Research Article Identification and Characterization...

Hindawi Publishing CorporationAdvances in MeteorologyVolume 2013 Article ID 380614 11 pageshttpdxdoiorg1011552013380614

Research ArticleIdentification and Characterization of Black Carbon AerosolSources in the East Baltic Region

Steigvil ByIenkien Vidmantas Ulevicius Vadimas Dudoitis and Julija Paurait

Center for Physical Sciences and Technology Savanoriu 231 LT-02300 Vilnius Lithuania

Correspondence should be addressed to Vidmantas Ulevicius ulevicvktlmiilt

Received 19 February 2013 Accepted 27 April 2013

Academic Editor Giulia Pavese

Copyright copy 2013 Steigvile Bycenkiene et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

One-year continuous measurements of aerosol black carbon (BC) at the background site Preila (55∘551015840N 21∘001015840E 5m aslLithuania) were performed Temporal and spatial evolution and transport of biomass burning (BB) and volcanic aerosols observedwithin this period were explained by the air mass backward trajectory analysis in conjunction with the fire detection data producedby theMODIS Rapid Response System andAERONETdatabaseThe surfacemeasurements and analysis of the Angstrom exponentof the absorption coefficient done separately for shorter and longer wavelengths (ie 120572 = 370ndash520 nm and 120572 = 660ndash950 nm)showed that high levels of aerosol BC were related to the transport of air masses rich in BB products fromUkraine caused by activegrass burning During the events the highest mean values of the Angstrom exponent of the absorption coefficients 120572

370ndash520 and120572590ndash950 nm were observed (24 plusmn 01 and 15 plusmn 01 resp) The ash plume of the Grimsvotn eruption on May 21 2011 offered an

exceptional opportunity to characterize the volcanic aerosols The largest ash plume (in terms of aerosol optical thickness) overLithuania was observed at May 2425 2011The highest values of the Angstrom exponent of the absorption coefficients 120572

370ndash520 and120572590ndash950 nm were reached (13 plusmn 01 and 14 plusmn 01 resp)

1 Introduction

Atmospheric aerosol particles which are very small and sus-pended in the air affect Earthrsquos climate directly by scatteringand absorbing the atmospheric radiation and indirectly byacting as cloud condensation nuclei and modifying the opti-cal properties of clouds Furthermore human exposure tofine particulate matter is associated with adverse impacts onhuman health [1ndash3] and the environment (climate visibilityand biogeochemical cycles) [4]

High concentrations of carbonaceous particles contain-ing BC or organic compounds (OC) (organic carbon) arefound mainly in the submicron size range The main naturalsources of BC are volcanic eruptions forest fires and so forthWildfires (biomass burning) emit a variety of trace gasesand particulates [5ndash9] The growing recognition of BB as awidespread and significant agent of change in Earthrsquos systemhas led to an ongoing need for fire data on the regionalcontinental and global scale Due to high sorption capacityand optical properties black carbon originated from BB is of

great importance to the atmosphere processes and is a goodindicator of anthropogenic air pollutants Black carbon fromBB alters chemical and physical properties of the atmosphereand snow albedo

The eruption of Grimsvotn in Iceland (May 21 2011)initiated a great number of research activities to characterizethe physical chemical and optical properties of airbornevolcanic material associated with this eruption The mainchallenge in measuring volcanic material in the surface airis that this material is always mixed with other pollutantspresent in the air A large fraction of carbonaceous compo-nent of the aerosol (in upper tropospherelower stratosphere)was identified by Murphy et al [10] and Martinsson et al[11] Volcanic eruptions inject large quantities of ash and gasesinto the atmosphere Despite the evident significances of car-bonaceous aerosol in air chemistry and physics informationconcerning their temporal variability is still quite incomplete

The long-range transport (LRT) of atmospheric particu-late matter (PM) is a transboundary problem that can havesignificant impacts on PM

10and PM

25levels in background

2 Advances in Meteorology

European areas [12] LRT can cause high PM25

concentrationpeaks when air masses arrive under suitable meteorologicalconditions from regions with high emissions of particlesandor their precursor gases Many research studies havebeen performed and reported for the background atmo-spheric monitoring of LRT [13ndash15] Long-range transport ofemissions from agricultural fires and volcanoes to Europewas recognized previously in USA Europe and Russia [16ndash20] In addition to European transport LRT of smoke fromRussian biomass burning has been recorded in MongoliaChina Japan and Korea [21] and studies have revealedthat hemispheric-scale transport is possible [16] As themass concentrations of fine particles are commonly lower inNorthern Europe countries compared with those in CentralEurope [22 23] LRT can have a notable impact on the aerosolparticle levels under favorable conditions [23 24] The LRTof emissions from wildfires from the Ukraine and Europeanpart of Russia frequently increases the particulate matterconcentrations in these countries during spring and summer[25ndash27] Volcanic dust plumes can also be transferred overlong distances by air masses affecting the air quality not onlyover Europe but in other parts of the world as well [28]

Combining the variety of terrestrial information oper-ationally derived from satellite data including those fromNASArsquosMODIS instrument enables linking the different tem-poral and spatial scales The use of the Angstrom exponent120572 has significantly increased in the last years because thisparameter is easily estimated using automated-surface sunphotometry while it is becoming increasingly accessible tosatellite retrievals [29] Moreover the value of the Angstromexponent is a qualitative indicator of the aerosol particle sizeor fine mode fraction [30]

In this work the pollution events due to biomass burningthat occurred in Lithuania onMarch 21ndash25 2011 andpollutionepisode caused by long-range transported volcanic materialfrom the Grimsvotn eruption that took place at the end ofMay 2011 in Iceland were investigated Air mass backwardtrajectories fire satellite observations dispersion modelingresults and emission source data were used to identify thesource The mass and number concentrations of PM

25in

conjunction with Angstrom exponent of the absorption coef-ficient and other optical properties of aerosol were studied

2 Methodology

21 Site Description The aerosol particle concentration mea-surements were performed at the Preila environmental pol-lution research station (55∘551015840N 21∘001015840E 5m above sea level)in the coastalmarine environment This station is located onthe Curonian Spit which separates the Curonian Lagoon andthe Baltic Sea and thus can be characterized as a regionallyrepresentative background area The Curonian Lagoon thelargest coastal bay in the Baltic Sea is a highly eutrophiedwater body One of the nearest industrial cities Klaipeda isat a distance of about 40 km to the north and the other majorcity Kaliningrad is 90 km to the south from Preila

22 Long-Range Transport and Air Mass Backward Trajecto-ries Air mass backward trajectory analysis provides a better

understanding of air flow and long-range transport patternsduring the events We analyzed the aerosol characteristicswith respect to categorized air mass backward trajectories forthe initial estimation of the wildfire and volcanoes potentialsource locations and quantitative contribution Backwardtrajectories were produced using the Flextra model (NILUInstitute of Meteorology and Geophysics Vienna) [31] andmeteorological data provided fromECMWF (EuropeanCen-tre for Medium Range Weather Forecast) The potentialsource areas of long-range transported pollutionwere studiedusing 7-day backward air mass trajectories with a startingheight of 500 1000 and 1500m above the ground levelTrajectories were produced at 6 h intervals for each eventThe height along the trajectories is indicated by colour(colourheight scale in the lower left corner of the plot) Each3-hour interval along the trajectory path is indicated by asmall legend and each 24-hour interval by a big legend

To identify possible events of regional transport hotspotfire the location of a thermal anomaly was detected byMODIS using data from the middle infrared and thermalinfrared bands [32] Each MODIS sensor achieves globalcoverage once per day and once per night every 24 hTherefore most fires detectable at a 1 km spatial resolutionhave the potential to have their FRP (Fire Radiative Power)measured four times a day Figure 2 shows the locations ofthe active fires identified from MODIS observations duringstudy events over the area of our interest The MODISmonthly fire maps available since 2001 showed that theseevents of biomass burning inMarch-April occurred annuallyHowever the duration geographical extent and emission ofthe smoke from these fires differ year by year

