lable at ScienceDirect

Atmospheric Environment xxx (2011) 1e8

Contents lists avai

Atmospheric Environment

journal homepage: www.elsevier .com/locate/atmosenv

Ultrafine particles pollution in urban coastal air due to ship emissions

Yenny González a,b, Sergio Rodríguez a,*, Juan Carlos Guerra García c, Juan Luis Trujillo a,b, Rosa García a,d

a Izaña Atmospheric Research Centre, AEMET, Joint Research Unit to CSIC ‘Studies on Atmospheric Pollution’, La Marina 20, Planta 6, 38071 Santa Cruz de Tenerife,Canary Islands, Spainb Sieltec Canarias S.L., San Cristóbal de La Laguna, Sta. Cruz de Tenerife, SpaincUniversity of La Laguna, Hydrometeorology Research Group (GRIHM), SpaindUniversity of Valladolid, Atmospheric Optics Group (GOA-UVA), Spain

a r t i c l e i n f o

Article history:Received 14 March 2011Received in revised form30 May 2011Accepted 2 June 2011

Keywords:Ultrafine particlesShip emissionsVehicle exhaustUrban air quality

* Corresponding author. Tel.: þ34 922151718; fax:E-mail address: srodriguezg@aemet.es (S. Rodrígu

1352-2310/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.atmosenv.2011.06.002

Please cite this article in press as: GonzáleEnvironment (2011), doi:10.1016/j.atmosenv

a b s t r a c t

Two years of experimental data (2008e2009) of particle number (�2.5 nm diameter) and black carbonconcentrations and of gaseous pollutants recorded in the ambient air of a coastal city were analysed inorder to assess the impact of ship emissions on the ultrafine particles (UFPs, diameter <100 nm)concentrations in urban ambient air. The observed relationship between particle number and the otherair pollutants, allowed segregating the contribution of vehicle exhaust and of ship emissions to the UFPconcentrations in the urban ambient air. Vehicle exhausts resulted in high concentrations of UFP, blackcarbon and NOx during the early morning, when UFPs showed concentrations 15e30 � 103 cm�3.Pollutants linked to this source rapidly decreased when inland sea breeze started to flow. However, thisairflow resulted in inland transport of ship plumes (emitted in the harbour and in the sea) into the cityand in high concentrations of SO2, NOx and UFP from mid morning to the evening. In this context, UFPsshowed concentrations 35e50 � 103 cm�3, being the 65e70% of these linked to ship emissions mostlyrelated to SO2 (gas phase precursor). UFPs pollution is a matter of concern due to adverse effects onhuman health. Up to the date, most of studies on urban air quality and UFPs have focused on vehicleexhaust emissions. This study shows how inland transport of ship plumes due to sea breeze blowingresults in UFPs pollution in coastal cities.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Air quality impairment due to ultrafine particles (UFPs; diam-eter <100 nm) pollution has become in a topic of great interestwithin the context of environmental sciences and public healthprotection. Although it is considered that exposure to UFP affectshuman health, there are large uncertainties about the global effects.Translocation of inhaled atmospheric particles in the human bodydepends, among other factors, on particle size. The smallest UFP(e.g. <10 nm) mostly deposit on the nasopharyngeal region, fromwhere they may be transported to the central nervous system(Oberdörster et al., 2004); whereas larger UFP (e.g. >20 nm) mayreach the alveoli and enter the blood stream (Geiser et al., 2005).Inhaled UPF may also reach other organs such as liver, heart andepithelial cells (Oberdörster et al., 2004). Once they have reachedthe blood stream, solid UPF may enhance atherosclerosis andcardiovascular ischemic events (Araujo and Nel, 2009; Domínguez-Rodríguez et al., in press).

þ34 922574475.ez).

All rights reserved.

z, Y., et al., Ultrafine particle.2011.06.002

UFPs typically account for 80e90% of the total particle numberconcentration in urban air (Wehner and Wiedensohler, 2003;Rodríguez et al., 2007; Mejía et al., 2008). For this reason, manystudies have been based on measurements of total numberconcentrations of particles coarser than 10 nm (Harrison and Jones,2005) or 3 nm (Shi et al., 1999). Most of studies attribute UFPs inurban ambient air to (very briefly, see details in Rodríguez andCuevas, 2007):

� Vehicle exhausts emissions. Fresh particles emitted by vehicleexhausts tend to be bimodal, exhibiting a nucleation (<20 nm)and a carbonaceous (50e200 nm)mode. The nucleationmode ismostly constituted bysulphuric acid droplets, formedduring thedilution and cooling of the exhaust and coated by condensedsulphate and/or hydrocarbons. The exhaust after-treatmentsystem, fuel composition, lubricant oil and environmentalparameters may influence on the nucleation mode formationrates. A fraction of the nucleationmodematerial is semi-volatileand its formation rate depends on thedilution conditions andonthe temperature and relative humidity of the ambient air. Thecarbonaceous mode (50e200 nm) is predominantly made up of

s pollution in urban coastal air due to ship emissions, Atmospheric

Y. González et al. / Atmospheric Environment xxx (2011) 1e82

soot, absorbed organic material and other trace elementsformed in the engine and directly emitted in the solid phase.

