View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

 AtmosphericPollutionResearch2(2011)151Ͳ157

 Atmospheric Pollution Research   www.atmospolres.com  Long–rangetransportofsoildustandsmokepollutionintheSouth Asianregion

BilkisA.Begum1,SwapanK.Biswas1,GauriG.Pandit2,I.VijayaSaradhi2,ShahidaWaheed3, NailaSiddique3,M.C.ShiraniSeneviratne4,DavidD.Cohen5,AndreasMarkwitz6,PhilipK.Hopke7

1BangladeshAtomicEnergyCommission(BAEC),AtomicEnergyCentre,Dhaka(AECD),P.O.Box:164,Dhaka, Bangladesh 2BhabaAtomicEnergyCentre,Trombay,Mumbai,400085, 3ChemistryDivision,DirectorateofScience,InstituteofNuclearScienceandTechnology(PINSTECH),P.O.Nilore,Islamabad,45650Pakistan 4AtomicEnergyAuthority,60/460,BaselineRoad,Orugodawatta,Wellampitiya,SriLanka 5AustralianNuclearScienceandTechnologyOrganisation(ANSTO)LockedBag2001,KirraweeDC,NSW, 6GNSScience,30GracefieldRoad,P.O.Box31Ͳ312,LowerHutt,NewZealand 7CentreforAirResourcesEngineeringandScience,DepartmentofChemicalandBiomolecularEngineering,ClarksonUniversity,Potsdam,NY136999Ͳ 5708,UnitedStates

ABSTRACT  TransboundarytransportofairpollutionintheSouthAsianregionhasbeenanissueofincreasingimportanceoverthe Keywords: pastseveraldecades.Long–rangetransportofanthropogenicpollutioniscontrastedwiththatofpollutionproduced Airborneparticulatematter by natural processes such as dust storms or natural forest fires. Airborne particulate matter datasets covering the LongͲrangetransport periodfrom2002to2007fromtheneighboringcountrieslikeBangladesh,India,PakistanandSriLankawereusedto Blackcarbon findthesourceareasthatareprimarilyresponsibleforlongrangetransportedpollutants.Allfourcountriescollected Soil sampleswiththesametypeofsamplerandfollowthesametechniqueformassandBCmeasurements.Itwasfound thathighfinesoilcontributionswerefromduststorms.Ontheotherhand,smokeinthisregionmainlycomesfrom ArticleHistory: northernIndiawhereagriculturalwasteisoftenburned. Received:10May2010 Revised:02June2010 Accepted:07June2010

CorrespondingAuthor: PhilipK.Hopke Tel:+1Ͳ3152683861 Fax:+1Ͳ315Ͳ268Ͳ4410 EͲmail:[email protected]

©Author(s)2011.ThisworkisdistributedundertheCreativeCommonsAttribution3.0License. doi:10.5094/APR.2011.020   1.Introduction (Pandve and Patil, 2008). There are two possible sources for the  particlesin.Theyareeithergeneratednaturally(e.g.seasalt, TheAsianBrownCloudisalayerofairpollutionthatcovers soil dust) or are man–made (e.g., sulfate and ). From an partsofSouthAsia,namelythenorthernIndianOcean,Indiaand aircraft,thehazeappearsbrownwhenthefractionofsootorsoil Pakistan (Ramanathan et al., 2001; Srinivasan and Gadgil, 2002) dust is large. The potent haze lying over the entire Indian and the cloud appears as a giant brown stain hanging in the air subcontinent –from Sri Lanka to Afghanistan – has led to some overmuchofSouthAsiaandtheIndianOceaneveryyearbetween erratic weather, sparking floods in Bangladesh, Nepal and JanuaryandMarch.Thevisibleimpactofairpollutionisthehaze,a northeastern IndiabutdroughtinPakistanandnorthwesternIndia layer of pollutants and particles from  burning and (Ramanathanetal.,2007). industrial emissions. This Asian Brown Cloud of pollutants has a  brownish color and this brown cloud phenomenon is a common Hazeisanatmosphericphenomenonwheredust,smokeand featureofindustrialandruralregionsaroundtheworld(UNEPand other dry particles affect visibility and obscure the clarity of the C4, 2002). Because of long–range transport of air pollutants, the sky. Sources for haze particles include farming, traffic, industry, mostlyurban(fossilfuelrelated)and/orrural(biomassburningand and . Haze often occurs when dust and smoke particles brickkilnrelated)phenomenonistransformedintoaregionalhaze accumulateinrelativelydryair.Subsequentincreasesinhumidity (orcloud)thatcanspanlargeareasincludinganentirecontinent.It can result in hygroscopic growth and decreased visibility.When is now becoming clear that the brown cloud mayehav  huge weather conditions block the dispersal of smoke and other impactsonagriculture,health,climateandthewaterbudgetofthe pollutants they concentrate and form a usually low–hanging planet(Ramanathan,2008;RamanathanandCarmichael,2008). shroudthatimpairsvisibilityandmaybecomearespiratoryhealth  threat.Industrialpollutioncanalsoresultindensehaze. The haze was really defined as a result of the INDOEX  measurement campaign (Ramanathan et al., 2001; UNEP and C4, The main objective of this study is to locate the sources of 2002).This haze consists of acombination of droplets and solid significant events of fine soil dust and smoke particles in the particles.Thedropletsinthehazearelessthan1.0Pminradius atmospheric aerosol that are major causes of haze in the South 152 Begumetal.–AtmosphericPollutionResearch2(2011)151Ͳ157