The AErosol RObotic NETwork (AERONET) is aground-based network of Cimel sun photometer that mea-sures the extinction aerosol optical depth at 7 wavelengths120582 (340 380 440 500 670 870 and 1020 nm) from thedirect sun measurements over locations The AOD measure-ments are accurate to within plusmn001 The size distribution ofaerosols can be estimated from spectral AOD (typically 440ndash870 nm) The negative slope of AOT with the wavelengthon a logarithmic scale is known as the Angstrom parameter(120572) This parameter (1) can be calculated from two or morewavelengths using the least squares fit Values of 120572 gt 20indicate that fine mode particles exist (eg smoke particlesand sulphates) while values of 120572 near zero indicate thepresence of coarse mode particles such as desert dust [33]

120572 = minus

119889 ln120591119886

119889 ln 120582= minus

ln (12059111988621205911198861)

ln (12058221205821)

(1)

where 120572 is the wavelength exponent depending on the ratioof the particle coarse mode to the accumulation (small)one (Angstrom parameter) 120591

119886is the AOD and 120582 is the

wavelengthThe Navy Aerosol Analysis and Prediction System

(NAAPS) model results were used to determine the dis-tribution of aerosols from fires and volcanoes (modeldescription and results are available from the web pagesof the Naval Research Laboratory Monterey CA USAhttpwwwnrlmrynavymilaerosol) The NAAPS model

Advances in Meteorology 3

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec0

1000

2000

3000

800012000

BC co

ncen

trat

ion

(ng m

minus3)

(a)

0 500 1000 1500 2000 2500 3000 3500 4000 4500

1

10

100

N p

er b

in

BC concentration (ng mminus3)

Equation

Data black carbonModel log normal

1147497 plusmn 22477A09 plusmn 002w2845 plusmn 90xc21 plusmn 12y0

R2 = 09

y = y0 + A(radic(2 lowast PI) lowast w lowast x) lowast exp(minus(ln(xxc))2(2 lowast w2))

1205942D ∘ F = 1628

(b)

Figure 1 (a) Box and whisker plots for comparison of mass concen-trations at Preila (b) BC concentration frequency distribution

has been modified to incorporate real-time observations ofbiomass burning based on the joint NavyNASANOAAFire Locating and Modeling of Burning Emissions system(FLAMBE httpwwwnrlmrynavymilaerosol) [34] Themethod has proven to be helpful in previous studies of long-range and regional transport of smoke [35]

23 Black Carbon and Aerosol Particle Size DistributionMeasurements A Magee Scientific Company AethalometerModel AE31 Spectrum manufactured by Optotek Sloveniawas deployed at the Vilnius site and provided real-time con-tinuous measurements of the BC mass concentrations Theoptical transmission of carbonaceous aerosol particles wasmeasured sequentially at seven wavelengths 120582 (370 450 520590 660 880 and 950 nm) The 880 nm is considered as thestandard channel for BC measurements as at this wavelengthBC is the principal absorber of lightTheAethalometer outputis calculated directly as the BC concentrations through aninternal conversion use assumed mass absorption efficiencyThe Aethalometer converts light attenuation to the BCmass using a fixed specific attenuation cross-section (120590) of166m2 gminus1 of BC (Aethalometer Operations manual MageeScientific) [36]

This is the default value set by the manufacturer for awavelength of 880 nmThe Aethalometer data recorded witha 5-minute time base were compensated for loading effects

using an empirical algorithm [36] The Aethalometer wasequipped with an additional impactor removing the particleslarger than 25120583m of the particle aerodynamic diameter Thestarting time referred in this paper is Greenwich Mean Time(for local time GMT+200)

The absorption coefficient 119887abs for airborne particles isdefined with Lambert-Beerrsquos law as follows

119868 = 1198680119890(minus119887abs119909) (2)

where 1198680is the intensity of the incoming light and 119868 is the

remaining light intensity after passing through a mediumwith thickness 119909

A power law fit is commonly applied to describe thewavelength dependence of aerosol absorption coefficient 119887abs

119887abs prop 120582minus120572

(3)

where 120582 is the wavelength and 120572 is the usually calledAngstrom exponent of the absorption coefficient and iscomputed by fitting an exponential curve [37]

The aerosol size distribution was measured by using theScanning Mobility Particle Sizer (SMPS) built by the LeibnizInstitute for Tropospheric Research Germany The SMPS iscomposed of a differential mobility analyzer and CPC UF-02protoThis system had the following general specificationssize range 8ndash900 nm scan time 5min resolution 71 sizechannels

3 Results

In general BC particles are mainly produced by anthro-pogenic activities such as fossil fuel burning (industries andtransport) and biomass burning (agriculture or wildfires)However the natural sources such as wildfires and volcaniceruptions also contribute to the BC mass concentration buttheir contributions are relatively much less We focused onthe strongest periods of wildfire and volcanic eventsThe firstevent occurred on April 24ndash30 2011 and the second on May25ndash27 2011 The concentrations observed during both eventsare clear outliers in their respective series both for hourlyconcentrations and 24 h means

31 Annual Variations of Aerosol Black Carbon Box andwhisker plots of BC concentrations and mass concentrationfrequency distribution are shown in Figures 1(a) and 1(b)Therange of the box depicts the bounds of the 25th and the 75thpercentiles of the data The whiskers extending from the boxrepresent the bounds of the 10th and the 90th percentiles

The median and outliers of each dataset are also shownIt can be clearly seen that higher concentrations of BCare observed during October to March when cold weatherconditions prevail during any year

The concentration shows a decrease in June (280plusmn230 ngmminus3) with the monthly mean BC concentration falling tomore than half the value that prevailed in cold monthsThis low value is maintained until September During thisperiod no influence of residential heating was observed atthe site The mean BC concentration during this period is


(a)

(a)

65

55

45

35

40minus30 minus20 minus10 0 10 20 30

Smoke surface concentration (ug mminus3)for 1800 April 26 2011

(b)

0600

Height (m asl)Preila500m1000m1500m1500 2500 3500 4500 5500 6500500 7500

April 27 2011

(c)

20110427 1200

(d)

Figure 2 Active fires detected during April 24ndash30 2011 during high BC concentration event (a) by theMODIS Rapid Response System ((b)(d)) sulphate dust and smoke surface concentration (120583gmminus3) for 1800 April 26 2011 and (c) 7-day air mass backward trajectories to Preila(at 500 1000 and 1500m AGL) during the event of grass fires

370 plusmn 300 ngmminus3 in contrast to the value of 750 plusmn 550 ngmminus3during cold months (in blue)

Taking 25 ngmminus3 value as frequency statistics step thefrequency distribution map of hourly average concentrationof BC aerosol is depicted in Figure 1(b) The frequencydistribution of hourly average BC aerosol does not havenormal characterization Using the log-normal distributionfunction to fit the frequency distribution characteristicsresults show that approximately 70 of the measurementshourly average concentrations of BC aerosols were foundin a mode in a low range of concentrations centered atapproximately 200 ngmminus3 indicating a direct impact of localemissions from combustion activities at the stationMean BCmass concentrations (580 ngmminus3) observed at Preila in 2011are relatively low comparable to that measured at the remotelocation Comparingwith the values reported from elsewherefor rural sitesmeanBC values at Preila are comparable to thatreported for Zappoli et al [38] detected BC concentrations of600 ngmminus3 at a background site in Hungary (K-Puszta)

32 Spectroscopic Properties of Carbonaceous Aerosol

321 April 24ndash30 2011 Event Fire Mapper products haveprovided data about the location and extent of fires duringtheevent days (Figure 2) Satellite images andmodel forecastsfrom NAAPS (Figures 2(b) 2(d)) for the study periodsuggested substantial emissions and LRT of smoke fromthe large fires burning in the southern part of Ukraine andKaliningrad at that time

Grass burning productswere redistributed over Lithuaniaby the large-scale and regional atmospheric circulation Theair mass backward trajectory analysis showed that particlesfrom BB with air masses in 3 days were brought to Lithuaniaand caused high peaks of aerosol number and BC concentra-tions BC and aerosol number concentrations rose to higherlevels over the whole territory of Lithuania During this eventthe 1-hour average aerosol particle number concentrationreached 4000 cmminus3 BC-3500 ngmminus3 while normally averageconcentration values are about 3100 cmminus3 and 580 ngmminus3


22

2

18

16

14

12

Ang

strom

par

amet

er

Angstrom parameterAOD

AOD

1

08

06

04

02

0

1

08

06

04

02

01 5 10 15 20 25 30

April 2011

(a)

