Physicochemical characteristics of aerosols measured in the spring

Atmospheric Environment 54 (2012) 545e556
Contents lists available at SciVerse ScienceDirect
Atmospheric Environment
journal homepage: www.elsevier.com/locate/atmosenv
Physicochemical characteristics of aerosols measured in the spring time in the
Mediterranean coastal zone
J. Piazzola a, *, K. Sellegri b, L. Bourcier c, M. Mallet d, G. Tedeschi a, T. Missamou a
a
Aix-Marseille Univ, Mediterranean Institute of Oceanography, Marseille, France
LAMP, Blaise Pascal University, France
c
Institute for Reference Materials and Measurements, JRC, B-2440 Geel, Belgium
d
Université de Toulouse, UPS, LA, 31400 Toulouse, France
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 15 October 2011
Received in revised form
27 January 2012
Accepted 13 February 2012
Aerosol particles in coastal areas result from a complex mixing between sea-spray aerosols locally
generated at the sea surface by breaking waves and a continental component arising from natural and/or
anthropogenic sources. This paper presents physicochemical characterization of aerosols observed
during meteorological conditions characteristics of coastal areas. In particular, we study the influence of
sea-breezes and land-breezes as well as the fetch variation, which superpose on larger synoptic
conditions, on aerosol properties. This was achieved using a physical, chemical and optical analysis of the
aerosol data acquired in May 2007 on the French Mediterranean coast. The aerosol distributions were
measured using a TSI SMPS 3081 model and the chemical characterization was made using an Ion
Chromatography analysis (IC) and a thermo-optical technique. In addition, aerosol optical characteristics
were provided by aethalometer (absorption) and nephelometer (scattering) measurements. For low wind
speeds, we detect high aerosol number concentrations as well as high NO
3 and carbonaceous
compounds contributions, which are observed even when the aerosol is sampled in pure maritime air
masses. These results indicate that air masses are strongly impacted by pollution transported over the
Mediterranean. In addition, the combination of low wind speeds and land/sea-breezes lead to the
production of new ultrafine particle formation events that seem to take place over the sea before being
transported back to the coast. Under higher wind speed conditions, aerosol number and mass concentrations of smaller sizes are significantly lowered due to the dispersion of anthropogenic pollutants.
Optical measurements reveal that mean scattering and absorbing coefficients are about 15.2 Mm1 and
3.6 Mm1, respectively. Associated mean aerosol single scattering albedo is found to be about 0.87 and
0.94 (at 520 nm) for continental and maritime influences.
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Coastal aerosols
Anthropogenic compounds
1. Introduction
The role of atmospheric aerosols in climate change is wellrecognized (Intergovernmental Panel on Climate Change (IPCC),
2007), but the estimation of their impact remains an important
scientific challenge. In particular, the uncertainties related to
aerosol radiative forcings remain very large (IPCC, 2007). This is
due to the heterogeneous spatial and temporal distribution of
tropospheric aerosol particles, their different origins (natural or
anthropogenic), and their physical and chemical transformation in
the free troposphere. In marine areas, the sea-sprays generated
at the airesea interface by wave breaking represent a major
* Corresponding author. Tel.: þ33 (0) 494142082; fax: þ33 (0) 494142448.
E-mail address: piazzola@univ-tln.fr (J. Piazzola).
1352-2310/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.atmosenv.2012.02.057
component of the natural aerosol mass (Jaenicke, 1984; Andreae,
1995; Yoon et al., 2007; Piazzola et al., 2009) and therefore are
important in the Earth radiative budget (Laskin et al., 2003; Mallet
et al., 2003; Mulcahy et al., 2008). In addition, they have a significant influence on the coastal urban air quality (Knipping and
Dabdub, 2003) through their ability to have chemical and physical interactions with other aerosol species and gases. In addition,
sea-spray aerosols transport a large variety of organic matter to
ecosystems of estuaries (Paerl et al., 2000). The size of sea-sprays
can vary over a wide range (0.02e50 mm diameter). Moreover,
they are also very soluble and hygroscopic, and increase to twice
their size for a relative humidity varying from 50 to 100% (Tang,
1996; Fitzgerald, 1989). Sea-spray is made primarily of sodium
chloride (NaCl) and small amounts of sulphate, calcium and
potassium, but they can also contain significant amounts of
organic carbon. On the basis of measurements carried out over
546
J. Piazzola et al. / Atmospheric Environment 54 (2012) 545e556
the North Atlantic coastal area, recent studies have evidenced
a strong seasonal variation in the chemical aerosol properties
connected to biological oceanic activity (Cavalli et al., 2004;
O’Dowd et al., 2004). In particular, a dominant inorganic sea-salt
contribution to sub-micron aerosol particle mass was observed
during winter, while, during periods of high biological activity, the
inorganic sea-salt signal was progressively replaced by organic
matter and nss-sulphate. The organic component was mainly
attributed to bubble bursting processes, due to the predominantly
insoluble and surface active character of organic carbon in the
marine aerosol particles (Ceburnis et al., 2008). However, the
contribution of secondary sources to organic matter is still under
debate.
Over the Mediterranean basin, the atmospheric aerosol
represents a mixing between particles produced by natural
processes (of both continental and marine origin) and particles of
anthropogenic origin released in dense urban areas. As a consequence, the sea-spray aerosol is often strongly impacted by
anthropogenic emissions. In the Mediterranean Sea, the amount
of aqueous phase pollutants are expected to be much higher than
in the Atlantic due to important inputs from major continental
rivers such as the Rhône (France) or Pô (Italy). Pollution in coastal
waters combined with the high intensity of solar radiation in the
Mediterranean area can cause an excess in nutrients, as phosphate
and nitrate compounds resulting in eutrophication processes
throughout the year (Jamet et al., 2001). The size distribution of
marine aerosol particles has also been proved to experience
significant seasonal variations (Yoon et al., 2007), which could also
be due to the biological activity. During laboratory measurements,
surfactants present in artificial sea water also proved to modify
the aerosol size distribution (Sellegri et al., 2005). However, the
atmospheric aerosol composition over the western part of the
Mediterranean Sea has rarely been characterized away from large
urban areas.
In addition, coastal areas are characterized by the occurrence
of local meteorological processes, specifically induced by the
landesea transition, as the land breeze, sea breeze and the fetch
occurring for offshore wind conditions. A better knowledge of the
influence of these processes on the aerosol properties is of great
importance for improvement of the models dealing with the
aerosol transport and transformation in the atmosphere. The main
objectives of the present paper are to (1) determine the extent to
which marine aerosol is impacted by anthropogenic activities in
pure maritime air masses representative and (2) to study the role
of the local meteorological conditions typical of coastal zone, in
particular the sea breeze and land-breeze conditions on the aerosol
properties. In this latter case, a better knowledge of the mixing
between sea-sprays produced for short fetch and continental
sources is provided. The study area is located on the French
Mediterranean coast, which represents a densely populated zone
where large amounts of aerosol particles of anthropogenic origin
are expected. If a strong contribution of sea-spray aerosols was
noted even for offshore winds at short fetches (Piazzola and
Despiau, 1997a; Piazzola et al., 2002), pollutants are much higher,
even for long fetches, than in the Atlantic atmosphere. Previous
work conducted by Sellegri et al. (2001) then showed the
concomitant presence of sea-sprays and high levels of anthropogenic species such as nitrate.