Asian region. Regional fine air particle data dated from January ThesamplerinColombo,SriLankawaslocatedattheAtomic 2002 to December 2007 that includes mass, black carbon and EnergyAgency(AEA)facility (6.933°N,79.833°E)andthesampler elemental concentrations in Bangladesh, India, Pakistan, and Sri wasplacedontheflatareaonthefirstflooroftheAEAbuilding. Lanka measured using appropriate nuclear analytical techniques Thisbuildingissetinaresidentialareaonthenorthwesternsideof were analyzed to understand the haze problem and long–range Colombo.Therearemoderatelybusyroadswithin300m. transportoffineparticles.   The sampler in Nilore, Islamabad, Pakistan (33.37°N and 2.Methodology 73.06°E)islocatedontheroofofabuildingatPINSTECHataheight  ofapproximately5–6maboveground.Theinletofthesampleris 2.1.Sampledescriptionandanalysis at a height of a1 m above roof level. The nearest road is 1 km  away. The air flow is unobstructed as the buildings are well AllthefourcountriescollectedsamplesusetheGentstacked separated.Thesurroundingareaconsistsofsomeresidencesand filtersampler(Hopkeetal.,1997),whichiscapableofcollectingair farms. The Nilore site is located on the outskirts of the city of particulatesamplesincoarse(2.5–10Pm)andfine(<2.5Pm)size Islamabad.Thisareawasmostlyfarmlanduntil10yearsago,but fractions. The sampling locations are shown in Figure 1 and the nowmanyhousingunitsarebeingdevelopedinthisarea.Theroad sampling sites description and analytical techniques used for leadingtoPINSTECHfromthemaincityisbeingwidenedthatwill sample analysis for individual countries are given in Table 1. In furtherincreasethisurbanizationprocess. Dhaka,Bangladesh,thesiteisatasemi–residentialarealocatedat  the Atomic Energy Centre, Dhaka (AECD) Campus (23.73°N, The samples were collected on Nuclepore filters with 8 Pm 90.40°E).Thesamplerwasplacedontheflatroofofthebuildingat poresforthecoarsefractionsamplesand0.4Pmporesforthefine aheightof5mabovetheground.Theintakenozzleofthesampler fractionsamples.Ingeneral,sampleswerecollectedforrepresenͲ is located 1.8m above the roof. The intake is about 80m away tative24–hourperiodsatleastonceaweek.Insomelocations,it from the roadside. The sample was placed so that the airflow wasnotpossibletocollectthesampleoverafull24hoursbecause arounditwasunobstructed. offilterclogging.Thenthesamplerwasoperatedwithalternating  timeonandtimeoff(e.g.onehouronfollowedbyonehouroff) overthecourseofthe24–hourperiodtoprovidearepresentative sample during that day. In all cases, the total sampling time was neverlessthan8–hoursequallydistributedover24–hours.  Themethodsusedforthegravimetricmassmeasurementand the black carbon measurements are same for the four countries (Hopke et al., 2008). The collected samples were analyzed using nuclear analytical methods including Ion Beam Analysis (IBA) (Cohen et al., 1996), X–ray Fluorescence (XRF) and Instrumental NeutronActivationAnalysis(INAA).Thesemethodsaredescribed indetailelsewhere(LandsbergerandCreatchmam,1999).  2.2.Sourcefingerprints  TheIBAorXRFanalysesofPMsamplesprovidedopportunities of detecting sufficient number of elemental concentrations to developfingerprintsforanumberofparticlesources.Itisusefulto combinesomeoftheseelementsandestimatetheconcentrations of the major compounds such as ammonium sulfate from the measured sulfur concentration. It is also possible to derive other combinations of the elements that represent signatures for interesting aerosol components. These combinations are called pseudo–elements such as “soil” (Malm et al., 1994). Thus, these Figure1.Maplocatingthesamplingsites. composite variables and pseudo–elements provide a better  understandingofthecompositionofthefineparticlesandleadto ThesamplerinMumbai,Indiawasplacedataheightof15m betterestimatesofpossiblesourcesandtheircontributionstothe above the ground on the terrace of a building in Vashi that is averageambientaerosol(Malmetal.,1994). situatedinNaviMumbai.NaviMumbaiisthelargestplannednew  city near Mumbai (19.07°N, 72.97°E). The city is very close to 2.3.Soil Thane Belapur area that was known as largest industrial belt in  India. The industrial estate is mainly comprised of chemical, bulk WindblownsoiliscomposedmainlyoftheoxidesofMg,Al,Si, drugs and intermediates, dye and dye intermediates, pharmaͲ Ca, Ti and Fe with many other trace elements. The average ceutical, pesticide, petrochemical, engineering goods and textile composition of sandstone and sedimentary rocks and the manufacturingindustries.Anationalhighwaypasses2kmfromthe summationofthe5majoroxidesofAl,Si,Ca,TiandFeaccountfor samplingsite. morethat85%ofthetotalcomposition(WeastandAstle,1982).  Sotheequationforsoilis: Table1.Informationofthesamplingsites  Site Sampling Analytical Country Latitude Longitude  (1) description period Technique Soil=2.20Al+2.49Si+1.63Ca+1.94Ti+2.42Fe