3

28

26

24

22

2

18

16

14

12

0 2 4 6 8 10 12 14 16 18 20 2223

370ndash950nm370ndash520nm590ndash950nm

120572

April 29 2011

(b)

Figure 3 (a)The spectral variation of instantaneous measurements of AOD for April 2011 showing the gradual buildup and decline of smokeconcentrations in this period (April 25ndash29 blue cycle) (b) values of the Angstrom exponent of the absorption coefficient (April 29 2011)showing dynamic variations as a result of both increasing (decreasing) AOD and BC concentration and greater domination of fine modeversus coarse mode optical depth

respectively It should be noted that during April 24ndash302011 urban sites belonging to the Lithuanian AutomaticUrban Network (EPA) (httpstotelesgamtalt) experiencedpeaks of PM

10concentration which coincided with extensive

wildfires in Ukraine exceeding the European limit value 15times

Data analysis of the BC combinedwith TERRAModerateResolution Imaging Spectroradiometer MODIS fire detec-tions (Figure 2(a)) and NAAPS aerosol optical depth pro-vided some insight into the aerosol regional transport ofsmoke As shown in Figure 2(a) theMODIS clearly illustratesthe fire location The model forecasts from NAAPS forthis period suggested substantial emissions and regionaltransport of smoke (Figure 2(c)) from the large fires burningin Kaliningrad and Ukraine regions at that time During thestudy the NAAPS model predicted the optical depth of 06 ata wavelength of 055 microns for three components sulfateand smoke over Europe due to smoke from the wildfires(Figure 2(d)) On the same days substantial concentrations ofBC were observed during the period after air masses arrivedfrom the fire location

During the air mass transport to Lithuania aerosoloptical properties changed due to both deposition andmixingwith local aerosols The similar events were detected andstudied by lidar networks covering the whole continent [3940] and also by a single lidar station [41] Smoke plumewas detected by the lidar and sun-photometric stations inMinsk Belarus Poland and Belsk [42] Belsk site is a jointAERONET site and it was possible to characterize the opticalproperties of aerosol particles at this site during our casestudy From the first general overview of the field data April

BB event can be recognized In this study the values ofthe Angstrom parameter were computed in the wavelengthinterval of 440ndash870 nm using the AOD data of the selectedAERONET Belsk site applying the least-squares methodAERONET profiles of the absorption Angstrom parameter(indicator of the AOD spectral dependence and related toparticle size predominance) and AOD values at 500 nmobtained from MODIS for a period between April 1 and30 2011 are presented in Figure 3(a) The AOD measuredin Belsk reached 090 plusmn 002 (these results are based on theAERONET cloud-screened and quality-assured data (level20)) It can be seen that the Angstrom parameter valuesare mostly in the 14ndash20 range suggesting the presence ofspecies that absorb more strongly at shorter wavelengthsthan black carbon which might be associated with organicsfrom biomass burning [43] It is known that typical values ofthe Angstrom parameter range sim2 for fresh smoke particles(accumulation mode) to sim0 for coarse mode particles [33]

The wavelength dependence of the light absorption canbe better approximated by separate exponential fits of theshorter (370ndash520 nm) and longer (660ndash950 nm) wavelengthsobtained by an exponential curve fit over all 7 wavelengthsTheAngstromexponentswere calculated by fitting 119887abs for thewhole available wavelength intervals the short (370ndash520 nm)and long (590ndash950 nm) wavelength intervals (Figure 3(b))

On the event day (April 29) the mean 120572 values weresignificantly larger for the longer wavelengths (120572590ndash960 =22 plusmn 01) compared to the shorter wavelengths (120572370ndash520 =14 plusmn 01) During the event day the highest values of120572590ndash960 were observed between 11 AM and 11 PM (27 plusmn01) while the highest values of 120572370ndash520 were observed


between midnight and morning (17 plusmn 01) On noneventdays the light absorption coefficients 120572370ndash950 120572370ndash520 and120572660ndash950 nm were within a narrow range of 12 with weakdiurnal cycles For example 120572 value up to 15 was observedat a high Alpine station during a Saharan dust event [44]

Kirchstetter et al [45] reported 120572 value of 22 for outdoorfirewood burning 18 for a savanna fire and 08ndash11 fortraffic-dominated sites (measurements with six wavelengthsbetween 370 and 850 nm) Schnaiter et al [46 47] reported 120572value of 11 for uncoated diesel soot (measurements with 120582 =450 550 and 700 nm) Previous investigations showed thatthe type of wood being burned also influenced the 120572 valueas shown by Day et al [48] where they measured fresh woodsmoke from seven types of forest wood with an Aethalometer(120582 = 370ndash950 nm) and reported120572 values between 09 and 22Hoffer et al [49] calculated 120572 = 6ndash7 for water-soluble humic-like substances (HULIS) isolated from the fine fraction of thebiomass burning aerosol spectrophotometer measurementwith 120582 = 300ndash700 nm)

These very high values of the absorption exponent areindicative of biomass combustion [37] in our casemdashwildfiresas indicated by the air mass backward trajectory analysis(Figure 1) [50 51] Though the increase in BC concentrationassociated with the BB event was instantaneous the gradualdecay after some days instantly recovered the BC concentra-tion to the preevent level despite the fact of long atmosphericlifetime of BC Reddy and Venkataraman [52] have found anannual average value of sim65 plusmn 48 days for the BC lifetimeGlobally the average lifetime of BC is about 8 days

The observed variations in the Angstrom exponent at590ndash950 nm derived from the BCmeasurements are not con-sistent with AERONET Angstrom parameter (Figures 3(a)3(b)) These comparisons show that there are differencesbetween these two techniques for obtaining column averageaerosol optical properties Every aerosol technique has itsadvantages and disadvantages Some are inherent in theinstrument design or measurement type (eg sun photome-ters cannot measure valid AOD during cloudy periods)

322 May 24ndash27 2011 Event Figure 4 shows the air massarriving at Preila at 1200 (02 GMT) on May 24ndash27 2011 Asillustrated in Figure 4 the air mass came from the northeastpassing over Grimsvotn from which it was two-three daysbefore arrival at the site Results confirm that the volcanicplume arrived during the night between May 25 and May26 2011 after 72 h of transportation The arrival elevation of500m agl locates within the inversion layer so that it isrepresentative of the air sampled at Preila

The sulphate concentration in the air was used as anindicator of volcanic pollution as sulphur dioxide is the thirdmost abundant gas in volcanic emissions after water vaporand carbon dioxide [53] One week after the eruption theSO2cloud reached Europe where it was detected by ground-

based instruments At the Preila station the plume was firstdetected on May 24-25 as SO

2concentrations peaked until

reaching a maximum value of 136mgmminus3 and made up14 of submicron aerosol particles [20] The advection ofvolcanic plumes from Grimsvotn was also confirmed usingthe satellite images For the UVvisible instrument GOME-2

the Differential Optical Absorption Spectroscopy was usedAfter the eruption the plume drifted to the southeast whereit formed a distinct loop due to atmospheric winds GOME-2 measured SO

2column amounts on May 29 of gt10DU

after 8 days after the eruption (httpsacsaeronomiebe)Time variations of sulphate AOD observed duringMay 17ndash312011 showed that the AOD values were significantly larger onMay 28-29 whereas extremely low sulphate concentrationsprevailed during the period from 25 to 26 May Thus weconclude that the high values of sulphate AODmeasurmentsmeasured during the first half of the observation period aftereruption were caused by the volcanic plumes transportedfrom Grimsvotn

A carbonaceous aerosol component has been observedin fresh volcanic clouds from several volcanoes [11] TheAngstrom exponent of the absorption coefficient for organicspecies depends on the wavelength Studies show 120572 value fororganic species is larger than for black carbon [45] whereas itis less dependent on the wavelength for BC The dependenceof 120572 on the wavelength can be seen in Figure 5 for the studycase