� New particle formation process in ambient air. This is a relativelycomplex process that involves nucleation of an initial cluster(<1 nm) and its subsequent growth to detectable sizes (severalnanometres diameter) due to condensation of vapours. It isconsidered that sulphuric acid gas molecules play a key role asinitial cluster, whereas vapours involved in the growthprocesses are linked to photo-oxidation of SO2 and/or organicvapours.

Most of studies on urban air quality and UFPs have focused onvehicle exhaust emissions, even if it is well known that othersources may also contribute. Moreover, the number of techniquesfor quantifying the sources and processes that contribute to UFP israther very low (Fernández-Camacho et al., 2010).

We have focused on the UFP pollution due to ship emissions.These emissions impact on human health (Corbett et al., 2007a)and contribute to the acidification and eutrophication of water andsoil in coasts due to sulphur and nitrogen deposition (Derwentet al., 2005), to climate cooling owing to the high sulphur contentof maritime fuel (Lauer et al., 2007) and to climate warming causedby the emissions of greenhouse gases (Stern, 2007) and absorbingblack carbon (Lack et al., 2009). Ship emissions are known as one ofthe least regulated source of pollutants (Eyring et al., 2005). In theEU, inland SO2 emissions have been decreasing since the 1980’s,whereas emissions in ships are the only growing source of sulphur.Ships may release primary particles and precursors of secondaryparticles. Primary particles emitted by ships are constituted bya complex mixture of elemental and organic carbon, sulphate

Fig. 1. A) Satellite Google� map view of Santa Cruz de Tenerife highlighting the location of tof the harbour (dock-2) from SCO. (For interpretation of the references to colour in this fig

Please cite this article in press as: González, Y., et al., Ultrafine particleEnvironment (2011), doi:10.1016/j.atmosenv.2011.06.002

condensed onto ashes (Petzold et al., 2004; Kasper et al., 2007) andtrace metals (e.g. V and Ni; Viana et al., 2009). Moreover, emissionsof volatile organic compounds and sulphur dioxide result in theformation of secondary particles by nucleation and condensationprocesses (Petzold and Schönlinner, 2004). Because of thesefeatures, the size distribution of particles within plume of shipsmay covers from the nucleation to the coarse mode (e.g. 0.005 to>3 mm; Petzold et al., 2004; Kasper et al., 2007).

Air quality impairment due to PM10 and PM2.5 pollution by shipsis alreadywell characterised (Agrawal et al., 2008;Moldanová et al.,2009). In this study we have focused on quantifying the contribu-tion of ship emissions to the UFP concentrations in the urbanambient air in coastal cities. Results show that, because of the dailyinland transport of ship’s plumes prompted by coastal breezes andbecause the high content of sulphur in these plumes, ship emis-sions emerge as a major source of UFPs in coastal cities, frequentlyexceeding the contribution of the vehicle exhaust emissions.

2. Methodology

2.1. Study area

The studywas performed in Santa Cruz de Tenerife (28�2802200 N16�1401900 W; Canary Islands, Spain). The main sources of pollutantsin the city are (Fig. 1A):

� Vehicle exhausts emissions occurring throughout the city.� Emissions of ships at the East of the city (Fig. 1B). Theseemissions occur during sailing, entries and exits in the harbour,when ships are anchored off shore, and in the port, owing

he measurement site (SCO) and the main roads of the city (yellow dotted lines). B) Viewure legend, the reader is referred to the web version of this article).

s pollution in urban coastal air due to ship emissions, Atmospheric

Y. González et al. / Atmospheric Environment xxx (2011) 1e8 3

to the engines keep activated for power supply. Annual(calendar year) maritime transport, number of merchant shipsand cruisers was 17,136,449 tons $ year�1, 16,341 ships year�1

and 341 cruisers year�1 during 2008e2009 as average, respec-tively (http://www.puertos.es).

� Acrude oil refinery located at the SWof the city, at about 3 km farfrom the measurement site. These emissions only reached themeasurement site under synoptically driven Southern winds,which occurs less than 10% of the days per year. Data collectedduring fumigations of the plumes of the refinery were identifiedand not included in the analysis performed in this study.

The pollutants transport in the urban area is highly influencedby the development of coastal breezes (Fig. 2D, discussed below).

Fig. 2. Hourly mean values of particles (N and BC) and gaseous pollutants concentrations (speed, the road traffic intensity/wind speed ratio, of the N/BC ratio, N1, N2, solar radiationwind direction.

Please cite this article in press as: González, Y., et al., Ultrafine particleEnvironment (2011), doi:10.1016/j.atmosenv.2011.06.002

From 10 to 17 GMT, the landward breeze circulation favours theinland transport of the pollutants emitted in and near the shore. Atnight, a slight seaward airflow occurs.