This paper presents an analysis of aerosols measured in the
Mediterranean coast in May 2007 on the island of Porquerolles,
which is located at about 5 km from the coastline. The results deals
with the physicochemical and optical properties of Mediterranean
aerosols for sea breeze and land breeze as well as offshore wind
conditions. The data are then compared with aerosol concentrations measured in other coastal locations.
2. Field site and experiments
The experiments took place from the 5th May to the 29th May
2007. The present paper deals more particularly with the aerosol
data recorded from the 23rd to the 28th of May. The study area is
the Toulon-Hyères bay (Fig. 1) located on the French Riviera,
between 6.15 and 6.25 east longitude and at 43 north latitude.
Measurements of aerosol particle size distributions and meteorological parameters were performed at the extreme west point of
the island of Porquerolles (Fig. 1). Fig. 1 shows that the sea state
depends on the wind’s trajectory over water, i.e., the fetch. The
station is exposed to air masses from the open sea, which corresponds to infinite fetches as defined by the criterion applied for
fully developed sea conditions (Hsu, 1986), as well as to air masses
originating over the European mainland, with a very short fetch
(about 5 km), that represent continentally polluted conditions.
Fetch limited conditions occur for local wind directions varying
between 290 and 30 , whereas the 160e220 wind direction
interval generally corresponds to open waters (Fig. 1). In Fig. 2, the
wind vectors recorded during the sampling period are reported.
2.1. Sampling and analytical procedures
Meteorological data measured at the site were wind speed, wind
direction, air and sea temperature and relative humidity. Size
distributions of particles in the 10e430 nm diameter range were
monitored using an SMPS-TSI model 3081. The aerosol counters
and the meteorological sensors were located at heights of 22 and
30 m, respectively, above the sea surface. We did not monitor the
relative humidity in the SMPS. However, although aerosol was not
dried before measurement for cascade impactor samplings, it was
presumably dry for SMPS samples since data counting was performed at higher temperatures than ambient temperature. This
was shown to dry the aerosol to humidity lower than 50% even in
very humid environments, such as cloudy environments as it was
reported at the Puy de Dôme station (Venzac et al., 2009).
For chemical characterization, aerosols were sampled directly
in the atmosphere with two low pressure cascade impactors
(Dekati) from the 5th of May to the 10th of May 2007. The first
impactor was a 30 lpm 13-stages low pressure cascade impactor
which cut-off aerodynamic diameters were 0.03, 0.06, 0.108, 0.17,
0.26, 0.4, 0.65, 1, 1.6, 2.5, 4.4, 6.8 and 9.97 mm, which was dedicated
to the organic carbon (OC) and elemental carbon (EC) analysis
while the second impactor operated at 20 lpm with approximately
the same size cut diameters for subsequent analysis with Ion
Fig. 1. Detailed view of the study area. The black square shows the location of the
experimental station. The grey part of the circle shows the wind directions which
correspond to fetch limited conditions, while the empty portion of the circle shows the
wind direction interval denoting infinite fetch conditions.
J. Piazzola et al. / Atmospheric Environment 54 (2012) 545e556
Episode 1 23-25 May (average 15 min)
a
Wind direction(deg) and strenght (ms-1)
6
5
4
3
2
1
0
-1
-2
-3
-4
23/05 0:00
12:00
24/05 1:00
13:00
25/05 1:00
Date and time
Episode 2 25-27 May (average 15 min)
Wind direction(deg) and strenght (ms-1)
10
8
6
4
2
0
-2
-4
-6
25/05 8:00
5:00
26/05 14:00
18:00
27/05 0:00
6:00
27/05 12:00
Date and time
c
Episode 3 27-28 May (average 15 min)
Wind direction(deg) and strenght (ms-1)
20
15
10
5
0
-5
-10
27/05 16:00 21:00
28/05 0:00
5:00
8:00
12:00
16:00
21:00
28/05 0:00
Date and Time
Fig. 2. (a) Wind direction and strength from the 23rd to the 25th of May, i.e., episode 1
as referred in the text. (b) Wind direction and strength from the 25th to the
28th of May, i.e., episode 2 as referred in the text. (c) Wind direction and strength the
28/05/07, i.e., episode 3 as referred in the text.
Chromatography. Sampling duration varied from 4 to 24 h. The
collection plates were custom-made out of aluminium foil
for Ionic Chromatography analysis and quartz Whatman for the
OC/EC analysis. Blank levels were performed on a sample-tosample basis. Collection plates were handled and extracted
under a laminar flow hood. Before sampling, the aluminium foil
were washed with MeOH and rinsed with Milli-Q water. Then
the collection plates were dried in a class 1000 clean room
(T ¼ 22 C 1; DP ¼ þ60 Pa; RH ¼ 50% 10; Clermont Ferrand,
France). Before and after sampling, the collection plates were
weighted, using a microbalance UMT2 Metler Toledo, after 24 h in
547
the clean room in order to reach the equilibrium temperature and
relative humidity. Impaction plates were stored at 4 C before
analysis.
As we will discuss in Section 4, the measurement of NO
3 is
delicate and may be affected by large uncertainties (Chang et al.,
2001). A possible artefact can occur by the adsorption of gaseous
nitric acid onto the filter material, which will be then accounted as
nitrate (Ten Brink et al., 2009). On the other hand, nitrate can be
evaporated in the form of semi volatile ammonium nitrate and
there is some nitrate losses during impactor sampling due to low
pressures in the instrument. However, these losses are generally
lower than those observed with filter samplers (Zhang and
McMurry, 1992; Wang and John, 1988). Wang and John (1988)
estimated evaporation losses up to 7%. For impactor sampling, the
same authors determined wall losses of nitrate of 1.1% for particles
larger than 2 mm, of 3.0% for particles between 0.5 and 2 mm and of
8.1% for particles smaller than 0.5 mm. Thus, in the present results,
nitrate data has been corrected only from the filter blank.