Figure 5 shows time series for the hourly mean values of120572 from May 24 to May 28 2011 Since at the UV wavelengththere could be some organic compound strongly absorbingthus leading to an overestimation of the 120572 parameter thewavelength range 370ndash950 nm has been considered for thebest-fit technique application and 120572 parameter estimationThe fact that120572370ndash550 becomes higher than120572590ndash950 shows thatthe influence of the organic aerosols exists but its influenceis limited probably connected to changes in variations inthe composition of air masses crossing the Preila site BCconcentration maximum values (at 880 nm) depending onthe day were varying between 150 and 1200 ngmminus3 In Preilamean 120572370ndash550 values (Figure 5) collected in May 25-26 2011have been found to be sim19 plusmn 05 while measurementscollected during May 24 and 27-28 2011 give mean alphavalue sim15 plusmn 05 Hence it can be concluded that low BCcontributed mainly to the light absorption which is wellconsistent with an average 120572370ndash550 and 120572590ndash950 close to 15and 18 respectively Explanation of the carbon content ofthe air other than direct volcanic emissions needs to beconsidered [54]

Parallel to the measurement of 120572 the aerosol absorptioncoefficient 119887abs (120582) was evaluated (Figure 6) The spectraldependence allows distinguishing carbonaceous aerosolsfrom different sources because of light absorbing OC whichin contrast to BC exhibits a stronger absorption at shorterwavelengths [43] In the period during the volcanic activitythe value of 120572370ndash550 was slightly lower (13 plusmn 03) thanthat of 120572590ndash950 (14 plusmn 02) and was contrary to the valueafter air masses passed the Preila site that is 120572370ndash550 and120572590ndash950 nm of 21 plusmn 04 and 24 plusmn 04 respectively This is astrong suggestion that the aerosol includes less absorbers ofthe shorter wavelengths during the volcanic event Accordingto values of the Angstrom exponent of the absorptioncoefficients 120572370ndash550 and 120572590ndash950 nm along with the air massbackward trajectories calculation we can assume that thedecrease in the organic carbon concentration was caused bythe volcano at Grimsvotn in Iceland


Height (m asl)

1500 2500 3500 4500 5500 6500500 7500

1200May 24 2011

(a)

1200

1500 2500 3500 4500 5500 6500500 7500

Height (m asl)

May 25 2011

(b)

Preila500m1000m1500m

1500 2500 3500 4500 5500 6500500 7500

1200

Height (m asl)

May 26 2011

(c)

1500 2500 3500 4500 5500 6500500 7500


1200

Height (m asl)

May 27 2011

(d)

Figure 4 Backward trajectories of the air masses from the volcano at Grimsvotn to Lithuania on 24ndash27 ((a)ndash(d)) May 2011 (1200 UTC)

It is to be noted that the Angstrom exponent could be arough indicator of the size distribution of the aerosol particleswhereas the Angstrom exponent of the absorption coefficienttells about the nature of absorbing aerosols A low 120572 value(lt1) indicates dominance of larger-size particles and a high 120572(gt1) indicates dominance of smaller-size particles [55] Thisassumption was also confirmed in the study by Kvietkus et al[20] when the particulate organic matter concentrations ofPM1were lower during the period of the volcanic erup-

tion but an increase in sulfate concentration from 113 to386 120583gmminus3 (90 of PM

1) on May 25 in the Vilnius area

was observed Baker et al [56] showed that in the volcaniccloud ofGrimsvotn chlorine radicals rapidly depleted organictrace gases to levels well below background concentrationsThis process could increase particulate carbon amount inthe volcanic cloud This could explain the increase in BCconcentration in May 27 and parallel decrease in organicfraction (Figure 5)

Volcanic particles are assumed to be composed of ashparticles and hygroscopic sulfates Volcanic particles areinjected in the atmosphere both as primary particles rapidlydeposited due to their large sizes on time scales of minutesto a few weeks in the troposphere and as secondary particlesmainly derived from the oxidation of sulphur dioxide [57]For a detailed analysis of the event the diurnal pattern ofthe time dependent size distribution for Preila during May24 2011 is depicted in Figure 7 New particle formationevents were found on 4 days (May 24ndash27) We will focushere on the event that we observed at the Preila stationon May 24 2011 In the morning of these days air massesarrived to Preila from the Baltic Sea area when higheraltitude (above 2000ndash3500m asl) air masses continued toarrive straight from the Grimsvotn the lower level (below500m asl) The aerosol also shows a high concentrationof SO

2during the experiment when air masses came from

the volcano accompanied with high sulphur (452 120583gmminus3)


3

25

2

15

1

05

0

120572

1200

1000

800

600

400

200

0

BC co

ncen

trat

ion

(ng m

minus3)

24 25 26 27 28May 2011

BC370ndash550590ndash950

Figure 5 Time series of the hourlymean Angstrom exponent of theabsorption coefficient on May 24ndash28 2011

22 times 10minus5

2 times 10minus5

18 times 10minus5

16 times 10minus5

14 times 10minus5

12 times 10minus5

1 times 10minus5

8 times 10minus6

6 times 10minus6

4 times 10minus6

2 times 10minus6

0

24120572370ndash550 = 11

120572590ndash950 = 15

25120572370ndash550 = 13

120572590ndash950 = 14

26120572370ndash550 = 14

120572590ndash950 = 18

27120572370ndash550 = 21

120572590ndash950 = 24

120582 (nm)

b abs

(mminus1)

370 450 520 590 660 880 950

Figure 6 The light Angstrom exponent of the absorption coeffi-cients 120572370ndash520 and 120572590ndash950 nm during May 24ndash27 2011

A very strong source of SO2located in the southernmaritime

area is suspected to have caused these extraordinary high SO2

concentrations As the sun is rising photochemical reactionslead gas phase SO

2to be oxidized to sulfuric acid

As shown in Figure 7 the nigh ttime 0000ndash0700AMwas associated with higher number concentration of accu-mulation mode particles (119863

119901sim 100 nm) During new

particle formation events the nucleation mode particlesshowed an obvious one-peak diurnal variation the numberconcentration was very low in the early morning started toincrease abruptly at about 900 in the morning and reachedits maximum to a value of 30000 cmminus3 On the contrarythe concentration of Aitken mode particles remained lowbetween 0000 and 900AM After this point the formation

0 3 6 9 12 15 18 21 2410

100

1000

Part

icle

dia

met

er (n

m)

Time (h)

10 100 1000 10000 100000Particle number concentration (cmminus3)

(a)

0 3 6 9 12 15 18 21 24Time (h)

105

104

103

102Part

icle

num

ber

conc

entr

atio

n (c

mminus3

)

(b)

Figure 7 Contour plot of new particle formation and growth on 24May 2011 colour represents logarithm of concentration in cmminus3

Table 1 The growth rate of new particles

Date24 May 25 May 26 May 27 May

Growth rate nmhminus1 536 255 558 2520

of new particles started and the concentration of nucleationmode particles increased Later the number of concentrationsreached its lowest level from 600 PM till the next morningThe growth rates (Table 1) of new particles were calculatedusing the method described by Kulmala et al [58]

The growth rates were in the range of 255ndash2520 nmhminus1the growth rates are in the range of typical observed particlegrowth rates (1ndash20 nmhminus1) [59]

4 Conclusions

A combination of ground-based and satellite observationswas used in this study to investigate different sources ofaerosol BC loadings over Lithuania during two different timeperiods April 24ndash30 and May 24ndash2 2011 The 7-wavelengthsetup of the Aethalometer has been exploited to obtaininformation on variation of spectroscopic properties of car-bonaceous particles Although BCmass concentration valuesdo not allow one to a priori determine which process pro-duces BC the wavelength dependence of the BC absorptiongives the option to follow the variation carbonaceous aerosolproperties and easily characterize the black carbonorganiccomponent

Near-source biomass burning smoke is observed rou-tinely in Lithuania while volcanic effluent plumes providedsome new information During these periods the measured


black carbon concentration during BB and volcanic activityevents (3500 and 1200 ngmminus3 resp) exceeded previouslyestablished average values of 580 ngmminus3 Compared to back-ground aerosols the sampled plumes have higher AODand contain particles having expected differences in theAngstrom exponent and size During the wildfire burningevent on April 24ndash30 2011 the highest mean values of theAngstrom exponent of the absorption coefficients 120572370ndash520and 120572590ndash950 nm were observed (14plusmn01 and 22plusmn01 resp)and during the volcanic event on May 25ndash27 2011 the peakvalues were determined (13 plusmn 03 and 14 plusmn 02 resp) This isa strong suggestion that the aerosol includes less absorbers ofthe shorter wavelengths during the volcanic event

