey, J., Pérez, N., Castillo, S., Viana, M., Moreno, T., Pandolfi, M., López-Sebastián, J.M., Alastuey, A.,Querol, X.: Geochemistry of regional background aerosols in the Western Mediterranean. Atmos Res 94,422–435 (2009)

64 J Atmos Chem (2012) 69:47–66

Pey, J., Pérez, N., Querol, X., Alastuey, A., Cusack, M., Reche, C.: Intense winter atmospheric pollutionepisodes affecting the Western Mediterranean. Sci Total Environ 408, 1951–1959 (2010)

Pio, C.A., Legrand, M., Oliveira, T., Afonso, J., Santos, C., Caseiro, A., Fialho, P., Barata, F., Puxbaum, H.,Sanchez-Ochoa, A., Kasper-Giebl, A., Gelencsér, A., Preunkert, S., Schock, M.: Climatology of aerosolcomposition (organic versus inorganic) at nonurban sites on a west–east transect across Europe. JGeophys Res (2007). doi:10.1029/2006JD008038

Plate, E., Schulz, M.: Coarse nitrate formation in a coastal area of the North Sea. J Aerosol Sci 28, 333–334(1997)

Putaud, J.-P., Raes, F., Van Dingenen, R., Brüggemann, E., Facchini, M.C., Decesari, S., Fuzzi, S., Gehrig, R.,Hüglin, C., Laj, P., Lorbeer, G., Maenhaut, W., Mihalopoulos, N., Müller, K., Querol, X., Rodriguez, S.,Schneider, J., Spindler, G., Ten Brink, H., Tørseth, K., Wiedensohler, A.: A European AerosolPhenomenology-2: chemical characteristics of particulate matter at kerbside, urban, rural and backgroundsites in Europe. Atmos Environ 38, 2579–2595 (2004)

Putaud, J.-P., Van Dingenen, R., Alastuey, A., Bauer, H., Birmili, W., Cyrys, J., Flentje, H., Fuzzi, S., Gehrig,R., Hansson, H.C., Harrison, R.M., Herrmann, H., Hitzenberger, R., Hüglin, C., Jones, A.M., Kasper-Giebl,A., Kiss, G., Kousa, A., Kuhlbusch, T.A.J., Löschau, G.,Maenhaut,W.,Molnar, A.,Moreno, T., Pekkanen, J.,Perrino, C., Pitz,M., Puxbaum, H., Querol, X., Rodriguez, S., Salma, I., Schwarz, J., Smolik, J., Schneider, J.,Spindler, G., Ten Brink, H., Tursic, J., Viana, M., Wiedensohler, A., Raes, F.: A European aerosol phenom-enology - 3: Physical and chemical characteristics of particulate matter from 60 rural, urban, and kerbside sitesacross Europe. Atmos Environ 44, 1308–1320 (2010)

Querol, X., Alastuey, A., Puicercus, J.A., Mantilla, E., Miro, J.V., Lopez-Soler, A., Plana, F., Artiñano, B.:Seasonal evolution of suspended particles around a large coal-fired power station: particulate levels andsources. Atmos Environ 32, 1963–1978 (1998)

Raga, G.B., Jonas, P.R.: On the link between cloud-top radiative properties and sub-cloud aerosol concen-trations. Quart J Roy Met Soc 119, 1419–1425 (1993)

Ricard, V., Jaffrezo, J.-L., Kerminen, V.M., Hillamo, R.E., Sillanpaa, S., Ruellan, S., Liousse, C., Cachier, H.:Two years of continuous aerosol measurements in northern Finland. J Geophys Res (2002). doi:10.1029/2001JD000952

Robbins, R.C., Cadle, R.D., Eckhardt, D.L.: The conversion of sodium chloride to hydrogen chloride in theatmosphere. J Meteorol 16, 53–56 (1959)

