ARTICLE IN PRESS

1352-2310/$ - se

doi:10.1016/j.at

�CorrespondE-mail addr

Atmospheric Environment 40 (2006) 6718–6729

www.elsevier.com/locate/atmosenv

Chemical apportionment of shortwave direct aerosol radiativeforcing at the Gosan super-site, Korea during ACE-Asia

Jiyoung Kima,b, Soon-Chang Yoona,�, Sang-Woo Kima, Fred Brechtelc,Anne Jeffersond, Ellsworth G. Duttond, Keith N. Bowere,

Steven Clifff, James J. Schauerg

aSchool of Earth and Environmental Sciences, Seoul National University, Seoul 151-747, South KoreabMeteorological Research Institute, Korea Meteorological Administration, Seoul 156-720, South Korea

cBrechtel Manufacturing Inc., Hayward, CA, USAdClimate Monitoring and Diagnostics Laboratory, National Oceanic and Atmospheric Administration,

325 Broadway, Boulder, CO 80305, USAeUniversity of Manchester, Institute of Science and Technology, Manchester M60 1QD, UK

fUniversity of California at Davis, Davis, CA, USAgUniversity of Wisconsin-Madison, 660 North Park Street, Madison, WI 53706, USA

Received 19 December 2005; received in revised form 7 June 2006; accepted 7 June 2006

Abstract

Shortwave direct aerosol radiative forcing (DARF) at the surface as well as aerosol optical depth (AOD) were estimated

and chemically apportioned on the basis of ground-based aerosol and radiation measurements at the Gosan super-site in

Korea during the Asian Pacific Regional Aerosol Characterization Experiment (ACE-Asia) in April 2001. An aerosol

optical model and a radiative transfer model (RTM) were employed to calculate the aerosol extinction coefficient and

radiative flux at the surface, respectively. The calculated scattering and absorption coefficients for Dpo10mm aerosols

agreed well with measured scattering and absorption coefficients with root mean square errors (RMSEs) of 23.6 and

3.0Mm�1, respectively. The modeled direct and diffuse irradiances at the surface were also in good agreement with the

measured direct and diffuse irradiances. In this study we found that the 17-day mean aerosol radiative forcing of

�38.3Wm�2 at the surface is attributable to mineral dust (45.7%), water-soluble components (sum of sulfate, nitrate,

ammonium, and water-soluble organic carbon (WSOC)) (26.8%), and elemental carbon (EC) (26.4%). However, sea salt

does not play a major role. For the cases of Asian dust and smoke episodic events on 26 April 2001, a diurnal averaged

forcing of �36.2Wm�2 was contributed by mineral dust (�18.8Wm�2), EC (�6.7Wm�2), and water-soluble components

(�10.7Wm�2). The results of this study suggest that water-soluble and EC components as well as a mineral dust

component are responsible for a large portion of the aerosol radiative forcing at the surface in the continental outflow

region of East Asia.

r 2006 Published by Elsevier Ltd.

Keywords: Surface aerosol radiative forcing; Extinction; Closure study; Chemical apportionment; Aerosol; ACE-Asia

e front matter r 2006 Published by Elsevier Ltd.

mosenv.2006.06.007

ing author. Tel.: +82 2 880 6717; fax: +82 2 883 4972.

ess: yoon@snu.ac.kr (S.-C. Yoon).

ARTICLE IN PRESSJ. Kim et al. / Atmospheric Environment 40 (2006) 6718–6729 6719

1. Introduction

Aerosols have a direct radiative forcing becausethey scatter and absorb solar and terrestrial radia-tion in the atmosphere (Yoon and Kim, 2006).Aerosols also alter the formation, lifetime, andprecipitation efficiency of liquid-water, ice andmixed-phase clouds, thereby causing an indirectradiative forcing associated with these changes incloud properties (Intergovernmental Panel on Cli-mate Change (IPCC), 2001). The quantification ofaerosol radiative forcing is more complex than thequantification of radiative forcing by greenhousegases because aerosol mass loading and chemicalcomposition are highly variable in space and time(IPCC, 2001; Yoon and Kim, 2006). Quantifyingand assessing the climatic impact of atmosphericaerosols require knowledge of their physical, che-mical, optical, and radiative properties, as well as oftheir spatial and temporal variability (Penner et al.,1994; Huebert et al., 2003; Yoon et al., 2005).