The soluble inorganic compounds were analysed by IC. The
collection plates were extracted in their storage bottle for few
þ
2þ
minutes using 10 mL of Milli-Q water. Cations (Naþ, NHþ
4 , K , Mg
2þ
and Ca ) were analysed with a Dionex ICS-1500 chromatograph,
using a CS16 column, a CG16 guard column and chemical regeneration was made with a CSRS ULTRA II autosuppressor and
a 0.20% MSA eluent. Concentrations of major water-soluble anions
2
2
(Cl, NO
3 , SO4 and C2 O4 ) were determined with a Dionex IC25
chromatograph, using an AS11 column, an AG11 guard column
and an ASRS ULTRA II autosuppressor. Injection was performed
a KOH gradient and an EG40 eluant generator. Anions and cations
are injected in parallel with an AS40 automated sampler with the
injection loop of 750 mL. The collection plate blank arithmetic
averages were subtracted from sample concentrations. The
contribution of blank collection plates varied according to the
chemical species, the highest contribution observed for the less
concentration compounds such as chloride and sodium was 3.6
and 4.6% respectively. For samples under the detection limit, we
used the average blank value to rise up their detection levels.
The carbonaceous fraction (OC and EC) was determined by
a thermo-optical technique with correction of pyrolysis by laser
transmission (the TOT method) on a Sunset Lab analyser (Birch
and Cary, 1996). The temperature ramp uses four steps up to
870 C under pure helium for the quantification of OC (Organic
Carbon), and four steps up to 900 C with a mixture of 98% He þ 2%
O2 for EC (elemental carbon) (Aymoz et al., 2007).
For optical characterization, a Cimel sun photometer was
installed near the station which provided the aerosol optical
depth (AOD) at four wavelengths (440, 670, 870 and 1020 nm) and
the associated Angstrom coefficient (spectral dependence of AOD).
The absorption coefficient (babs) has been obtained using a multichannel Aethalometer (Model AE-31, Magee Scientific) at seven
wavelengths (370, 470, 520, 590, 660, 880 and 950 nm) using the
raw absorbance data following Weingartner et al. (2003). The
methodology used to estimate the aerosol absorbing coefficient is
detailed in Saha et al. (2008). Scattering coefficient of aerosols
(bscat) has been measured using a Nephelometer (Model M9004,
Ecotech) at 520 nm. Nephelometer draws the ambient air through
an inlet tube, which is then illuminated by a flash lamp and the
scattered light intensity is measured by a photo-multiplier tube.
Temperature and pressure measurements made in the scattering
chamber are used to calculate the scattering by air molecules,
which are then subtracted from the total scattering to get the
scattering due to aerosols. The instrument was calibrated regularly
using standard procedure, in which the nephelometer is adjusted to
read zero by passing particle free air and the span calibration was
done in the laboratory using CO2 gas. The scattering data were
548
J. Piazzola et al. / Atmospheric Environment 54 (2012) 545e556
corrected for truncation errors following the work of Heintzenberg
et al. (2006) for Ecotech Nephelometer.
2.2. Campaign overview
The present study focuses on three particular episodes, which
are characteristics of the general meteorology of the French
Mediterranean coast. In addition to the local wind speed and
direction as measured during the sampling period (Fig. 2aec),
numerical calculations of the air mass trajectory were performed
using the hysplit model (Draxler and Rolph, 2003), as shown in
Fig. 3aec. Furthermore, to help the data analysis (see Section 3), the
Mixed Layer height, as calculated using the hysplit model, was
reported in Fig. 4.
First, the so-called “episode 1” covering the period 23/05/07 to
25/05/07, 08:00 LT deals with typical coastal conditions with air
Fig. 3. Calculated air mass trajectories for (a) episode 1 (b) episode 2 and (c) episode 3.
J. Piazzola et al. / Atmospheric Environment 54 (2012) 545e556
549
Fig. 4. The mixed layer height calculated using the Hysplit model for the period covering the Porquerolles experiments.
masses coming from the north of Italy. This low wind speed period
is characterized by the occurrence of sea-breezes and land-breezes
which are superimposed on the synoptic conditions (Fig. 2a). In
particular, both the 23rd and 24th of May correspond to a very clear
diurnal variation with land breeze from 00:00 LT to 08:30 LT and
a sea breeze from 10:00 LT to 19:00 LT. Under these conditions, we
expect a strong continental background aerosol mixed with particles of marine origin.
Episode 2 covers the period from the 25 May 08:00 LT to the 27
May 2007 at 12:00 LT. This episode corresponds to pure marine air
mass conditions with winds coming from the Mediterranean Sea,
as shown in Fig. 3b. This corresponds to a local wind direction
varying from 220 to 270 in the study area. The fetch for these wind
directions can range from 100 km to unlimited fetch (open water
conditions). The 26th of May, Fig. 2b shows a local wind direction
shift from land to sea breeze also in the mid-morning (around
09:00 LT). Again, sea breeze and land breeze add to the synoptic air
mass path (Fig. 2b).
Episode 3 took place the 28th of May 2007 and deals with air
masses originate over the Rhône valley, northwest of the study
area, and locally correspond to a short 25 km fetch on the island of
Porquerolles (Fig. 2c). These conditions are typical of the study area,
and defined as “Mistral” conditions. They are associated to high
wind speed of northwestern origin (Fig. 3c). For episode 3, local
winds shift from northwest (00:00 LT to 12:00 LT) to southwest
(12:00 LT to 23:00 LT) around noon.
show little variations during the whole period of episode 1, and
more particularly the 24th May (Fig. 3a). Such variation has probably a minor influence on the aerosol concentration.
Fig. 6 shows averaged size distributions measured during land
and sea breeze conditions, respectively. Each aerosol size spectrum
corresponds to average of about one hundred distributions. In
addition, the aerosol size spectra were fitted using lognormal
distributions and their parameters are listed in Table 1. We can
note that the accumulation mode mean diameter and integrated
concentration are rather constant from one local wind condition to
the other at around 100 nm and 4500 cm3. The differences in the
land/sea breeze size distributions occur mainly in the particle burst
mode (less than 10 nm) which are more numerous during nighttime and on the grown nucleation mode (around 50 nm) which are
observed during the sea breeze period. As a consequence, the total
number concentrations do not vary substantially from night to day,
i.e., 10,150 and 9285 cm3, respectively.
A typical daily variation of aerosol size distributions recorded
during episode 2 (pure marine air mass conditions at low wind) is
shown in Fig. 7. Local contaminations are not observed in the
Mediterranean Sea sector air masses. Fig. 7 shows larger aerosol
concentrations during nighttime, which corresponds to periods
of local land-breezes, than during daytime which correspond to
3. The aerosol size distributions
This section deals with the aerosol size distributions measured
using the SMPS system for the three episodes as described in
Section 2. First of all, Fig. 5 shows daily variation of the aerosol size
distribution recorded during periods of low wind speed typical of
episode 1 (coastal conditions). During sea breeze or land breeze,
day or night, fine particle bursts were observed, corresponding
probably to coastal anthropogenic activities which take place
nearby on the urbanized coast. These local anthropogenic activities
are more frequent during the night/land breeze. The end of an
aerosol formation event, which is characterized by the growth of
ultrafine particles, might be visible in the afternoon. This could be
due to grown particles nucleating over the sea, which strongly
impact the daytime aerosol size spectrum. The mixed layer depth
Fig. 5. Daily variation of the aerosol size distribution recorded using the SMPS system
under light coastal winds typical of episode 1.