The observational data and analysis presented heredemonstrate that nucleation and subsequent growth can bederived from volcanic eruption gaseous released and that thisnew secondary particle formation event could occur withinthe lower troposphere at a large distance from the eruptiveactivity

Acknowledgments

This work was supported by the Lithuanian-Swiss Coop-eration Programme ldquoResearch and Developmentrdquo projectAEROLIT (Nr CH-3-SMM-0108) The authors thank PiotrSobolewski Brent Holben and Aleksander Pietruczuk andtheir staff for establishing andmaintaining the Belsk site usedin this investigation The authors declare that they have nocompeting interests as defined by Advances in Meteorologyor other interests that might be perceived to influence theresults and discussion reported in this paper

References

[1] M L Bell J M Samet and F Dominici ldquoTime-series studies ofparticulate matterrdquo Annual Review of Public Health vol 25 pp247ndash280 2004

[2] C A Pope andDWDockery ldquoHealth effects of fine particulateair pollution lines that connectrdquo Journal of the Air and WasteManagement Association vol 56 no 6 pp 709ndash742 2006

[3] J P Schwarz J R Spackman D W Fahey et al ldquoCoatingsand their enhancement of black carbon light absorption in thetropical atmosphererdquo Journal of Geophysical Research vol 113no 3 Article ID D03203 2008

[4] V Ramanathan F Li M V Ramana et al ldquoAtmosphericbrown clouds hemispherical and regional variations in long-range transport absorption and radiative forcingrdquo Journal ofGeophysical Research vol 112 no 22 Article ID D22S21 2007

[5] P J Grutzen and M O Andreae ldquoBiomass burning in thetropics impact on atmospheric chemistry and biogeochemicalcyclesrdquo Science vol 250 no 4988 pp 1669ndash1678 1990

[6] P G Simmonds A J Manning R G Derwent et al ldquoA burningquestion can recent growth rate anomalies in the greenhousegases be attributed to large-scale biomass burning eventsrdquoAtmospheric Environment vol 39 no 14 pp 2513ndash2517 2005

[7] M O Andreae and P Merlet ldquoEmission of trace gases andaerosols from biomass burningrdquo Global Biogeochemical Cyclesvol 15 no 4 pp 955ndash966 2001

[8] Y J Kaufman D Tanre and O Boucher ldquoA satellite view ofaerosols in the climate systemrdquo Nature vol 419 pp 215ndash2232002

[9] H Huntrieser J Heland H Schlager et al ldquoIntercontinentalair pollution transport from North America to Europe exper-imental evidence from airborne measurements and surfaceobservationsrdquo Journal of Geophysical Research vol 110 no 1 pp1ndash22 2005

[10] D M Murphy D S Thomson and M J Mahoney ldquoIn situmeasurements of organics meteoritic material mercury andother elements in aerosols at 5 to 19 kilometersrdquo Science vol282 no 5394 pp 1664ndash1669 1998

[11] B G Martinsson C A M Brenninkmeijer S A Cam et alldquoInfluence of the 2008 Kasatochi volcanic eruption on sulfurousand carbonaceous aerosol constituents in the lower strato-sphererdquo Geophysical Research Letters vol 36 no 12 Article IDL12813 2009

[12] S S Abdalmogith and R M Harrison ldquoThe use of trajectorycluster analysis to examine the long-range transport of sec-ondary inorganic aerosol in the UKrdquoAtmospheric Environmentvol 39 no 35 pp 6686ndash6695 2005

[13] ldquoTF-HTAP Hemispheric Transport of Air Pollutionrdquo in AirPollution Studies no 16UnitedNations EconomicCommissionfor Europe NewYork NYUSA andGeneva Switzerland 2007httpwwwhtaporg

[14] J F Liu D L Mauzerall and L W Horowitz ldquoSource-receptorrelationships of trans-pacific transport of East Asian sulfaterdquoAtmospheric Chemistry and Physics vol 8 pp 5537ndash5561 2008

[15] E Saikawa V Naik L W Horowitz J Liu and D L MauzerallldquoPresent and potential future contributions of sulfate black andorganic carbon aerosols fromChina to global air quality prema-turemortality and radiative forcingrdquoAtmospheric Environmentvol 43 no 17 pp 2814ndash2822 2009

[16] R Damoah N Spichtinger C Forster et al ldquoAround the worldin 17 daysmdashHemispheric-scale transport of forest fire smokefrom Russia in May 2003rdquo Atmospheric Chemistry and Physicsvol 4 no 5 pp 1311ndash1321 2004

[17] J V Niemi H Tervahattu H Vehkamaki et al ldquoCharacter-ization and source identification of a fine particle episode inFinlandrdquo Atmospheric Environment vol 38 no 30 pp 5003ndash5012 2004

[18] D Muller I Mattis U Wandinger A Ansmann D Althausenand A Stohl ldquoRaman lidar observations of aged Siberian andCanadian forest fire smoke in the free troposphere over Ger-many in 2003 microphysical particle characterizationrdquo Journalof Geophysical Research vol 110 no 17 Article ID D17201 pp75ndash90 2005

[19] B Langman A Folch M Hensch and V Matthias ldquoVolcanicash over Europe during the eruption of Eyjafjallajokull onIceland April-May 2010rdquoAtmospheric Environment vol 48 pp1ndash8 2012

[20] K Kvietkus J Sakalys J Didzbalis I Garbariene N Spirk-auskaite and V Remeikis ldquoAtmospheric aerosol episodes overLithuania after the May 2011 volcano eruption at GrimsvotnIcelandrdquo Atmospheric Research vol 122 pp 93ndash101 2013

[21] K H Lee J E Kim Y J Kim J Kim and W Von Hoyningen-Huene ldquoImpact of the smoke aerosol from Russian forest fireson the atmospheric environment over Korea duringMay 2003rdquoAtmospheric Environment vol 39 no 1 pp 85ndash99 2005

[22] M Sillanpaa R Hillamo S Saarikoski et al ldquoChemical compo-sition andmass closure of particulatematter at six urban sites in


Europerdquo Atmospheric Environment vol 40 no 2 pp 212ndash2232006

[23] P Anttila U Makkonen H Hellen et al ldquoImpact of the openbiomass fires in spring and summer of 2006 on the chemicalcomposition of background air in south-eastern FinlandrdquoAtmospheric Environment vol 42 no 26 pp 6472ndash6486 2008

[24] A Karppinen J Harkonen J Kukkonen P Aarnio and TKoskentalo ldquoStatistical model for assessing the portion of fineparticulate matter transported regionally and long range tourban airrdquo Scandinavian Journal of Work Environment andHealth vol 30 no 2 pp 47ndash53 2004

[25] M Sillanpaa A Frey R Hillamo A S Pennanen and R OSalonen ldquoOrganic elemental and inorganic carbon in particu-late matter of six urban environments in Europerdquo AtmosphericChemistry and Physics vol 5 no 11 pp 2869ndash2879 2005

[26] J Ovadnevaite K Kvietkus and A Marsalka ldquo2002 summerfires in Lithuania impact on the Vilnius city air quality and theinhabitants healthrdquo Science of the Total Environment vol 356no 1ndash3 pp 11ndash21 2006

[27] S Saarikoski M Sillanpaa M Sofiev et al ldquoChemical compo-sition of aerosols during a major biomass burning episode overnorthern Europe in spring 2006 experimental and modellingassessmentsrdquoAtmospheric Environment vol 41 no 17 pp 3577ndash3589 2007

[28] V-M Kerminen J V Niemi H Timonen et al ldquoCharac-terization of volcanic ash in southern Finland caused by theGrimsvotn eruption in May 2011rdquo Atmospheric Chemistry andPhysics vol 11 pp 12227ndash12239 2011

[29] G L Schuster O Dubovik and B N Holben ldquoAngstromexponent and bimodal aerosol size distributionsrdquo Journal ofGeophysical Research vol 111 no 7 Article ID D07207 2006

[30] Y J Kaufman A Gitelson A Karnieli et al ldquoSize distributionand scattering phase function of aerosol particles retrieved fromsky brightness measurementsrdquo Geophysical Research Lettersvol 99 pp 10341ndash10356 1994