Bangladesh Residential 23.73°N 90.40°E IBA  Equation (1) assumes that the two common oxides of iron India Suburban 19.05°N 73.02°E XRF,IBA Fe2O3andFeOoccur inequal proportions.Thefactorof2.42for 2002Ͳ2007 iron also includes the estimate for K2O in soil through the Pakistan Suburban 33.72°N 73.05°E IBA (K/Fe)=0.6ratioforsedimentarysoils.  SriLanka Residential 6.82°N 79.85°E IBA   Begumetal.–AtmosphericPollutionResearch2(2011)151Ͳ157153

2.4.Smoke Biomass burning in low quality cook stoves is an important  sourceofmethaneandupto6%ofthemethaneemissionsaredue Smokeisthecollectionofairbornesolidandliquidparticulates to biomass burning (Islam, 2002). BC may also originate from andgases(Mulholland,1995)emittedwhenamaterialundergoes vehicularcombustionsources. combustion or pyrolysis, together with the quantity of air that is  entrainedorotherwisemixedintothemass.Itisanunwantedby– 2.5.Meteorology product of fires (including stoves, candles, oil lamps and  fireplaces).Smoke is also a component of internal combustion Bangladesh has a climate of tropical , mild winter engineexhaustgas,particularlydieselexhaust.Thecompositionof (October to March), hot, humid summer (March to June) and smoke depends on the  of burning fuel and conditions of humid warm rainy monsoon (June to October). January is the combustion. Fine potassium is an accepted indicator for smoke coolest month with temperature averaging 26°C and April is the frombiomassburning/brickkiln(Baxlaetal.,2009;Begumetal., warmest month with temperature ranging from 33°C to 36°C. It 2004; Begum et al., 2009). In order to obtain a reliable smoke rainsmostlyduringJunetoOctober. indicator from the fine potassium, it is necessary to subtract the  finepotassiumassociatedwith soilandseasaltcomponentfrom Mumbai(India)hasatropicalwetanddryclimate.Mumbai's totalK(WeastandAstle,1982).Hence,smokeisobtainedbyusing climatecanbebestdescribedasmoderatetemperatureswithhigh thefollowingequation: levelofhumidity.Itscoastalnatureandtropicallocationensures  moderate temperatures throughout the year, average of 27.2°C and average precipitation of 2422 mm. The temperature on Smoke=(Ktot–0.036Na–0.6Fe) (2) average is about 30°C in summer and 18°C in winter. Mumbai's  experiences four distinct seasons winter: (December–February); The haze is seen from the southern edge of the Himalaya summer: (March–May); monsoon (June–September) and Post Mountains southward over the Bay of Bengal every year. The Monsoon(October–December). durationofthishazevariesfromcountrytocountrybutwilloften  beobservedfromearlyDecembertoearlyMarch.Thehaze,mostly Pakistan has four distinct seasons: a cool, dry winter from a mixture of urban and industrial pollution, often collects at the DecemberthroughFebruary;ahot,dryspringfromMarchthrough baseofthemountainsinthewintertime.Thiscloudofhazethat May; the summer rainy season, or southwest monsoon period, frequently lingers over parts of Asia from Pakistan to China and fromJunethroughSeptember;andtheretreatingmonsoonperiod even the Indian and Pacific Oceans has been called the “Asian of October and November. The onset and duration of these Brown Cloud”. The brownish haze consists of a three kilometers seasons vary somewhat according to location. The climate in the thick mixture of anthropogenic sulfate, nitrate, organics, black capitalcityofIslamabadvariesfromanaveragedailylowof2°Cin carbon,dustandflyashparticlesandnaturalaerosolssuchassea Januarytoanaveragedailyhighof40°CinJune.Halfoftheannual saltandmineraldust.Thebrownishcolorisduetotheabsorption rainfalloccursinJulyandAugust,averagingabout255millimeters andscatteringofsolarradiationbyanthropogenicblackcarbon,fly in each of those two months. The remainder of the year has significantly less rain, amounting to about 50 mm per month. ash, part of the soil dust and NO2 (Ramanathan and Ramana, 2003). Hailstormsarecommoninthespring.   The brown haze often exists over large South Asian cities in SriLanka’sclimateissimilartothatexperiencedinsouthern part due to particulate emissions within the urban area, dust India, and involves two monsoon seasons and a dry season. storm, and distant fires. Emitted soil dust and gases have the Colombo’sweather,andmuchoftheweatherofthesouthwestern capacity to be transported over long distances, sometimes many sectionoftheisland,isimpactedbytheYalamonsoonduringthe hundredsofkilometersandmay giverisetodepositioninanother periodfromMaytoAugust.ThedryseasonrunsfromDecemberto country. In addition to the respiratory problems the persistent April. In the northern and eastern sections of the island, the hazecancause,italsoappearstohindercropsbyblockingsunlight weather patterns are different: the monsoon season is from andcouldbealteringregionalweather.Hazecanhaveimpactson October to January and the dry season lasts from May to agricultural productivity through direct and indirect effects. The September.DuringthemonthsofOctoberandNovemberallparts directeffectsinclude: of Sri Lanka experience occasional heavy downpours and (1) Reduction of total solar radiation (sum of direct and thunderstorms. diffused) in the photo–synthetically active part of the spectrum  (0.4to0.7micron)reducesphotosynthesis. Therefore,itwasobservedthatfromearlyDecembertoearly (2)Settlingofparticles(e.g.flyash,blackcarbon,anddust)on March, the winter season exist within this region and the wind theplantscanshieldtheleavesfromsolarradiation. directionismainlynorthwesterly. (3) In addition, particle deposition can increase acidity and  causeplantdamage. 2.6.Backtrajectorycalculation   Theindirecteffectsinclude: Using models of atmospheric transport, a trajectory model (1)Changesinsurfacetemperaturethatcandirectlyaffectthe calculatesthepositionoftheairbeingsampledbackwardintime growing season. In the tropics, surface cooling(such asexpected fromthereceptorsitefromvariousstartingtimesthroughoutthe fromparticles)canextendthegrowingseason(whileagreenhouse samplinginterval.Thetrajectoriesarepresentedasasequenceof warmingcandecreaseit). latitude and longitude values for the endpoints of each segment (2)Changesinrainfallorsurfaceevaporationcanhavealarge representing each specific time interval being modeled. The impact(UNEPandC4,2002). verticalmotionofairparcelsisconsideredduringthismodel.The  NOAA Hybrid Single Particle Lagrangian Integrated Trajectory Atmosphericblackcarbon(BC)canactintwoways.Firstasa (HYSPLIT–4)(DraxlerandRolph,2003)modelwasusedtocalculate directabsorberofvisiblelightandthatprovidesdirectwarmingin the air mass backward trajectories for those days when fine theloweratmosphere(RamanathanandCarmichael,2008).SecͲ particlesweresampled.ArchivedREANALYSISmeteorologicaldata ondly,thedepositionofblackcarbononiceorsnowsuchasinthe wereusedasinput.Thelatitude/longitudewasuseddependingon Himalayanglaciersispartofwhatiscausingthemtomeltquickly the site of location of four countries and trajectories were (Kehrwald et al., 2008). Thus, there are good reasons to underͲ computedbackwardintimeupto240hours(10days).Tickmarks standtheextentandsourcesofBC. onthetrajectoryplotsindicate6–hourmovementlocations.   154 Begumetal.–AtmosphericPollutionResearch2(2011)151Ͳ157