550
J. Piazzola et al. / Atmospheric Environment 54 (2012) 545e556
100000
dN/dlogD (cm-3)
10000
Fig. 7. Daily variation of the aerosol size distribution recorded using the SMPS system
for pure marine air mass conditions of episode 2.
1000
South at 10:00 LT to stay South until 18:30 LT. Average size distributions were reported in Fig. 8 for both the land and the sea breeze
periods. Again, the accumulation mode particles concentrations do
not vary much from night to daytime although their geometric
mean diameter is larger during nighttime (120 nm) compared to
daytime (105 nm). The Aitken mode particles concentrations
increase from 700 to 2700 cm3 between the daytime (sea breeze)
and nighttime (land breeze) conditions, and their geometric mean
diameter are also larger during nighttime (75 nm instead of 50 nm).
100
Episode 1 Land Breeze
Episode 1 Sea Breeze
10
0.001
100000
0.01
0.1
1
D (µm)
Fig. 6. Average size distributions measured during land and sea breeze conditions
occurring during episode 1.
10000
Table 1
Characteristics of lognormal distributions fitted on the experimental aerosol spectra
as recorded using the SMPS-TSI for each mode diameter Dp (in nanometre). The
number concentration, N (in cm3) and the standard deviation of the lognormal
function, s are reported.
Dp (nm)
Episode 1
Episode 2
Episode 3
N (cm3)
Episode 1
Episode 2
Episode 3
s
Episode 1
Episode 2
Episode 3
Land breeze
Sea breeze
Land breeze
Sea breeze
Land breeze
Sea breeze
Land breeze
Sea breeze
Land breeze
Sea breeze
Land breeze
Sea breeze
Land breeze
Sea breeze
Land breeze
Sea breeze
Land breeze
Sea breeze
Mode1
Mode2
Mode3
7
7
7
7
22
27
11,000
4000
2000
100
1200
155
1.65
1.65
1.65
1.65
1.65
1.5
25
50
71
50
65
65
2500
2800
2700
700
1100
125
1.75
1.3
1.3
1.5
1.5
1.5
100
110
120
105
145
170
4300
4800
3600
3000
200
40
1.65
1.6
1.6
1.6
1.5
1.5
dN/dlogD (cm-3)
cleaner southern sea breeze conditions. Indeed, total concentrations decrease from 8050 cm3 during the night to 3930 cm3
during the day. Such a variation in the aerosol concentration is
clearly related to the change in local wind direction from North to
1000
100
Episode 2 Land Breeze
Episode 2 Sea Breeze
10
0.001
0.01
0.1
1
D (µm)
Fig. 8. Average size distributions measured during land and sea breeze conditions
occurring during episode 2.
J. Piazzola et al. / Atmospheric Environment 54 (2012) 545e556
Fig. 9. Daily variation of the aerosol size distribution under high northwest wind
conditions, i.e., Mistral episode (episode 3).
10000
1000
dN/dlogD (cm-3)
A nucleation mode is observed in the nighttime average aerosol
size spectra but Fig. 7 shows that a new particle formation event
might have occurred during the morning local wind shift, i.e., at
about 9:00 LT.
Fig. 9 shows the variation of aerosol size distributions recorded
during episode 3, i.e., during high wind speeds from the northwest
sector, which correspond to a 25 km fetch. At 12:00 am, the wind
shifts from a northwest to a southwest direction. This induces
a sudden decrease of aerosol concentrations from 2450 to 320 particles cm3. Average size distributions are reported in Fig. 10 for land
and sea breeze conditions. The shift from land breeze to sea breeze is
more marked than in episode 1. During the land breeze, the aerosol
size distribution is bimodal with a first mode which mean diameter
is 22 nm and the second one at 65 nm. During the sea breeze
conditions, average modal diameters are found at 27 nm and 65 nm,
with an additional accumulation mode at 170 nm. The differences in
the aerosol size distributions reported in Fig. 10 are probably also
due to the variation of the Mixed Layer height between the night
and the daytime (Fig. 4). Indeed, we can see in Fig. 4 a strong
gradient of the ML height between the night and the day.
The particle size distributions measured during sea breeze
conditions of episode 1 and episode 2 can be compared to aerosol
data recorded at the coastal site of Finokalia located at the Eastern
Mediterranean region. and on board of the R/V “Aegaeon” by
Eleftheriadis et al. (2006) during the SUB-AERO experiment. These
measurements conducted during these studies showed the occurrence of monomodal particle size distributions centred on a mean
diameter ranging from 98 to 144 nm with a standard deviation
varying from 1.56 to 1.9. A large standard deviation on a single
mode could mask one or two modes, which would be more in
agreement with our results.
In term of size, the aerosols sampled on the island of Porquerolles during the Mistral conditions (episode 3) seems to be
substantially smaller than the aerosol sampled on board of the R/V
“Aegaeon”, but they showed similar mean diameters during the
Mediterranean conditions.
A comparison has been made with aerosol size spectra
measured at the coastal site of Mace Head, Ireland, which can be
designated as “clean marine” in character. In this case, particles in
the Aitken mode have modal median diameters at 31 nm and
49 nm during winter and summer respectively, in the accumulation
mode at 103 nm and 177 nm during winter and summer respectively. Hence, the aerosol sampled during the Mediterranean air
masses and sea breeze conditions also show similar geometric
mean diameters to the aerosols sampled in the Atlantic coastal area
during summer.
551
100
10
Episode 3 Land Breeze
Episode 3 Sea Breeze
1
0.001
0.01
0.1
1
D (µm)
Fig. 10. Average size distributions measured during land and sea breeze conditions
occurring, after episode 3.