2.2. Measurements

Measurements were performed at the so-called “Santa CruzAtmospheric Observatory” (SCO; www.izana.org). The measurementsite is placed in the roof of the Izaña Atmospheric Research Centre(28.472373�N,16.246962�E, 60m above sea level). Data of trace gasesconcentrations (NOx, SO2, O3 and CO), of particle number and blackcarbon concentrations and of meteorological parameters (radiation,wind speed and direction, temperature, relative humidity, precipita-tion and pressure) collected from April 2008 to December 2009 were

NOx, SO2 and O3), and of the road traffic intensity (number of vehicles hour�1), windand wind speed and direction for working days and weekends. WS: wind speed. WD:

s pollution in urban coastal air due to ship emissions, Atmospheric

Y. González et al. / Atmospheric Environment xxx (2011) 1e84

used. Trace gases were monitored using reference European methods(EU Directive 2008/50/EC). A daily zero-check was performed. Five-point calibrations were done every 3 months using certified gasconcentrations. Furthermore, O3 and CO analysers were auditedby the World Calibration Centre for Surface Ozone (EMPA; http://www.empa.ch/) in 2009. Black Carbon (BC) and particle number(diameter �2.5 nm; N) concentrations were measured with a MAAP(Thermo�) and a UCPC (TSI�, model 3776), respectively. A massabsorption efficiency s ¼ 9.79 (�0.45, r2 ¼ 0.89) m2 � g�1 was usedfor determining BC mass concentrations from absorption units. Thisvalue was used based on experimental determinations (described inReche et al., 2011). The used instruments setup was similar to thatdescribed by Fernández-Camacho et al. (2010). The two equipmentswere inter-compared at the University of Huelva with similarinstruments; differences between the CPCs andMAAPs were equal to0.2% and 7% for 10 min averaged data (Fernández-Camacho et al.,2010). Data management is performed using Greenwich MeridianTime (GMT) as reference, which is similar to local time in winter(October to March) and 1 h less than local time in summer (March toOctober).

Data availability for the period 2008e2009was 67% for NOx, 98%for SO2, 98% for O3, 70% for N and 83% for BC. Additionally, roadtraffic intensity (number of vehicles/hour) was monitored by thecity council staff at the Anaga Avenue.

3. Results and discussion

3.1. Daily evolution

Fig. 2AeE shows hourly mean values of particles (N and BC)and gaseous (SO2 and NOx) pollutants, road traffic intensity(number of vehicles/hour), road traffic intensity/wind speed ratio,N/BC ratio and meteorological parameters for working days,Saturdays and Sundays. Concentrations of BC and NOx reachedthe highest values during the morning rush hours (07e09 GMT)of working days (Fig. 2AeB). The sharp decrease in BC and NOxafter the morning rush hours (not observed in the road trafficintensity, Fig. 2C) is associated with the entry of easterly marineair in the city due to the inland sea breeze blowing (Fig. 2D). TheNOx and BC weekly evolution is similar to that observed in theroad traffic intensity/wind speed ratio (Fig. 2C), which indicatesthat fresh vehicle exhaust emissions and dilution/ventilation

-

N, c

m-3

N, c

m-3

150 103

100 103

150 103

50 103

100 103

0

50 103

0

7-9 GMT

S1=7.9

S2

0 2000 4000 60

10-17 GMT

S1=14.5

S2

·

·

·

·

·

·

BC, ng·m 3

A

C

Fig. 3. Hourly mean values of N versus BC concentrations at different times of the

Please cite this article in press as: González, Y., et al., Ultrafine particleEnvironment (2011), doi:10.1016/j.atmosenv.2011.06.002

conditions (prompted by breeze) modulate the regular variationsof NOx and BC.

The behaviour of SO2 is different to that of BC and NOx. High SO2concentrations were recorded from 10 to 17 GMT, associated withthe inland sea breeze blowing (Fig. 2E), and caused by the entry ofship plumes in the city. This sea breeze is associated with anincrease in wind speed that results in the air mass renovation andin increase in ozone concentrations (Fig. 2D). These plumes alsoinfluence on NOx concentrations (Williams et al., 2009), although asaverage this pollutant showed much higher values during themorning rush hours.

In the morning (07e09 GMT), N experienced a sharp increase(correlated with BC and NOx) linked to fresh vehicle exhaust emis-sions (Fig. 2AeC). The N/BC ratio exhibited a correlated evolutionwith SO2, wind speed, solar radiation and ozone concentrations(Fig. 2DeE). This suggests that amechanism linkedwith the entry ofmarine air enriched in SO2, probably also involving photochemicalprocesses, is contributing to UFP concentrations in the urban airduring the inland sea breeze blowing period.