Back trajectories were calculated for periods of 10 days. Although the error in positions of the individual trajectory 40000 endpointsincreasewithincreasingtrajectoryduration(Kahl,1996), Bangladesh 10–daybacktrajectorieshaveprovenusefulinassessingthelikely 20000 sourceregionsforparticulatepollution(Polissaretal.,2001).

 ) 0 3.ResultsandDiscussion 3  26000 India Although airborne particles are generally associated with 13000 global cooling effects, recent studies have shown that they can actually have a positive radiative forcing effect particularly in 0 certainregionssuchasthe(Ramanathanetal.,2007).In 36000 South Asia, many countries recognize air pollution as a major 27000 Pakistan publichealthconcernandhaveundertakenprojectsorregulatory 18000 programs to control air pollution, but data provided by some of 9000 these countries indicate that in many cities, air quality still falls 0 below world standards for acceptable air quality (Hopke et al., Fine Soil Concentration (ng/m 12000 2008).Thesestudiesshowedthatmajorsourcesofpollutantsare 9000 Sri Lanka soil dust, smoke and traffic, which form haze in this continent 6000 (Pandve,2008).Thus,thedataforthesesourcesarediscussed. 3000  0 3.1.Long–rangetransportofsoil 02 02 3 04 05 05 06 07 08 01/ 01/ 01/0 01/ 01/ 01/ 01/ 01/ 01/  01/ 10/ 07/ 04/ 01/ 10/ 07/ 04/ 01/ Inordertoexcludelowlocalsourcecontributions,onlydaily eventswithconcentrationsthataretwostandarddeviationsabove  themeanvalueforameasuredspecieswereconsidered.Table2 Figure2.TimeseriesplotsoffinesoilconcentrationsinBangladesh,India, showsthemean,standarddeviation,andthethresholdvaluesfor Pakistan,andSriLanka. soil, smoke and black carbon concentrations. The meteorological  conditioninthisregionduringwinterleadstoprevailingnortherly Fortheobservedpotentialtransboundaryepisoderesultingin and northwesterly winds. There is then the possibility of the highest soil contribution (Figure 2), the elemental data were transboundary events (Adhikary et al., 2007) affecting local air examined and it was seen that the concentration of silicon, a quality. Figure 2 shows the time series plot for the fine soil signatureforsoilwasalsoveryhighforthosesamples.Toexplore concentrationsforallfourcountriesduringthestudyperiod.Itwas thelong–rangetransportofsoil,backtrajectorieswerecalculated found that there was a large peak on 5 March 2003 in case of on these days. Thus, back trajectories of air parcels at different Bangladesh and the value was 38 Pg/m3. On the other hand, for heights starting from 300m, 500m, 1000m and 1500m on 5 India,thelargepeakwason20February2003andthevaluewas March2003(10days)forDhaka,Bangladeshand20February2003 24 Pg/m3. In case of Pakistan, there is no large peak in the time (7 days) for Mumbai, India were tried. Figure 3 shows interval from February 2003 to March 2003. For Sri Lanka, there representative trajectories for 500m and 1000m height. They arenodataavailableduringthisperiod.Fromthemeteorological show that the air parcels come over Iran and Pakistan before data,itwasobservedthatwinterseasonpredominatesduringthe reachingthesamplingsiteatDhakaandpossiblyoverNorthAfrica monthofDecembertoFebruaryinthisregionandthedirectionof before reaching the sampling site at Mumbai. In case of Dhaka wind is usually northwesterly. Beginning in March, the premonͲ Bangladesh,itwasfoundthatfor1500mstartingheight,theair soon starts and the average wind speed becomes higher than in parcels came from northwest direction. There was a thick dust thewintertime. plume (light brown) that blew westward and then routed  northwardbystrongsoutherlywinds(NASAwebsite). Table2.Themean,standarddeviationandpeakvaluesofsoil,smokeand  blackcarbonconcentrations(Pg/m3)duringthestudyingperiod