The aerosol size distributions measured during the Porquerolles
campaign show that the number concentrations sampled during
episodes 1 and 2 are of the same order of magnitude measured in
the same study area by Piazzola and Despiau (1997a). During the
(cleaner) episode 3, total number concentrations are found to be of
the same order of magnitude than at Mace Head during the sea
breeze conditions. At Mace Head, total concentrations varied from
253 cm3 during winter to 469 cm3 during summer (Yoon et al.,
2007). However, number concentrations are one order of magnitude higher than in the Atlantic when the synoptic conditions are
Mediterranean (episode 2), and 20 times higher when air masses
pass over the French and Italian coast at low wind speed before
reaching the study area (episode 1). These high concentrations
are even higher than the previous measurements performed by
Eleftheriadis et al. (2006) during SUB-AERO experiment, who found
average concentrations of 1340 cm3 when air mass back trajectory
paths mainly remained over the Mediterranean Sea, and 3470 cm3
over the Aegean Sea when air mass back trajectories passed over
the southern European coastal zone. Our measurements are located
closer to the coast for both air mass trajectory type, compared to
those made during the SUB-AERO experiment. In this latter case,
the continental pollution is inclined to be less diluted and this
would explain the higher concentrations that we observed. In
addition, the high concentrations measured during episodes 1 and
2 are largely due to the nucleation mode particles, which were
not reported in the Eleftheriadis et al. (2006) study. Spatial and
temporal inhomogeneity between the two rather short data sets
(this study and the one of Eleftheriadis et al. (2006)) can also
552
J. Piazzola et al. / Atmospheric Environment 54 (2012) 545e556
explain the differences. However, the concentrations found in the
Mediterranean originating air masses under sea breeze conditions
should indicate a rather homogeneous background concentration
over the Mediterranean Sea at this period of the year, also found by
Eleftheriadis et al. (2006). This was already noted in the results
published on the basis of data recorded in the same study area
(Piazzola and Despiau, 1997a, 1997b).
4. Chemical analysis
The experimental measurements made in May 2007 on the
island of Porquerolles allowed for a survey of different anthropogenic and natural compounds of atmospheric aerosols for local
meteorological conditions of the Mediterranean coastal area.
Among 11 impactor samples, only those acquired during clear and
stable meteorological conditions described in Section 2 were
selected. The episode 1 deals with 3 samples, while episode 2 and
episode 3 correspond to 2 samples each. Total concentrations
recorded for major species are reported in Table 2. Size-segregated
percentage contribution of chemical species in aerosol sampled
during each episode (as defined in Section 2) is shown in Fig. 11. The
size distributions are plotted versus the aerodynamic diameter,
which is the diameter of a sphere of unit density that has the same
gravitational settling velocity. Sodium and chloride concentrations
show the highest concentrations for high northwest wind episode
(episode 3), despite the fact that it corresponds to short fetch
conditions, i.e., 25 km. Hence, we can conclude that a 25 km fetch is
enough to efficiently load the air mass with sea-sprays up to an
equilibrium determined by wind speed. The Naþ and Cl concentrations are comparable to those found by Sellegri et al. (2001)
during the FETCH experiment under high wind speed conditions
of southern sectors in aerosols sampled on board of a vessel sailing
off the French Mediterranean coast (2.37 < [Naþ] < 4.51 mg m3)
for larger fetch, i.e., approximately 80 km. Similar Naþ concentrations were found during summer and winter recordings at an
eastern Mediterranean coastal site by Bardouki et al. (2003), i.e.,
2.03 0.28 mg m3 in summer time and 1.20 18 mg m3 for winter
experiments.
2
Anthropogenic tracers such as NHþ
and NO
3 also
4 , nss-SO4
showed concentrations in the range of those measured during the
FETCH experiment, with however, slightly lower concentrations for
NHþ
4 in the present study. It seems that sulphate concentrations
measured at Porquerolles are rather low compared to those
acquired during eastern Mediterranean measurements in the
Aegean Sea by Eleftheriadis et al. (2006) and in the coastal site of
Finokalia by Bardouki et al. (2003), i.e., 10.12 1.10 mg m3 and
6.88 0.96 mg m3, respectively. While sulphate was the dominant
species in the sub-micron range during coastal conditions, as
expected from the most polluted air mass back trajectories, nitrate
was maximum for the Mediterranean air masses, which were
expected to be corresponding to the cleanest conditions (highest
fetches). This could be due to the fact that the fixation of nitrogen
species is favoured by the presence of sea-sprays. Indeed, in the
literature, nitrate has often been found in the super micron range
when sea-salt is present (Sellegri et al., 2001; Bardouki et al., 2003)
and hence suspected to originate from the reaction of NaCl with
HNO3 on the particles surface. We also found the most important
contribution of nitrate to be in the super micron range, and especially in the Mediterranean air masses. Under Mistral conditions,
we observe the smallest concentration of anthropogenic tracers,
which could be due to a strong dispersion of the pollutants typical
of large wind speed episodes (Piazzola and Despiau, 1997b).
The very low concentrations of particle mass in the sub-micron
mode confirm the SMPS size distribution showing a much cleaner
atmosphere. As already noted in Section 3, the measurement of
NO
3 is delicate and may be affected by large uncertainties (Chang
et al., 2001). Nitrate can be evaporated in the form of semi volatile ammonium nitrate and there is some nitrate losses during
impactor sampling due to low pressures in the instrument.
However, these losses are generally lower than those observed with
filter samplers (Zhang and McMurry, 1992; Wang and John, 1988)
and do not really change the shape of the aerosol size distributions
of the present paper.
The major contribution of the carbonaceous compounds is
found in the sub-micron mode (Fig. 11). Organics contribute by 39%,
14% and 34% to the particle mass for coastal (episode 1), Mediterranean (episode 2) and mistral conditions (episode 3), respectively.
The OC contribution in the Mediterranean air mass is similar to the
one observed in the Eastern Mediterranean for summer and winter
seasons (16% at Finokalia in the submicronic mode, Bardouki et al.,
2003). During episodes 1 and 3, the OC contribution was significantly higher and closer to the ones observed in the sub-micron
mode during the phytoplanctonic blooming period in the Northern
Atlantic coastal area (Cavalli et al., 2004), although in our case, OC is
likely rather composed of anthropogenic species. EC/OC ratio is
ranging from 0.02 for Mistral conditions (episode 3) to 0.07e0.08
for both coastal and Mediterranean conditions (episode 1 and
episode 2, respectively), which is about three times lower than the
ratios found by Bardouki et al. (2003). As the EC elements, Potassium is also emitted during biomass burning specifically in flaming
combustion (e.g., Leslie, 1981). We can note that Potassium
concentrations were also quite low and we can conclude that
biomass burning contributed little to the particulate matter during
the sampling period. This is rather logical since the forest fires in
the Mediterranean occur in the summer time (between June and
September) and can explain why their influence is low in May in
the study area.
5. The optical characteristics
The absorption coefficient (babs) has been obtained using
a multi-channel Aethalometer (Model AE-31, Magee Scientific) at
seven wavelengths (370, 470, 520, 590, 660, 880 and 950 nm) using
the raw absorbance data following Weingartner et al. (2003).