3.2. Relationship between particle number and blackcarbon concentrations

Previous studies indicate that the relationship between particle Nand BC concentrations provides very useful information about thesources and processes contributing to UFPs in urban ambient(Rodríguez and Cuevas, 2007; Fernández-Camacho et al., 2010). In thecoastal study city, N and BC were also correlated. In Fig. 3, the scatterplots of N versus BC data registered at different times are shown(07e09, 10e17, 18e21, and 01e03 GMT). At any time, the N-vs-BCdata set is comprised between two well defined linear borders withslopes S1 and S2, representing the minimum and maximum N/BCratios, respectively. During the morning rush hours (07e09 GMT,when vehicle exhaust emissions predominate), a value ofS1¼7.9� 106 particles ngBC�1 was found. This S1 slope is interpretedas “the minimum number of particles directly entering in the air pereach nanogram of BC emitted by vehicle exhaust”. N and BC dataresulting inN/BC ratios higher than S1 are caused by enhancements inthe new particle formation processes during the dilution and coolingof the vehicle exhaust, e.g. favoured by low temperature, low dilutionconditions due to low wind speed or low concentrations of pre-existing particles (Casati et al., 2007). The much higher “minimum

-

18 -21 GMT

S1=11.1

S2

00 0 2000 4000 6000

1-3 GMT

S1=7.9

S2

BC, ng·m 3

B

D

day in Santa Cruz City. S1 (106 particles ngBC�1): minimum N-vs.-BC slope.

s pollution in urban coastal air due to ship emissions, Atmospheric

Table 1Factor loading of the Principal Components Analysis (followed by a varimax rota-tion) obtained with hourly data in the morning (07:00e09:00 GMT) and noon-afternoon (10:00e17:00 GMT). Factor loadings with absolute values �0.7 are inbold. RTI/WS: road traffic intensity/wind speed.

07e09 GMT 10e17 GMT

PC 1 PC 2 PC 1 PC 2

vehicle exhaust 1 vehicle exhaust 2 ships vehicle exhaust

N1 0.81 0.44 0.78 0.35N2 �0.05 0.89 0.79 �0.12CO 0.90 0.04 0.00 0.80SO2 0.21 0.86 0.90 �0.08NOx 0.83 0.46 0.86 0.24RTI/WS 0.82 �0.18 0.09 0.76Variance 48% 33% 46% 24%

Y. González et al. / Atmospheric Environment xxx (2011) 1e8 5

number of particles per each nanogram of BC” found during the10e17 GMT period (S1 ¼ 14.5 � 106 particles ngBC�1; Fig. 3C)evidences that other sources and/or mechanisms are contributing toUFP concentrations in ambient air during the inland sea breezeblowing period. This enhancement in the new particle formationprocesses also accounts for the higher S2 values during the inland seabreeze blowing period (S2¼ 62.0�106 particles ngBC�1; Fig. 3C) thanduring the morning rush hours (S2 ¼ 30.3 � 106 particles ngBC�1;Fig. 3A).

Fig. 4. Hourly mean daily evolution of N2, N1, SO2 and NOx for events of different SO2 conce10e30 mg m�3 (A3eB3), 30e50 mg m�3 (A4eB4), 50e250 mg m�3 (A5eB5).

Please cite this article in press as: González, Y., et al., Ultrafine particleEnvironment (2011), doi:10.1016/j.atmosenv.2011.06.002

The relationship between N and BC we observed in the studycity (Fig. 3) is typically observed in urban ambient air, as shownstudies performed in other cities, such as Huelva, Barcelona, Milan,Bern, Lugano and London (Rodríguez and Cuevas, 2007; Fernández-Camacho et al., 2010; Reche et al., 2011).

3.3. Components contributing to UFPs

In order to quantify the contribution of the different sources andprocesses to the UFP concentrations in the urban ambient air,the methodology developed by Rodríguez and Cuevas (2007) wasapplied. In this technique, the number concentration N is split intwo components:

N1 ¼ S1$BC (1)

N2 ¼ N� N1 (2)

where the S1 value observed in the morning (7.9� 106 particles ngBC�1), when vehicle exhaust is the dominant sourceof UFPs, was used. Component N1 accounts for those “compoundsdirectly emitted (by the vehicle exhaust) in the particle phase andthose compounds nucleating immediately after the emission”. Thiscomponent includes incomplete fuel combustion products (e.g. blackcarbon, long-chain organic matter compounds or PAH), condensedtrace metals, unburned oil and a fraction of sulphate and organic

ntration ranges (during the 10e17 GMT): 0e3 mg m�3 (A1eB1), 3e10 mg m�3 (A2eB2),

s pollution in urban coastal air due to ship emissions, Atmospheric

Y. González et al. / Atmospheric Environment xxx (2011) 1e86

compounds nucleating/condensing immediately after emission(Kittelson,1998;Burtscher, 2005;Arnold et al., 2006;Rose et al., 2006).Component N2 accounts for those compounds resulting in the afore-mentioned increases in the N/BC ratio. This occurs by enhancementin the new particle formation rates due to increased nucleationand/or growth rates to detectable sizes (�2.5 nm in our case) indifferent contexts: during the dilution and cooling of the vehicleexhaust emissions (Charron andHarrison, 2003; Casati et al., 2007), inambient air due to photochemistry (Woo et al., 2001; Wehner et al.,2002), and/or in aerosol precursor plumes (Fernández-Camachoet al., 2010).