Parameter Statistics Bangladesh India Pakistan SriLanka Soil Mean 3.51 6.82 3.17 2.00

Dhaka (Yellow dots at Median 2.69 5.03 1.98 1.83 500m and Red dots at 1000m) STD 3.05 6.53 3.63 1.48

ThresholdValue 9.60 19.9 10.4 4.95

Smoke Mean 0.31 0.33 0.15 0.06 Mumbai (Yellow dots at 500m and Red dots at 1000m) Median 0.24 0.12 0.10 0.05

STD 0.24 0.81 0.22 0.12

ThresholdValue 0.78 1.96 0.59 0.31 Figure3.Airparcelbacktrajectoriesshowinglongrangetransportofsoil dustinBangladeshandIndiainFebruary2003. BC Mean 8.97 7.86 2.62 10.8   Median 7.54 6.70 2.34 10.3 EveryyearstartingfromFebruarytoMarchduetothechange  of season several dust stormsoccur and the dust particles travel  STD 6.34 4.63 1.61 4.15 thousands of miles. The trajectories calculated at lower heights  (300mand500m)showareasofIranandOman.Itisalsoseenin  ThresholdValue 21.7 17.1 5.84 19.1 thesatelliteimagethatthereweretwothickplumesofdesertdust blew over Oman and across the Gulf of Oman toward Iran and Pakistan on 18 February 2003 (see the Supporting Material–SM,  Begumetal.–AtmosphericPollutionResearch2(2011)151Ͳ157155

Figure S1). Therefore, it can be concluded that the high fine soil Colombo on 10 November, 03 November, 15 November and 12 contribution is most likely due to long–range transport of the December2003,respectively(Figure8).Thetrajectoryplotsshow desertdustfromtheseregions.ForIndia,ithasbeenfoundthat theareaofnorthwestpartofSouthAsia. the trajectory plots at starting heights of 1000m and 500m overlapandshowtheesam originsofdesertdust.FromtheNASA 50000 websiteitwasfoundthattherewereseveralduststormsinearly )

February2003thatblewinanortheasterlydirection.Ithasbeen 3 found from calculated trajectories that the dust particles travel 40000 about 7000 km to reach both of the receptor sites (Dhaka and Mumbai).  30000 High concentrations of fine soil were found on 21 and 26 February 2004 in Colombo, Sri Lanka and Islamabad, Pakistan 20000 respectively. The backward trajectories of air parcels at different startingheightsof300m,500m,1000mon21and26February 2004in Colombo,SriLanka(10daysback)andIslamabad,Pakistan 10000

(5daysback)respectivelywerecalculated(Figure4).Itwasfound BC concentration(ng/m thatthetrajectoryplotsforbothreceptorsitesat300mand500m weresimilarandoverlap.Therefore,thetrajectorieswitha300m 0 0 200 400 600 800 1000 1200 1400 starting height were discarded for clarity. The trajectory plots 3 (500m and 1000 m) showed that the air parcels came from a Smoke concentration (ng/m )  northwesterly direction and traveled about 4000 km and 12000 Figure5.ScatterplotbetweenBCandsmokeconcentrations. kmtoreachIslamabadandColombo,respectively.Therefore,the  originofdustsourcewasnotsame.FromNASAwebsite(seethe SM,FigureS2)itwasfoundthattherewereseveralduststormsin Sahara, Jordan and Iran in February 2004 which blew over 60000 Bangladesh northeasternpartofSouthernAsia. 40000  20000

0 28000 India ) 3 Islamabad (Red dots at 14000 500m and Yellow dots at 1000m) 0 10000 Pakistan