Scattering coefficient of aerosols (bscat) has been measured using
a Nephelometer (Model M9004, Ecotech) at 520 nm. Results on the
optical properties of different aerosol species were published in
Mallet et al. (2011). Fig. 12 reports the time evolution of the scattering coefficient during the experiment. Results display bscat values
comprised between 2 and 30 Mm1 (mean of 15.2 Mm1). Such
values are found to be logically lower than those obtained by Saha
et al. (2008) over the city of Toulon with higher bscat ranging
Table 2
Concentrations of major species (mg m3).
Episode 1
Episode 2
Episode 3
Cl
NO3
SO4
Oxalate
Na
NH4
K
Mg
Ca
OC
EC
0.84
1.84
5.14
1.52
3.48
0.24
3.1
2.67
1.18
0.17
0.37
0.02
0.79
1.79
2.99
0.48
0.21
0.08
0.13
0.16
0.12
0.13
0.35
0.41
0.3
1.78
0.1
4.97
2.07
5.37
0.33
0.17
0.13
J. Piazzola et al. / Atmospheric Environment 54 (2012) 545e556
553
Fig. 11. (a) Size-segregated percentage contribution of chemical species ( mg m3) in aerosol sampled during coastal air mass conditions (episode 1). Da is the aerodynamic diameter
(see Section 4). (b) Size-segregated percentage contribution of chemical species ( mg m3) in aerosol sampled during episode 2. (c) Size-segregated percentage contribution of
chemical species ( mg m3) in aerosol sampled during a high wind speed period of northwest direction (episode 3).
from 20 to 120 Mm1, due to the proximity of polluted
aerosol sources. Similar results are obtained for absorbing coefficients (Fig. 12) with lower values observed at Porquerolles
(1 < babs < 10 Mm1, associated mean of 3.6 Mm1) compared to
the Toulon site, i.e., 10 < babs < 50 Mm1 (Saha et al., 2008).
Our results show that scattering coefficients present higher
values (between 10 and 30 Mm1) during the 23rd to the 26th
period (episode 1 and beginning of episode 2), followed by a rapid
decrease due to a change in wind direction (South, East/South
wind) for episode 2 which begins the 26th May. Maritime clean air
masses characteristic of episode 2 exhibited scattering coefficients,
with values close to 5 Mm1. Low values of bscat are also observed
during and after the strong “mistral” case (28th), cleaning the lower
troposphere. Fig. 12 shows the associated aerosol single scattering
albedo (SSA) at 520 nm, our results indicate values comprised
between 0.80 and 0.90 (at 550 nm) for the period from the 23rd
to the 26th May. Such values reveal an absorbing aerosol in cases
of continental air masses generally associated with higher
554
J. Piazzola et al. / Atmospheric Environment 54 (2012) 545e556
Fig. 12. Time evolution of (a) the scattering coefficient (b) the absorbing coefficient and (c) the associated aerosol single scattering albedo measured during the whole campaign.
concentration of EC, as shown in Fig. 11a in Section 4. For oceanic air
masses, SSA values rapidly increase to reach up to 0.95e0.98, which
is consistent with the absence of continental polluted absorbing
aerosols.
6. Discussion and conclusion
The aim of this work was to provide a physicochemical analysis
of aerosols measured for meteorological conditions characteristics
of coastal areas. The present data, recorded in May 2007 on the
island of Porquerolles, confirm previous results obtained in the
study area in the nineties by Piazzola and Despiau (1997a). First of
all, both data sets are roughly in the same order of magnitude in
terms of aerosol concentration. In addition, in accordance with
Piazzola and Despiau (1997a), we noted a substantial contribution
of sea-salts in the aerosol concentrations measured during high
wind speed periods of Northwest direction (episode 3), which
corresponds to short fetch conditions. In addition, a rather constant
contribution of anthropogenic particles is observed for low wind
speed periods (episode 1 and episode 2). This shows that at high
wind speeds, the sea surface production is the dominant mechanism while low wind speed periods allow accumulation of high
continental aerosol concentrations, even for marine air, due to
a slower atmospheric dispersion. As a consequence, episode 3 is
characterized by an absence of continental polluted absorbing
aerosols, as shown by the optical analysis reported in Section 5.
The concentrations measured on the island of Porquerolles were
also compared to those recorded in the Atlantic and in other
Mediterranean locations. The present aerosol data recorded during
episode 2, i.e., when the synoptic conditions are Mediterranean,
exhibit values one order of magnitude larger than those recorded in
the Atlantic and 20 times higher when air masses are transported
above the French and Italian coasts at low wind speed (episode 1).
These concentrations are even larger than the previous measurements performed by Eleftheriadis et al. (2006), who found average
concentrations of 1340 cm3 when air mass back trajectory paths
mainly remained over the Mediterranean Sea, and 3470 cm3 over
the Aegean Sea when air mass back trajectories were transported
over the southern European coastal zone. Larger concentrations on
the island of Porquerolles can be explained by the vicinity of the
French urban coasts, as already noted in Section 3, expected to
release large amount of anthropogenic aerosols, in particular in the
form of new particle formation events. These events seem to take
place away from the coast and after precursors have been transported over the sea by land breezes where they can be photochemically transformed to low vapour pressure compounds.
Besides these new particle formation events, the concentrations
found under sea breeze conditions during episode 2 (Mediterranean air masses) could indicate a rather homogeneous background
concentration over the Mediterranean Sea at this period of the year,
since this corresponds to aerosol concentrations in the same order
of magnitude of those found by Eleftheriadis et al. (2006).
The variation of the height of the boundary layer between the
daytime and the nighttime could be responsible for the aerosol
concentration variations due to dilution effects. However, for
low wind speed periods encountered during episode 1 and episode
2, the calculations of the mixed layer depth using the Hysplit
model, show little variations. The mixed layer height varies from
J. Piazzola et al. / Atmospheric Environment 54 (2012) 545e556
approximately 50 m height to a maximum value of about 300 m
height, typically reached at noon as already noted by Nilsson et al.
(2001). This has probably a minor influence on the aerosol
concentration variations during episode 1, which, by the way, does
not vary a lot from night to day (see Section 3).
From the chemical point of view, the results show that the Naþ
and Cl concentrations are of the same order of magnitude than
what it was measured in Mediterranean areas by Sellegri et al.