[31] A Stohl G Wotawa P Seibert and H Kromp-Kolb ldquoInterpo-lation errors in wind fields as a function of spatial and temporalresolution and their impact on different types of kinematictrajectoriesrdquo Journal of Applied Meteorology vol 34 no 10 pp2149ndash2165 1995

[32] L Giglio ldquoCharacterization of the tropical diurnal fire cycleusing VIRS and MODIS observationsrdquo Remote Sensing ofEnvironment vol 108 no 4 pp 407ndash421 2007

[33] T F Eck B N Holben J S Reid et al ldquoWavelength dependenceof the optical depth of biomass burning urban and desert dustaerosolsrdquo Journal of Geophysical Research vol 104 no 24 pp31333ndash31349 1999

[34] J S Reid E M Prins D LWestphal et al ldquoReal-timemonitor-ing of South American smoke particle emissions and transportusing a coupled remote sensingbox-model approachrdquo Geo-physical Research Letters vol 31 no 6 Article ID L06107 5pages 2004

[35] R E Honrath R C Owen M Val Martın et al ldquoRegional andhemispheric impacts of anthropogenic and biomass burningemissions on summertime CO and O

3

in the North Atlanticlower free troposphererdquo Journal of Geophysical Research vol109 no 24 pp 1ndash17 2004

[36] A Virkkula T Makela R Hillamo et al ldquoA simple procedurefor correcting loading effects of aethalometer datardquo Journal ofthe Air and Waste Management Association vol 57 no 10 pp1214ndash1222 2007

[37] J Sandradewi A S H Prevot E Weingartner R Schmid-hauser M Gysel and U Baltensperger ldquoA study of woodburning and traffic aerosols in an Alpine valley using a multi-wavelength Aethalometerrdquo Atmospheric Environment vol 42no 1 pp 101ndash112 2008

[38] S Zappoli A Andracchio S Fuzzi et al ldquoInorganic organicand macromolecular components of fine aerosol in differentareas of Europe in relation to their water solubilityrdquoAtmosphericEnvironment vol 33 no 17 pp 2733ndash2743 1999

[39] A Ansmann J Bosenberg A Chiakovsky et al ldquoLong-rangetransport of Saharan dust to northern Europe the 11ndash16October2001 outbreak observed with EARLINETrdquo Journal of Geophysi-cal Research vol 108 no 24 pp 1ndash15 2003

[40] J G Goldammer ldquoThe wildland fire season 2002 in the RussianFederation an assessment by the global fire monitoring centre(GFMC)rdquo International Forest Fire News vol 28 pp 2ndash14 2003

[41] A E Kardas K M Markowicz S P Malinowski et al ldquoSAWAexperimentmdashproperties of mineral dust aerosol as seen by syn-ergic lidar and sun-photometer measurementsrdquo AtmosphericChemistry and Physics Discussions vol 6 pp 12155ndash12178 2006

[42] A Pietruczuk and A P Chaikovsky ldquoProperties of fire smokein Eastern Europe measured by remote sensing methodsrdquo inRemote Sensing of Clouds and the Atmosphere vol 6745 ofProceedings of SPIE October 2007

[43] H Moosmuller R K Chakrabarty and W P Arnott ldquoAerosollight absorption and its measurement a reviewrdquo Journal ofQuantitative Spectroscopy and Radiative Transfer vol 110 pp844ndash878 2009

[44] M Collaud-Coen E Weingartner D Schaub et al ldquoSaha-ran dust events at the Jungfraujoch detection by wavelengthdependence of the single scattering albedo and first climatologyanalysisrdquo Atmospheric Chemistry and Physics vol 4 no 11-12pp 2465ndash2480 2004

[45] T W Kirchstetter T Novakov and P V Hobbs ldquoEvidencethat the spectral dependence of light absorption by aerosols isaffected by organic carbonrdquo Journal of Geophysical Research vol109 no 21 Article ID D21208 12 pages 2004

[46] M Schnaiter H Horvath O Mohlter K H Naumann HSaafhoff and O W Schock ldquoUVndashVISndashNIR spectral opticalproperties of soot and soot-containing aerosolsrdquo Journal ofAerosol Science vol 34 pp 1421ndash1444 2003

[47] M Schnaiter C Linke O Mohler et al ldquoAbsorption amplifica-tion of black carbon internally mixed with secondary organicaerosolrdquo Journal of Geophysical Research vol 110 no 19 ArticleID D19204 pp 1ndash11 2005

[48] D E Day J L Hand C M Carrico G Engling and W CMalm ldquoHumidification factors from laboratory studies of freshsmoke from biomass fuelsrdquo Journal of Geophysical Research vol111 no 22 Article ID D22202 2006

[49] A Hoffer A Gelencser P Guyon et al ldquoOptical properties ofhumic-like substances (HULIS) in biomass-burning aerosolsrdquoAtmospheric Chemistry and Physics vol 6 no 11 pp 3563ndash35702006

[50] VUlevicius S Bycenkiene V Remeikis et al ldquoCharacterizationof aerosol particle episodes in Lithuania caused by long-rangeand regional transportrdquo Atmospheric Research vol 98 no 2ndash4pp 190ndash200 2010

[51] V Ulevicius S Bycenkiene N Spirkauskaite and S KecoriusldquoBiomass burning impact on black carbon aerosol mass con-centration at a coastal site case studiesrdquo Lithuanian Journal ofPhysics vol 50 no 3 pp 335ndash344 2010


[52] M S Reddy and C Venkataraman ldquoDirect radiative forcingfrom anthropogenic carbonaceous aerosols over IndiardquoCurrentScience vol 76 no 7 pp 1005ndash1011 1999

[53] R von Glasow N Bobrowski and C Kern ldquoThe effectsof volcanic eruptions on atmospheric chemistryrdquo ChemicalGeology vol 263 no 1-4 pp 131ndash142 2009

[54] S M Andersson B G Martinsson J Friberg et al ldquoCom-position and evolution of volcanic aerosol from eruptions ofKasatochi Sarychev and Eyjafjallajokull in 2008ndash2010 based onCARIBIC observations during air space closure in April andMay 2010rdquoAtmospheric Chemistry and Physics vol 13 pp 1781ndash1796 2013

[55] O Dubovik B Holben T F Eck et al ldquoVariability of absorptionand optical properties of key aerosol types observed in world-wide locationsrdquo Journal of the Atmospheric Sciences vol 59 no3 pp 590ndash608 2002

[56] A K Baker A Rauthe-Schoch T J Schuck et al ldquoInvestigationof chlorine radical chemistry in the Eyjafjallajokull volcanicplume using observed depletions in non-methane hydrocar-bonsrdquo Geophysical Research Letters vol 38 Article ID L138012011

[57] U Schumann B Weinzierl O Reitebuch et al ldquoAirborneobservations of the Eyjafjalla volcano ash cloud over Europeduring air space closure in April and May 2010rdquo AtmosphericChemistry and Physics Discussions vol 10 no 9 pp 22131ndash22218 2010

[58] M Kulmala T Petaja P Monkkonen et al ldquoOn the growthnucleation mode particles source rates of condensable vapor inpolluted and clean environmentsrdquo Atmospheric Chemistry andPhysics vol 5 no 2 pp 409ndash416 2005

[59] M Kulmala H Vehkamaki T Petaja et al ldquoFormation andgrowth rates of ultrafine atmospheric particles a review ofobservationsrdquo Journal of Aerosol Science vol 35 no 2 pp 143ndash176 2004

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

ClimatologyJournal of

EcologyInternational Journal of


EarthquakesJournal of


Hindawi Publishing Corporationhttpwwwhindawicom

Applied ampEnvironmentalSoil Science

Volume 2014

Mining


Journal of

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal of

Geophysics

OceanographyInternational Journal of


Journal of Computational Environmental SciencesHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering


GeochemistryHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Atmospheric SciencesInternational Journal of


OceanographyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in


MineralogyInternational Journal of


MeteorologyAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Paleontology JournalHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Geological ResearchJournal of