5000

0 27000 Sri Lanka Colombo (Red dots at 500m and Yellow dots 18000 at 1000m) (ng/m BC concentration 9000  0 Figure4.Airparcelbacktrajectoriesshowinglongrangetransportofsoil 1/02 1/02 1/03 1/04 1/05 1/05 1/06 1/07 1/08 dustinPakistanandSriLankainFebruary2004. 01/0 10/0 07/0 04/0 01/0 10/0 07/0 04/0 01/0   3.2.Long–rangetransportofsmoke Figure 6. Time series plots of black carbon (BC) concentrations in  Bangladesh,India,Pakistan,andSriLanka. The scatter plot for BC and smoke concentrations (Figure 5)  showsthatbothcomefromthesamesource.Figures6and7show The smoke signature reaches at Dhaka by traveling about the time series of BC and smoke concentrations and the smoke 2200kmto5000kmfromthenorthwestdirectionandshowsthe concentrationscalculatedfromEquation(2)forthefourcountries. areaofIndia,IranandIraq.IncaseofMumbai,thesourceregionis In order to remove the local contributions to smoke, similar northern India.For Sri Lanka, the source regions are Bangladesh, criterionvaluesasdescribedpreviouslywerecalculatedtoidentify Afghanistan,Pakistan,andVietnamandtravelsabout2500kmto the likely long–range transport episodes. From the Dhaka time 8000 km to reach Colombo. For Pakistan, the source regions of series, the peak concentration above threshold value, which was smoke signature are west part of India and Kagakhantan and found on 10 November 2003, to be 1.13Pg/m3, and the BC travelsabout800kmto2200kmtoreachatIslamabad. concentrationonthatdaywas25.4Pg/m3.IncaseofIndia,on5  November2003,thehighconcentrationsforsmokeandBCwere Thereweretwosatelliteimages(seetheSM,FiguresdS3an  2.66Pg/m3 and 13.6Pg/m3, respectively. For Pakistan, on 15 S4) that show the potential source areas in northwestern India. November2003,thehighconcentrationsforsmokeandBCwere These two satellite images show the evidence of haze that 1.15Pg/m3 and 8.93Pg/m3, respectively. For Sri Lanka, on 12 occurred due to smoke from agricultural fires in northern India December 2003, the high concentrations of smoke and BC were starting from 3 November 2003 and backing up against the 0.48Pg/m3and16.2Pg/m3,respectively. HimalayanMountainsuntil6November2003.Hazeiscreatedbya  range of airborne particles and pollutants released from Theverticallymixedmodelstartingatdifferentheightsfrom incomplete burning of biomass, agricultural waste, or traffic 300m, 500m and 1000m above the ground level was used to related fuel. The cloud is associated with the winter monsoon calculatetheseven,ten,tenandsevendaysbackwardtrajectories (November/December to April) during which there is no rain to arriving at the receptor sites in Dhaka, Mumbai, Islamabad and wash pollutants from the air. The increasing aerosol loading in 156 Begumetal.–AtmosphericPollutionResearch2(2011)151Ͳ157

Asian countries associated with industrialization during last two 4.Conclusion decades has caused major health–related problems associated  with worsening air quality and has also had impact on aviation Haze consists of a combination of suspended water droplets safety (Lau et al., 2008). Studies over South Asian region have and minute particles in the atmosphere. From this study, it was found that anthropogenic aerosols may significantly change the observedthattheairqualityproblemislargestduringthewinter energy balance of the atmosphere and earth’s surface monsoon(thedryseason)inmeteorologicallystableandcloudfree (Ramanathan et al., 2001; UNEP and C4, 2002). These atmospheric conditions. It was also seen that the peak concenͲ anthropogenicaerosolforcingcouldinfluencetheseasonalrainfall trationsofsmokeobservedinBangladeshduringNovember2003 distributioninthemonsoonregionsoverSouthAsia(Ramanathan islikelyduetoagriculturalfiresinnorthernIndiaduringthatperiod et al., 2001; Chung et al., 2005; Ramanathan et al., 2005). The and the high fine soil contribution is likely due to long–range observedweakeningIndianmonsoonandinChina,droughtinthe transportofthedesertdustwhichwasblewoverOmanandacross northandfloodinginthesouthareinfluencedbythehaze(Lauet theGulfofOmantowardsIrandan Pakistanon18February2003. al.,2008).Asianglacialmeltingcouldleadtowatershortagesand  floodsforthehundredsofmillionsofpeoplewholivedownstream. Acknowledgment   The work reported here was supported in part by the 1500 1200 Bangladesh International Atomic Energy Agency (IAEA) under the Regional 900 Cooperative Agreement (RCA) project RAS/7/015. The authors 600 gratefullyacknowledgetheNOAAAirResourcesLaboratory(ARL) )

3 300 for the provision of the HYSPLIT transport and dispersion model 0 and/ortheREADYwebsite. 3600 India  2700 SupportingMaterialAvailable 1800  900 ThickplumesofdesertdustblewoverOmanandacrossthe 0 Gulf of Oman toward Iran and Pakistan on February 18 2003 1600 Pakistan (Figure S1), An intense dust storm blew through the skies over JordanonFebruary22,2004(FigureS2),Smokefromagricultural 800 firesinnorthernIndia,continuestobackup againsttheHimalayan