These interpretations are also supported by the results of thisstudy. Observed in Fig. 2FeG (where the weekly evolution N1, N2,NOx and SO2 is shown) how the daily evolution of N1 was similar tothat of NOx, with a maximum value from 07 to 09 GMT on workingdays, whereas N2 was similar to that of SO2, with high valuesduring the inland sea breeze blowing. Table 1 shows the resultsobtained in a Principal Component Analyses (PCA) followedby varimax rotations performed with 1-h average data. PCAwere performed in two periods: 07e09 GMT and 10e17 GMT.Eigenvalues > 1 and a maximum number of principal componentsequal to 4 were set. Only two principal components (PCs) wereobtained. In the morning (07e09 GTM):

- a PC associated with N1, CO, NOx and road traffic intensity/windspeed is observed. This profile points to vehicle exhaust emis-sions of particulate carbonaceous material (Arnold et al., 2006).

- the PC associated with N2 and SO2 points to the formation ofnucleation sulphuric acid particles (Burtscher, 2005).

In the inland sea breeze blowing period (10e17 GMT):

Fig. 5. Hourly mean values of total number concentration (N ¼ N1þN2) classified from the hN, in absolute (cm�3; A1 and A2) and relative (%; B1 and B2) concentrations, are highlightedvalues (from 100th to 0th) are plotted.

Please cite this article in press as: González, Y., et al., Ultrafine particleEnvironment (2011), doi:10.1016/j.atmosenv.2011.06.002

- PC-1, associated with N1, N2, NOx and SO2, is attributed to theinland transport of ship plumes,

- PC-2 evidence the impact of vehicle exhausts emissions(modulated by wind speed) on CO concentrations.

3.4. Contribution of the ship emissions to UFP concentrationsin urban ambient air

The impact of ship emissions on the UFP particles concentrationsin the urban ambient air was quantified using the SO2 transportedfrom the east by the inland sea breeze as a tracer. We studied thedaily evolution of the hourly values of N, N1, N2, SO2 andNOx duringevents inwhich SO2 concentrations (in the 10e17 GMT period) werewithin the following ranges: 0e3 mg m�3 (Fig. 4A1 and B1),3e10 mg m�3 (Fig. 4A2 and B2), 10e30 mg m�3 (Fig. 4A3 and B3),30e50 mg m�3 (Fig. 4A4 and B4) and 50e250 mg m�3 (Fig. 4A5 andB5; this latter value was the maxima recorded). This analysis high-lights several important features:

� For any SO2 concentration range, N1 shows a maximum duringthe morning rush hours (Fig. 4A1eA5), indicating that thiscomponent is dominated by vehicle exhaust emissions, even ifships may emit BC particles.

� During events of SO2 concentrations<10 mgm�3, the behaviourof SO2, NOx and N is predominantly influenced by vehicleexhaust emissions. For this reason these pollutants showa maximum during the morning rush hours (Fig. 4A1eB2).

� During events of SO2 concentrations>10 mgm�3, the behaviourof SO2, NOx and N is predominantly influenced by ship emis-sions. Observe that these pollutants reach the highest

ighest (100th) to the lowest (0th) value. The contributions of N1 (black) and N2 (grey) to. NOx (C1 and C2) and SO2 (D1 and D2) concentrations associated with the decreasing N

s pollution in urban coastal air due to ship emissions, Atmospheric

0210 8040NOx, µg/m 3

0

12080

40

SO2, µg/m 3

0 40NOx, µg/m 3

80 100200 10

SO2, µg/m 3

15 20 065

N1,

cm

-3N

2, c

m-3

inland sea breeze blowing - shipsmorning rush hours - vehicle exhaust

A1 A2

B1 B2

20·103

10·103

0

60·103

20·103

10·103

0

30·103

20·103

10·103

0

30·103

0

30·103

40·103

20·103

80·103

Fig. 6. N1 versus NOx concentrations during the morning rush hours (06e09 GMT; A1) and during the inland sea breeze period (10e17 GMT; A2). N2 versus SO2 concentrationsduring the morning rush hours (B1) and during the inland sea breeze period (B2). In order to reduce the variability due to other influencing parameters, data of NOx and N1 wereaveraged in bins of 10 mg m�3 of NOx, whereas data of SO2 and N2 were averaged in bins of 3 and 10 mg m�3 of SO2 during the morning and inland sea breeze period, respectively.Mean � 1 standard deviation is included in the plot.

Y. González et al. / Atmospheric Environment xxx (2011) 1e8 7

concentrations during the inland sea breeze blowing period(Fig. 4A3eB5).