Geology Advances in


European areas [12] LRT can cause high PM25

concentrationpeaks when air masses arrive under suitable meteorologicalconditions from regions with high emissions of particlesandor their precursor gases Many research studies havebeen performed and reported for the background atmo-spheric monitoring of LRT [13ndash15] Long-range transport ofemissions from agricultural fires and volcanoes to Europewas recognized previously in USA Europe and Russia [16ndash20] In addition to European transport LRT of smoke fromRussian biomass burning has been recorded in MongoliaChina Japan and Korea [21] and studies have revealedthat hemispheric-scale transport is possible [16] As themass concentrations of fine particles are commonly lower inNorthern Europe countries compared with those in CentralEurope [22 23] LRT can have a notable impact on the aerosolparticle levels under favorable conditions [23 24] The LRTof emissions from wildfires from the Ukraine and Europeanpart of Russia frequently increases the particulate matterconcentrations in these countries during spring and summer[25ndash27] Volcanic dust plumes can also be transferred overlong distances by air masses affecting the air quality not onlyover Europe but in other parts of the world as well [28]

Combining the variety of terrestrial information oper-ationally derived from satellite data including those fromNASArsquosMODIS instrument enables linking the different tem-poral and spatial scales The use of the Angstrom exponent120572 has significantly increased in the last years because thisparameter is easily estimated using automated-surface sunphotometry while it is becoming increasingly accessible tosatellite retrievals [29] Moreover the value of the Angstromexponent is a qualitative indicator of the aerosol particle sizeor fine mode fraction [30]

In this work the pollution events due to biomass burningthat occurred in Lithuania onMarch 21ndash25 2011 andpollutionepisode caused by long-range transported volcanic materialfrom the Grimsvotn eruption that took place at the end ofMay 2011 in Iceland were investigated Air mass backwardtrajectories fire satellite observations dispersion modelingresults and emission source data were used to identify thesource The mass and number concentrations of PM

25in

conjunction with Angstrom exponent of the absorption coef-ficient and other optical properties of aerosol were studied

2 Methodology

21 Site Description The aerosol particle concentration mea-surements were performed at the Preila environmental pol-lution research station (55∘551015840N 21∘001015840E 5m above sea level)in the coastalmarine environment This station is located onthe Curonian Spit which separates the Curonian Lagoon andthe Baltic Sea and thus can be characterized as a regionallyrepresentative background area The Curonian Lagoon thelargest coastal bay in the Baltic Sea is a highly eutrophiedwater body One of the nearest industrial cities Klaipeda isat a distance of about 40 km to the north and the other majorcity Kaliningrad is 90 km to the south from Preila

22 Long-Range Transport and Air Mass Backward Trajecto-ries Air mass backward trajectory analysis provides a better

understanding of air flow and long-range transport patternsduring the events We analyzed the aerosol characteristicswith respect to categorized air mass backward trajectories forthe initial estimation of the wildfire and volcanoes potentialsource locations and quantitative contribution Backwardtrajectories were produced using the Flextra model (NILUInstitute of Meteorology and Geophysics Vienna) [31] andmeteorological data provided fromECMWF (EuropeanCen-tre for Medium Range Weather Forecast) The potentialsource areas of long-range transported pollutionwere studiedusing 7-day backward air mass trajectories with a startingheight of 500 1000 and 1500m above the ground levelTrajectories were produced at 6 h intervals for each eventThe height along the trajectories is indicated by colour(colourheight scale in the lower left corner of the plot) Each3-hour interval along the trajectory path is indicated by asmall legend and each 24-hour interval by a big legend

To identify possible events of regional transport hotspotfire the location of a thermal anomaly was detected byMODIS using data from the middle infrared and thermalinfrared bands [32] Each MODIS sensor achieves globalcoverage once per day and once per night every 24 hTherefore most fires detectable at a 1 km spatial resolutionhave the potential to have their FRP (Fire Radiative Power)measured four times a day Figure 2 shows the locations ofthe active fires identified from MODIS observations duringstudy events over the area of our interest The MODISmonthly fire maps available since 2001 showed that theseevents of biomass burning inMarch-April occurred annuallyHowever the duration geographical extent and emission ofthe smoke from these fires differ year by year

The AErosol RObotic NETwork (AERONET) is aground-based network of Cimel sun photometer that mea-sures the extinction aerosol optical depth at 7 wavelengths120582 (340 380 440 500 670 870 and 1020 nm) from thedirect sun measurements over locations The AOD measure-ments are accurate to within plusmn001 The size distribution ofaerosols can be estimated from spectral AOD (typically 440ndash870 nm) The negative slope of AOT with the wavelengthon a logarithmic scale is known as the Angstrom parameter(120572) This parameter (1) can be calculated from two or morewavelengths using the least squares fit Values of 120572 gt 20indicate that fine mode particles exist (eg smoke particlesand sulphates) while values of 120572 near zero indicate thepresence of coarse mode particles such as desert dust [33]

120572 = minus

119889 ln120591119886

119889 ln 120582= minus

ln (12059111988621205911198861)

ln (12058221205821)

(1)

where 120572 is the wavelength exponent depending on the ratioof the particle coarse mode to the accumulation (small)one (Angstrom parameter) 120591

119886is the AOD and 120582 is the

wavelengthThe Navy Aerosol Analysis and Prediction System

(NAAPS) model results were used to determine the dis-tribution of aerosols from fires and volcanoes (modeldescription and results are available from the web pagesof the Naval Research Laboratory Monterey CA USAhttpwwwnrlmrynavymilaerosol) The NAAPS model



1000

2000

3000

800012000

BC co

ncen

trat

ion

(ng m

minus3)

(a)

0 500 1000 1500 2000 2500 3000 3500 4000 4500

1

10

100

N p

er b

in


Equation



R2 = 09


1205942D ∘ F = 1628

(b)







119868 = 1198680119890(minus119887abs119909) (2)




119887abs prop 120582minus120572

(3)



3 Results






(a)

(a)

65

55

45

35



(b)

0600


April 27 2011

(c)

20110427 1200

(d)








22

2

18

16

14

12

Ang

strom

par

amet

er


AOD

1

08

06

04

02

0

1

08

06

04

02

01 5 10 15 20 25 30

April 2011

(a)

3

28

26

24

22

2

18

16

14

12

0 2 4 6 8 10 12 14 16 18 20 2223


120572

April 29 2011

(b)



























Height (m asl)

1500 2500 3500 4500 5500 6500500 7500

1200May 24 2011

(a)

1200

1500 2500 3500 4500 5500 6500500 7500

Height (m asl)

May 25 2011

(b)


1500 2500 3500 4500 5500 6500500 7500

1200

Height (m asl)

May 26 2011

(c)

1500 2500 3500 4500 5500 6500500 7500


1200

Height (m asl)

May 27 2011

(d)










3

25

2

15

1

05

0

120572

1200

1000

800

600

400

200

0

BC co

ncen

trat

ion

(ng m

minus3)

24 25 26 27 28May 2011



22 times 10minus5

2 times 10minus5

18 times 10minus5

16 times 10minus5

14 times 10minus5

12 times 10minus5

1 times 10minus5

8 times 10minus6

6 times 10minus6

4 times 10minus6

2 times 10minus6

0

24120572370ndash550 = 11

120572590ndash950 = 15

25120572370ndash550 = 13

120572590ndash950 = 14

26120572370ndash550 = 14

120572590ndash950 = 18

27120572370ndash550 = 21

120572590ndash950 = 24

120582 (nm)

b abs

(mminus1)

370 450 520 590 660 880 950









0 3 6 9 12 15 18 21 2410

100

1000

Part

icle

dia

met

er (n

m)

Time (h)


(a)

0 3 6 9 12 15 18 21 24Time (h)

105

104

103

102Part

icle

num

ber

conc

entr

atio

n (c

mminus3

)

(b)







4 Conclusions






Acknowledgments


References






































3




































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in



1000

2000

3000

800012000

BC co

ncen

trat

ion

(ng m

minus3)

(a)

0 500 1000 1500 2000 2500 3000 3500 4000 4500

1

10

100

N p

er b

in


Equation



R2 = 09


1205942D ∘ F = 1628

(b)







119868 = 1198680119890(minus119887abs119909) (2)




119887abs prop 120582minus120572

(3)



3 Results






(a)

(a)

65

55

45

35



(b)

0600


April 27 2011

(c)

20110427 1200

(d)








22

2

18

16

14

12

Ang

strom

par

amet

er


AOD

1

08

06

04

02

0

1

08

06

04

02

01 5 10 15 20 25 30

April 2011

(a)

3

28

26

24

22

2

18

16

14

12

0 2 4 6 8 10 12 14 16 18 20 2223


120572

April 29 2011

(b)



