Smoke ConcentrationSmoke (ng/m MountainsonNovember6,2003,itstartedfromNovember32003 0 800 Sri Lanka (FigureS3),Tothesoutheastofadenseclusterofagriculturalfires 600 (red dots), a pall of smoke hangs over northwestern India 400 (AgriculturalfiresinnorthernIndia,October23,2003)(FigureS4). 200 This information is availableefre  of charge via the Internet at 0 http://www.atmospolres.com. /02 /03 /04 /05 /06 /07 /08  01/01 01/01 01/01 01/01 01/01 01/01 01/01  References Figure 7. Time series plots of fine smoke concentrations in Bangladesh,  India,Pakistan,andSriLanka. Adhikary,B.,Carmichael,G.R.,Tang,Y.,Leung,L.R.,Qian,Y.,Schauer,J.J.,  Stone,E.A.,Ramanathan,V.,Ramana,M.V.,2007.Characterizationof the seasonal cycle of south Asian aerosols: a regionalͲscale modeling analysis.JournalofGeophysicalResearchD:Atmospheres112,art.no. D22S22. Baxla,S.P.,Roy,A.A.,Gupta,T.,Tripathi,S.N.,Bandyopadhyaya,R.,2009. Islamabad (Orange dots at 1000m and Brown dots at 500m) Analysis of diurnal and seasonal variation of submicron outdoor aerosol mass and size distribution in a Northern Indian City and its correlation to black carbon. Aerosol and Air Quality Research 9, 458Ͳ Dhaka (Green dots at 1000m and Red dots at 500m) 469. Mumbai (Yellow dots at 500m) Begum, B.A., Paul, S.K., Hossain, M.D., Biswas, S.K., Hopke, P.K., 2009. Indoor air pollution from particulate matter emissions in different householdsinruralareasofBangladesh.BuildingandEnvironment44, 898Ͳ903. Begum, B.A., Biswas, S.K., Hopke, P.K., 2008. Assessment of trends and

Colombo (Black dots at present ambient concentrations of PM2.2 and PM10 in Dhaka, 1000m and Blue at 300m) Banglaesh.AirQuality,AtmosphereandHealth1,125Ͳ133.  Begum, B.A., Kim, E., Biswas, S.K., Hopke, P.K., 2004. Investigation of Figure 8. Air parcel back trajectories showing the likely source areas for sources of atmospheric aerosol at urban and semiͲurban areas in smokeinIndia,Bangladesh,Pakistan,andSriLanka. Bangladesh.AtmosphericEnvironment38,3025Ͳ3038.  Chung, C.E., Ramanathan, V., Kim, D., Podgorny, A.I., 2005. Global Everyyearduringthewinter,startingfromDecembertoearly anthropogenicaerosoldirectforcingderivedfromsatelliteandgroundͲ March, there are high air pollution episodes in this region and it based observations. Journal of Geophysical Research D: Atmospheres oftencollectsatthebaseofthemountains.Thehighairpollutionis 110,art.no.D24207. due to anthropogenic activities and it has been found that BC concentrations are high during this season (Nair et al., 2007; Cohen,D.D.,Bailey,G.M.,Kondepudi,R.,1996.ElementalanalysisbyPIXE Begum et al., 2008). For South Asia, general circulation climate andotherIBAtechniquesandtheirapplicationtosourcefingerprinting models (GCMs) simulations suggest that a two–to three–fold of atmospheric fine particle pollution. Nuclear Instruments and increase in soot loading is sufficient to substantially weaken the Methods in Physics Research Section B: Beam Interactions with monsoon circulation, decrease rainfall by more than 25% and MaterialsandAtoms109/110,218Ͳ226. increasedroughtfrequencysignificantly(Menonetal.,2002).    Begumetal.–AtmosphericPollutionResearch2(2011)151Ͳ157157