These results are in agreement with those obtained in other EUcities (e.g. Barcelona, Huelva, London, Bern and Lugano), thatshowed that high N1 concentrations are associated with vehicleexhaust emissions, whereas N2 may be affected, in addition ofvehicle exhaust, by other sources (Fernández-Camacho et al., 2010;Reche et al., 2011).

The contribution of N1 and N2 to N concentrations were verydifferent in events dominated by “vehicle exhaust” (low SO2 levels)and in events dominatedby “ships” (highSO2 levels) emissions (Fig. 5).The highest contribution of N1 to N occurred during the morningrush hours (07-09 GMT), when N was dominated by vehicle exhaustemissions and it showedvalueswithin the range15e30�103 cm�3. Inthis context, the contribution of N1 andN2 toNwas very close, each ofthem typically accounting for 55e65% of total UFP each (N). Thehighest contribution of N2 to N occurred during the inland sea breezeblowing period (10e17 GMT), when both, vehicle exhausts and shipemissions, may contribute to UFP concentrations. In this period meanN concentrations showed a monotonic increase throughout the SO2sequence shown in Fig. 4B1eB5, when mean � standard deviationshowed the following values: 10,900 � 9400, 19,250 � 13,870,35,400 � 21,600, 37,160 � 26,300 and 48,350 � 37,500 cm�3.This increase in N concentrations is prompted by an increase inthe N2 contribution: 53 � 3, 53 � 2, 66 � 5, 68 � 4 and 70 � 4%,respectively.

The high impact of ship emissions on UFP concentrations in urbanambient air is easily recognized in Fig. 5. In this figure, the hourlymean N data are classified from the highest to the lowest value,highlighting the contribution of N1 and N2 to N (Fig. 5A1eB2). Theassociated SO2 andNOx concentrations (Fig. 5C1eD2) recorded duringthe morning rush hours and during the inland sea breeze blowingperiod are also shown.Under thepredominant influenceof the vehicleexhaust emissions in the morning, N shows an increasing trend with

Please cite this article in press as: González, Y., et al., Ultrafine particleEnvironment (2011), doi:10.1016/j.atmosenv.2011.06.002

NOx (Fig. 5A1 and5C1). In this period, the contribution ofN1 andN2 toN is very close, ranging between 40 and 60% each for the wholepercentile range (Fig. 5B1). Under the influence of ship emissionsduring the inland sea breeze blowing period, N also shows anincreasing trendwith SO2 andNOx (Fig. 5A2, 4C2 and 5D2), and itmayreach much higher concentrations than in the morning (wtwice;Fig. 5A2). Observed in Fig. 6 howN1 andN2 tend to increasewith NOxand SO2, respectively. The increase inN concentrations is prompted byan enhancement in N2, whose contribution accounts for 46 � 16%,58 � 14% and 70 � 14% for the N percentiles ranges <20th, 20e80thand >80th (Fig. 5B2). These much higher N2 concentrations areprompted by themuch higher SO2 concentrations under the influenceof ship plumes than vehicle exhausts (up to 10 time higher; Fig. 5D1and D2).

4. Conclusions

Air quality impairment due to ultrafine particles pollution hasbecome in a matter of concern due to adverse effects of theseparticles on human health. Up to the date, most of studies on ultra-fine particles in urban air have focused on vehicle exhaust emissions.It this study we have focused on the impact of ship emissions on theultrafine particle concentrations in downwind urban areas.

A data set of number of particles coarser than 2.5 nm (N),particulate black carbon concentrations (BC), gaseous pollutants(SO2, NOx and O3) and meteorological parameters recorded ina coastal city has been analysed. Results shows that the impactof vehicle exhaust and ship emissions can be segregated. Theparticle number concentration (N) was split into two components:N ¼ N1 þ N2, where N1 is linked to vehicle exhaust emission ofcarbonaceous material (including light absorbing matter) andcomponents nucleating immediately after the emission, whereasN2 is linked to new particle formation due to SO2 emissions (fromvehicle exhaust and ship).

s pollution in urban coastal air due to ship emissions, Atmospheric

Y. González et al. / Atmospheric Environment xxx (2011) 1e88

During the morning rush hours of the working days, whenultrafine particles are linked to vehicle exhaust emissions, Nshowed concentrations within the range 15e30 � 103 cm�3. Inthis context, N2 accounted for 55e65% of N, being the remainingfraction linked to N1. The inland sea breeze blowing results ina drop of the concentrations gaseous and particulate pollutantsemitted by vehicle exhausts. However, this airflow results in theinland transport of the ship plumes into the city. This results inan increase in the concentrations of ultrafine particles and SO2during the inland sea breeze blowing. In this period, N typicallyshows concentrations within the range 35e50 � 103 cm�3,when N2 accounts for 65e70% of N. This N2 fraction clearlyshows an increasing trend with SO2 concentrations. The resultsof this study show that ship emissions may result in muchhigher concentrations of ultrafine particles that vehicle exhaustemissions.