(2001) and Bardouki et al. (2003). Anthropogenic tracers such as
2
NHþ
4 , nss-SO4 and NO3 also showed concentrations in the range of
those measured in the Mediterranean by Sellegri et al. (2001), with
slightly lower concentrations for NHþ
4 in the present study. If
sulphate concentrations are rather low at Porquerolles compared to
the eastern Mediterranean measurements, the contribution of
organics are significantly higher than the ones observed in Eastern
Mediterranean for summer and winter seasons (Bardouki et al.,
2003) but rather close to those observed in the sub-micron mode
during the phytoplanctonic blooming period in the Northern
Atlantic coastal area (Cavalli et al., 2004). Indeed, the results
reported in Section 4 show that organics contribute by 39%, 14%
and 34% to the particle mass for coastal influence (episode 1),
Mediterranean (episode 2) and mistral conditions (episode 3),
respectively. Episode 1 then corresponds to the major contribution
of the carbonaceous compounds, more particularly found in the
sub-micron mode (Fig. 11). The OC contribution in the Mediterranean air mass is similar to the one observed in the Eastern Mediterranean for summer and winter seasons (16% at Finokalia in
the submicronic mode, Bardouki et al., 2003). In the meantime
scattering coefficients present higher values (between 10 and
30 Mm1) and the associated aerosol single scattering albedo (SSA)
at 520 nm recorded during episode 1 indicate values comprised
between 0.80 and 0.90 (at 550 nm) which reveal an absorbing
aerosol which is generally associated with higher concentration of
carbon particles, as confirmed by data reported in Fig. 11.
Although it is suspected that a large fraction of organic
compounds found in the aerosol phase comes from anthropogenic
sources, the relatively large concentrations of organics compounds
can also be partly due to larger amounts of organics in the sea
water. These organics can be produced either biologically or be
already present in the sea water as primary pollutants. On one side,
biological activity is primarily induced by light and furthermore
depends strongly on the input of nutrients into the sea. High
concentrations of pollutants have been observed in the Mediterranean area, in particular from the run-off after a rain which may
cause an excess of phosphate and nitrate compounds resulting in
eutrophication. Biological studies conducted in the Toulon bay and
in the Hyeres bay, have indeed revealed an excess in nutrients and
eutrophication processes throughout the year (Jamet et al., 2001).
In addition, during the marine air masses (episode 2), particles
had not travelled over land for 3 days, which indicates that air
masses are strongly impacted by pollution over the whole Mediterranean basin. A large fraction of the sub-micron mode was
composed of organics, for aerosols sampled during purely marine
air mass episodes, closely followed by sulphate. Both compounds
are presumably of anthropogenic origin, but this hypothesis should
be checked in future work. Another indication of anthropogenic
influence during marine air mass periods is the strong contribution
of nitrate in the super micron mode. Sub-micron sulphate and
organics, and super micron nitrate contribute as much to the
aerosol mass during pure marine air masses than within air masses
originating from the densely populated Mediterranean coast and
Northern Italy.
The chemical analysis, combined with the high aerosol
concentrations measured on the island of Porquerolles, make think
that air masses are strongly impacted by pollution which is present
555
over the Mediterranean basin. However, in spite of the large
number of anthropogenic aerosol sources of such urbanized
coastal sites with large population densities, as the Mediterranean
study area, the anthropogenic compounds seem to be efficiently
dispersed within the lower atmosphere resulting in the dominant
signature of sea-salts for Mistral conditions. In this case, the
absence of continental polluted absorbing aerosols is confirmed by
the optical analysis.
Acknowledgements
The experimental campaign on Porquerolles Island was
sponsored by the national PEA MIRA program in partnership
with the Parc National de Port-Cros. The authors wish to express
their gratitude to Dr. Alain Barcelo, for his technical and administrative help in the implementation of the experimental campaign
in a protected area of the island of Porquerolles.
References
Andreae, M.O., 1995. Climate effects of changing atmospheric aerosol levels. In:
Henderson-Sellers, A. (Ed.), World Survey of Climatology. Future Climate of the
World, vol. 16, pp. 341e392.
Aymoz, G., Jaffrezo, J.L., Chapuis, D., Cozic, J., Maenhaut, W., 2007. Seasonal variation
of PM10 main constituents in two valleys of the French Alps. I: EC/OC fractions.
Atmospheric Chemistry Physics 7, 661e675.
Bardouki, H., Liakakou, H., Economou, C., Scaiare, J., Smolik, J., Zdimal, V.,
Eleftheriadis, K., Lazaridis, M., Dye, C., Mihalopoulos, N., 2003. Chemical
composition of size resolved atmospheric aerosols in the eastern Mediterranean during summer and winter. Atmospheric Environment 37, 195e208.
Birch, M.E., Cary, R.A., 1996. Elemental carbon based method for monitoring
occupational exposure to particulate diesel exhaust. Aerosol Science and
Technology 25, 221e241.
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., 2004. Advances in characterization of size-resolved organic matter in marine aerosol over the North
Atlantic. Journal Geophysical Research 109, D24215. doi:10.1029/
2004JD005137.
Ceburnis, D., O’Dowd, C.D., Jennings, G.S., Facchini, M.C., Emblico, L., Decesari, S.,
Fuzzi, S., Sakalys, J., 2008. Marine aerosol chemistry gradients: elucidating
primary and secondary processes and fluxes. Geophysical Research Letters 35,
L07804. doi:10.1029/2008GL033462.
Chang, M., Sioutas, C., Cassee, F.R., Fokkens, P.H.B., 2001. Field evaluation of a mobile
high-capacity particle size classifier (HCPSC) for separate collection of coarse,
fine and ultrafine particles. Journal of Aerosol Science 32, 139e156.
Draxler, R.R., Rolph, G.D., 2003. HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory) Model. Silver Spring, MD. NOAA Air Resources Laboratory.
Available at http://www.arl.noaa.gov/ HYSPLIT info.php.
Eleftheriadis, K., Colbeck, I., Housiadaa, C., Lazaridis, M., Mihalopoulos, N.,
Mitsakou, C., Smolık, J., Zdımal, V., 2006. Size distribution, composition and
origin of the submicron aerosol in the marine boundary layer during the
eastern Mediterranean “SUB-AERO” experiment. Atmospheric Environment 40,
6245e6260.
Fitzgerald, J.W., 1989. Model of the aerosol extinction profile in a well-mixed marine
boundary layer. Applied Optics 28 (16), 3534e3538.
Heintzenberg, J., Wiedensohler, A., Tuch, M., Covert, D.S., Sheridan, P., Ogren, J.A.,
Gras, J., Nessler, R., Kleefeld, C., Kalivitis, N., Aaltonen, V., Wilhelm, R.-T.,
Havlicek, M., 2006. Intercomparisons and aerosol calibrations of 12 commercial
integrating nephelometers of three manufacturers. Journal Atmospheric and
Oceanic Technology 23, 902. doi:10.1175/JTECH1892.1.
Hsu, S.A., 1986. A mechanism for the increase of wind stress (drag) coefficient
with wind speed over water surfaces: a parametric model. Journal Physical
Oceanography 16, 144e150.
Intergovernmental Panel on Climate Change, 2007. Changes in atmospheric
constituents and in radiative forcing. In: Solomon, S., et al. (Eds.), Climate
Change2007: The Physical Science Basis, Contribution of Working Group I to the
Fourth Assessment Report of the Intergovernmental Panel on Climate Change.