Rodriguez, S., Querol, X., Alastuey, A., Plana, F.: Sources and processes affecting levels and composition ofatmospheric aerosol in the western Mediterranean. J Geophys Res (2002). doi:10.1029/2001JD001488

Rodriguez, S., Querol, X., Alastuey, A., Alarcón, M.M., Mantilla, E., Ruiz, C.R.: Comparative PM10-PM2.5

source contribution study at rural, urban and industrial sites Turing PM episodes in Eastern Spain. SciTotal Environ 328, 95–113 (2004)

Rolph, G.D.: Real-time Environmental Applications and Display sYstem (READY) Website (http://www.arl.noaa.gov/ready/hysplit4.html), NOAA Air Resources Laboratory, Silver Spring, MD (2003)

Röösli, M., Theis, G., Künzli, N., Staehelin, J., Mathys, P., Oglesby, L., Camenzind, M., Braun-Fahrländer,Ch: Temporal and spatial variation of the chemical composition of PM10 at urban and rural sites in theBasel area, Switzerland. Atmos Environ 35, 3701–3713 (2001)

Saliba, N.A., Kouyoumdjian, H., Roumié, M.: Effect of local and long-range transport emissions on theelemental composition of PM10-2.5 and PM2.5 in Beirut. Atmos Environ 41, 6497–6509 (2007)

Salvador, P., Artiñano, B., Pio, C., Afonso, J., Legrand, M., Puxbaum, H., Hammer, S.: Evaluation of aerosolsources at European high altitude backgroung sites with trajectory statistical methods. Atmos Environ 44,2316–2329 (2010)

Sellegri, K.: Etude du processus d’activation des particules d’aérosol en gouttelettes de nuage: implications enchimie multiphases. PhD thesis, University Joseph Fourrier, Grenoble, France (2002)

Sellegri, K., Laj, P., Peron, F., Dupuy, R., Legrand, M., Preunkert, S., Putaud, J.-P., Cachier, H., Ghermandi,G.: Mass balance of free tropospheric aerosol at the puy de Dôme (France) in winter. J Geophys Res(2003). doi:10.1029/2002JD002747

Song, C.H., Carmichael, G.R.: The aging process of naturally emitted aerosol (sea salt and mineral aerosol)during long range transport. Atmos Environ 33, 2203–2218 (1999)

Sørensen, L.L., Pryor, S.C., de Leeuw, G., Schulz, M.: Flux divergence of nitric acid in the marineatmospheric surface layer. J Geophys Res (2005). doi:10.1029/2004JD005403

Sullivan, R.C., Guazzotti, S.A., Sodeman, D.A., Prather, K.A.: Direct observations of the atmosphericprocessing of Asian mineral dust. Atmos Chem Phys 7, 1213–1236 (2007)

Ten Brink, H., Otjes, R., Jongejan, P., Kos, G.: Monitoring of the ratio of nitrate to sulphate in size-segregatedsubmicron aerosol in the Netherlands. Atmos Res 92, 270–276 (2009)

Tolocka, M.P., Lake, D.A., Johnston, M.V., Wexler, A.S.: Ultrafine nitrate particle events in Baltimoreobserved by real-time single particle mass spectrometry. Atmos Environ 38, 3215–3223 (2004)

J Atmos Chem (2012) 69:47–66 65

Tunved, P., Hansson, H.-C., Kerminen, V.-M., Ström, J., Dal Maso, M., Lihavainen, H., Viisanen, Y., Aalto,P.P., Komppula, M., Kulmala, M.: High natural aerosol loading over Boreal Forests. Science 312(5771),261–263 (2006)

Van Dingenen, R., Raes, F., Putaud, J.-P., Baltensperger, U., Charron, A., Facchini, M.-C., Decesari, S., Fuzzi,S., Gehrig, R., Hansson, H.-C., Harrison, R.M., Hüglin, C., Jones, A.M., Laj, P., Lorbeer, G., Maenhaut,W., Palmgren, F., Querol, X., Rodriguez, S., Schneider, J., Ten Brink, H., Tunved, P., Tørseth, K.,Wehner, B., Weingartner, E., Wiedensohler, A., Wåhlin, P.: A European aerosol phenomenology – 1:physical characteristics of particulate matter at kerbside, urban, rural and background sites in Europe.Atmos Environ 38, 2561–2577 (2004)