Acknowledgements

This study has been developed within the framework of theprojects supported by the Ministry of Science and Innovation ofSpain (GRACCIE CSD2007-006 and REDMAAS CGL2010-11095-E)and the Department of Innovation, Science and Enterprise ofGovernment of Andalusia (AER-REG, P07-RNM-03125). We thankJose-Manuel Rodríguez-Sánchez for his support for managing andtreatment of the files of the picture. We also thank to the twoanonymous referees for his detailed reading of the manuscript andtheir constructive comments.

References

Agrawal, H., Malloy, Q.G.J., Welch, W.A., Miller, J.W., Cocker III, D.R., 2008. In-usegaseous and particulate matter emissions from a modern ocean going containervessel. Atmospheric Environment 42, 5504e5510.

Araujo, J.A., Nel, A.E., 2009. Particulate matter and atherosclerosis: role of particlesize, composition and oxidative stress. Particle and Fibre Toxicology 6 (24).doi:10.1186/1743-8977-6-24.

Arnold, F., Pirjola, L., Aufmhoff, H., Schuck, T., Lähde, T., Hämeri, K., 2006. Firstgaseous sulfuric acid measurements in automobile exhaust: implications forvolatile nanoparticle formation. Atmospheric Environment 40, 7097e7105.

Burtscher, H., 2005. Physical characterization of particulate emissions from dieselengines: a review. Aerosol Science and Technology 36, 896e932.

Casati, R., Scheer, V., Vogt, R., Benter, T., 2007. Measurement of nucleation andsoot mode particle emission from a diesel passenger car in real world andlaboratory in situ dilution. Atmospheric Environment 41, 2125e2135.

Charron, A., Harrison, R.M., 2003. Primary particle formation from vehicle emissionsduring exhaust dilution in the roadside atmosphere. Atmospheric Environment37, 4109e4119.

Corbett, J.J., Winebrake, J.J., Green, E.H., Kasibhatla, P., Eyring, V., Lauer, A., 2007a.Mortality from ship emissions: a global assessment. Environmental Science andTechnology 41 (24), 8512e8518. doi:10.1021/es071686z.

Derwent, R., Stevenson, D.S., Doherty, R.M., Collins, W.J., Sanderson, M.G.,Amann, M., Dentener, F., 2005. The contribution from ship emissions to airquality and acid deposition in Europe. Ambio 34, 54e59.

Domínguez-Rodríguez, A., Abreu-Afonso, J., Rodríguez, S., Juarez-Prera, R.A.,Arroyo-Ucar, E., Jimenez-Sosa, A., Gonzalez, Y., Abreu-Gonzalez, P., Avanzas,P., 2011. Comparative study of ambient air particles in patients hospitalizedfor heart failure and acute coronary syndrome. Revista Española de Car-diología, in press, doi:10.1016/j.recesp.2010.12.017.

Eyring, V., Köhler, H.W., van Aardenne, J., Lauer, A., 2005. Emissions from interna-tional shipping: 1. The last 50 years. Journal of Geophysical Research 110,D17305. doi:10.1029/2004JD005619.

Fernández-Camacho, R., Rodríguez, S., de la Rosa, J., Sánchez de la Campa, A.M.,Viana, M., Alastuey, A., Querol, X., 2010. Ultrafine particle formation in theinland sea breeze airflow in Southwest Europe. Atmospheric Chemistry andPhysics 10, 9615e9630.

Please cite this article in press as: González, Y., et al., Ultrafine particleEnvironment (2011), doi:10.1016/j.atmosenv.2011.06.002

Geiser, M., Rothen-Rutishauser, B., Kapp, N., Schürch, S., Kreyling, W., et al., 2005. Ultra-fine particles cross cellular membranes by nonphagocytic mechanisms in lungs andin cultured cells. Environmental Health Perspect 113 (11). doi:10.1289/ehp.8006.

Harrison, R.M., Jones, A.M., 2005. Multisite study of particle number concentrationsin urban air. Environmental Science and Technology 39, 6063e6070.

Kasper, A., Aufdenblatten, S., Forss, A., Mohr, M., Burtscher, H., 2007. Particulateemissions from a low-speed marine diesel engine. Aerosol Science and Tech-nology 41, 24e32.

Kittelson, D.B., 1998. Engines and nanoparticles: a review. Journal of Aerosol Science29, 575e588.

Lack, D., Corbett, J.J., Onasch, T., Lerner, B., Massoli, P., Quinn, P.K., Bates, T.S.,Covert, D.S., Coffman, D., Sierau, B., Herndon, S., Allan, J., Baynard, T., Lovejoy, E.,Ravishankara, A.R., Williams, E., 2009. Articulate emissions from commercialshipping: chemical, physical, and optical properties. Geophysical ResearchLetters 114, D00F04. doi:10.1029/2008JD011300.