Cambridge Univ. Press, Cambridge, U.K, pp. 129e234.
Jaenicke, R., 1984. Physical aspects of atmospheric aerosol. In: Gerbard, H.E.,
Deepak, A. (Eds.), Aerosols and Their Climatic Effects, pp. 7e34.
Jamet, J.L., Bogé, G., Richard, S., Geneys, C., Jamet, D., 2001. The zooplankton
community in bays of Toulon area (northwest Mediterranean sea, France).
Hydrobiologia 457, 155e165.
Knipping, E.M., Dabdub, D., 2003. Impact of chlorine emissions from sea-salt aerosol
on coastal urban ozone. Environmental Science Technology 37 (2), 275e284.
Laskin, A., Gaspar, D.J., Wang, W., Hunt, S.W., Cowin, J.P., Colson, S.D., FinlaysonPitts, B.J., 2003. Science Reactions at interfaces as a source of sulfate formation
in sea-salt particles. Science 301 (5631), 340e344.
556
J. Piazzola et al. / Atmospheric Environment 54 (2012) 545e556
Leslie, A.C.D., 1981. Aerosol emissions from forest and grassland burnings in the
southern amazon basin and central Brazil. Nuclear Instruments and Methods
181, 345e351.
Mallet, M., Roger, J.C., Despiau, S., Dubovik, O., Putaud, J.P., 2003. Microphysical
and optical properties of aerosol particles in urban zone during ESCOMPTE.
Atmospheric Research 69 (1e2), 73e97.
Mallet, M., Gomes, L., Solmon, F., Sellegri, K., Pont, V., Roger, J.-C., Missamou, T.,
Piazzola, J., 2011. Calculation of key optical properties of the main anthropogenic aerosols over the Western French coastal Mediterranean Sea.
Atmospheric Research 101 (1e2), 396e411.
Mulcahy, J.P., O’Dowd, C.D., Jennings, S.G., Ceburnis, D., 2008. Significant enhancement of aerosol optical depth in marine air under high wind conditions.
Geophysical Research Letters 35, L16810. doi:10.1029/2008GL034303.
Nilsson, E.D., Rannik, U., Kulmala, M., Buzorius, G., O’Dowd, C.D., 2001. Effects of
continental boundary layer evolution, convection, turbulence and entrainment,
on aerosol formation. Tellus 53B, 441e461.
O’Dowd, C.D., Facchini, M.C., Cavalli, F., Ceburnis, D., Mircea, M., Decesari, S.M.,
Fuzzi, S., Yoon, Y.J., Putaud, J.P., 2004. Biogenically driven organic contribution to
marine aerosol. Nature 431, 676e680.
Paerl, H.W., Boynton, W.R.., Dennis, R.L., Driscoll, C.T., Greening, H.S., Kremer, J.N.,
Rabalais, N.N., Seitzinger, S.P., 2000. Atmospheric deposition of nitrogen in
coastal waters: biogeochemical and ecological implications. In: Valigura, R.A.,
Alexander, R.B., Castro, M.S., Meyers, T.P., Paerl, H.W., Stacey, P.E., Turner, R.E.
(Eds.), Nitrogen Loading in Coastal Water Bodies: an Atmospheric Perspective.
American Geophysical Union, Washington DC (USA), pp. 11e52.
Piazzola, J., Despiau, S., 1997a. Contribution of marine aerosols in the particle
size distributions observed in Mediterranean coastal zone. Atmospheric
Environment 31 (18), 2991e3009.
Piazzola, J., Despiau, S., 1997b. Vertical distribution of aerosol particles near
the air-sea interface in coastal zone. Journal Aerosol Science 28 (8),
1579e1599.
Piazzola, J., Forget, P., Despiau, S., 2002. A sea spray generation function for fetchlimited conditions. Annales Geophysicae 20 (11), 121e131.
Piazzola, J., Forget, P., Lafon, C., Despiau, S., 2009. Spatial variation of sea-spray
fluxes over a Mediterranean coastal zone using a sea-state model. BoundaryLayer Meteorology 132 (1), 167e183.
Saha, A., Mallet, M., Dubuisson, P., Piazzola, J., Despiau, S., 2008. One year
measurements of aerosol optical properties over an urban coastal site: effect on
local direct radiative forcing. Atmospheric Research 90, 195e202.
Sellegri, K., Gourdeau, J., Putaud, J.-P., Despiau, S., 2001. Chemical composition of
marine aerosol in a Mediterranean coastal zone during the FETCH experiment.
J. Geophys. Res. 106 (D11), 12023e12038. doi:10.1029/2000JD900629 (AGU).
Sellegri, K., Yoon, Y.J., Jennings, S.G., O'Dowd, C.D., Pirjola, L., Cautenet, S., Chen, H.,
Hoffmann, T., 2005. Quantification of coastal new ultra-fine particles formation
from in situ and chamber measurements during the BIOFLUX campaig. Environ.
Chem. 2 (4), 260e270. doi:10.1071/EN05074 (CSIRO Publishing).
Tang, I.N., 1996. Chemical and size effects of hygroscopic aerosols. Journal of
Geophysical Research 101 (D14), 19,245e19,250.
Ten Brink, H., Otjes, R., Jongejan, P., Kos, G., 2009. Monitoring of the ratio of
nitrate to sulphate in size-segregated submicron aerosol in the Netherlands.
Atmospheric Research 92, 270e276.
Venzac, H., Sellegri, K., Villani, P., Picard, D., Laj, P., 2009. Seasonal variation of aerosol
size distributions in the free troposphere and residual layer at the puy de Dôme
station, France. Atmospheric Chemistry and Physics 9 (4), 1465e1478.
Wang, H.-C., John, W., 1988. Characteristics of the berner impactor for sampling
inorganic ions. Aerosol Science and Technology 8, 157e172.
Weingartner, E., Saathoff, H., Schnaiter, M., Streit, N., Bitnar, B., Baltensperger, U.,
2003. Absorption of light by soot particles: determination of the absorption
coefficient by means of aethalometers. Journal of Aerosol Science 34,
1445e1463. doi:10.1016/S0021-8502(03)00359-8.
Yoon, Y.J., Ceburnis, D., Cavalli, F., Jourdan, O., Putaud, J.P., Facchini, M.C., Decesari, S.,
Fuzzi, S., Sellegri, K., Jennings, S.G., O’Dowd, C.D., 2007. Seasonal characteristics
of the physicochemical properties of North Atlantic marine atmospheric aerosols. Journal of Geophysical Research 112, D04206. doi:10.1029/2005JD007044.
Zhang, X.Q., McMurry, P.H., 1992. Evaporation losses of fine particulate nitrates
during sampling. Atmospheric Environment 26, 3305e3312.