Draxler,R.R.,Rolph,G.D.,2003.HYSPLIT(HybridSingleͲParticleLagrangian Pandve,H.T.,2008.TheAsianbrowncloud.IndianJournalofOccupational Integrated Trajectory) Model access via NOAA ARL READY Website andEnvironmentalMedicine12,93Ͳ95. (http://www.arl.noaa.gov/ready/hysplit4.html). NOAA Air Resources Polissar,A.,Hopke,P.K.,Harris,J.M.,2001.Sourceregionsforatmospheric Laboratory,SilverSpring,MD. aerosol measured at Barrow, Alaska. Environmental Scienced an  Hopke, P.K., Cohen, D.D., Begum, B.A., Biswas, S.K., Ni, B., Pandit, G.G., Technology35,4214Ͳ4226. Santoso,M.,Chung,Y.S.,Davy,P.,Markwitz,A.,Waheed,S.,Siddique, Ramanathan,V.,2008.AsianBrownCloud,http://wwwͲramanathan.ucsd. N.,Santos,F.L.,Pabroa,P.C.B.,Seneviratne,M.C.S.,Wimolwattanapun, edu/files/AMS16May2008.pdf W.,Bunprapob,S.,Vuong,T.B.,Hien,P.D.,Markowicz,A.,2008.Urban Ramanathan,V.,Carmichael,G.,2008.Globalandregionalclimatechanges airqualityintheAsianregion.Science oftheTotalEnvironment 404, duetoblackcarbon.NatureGeoscience1,221Ͳ227. 103Ͳ112. Hopke,P.K.,Xie,Y.,Raunemaa,T.,Biegalski,S.,Landsberger,S.,Maenhaut, Ramanathan,V.,Ramana,V.M.,Roberts,G.,Kim,D.,Corrigan,C.,Chung,C., Winker, D., 2007. Warming trends in Asia amplified by brown cloud W.,Artaxo,P.,Cohen,D.,1997.CharacterizationoftheGentstacked solarabsorption.Nature448,575Ͳ578. filterunitPM10sampler.AerosolScienceandTechnology27,726Ͳ735. Islam,M.,2002.Impactsofbiomasscookstoveuseonairpollution,Global Ramanathan, V., Chung, C., Kim, D., Bettege, T., Buja, L., Kiehl, J.T., Washington, W. M., Fu, Q., Sikka, D.R., Wild, M., 2005. Atmospheric warming and human health in rural Bangladesh. Bangladesh Brown Clouds: Impact on South Asian climate and hydrologic cycle. Environment1,69Ͳ80. ProceedingsoftheNationalAcademyofSciencesoftheUnitedStateof Kahl, J.D.W., 1996. On the prediction of trajectory model error, America102,5326Ͳ5333. AtmosphericEnvironment30,2945Ͳ2957. Ramanathan, V., Ramana, M.V., 2003. Atmospheric Brown Clouds: Long Kehrwald, N.M., Thompson, L.G., Tang Dong, Y., MosleyͲThompson, E., RangeTransportandClimateImpacts.EM:AirandWasteManagement Schotterer, U., Alfimov, V., Beer, J., Eikenberg, J., Davis, M.E., 2008. Association's Magazine for Environmental Managers December 2003 Mass loss on Himalayan glacier endangers water resources. Issue,28Ͳ33. GeophysicalResearchLetters35,art.no.L22503. Ramanathan, V., Crutzen, P.J., Lelieveld, J., Mitra, A.P., Althausen, D., Landsberger, S., Creatchmam, M., 1999. Elemental Analysis of Airborne Anderson,, J.  Andreae, M.O., Cantrell, W., Cass, G.R., Chung, C.E., Particles,GordonandBreachSciencePublisher,Malaysia. Clarke, A.D., Coakley, J.A., Collins, W.D., Conant, W.C., Dulac, F., Lau K.M., Ramanathan, V., Wu, G.X., Li, Z., Tsay, S.C., Hsu, C., Sikka, R., Heintzenberg, J., Heymsfield, A.J., Holben, B., Howell, S., Hudson, J., Holben,B.,Lu,D.,Tartari,G.,Chin,M.,Koudelova,R.,Chen,H.,Ma,Y., Jayaraman,A.,Kiehl,J.T.,Krishnamurti,T.N.,Lubin,D.,McFarquhar,G., Huang, J., Taniguchi, K., Zhang, R., 2008. The joint aerosolͲmonsoon Novakov, T., Ogren, J.A., Podgorny, I.A., Prather, K., Priestley, K., experimentͲanewchallengeformonsoonclimateresearch.Bulletinof Prospero,J.M.,Quinn,P.K.,Rajeev,K.,Rasch,P.,Rupert,S.,Sadourny, theAmericanMeteorologicalSociety89,369Ͳ383. R.,Satheesh,S.K.,Shaw,G.E.,Sheridan,P.,Valero,F.P.J.,2001.Indian Malm,W.C.,Sisler,J.F.,Huffman,D.,Eldred,R.A.,Cahill,T.A.,1994.Spatial Ocean experiment: an integrated analysis of the climate forcing and andseasonaltrendsinparticleconcentrationandopticalextinctionin effectsofthegreatIndoͲAsianhaze.JournalofGeophysicalResearchͲ the United States. Journal of Geophysical ResearchͲAtmospheres 99, Atmospheres106,28371Ͳ28398. 1347Ͳ1370. Srinivasan,J.,Gadgil,S.,2002.AsianbrowncloudͲfactandfantasy.Current Mason,B.,Moore,C.B.,1982.PrincipleofGeochemistry,4thedition,Wiley, Science83,586Ͳ592. NewYork. United Nations Environmental Programme (UNEP) and Center for Clouds, Menon, S., Hansen, J., Nazarenko, L., Luo, Y.F., 2002. Climate effects of ChemistryandClimate(C4),2002.TheAsianBrownCloud:Climateand blackcarbonaerosolsinChinaandIndia.Science297,2250Ͳ2253. OtherEnvironmentalImpacts,URL:http://www.rrcap.unep.org. rd Mulholland,G.W.,SmokeProductionandPropertiesͲSFPEHandbookofFire Weast,R.C.,Astle,M.J.,1982.CRCHandbookofChemistryandPhysics,63  ProtectionEngineering,2ndEdition,Section2.P.J.e.a.DiNenno,Ed. Edition,CRCPressInc.,BocaRaton,Florida. (1995),pp.217Ͳ227.  Nair,  V.S., Moorthy, K.K., Alappattu, D.P., Kunhikrishnan, P.K., George, S.,  Nair,P.R.,Babu,S.S.,Abish,B.,Satheesh,S.K.,Tripathi,S.N.,Niranjan, K.,Madhavan,B.L.,Srikant,V.,Dutt,C.B.S.,Badarinath,K.V.S.,Reddy, R.R.,2007.WintertimeaerosolcharacteristicsovertheIndoͲGangetic Plain(IGP):impactsoflocalboundarylayerprocessesandlongrange transport. Journal of Geophysical ResearchͲAtmospheres 112, art.no. D13205.