Lauer, A., Eyring, V., Hendricks, J., Jöckel, P., Lohmann, U., 2007. Global modelsimulations of the impact of ocean-going ships on aerosols, clouds, and theradiation budget. Atmospheric Chemistry and Physics 7, 5061e5079. http://www.atmos-chem-phys.net/7/5061/2007/.

Mejía, J.F., Morawska, L., Mengersen, K., 2008. Spatial variation in particle numbersize distributions in a large metropolitan area. Atmospheric Chemistry andPhysics 8, 1127e1138. doi:10.5194/acp-8-1127-2008.

Moldanová, J., Fridell, E., Popovicheva, O., Demirdjian, B., Tishkova, V., Faccinetto, A.,Focsa, C., 2009. Characterisation of particulate matter and gaseous emissionsfrom a large ship diesel engine. Atmospheric Environment 43, 2632e2641.

Oberdörster, G., Sharp, Z., Atudorei, V., Elder, A., Gelein, R., Kreyling, W., Cox, C.,2004. Translocation of inhaled ultrafine particles to the brain. Inhalation Toxi-cology 16, 1e9.

Petzold, A., Schönlinner, M., 2004. Multi-angle absorption photometryda newmethod for the measurement of aerosol light absorption and atmospheric blackcarbon. Journal of Aerosol Science 35, 421e441.

Petzold, A., Feldpausch, P., Fritze, L., Minikin, A., Lauer, P., Kurok, C., Bauer, H., 2004.Particle emissions from ship engines. Journal of Aerosol Science. http://www.puertos.es Abstracts of the European conference, S1095eS1096. Puertos delEstado, Official Manager of Ports in Spain.

Reche, C., Querol, X., Alastuey, A., Viana, M., Pey, J., Moreno, T., Rodriguez, S., Fer-nández-Camacho, R., González, y., de la Rosa, J., Dall’Osto, M., Prevot, A.S.H.,Hueglin, C., Harrison, R.M., Quincey, P., 2011. Variability of levels of PM, blackcarbon and particle number concentration in selected European cities. Atmo-spheric Chemistry and Physics Discussions 11, 8665e8717.

Rodríguez, S., Cuevas, E., 2007. The contributions of ‘minimum primary emissions’and ‘new particle formation enhancements’ to the particle number concen-tration in urban air. Journal of Aerosol Science 38, 1207e1219.

Rodríguez, S., Van Dingenen, R., Putaud, J.P., Dell’Acqua, A., Pey, J., Querol, X.,Alastuey, A., Chenery, S., Ho, K.-F., Harrison, R., Tardivo, R., Scarnato, B.,Gemelli, V., 2007. A study on the relationship between mass concentrations,chemistry and number size distribution of urban fine aerosols in Milan, Bar-celona and London. Atmospheric Chemistry and Physics 7, 2217e2232.doi:10.5194/acp-7-2217-2007.

Rose, D., Wehner, B., Ketzel, M., Engler, C., Voigtländer, J., Tuch, T., Wiedensohler, A.,2006. Atmospheric number size distributions of soot particles and estimation ofemission factors. Atmospheric Chemistry and Physics 6, 1021e1031.

Shi, J.P., Khan, A.A., Harrison, R.M., 1999. Measurements of ultrafine particleconcentration and size distribution in the urban atmosphere. Science of theTotal Environment 235, 51e64.

Stern, N., 2007. Stern Review on the Economics of Climate Change. HM Treasury,The Stationary Office, London.

Viana, M., Amato, F., Alastuey, A., Querol, X., Moreno, T., García do Santos, S.,Herce, M.D., Fernández-Partier, R., 2009. Chemical tracers of particulate emis-sions from commercial shipping. Environmental Science and Technology 43,7472e7477.

Wehner, B., Wiedensohler, A., 2003. Long term measurements of submicrometerurban aerosols: statistical analysis for correlations with meteorological condi-tions and trace gases. Atmospheric Chemistry and Physics 3, 867e879.doi:10.5194/acp-3-867-2003.

Wehner, B., Birmili, W., Gnauk, T., Wiedensohler, A., 2002. Particle number sizedistributions in a street canyon and their transformation into the urban-airbackground: measurements and a simple model study. Atmospheric Environ-ment 36, 2215e2223.

Williams, E.J., Lerner, B.M., Murphy, P.C., Herndon, S.C., Zahniser, M.S., 2009.Emissions of NOx, SO2, CO, and HCHO from commercial marine shipping duringTexas Air Quality Study (TexAQS) 2006. Journal of Geophysical Research 114,D21306. doi:10.1029/2009JD012094.

Woo, K.S., Chen, D.R., Pui, DY.H., Mc Murry, P.H., 2001. Measurement of Atlantaaerosol size distributions: observation of ultrafine particle events. AerosolScience and Technology 34, 75e87.

s pollution in urban coastal air due to ship emissions, Atmospheric