Vecchi, R., Bernardoni, V., Fermo, P., Lucarelli, F., Mazzei, F., Nava, S., Prati, P., Piazzalunga, A., Valli, G.:4-hours resolution data to study PM10 in a “hot spot” area in Europe. Environ Monit Assess 154, 283–300 (2009)

Venzac, H., Sellegri, K., Villani, P., Picard, D., Laj, P.: Seasonal variation of aerosol size distributions in thefree troposphere and residual layer at the puy de Dôme station, France. Atmos Chem Phys 9(4), 1465–1478 (2009)

Viana, M., Querol, X., Ballester, F., Llop, S., Esplugues, A., Fernández Patier, R., García Dos Santos, S.,Herce, M.D.: Characterising exposure to PM aerosols for an epidemiological study. Atmos Environ 42,1552–1568 (2008)

Vignati, E., de Leeuw, G., Berkowicz, R.: Modeling coastal aerosol transport and effects of surf-producedaerosols on processes in the marine atmospheric boundary layer. J Geophys Res 106, 20 225–20 238(2001)

Wall, S.M., John, W., Ondo, J.L.: Measurement of aerosol size distributions for nitrate and major ionicspecies. Atmos Environ 22, 1649–1656 (1988)

Wang, H.-C., John, W.: Characteristics of the berner impactor for sampling inorganic ions. Aerosol Sci Tech 8,157–172 (1988)

Weber, R., Orsini, D., Duan, Y., Baumann, K., Kiang, C.S., Chameides, W., Lee, Y.N., Brechtel, F., Klotz, P.,Jongejan, P., Ten Brink, H.M., Slanina, J., Boring, C.B., Genfa, Z., Dasgupta, P., Hering, S., Stolzenburg,M., Dutcher, D.D., Edgerton, E., Hartsell, B., Solomon, P., Tanner, R.: Intercomparison of near real timemonitors of PM2.5 nitrate and sulphate at the U.S. Environmental Protection Agency Atlanta Supersite. JGeophys Res (2003). doi:10.1029/2001JD001220

Weingartner, E., Burtscher, H., Baltensperger, U.: Hygroscopic properties of carbon and diesel soot particles.Atmos Environ 31(15), 2311–2327 (1997)

Wilson, T.R.S.: Salinity and the major elements of sea-water. In: Riley, J.P., Skirrow, G. (eds.) Chemicaloceanography. Academic Press, New York (1975)

Wobrock, W., Flossmann, A.I., Monier, M., Pichon, J.-M., Cortez, L., Fournol, J.-F., Schwarzenböck, A.,Mertes, S., Heintzenberg, J., Laj, P., Orsi, G., Ricci, L., Fuzzi, S., Brink, H.T., Jongejan, P., Otjes, R.: TheCloud Ice Mountain Experiment (CIME) 1998: experiment overview and modelling of the microphysicalprocesses during the seeding by isentropic gas expansion. Atmos Res 58(4), 231–265 (2001)

Yan, Z., Bate, S., Chandler, R.E., Isham, V., Wheater, H.: An analysis of daily maximum wind speed inNorthwestern Europe using generalized linear models. J Climate 15, 2073–2088 (2002)

Zhang, X.Q., McMurry, P.H.: Theoretical analysis of evaporative losses from impactor and filter surfaces.Atmos Environ 21, 1717–1779 (1987)

Zhang, X.Q., McMurry, P.H.: Evaporation losses of fine particulate nitrates during sampling. Atmos Environ26, 3305–3312 (1992)

66 J Atmos Chem (2012) 69:47–66