Height (m asl)

1500 2500 3500 4500 5500 6500500 7500

1200May 24 2011

(a)

1200

1500 2500 3500 4500 5500 6500500 7500

Height (m asl)

May 25 2011

(b)


1500 2500 3500 4500 5500 6500500 7500

1200

Height (m asl)

May 26 2011

(c)

1500 2500 3500 4500 5500 6500500 7500


1200

Height (m asl)

May 27 2011

(d)










3

25

2

15

1

05

0

120572

1200

1000

800

600

400

200

0

BC co

ncen

trat

ion

(ng m

minus3)

24 25 26 27 28May 2011



22 times 10minus5

2 times 10minus5

18 times 10minus5

16 times 10minus5

14 times 10minus5

12 times 10minus5

1 times 10minus5

8 times 10minus6

6 times 10minus6

4 times 10minus6

2 times 10minus6

0

24120572370ndash550 = 11

120572590ndash950 = 15

25120572370ndash550 = 13

120572590ndash950 = 14

26120572370ndash550 = 14

120572590ndash950 = 18

27120572370ndash550 = 21

120572590ndash950 = 24

120582 (nm)

b abs

(mminus1)

370 450 520 590 660 880 950









0 3 6 9 12 15 18 21 2410

100

1000

Part

icle

dia

met

er (n

m)

Time (h)


(a)

0 3 6 9 12 15 18 21 24Time (h)

105

104

103

102Part

icle

num

ber

conc

entr

atio

n (c

mminus3

)

(b)







4 Conclusions






Acknowledgments


References






































3




































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in


(a)

(a)

65

55

45

35



(b)

0600


April 27 2011

(c)

20110427 1200

(d)








22

2

18

16

14

12

Ang

strom

par

amet

er


AOD

1

08

06

04

02

0

1

08

06

04

02

01 5 10 15 20 25 30

April 2011

(a)

3

28

26

24

22

2

18

16

14

12

0 2 4 6 8 10 12 14 16 18 20 2223


120572

April 29 2011

(b)



























Height (m asl)

1500 2500 3500 4500 5500 6500500 7500

1200May 24 2011

(a)

1200

1500 2500 3500 4500 5500 6500500 7500

Height (m asl)

May 25 2011

(b)


1500 2500 3500 4500 5500 6500500 7500

1200

Height (m asl)

May 26 2011

(c)

1500 2500 3500 4500 5500 6500500 7500


1200

Height (m asl)

May 27 2011

(d)










3

25

2

15

1

05

0

120572

1200

1000

800

600

400

200

0

BC co

ncen

trat

ion

(ng m

minus3)

24 25 26 27 28May 2011



22 times 10minus5

2 times 10minus5

18 times 10minus5

16 times 10minus5

14 times 10minus5

12 times 10minus5

1 times 10minus5

8 times 10minus6

6 times 10minus6

4 times 10minus6

2 times 10minus6

0

24120572370ndash550 = 11

120572590ndash950 = 15

25120572370ndash550 = 13

120572590ndash950 = 14

26120572370ndash550 = 14

120572590ndash950 = 18

27120572370ndash550 = 21

120572590ndash950 = 24

120582 (nm)

b abs

(mminus1)

370 450 520 590 660 880 950









0 3 6 9 12 15 18 21 2410

100

1000

Part

icle

dia

met

er (n

m)

Time (h)


(a)

0 3 6 9 12 15 18 21 24Time (h)

105

104

103

102Part

icle

num

ber

conc

entr

atio

n (c

mminus3

)

(b)







4 Conclusions






Acknowledgments


References






































3




































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in


22

2

18

16

14

12

Ang

strom

par

amet

er


AOD

1

08

06

04

02

0

1

08

06

04

02

01 5 10 15 20 25 30

April 2011

(a)

3

28

26

24

22

2

18

16

14

12

0 2 4 6 8 10 12 14 16 18 20 2223


120572

April 29 2011

(b)



























Height (m asl)

1500 2500 3500 4500 5500 6500500 7500

1200May 24 2011

(a)

1200

1500 2500 3500 4500 5500 6500500 7500

Height (m asl)

May 25 2011

(b)


1500 2500 3500 4500 5500 6500500 7500

1200

Height (m asl)

May 26 2011

(c)

1500 2500 3500 4500 5500 6500500 7500


1200

Height (m asl)

May 27 2011

(d)










3

25

2

15

1

05

0

120572

1200

1000

800

600

400

200

0

BC co

ncen

trat

ion

(ng m

minus3)

24 25 26 27 28May 2011



22 times 10minus5

2 times 10minus5

18 times 10minus5

16 times 10minus5

14 times 10minus5

12 times 10minus5

1 times 10minus5

8 times 10minus6

6 times 10minus6

4 times 10minus6

2 times 10minus6

0

24120572370ndash550 = 11

120572590ndash950 = 15

25120572370ndash550 = 13

120572590ndash950 = 14

26120572370ndash550 = 14

120572590ndash950 = 18

27120572370ndash550 = 21

120572590ndash950 = 24

120582 (nm)

b abs

(mminus1)

370 450 520 590 660 880 950









0 3 6 9 12 15 18 21 2410

100

1000

Part

icle

dia

met

er (n

m)

Time (h)


(a)

0 3 6 9 12 15 18 21 24Time (h)

105

104

103

102Part

icle

num

ber

conc

entr

atio

n (c

mminus3

)

(b)







4 Conclusions






Acknowledgments


References






































3




































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in


















Height (m asl)

1500 2500 3500 4500 5500 6500500 7500

1200May 24 2011

(a)

1200

1500 2500 3500 4500 5500 6500500 7500

Height (m asl)

May 25 2011

(b)


1500 2500 3500 4500 5500 6500500 7500

1200

Height (m asl)

May 26 2011

(c)

1500 2500 3500 4500 5500 6500500 7500


1200

Height (m asl)

May 27 2011

(d)










3

25

2

15

1

05

0

120572

1200

1000

800

600

400

200

0

BC co

ncen

trat

ion

(ng m

minus3)

24 25 26 27 28May 2011



22 times 10minus5

2 times 10minus5

18 times 10minus5

16 times 10minus5

14 times 10minus5

12 times 10minus5

1 times 10minus5

8 times 10minus6

6 times 10minus6

4 times 10minus6

2 times 10minus6

0

24120572370ndash550 = 11

120572590ndash950 = 15

25120572370ndash550 = 13

120572590ndash950 = 14

26120572370ndash550 = 14

120572590ndash950 = 18

27120572370ndash550 = 21

120572590ndash950 = 24

120582 (nm)

b abs

(mminus1)

370 450 520 590 660 880 950









0 3 6 9 12 15 18 21 2410

100

1000

Part

icle

dia

met

er (n

m)

Time (h)


(a)

0 3 6 9 12 15 18 21 24Time (h)

105

104

103

102Part

icle

num

ber

conc

entr

atio

n (c

mminus3

)

(b)







4 Conclusions






Acknowledgments


References






































3




































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in


3

25

2

15

1

05

0

120572

1200

1000

800

600

400

200

0

BC co

ncen

trat

ion

(ng m

minus3)

24 25 26 27 28May 2011



22 times 10minus5

2 times 10minus5

18 times 10minus5

16 times 10minus5

14 times 10minus5

12 times 10minus5

1 times 10minus5

8 times 10minus6

6 times 10minus6

4 times 10minus6

2 times 10minus6

0

24120572370ndash550 = 11

120572590ndash950 = 15

25120572370ndash550 = 13

120572590ndash950 = 14

26120572370ndash550 = 14

120572590ndash950 = 18

27120572370ndash550 = 21

120572590ndash950 = 24

120582 (nm)

b abs

(mminus1)

370 450 520 590 660 880 950









0 3 6 9 12 15 18 21 2410

100

1000

Part

icle

dia

met

er (n

m)

Time (h)


(a)

0 3 6 9 12 15 18 21 24Time (h)

105

104

103

102Part

icle

num

ber

conc

entr

atio

n (c

mminus3

)

(b)







4 Conclusions






Acknowledgments


References






































3




































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in




Acknowledgments


References






































3




































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in
















3




































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in



















Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in

Research Article Identification and Characterization...

Documents

Transcript of Research Article Identification and Characterization...