During the last decade, there has been a remark-able improvement in the estimation of direct aerosolradiative forcing (DARF) by means of box models(Charlson et al., 1992; Penner et al., 1994; Nemesureet al., 1995), various measurements at the surfaceand onboard aircraft (Bush and Valero, 2003; Xuet al., 2003; Conant et al., 2003; Markowicz et al.,2003; Won et al., 2004; Yoon and Kim, 2006),remote sensing by satellites (Christopher andZhang, 2002; Zhang et al., 2005), and chemicaltransport models (CTMs) (IPCC, 2001; Takemuraet al., 2003; Yu et al., 2004; Yoon et al., 2005).However, there is still a large degree of uncertaintyin the aerosol radiative forcing estimation becausenot only the physico-chemical properties of aerosolparticles are not well established but also thecomplex interactions between atmospheric aerosolsand radiation are poorly understood (Penner et al.,1994; IPCC, 2001). The Asian Pacific RegionalAerosol Characterization Experiment (ACE-Asia)campaign was designed to characterize the physical,chemical, and optical properties of aerosols in theatmosphere of the East Asian region and estimatetheir effect on regional radiative imbalance(Huebert et al., 2003).

Estimation of the light-extinction, optical depth,and radiative forcing due to atmospheric aerosolswith chemical apportionment is important towardquantitatively assessing the aerosol impact onvisibility degradation and reduction of the incomingsolar radiation at the surface (McInnes et al., 1998;

Bush and Valero, 2003) and for creating optimalcontrol strategies to mitigate adverse aerosol effectson climate. Up to now chemically specified estima-tion of aerosol radiative forcing has mainly relied onCTMs (IPCC, 2001; Takemura et al., 2003; Yuet al., 2004; Yoon et al., 2005). In this paper, wepresent results of aerosol optical closure andradiative forcing studies, performed using aerosolchemical, physical, and optical measurements atGosan during ACE-Asia together with an aerosoloptical model and a radiative transfer model(RTM).

2. Experiments

2.1. Aerosol optical properties and atmospheric

radiation measurements

Aerosols and atmospheric radiation measure-ments were performed at the Gosan super-site,South Korea (1261100E, 331170N, altitude: �50mabove mean sea level), which was the most heavilyinstrumented ground site during ACE-Asia(Huebert et al., 2003). Aerosols were sampled usingan inlet with a height of �10m above ground level.Sample air was heated to achieve a relative humidity(RH) less than 50%. An impactor system was usedto enable measurements of two aerosol populations:particles with aerodynamic particle diameters (Dp)less than 1.0 and less than 10.0 mm. Aerosol lightscattering (ssp) and absorption (sap) coefficients at550 nm were measured using a TSI Model 3563nephelometer and a Particle Soot AbsorptionPhotometer (Radiance Research Inc.), respectively.The uncertainties in the 1-min average optical dataare �10% for ssp ¼ 33Mm�1 and �28% for sap ¼5Mm�1 (Kim et al., 2005a).

Direct and diffuse solar irradiances at the site inthe broadband wavelength range (0.285–4.0 mm)were measured using a pyrheliometer (EppleyNIP) and shaded pyranometers (Eppley PSP),respectively. Thermal offset errors in the diffusePSP were corrected. The radiative flux measure-ments are accurate within approximately 10Wm�2

for the 1-min averages recorded at a sampling rateof 1Hz. In addition, spectral aerosol optical depths(AODs) at 368, 412, 500, 610, 675, 778 and 862 nmwere determined once per minute by two Carter–Scott sunphotometers with an accuracy of �0.02(Kim et al., 2005b). Cloud screening of the solarradiation data was accomplished by examining thestability of the continuously recorded solar signal so

ARTICLE IN PRESSJ. Kim et al. / Atmospheric Environment 40 (2006) 6718–67296720

that thin or even invisible cirrus were identified(Dutton and Bodhaine, 2001).

2.2. Aerosol chemical composition

TSP, PM10, PM2.5, and PM1.0 filters wereacquired daily at Gosan during the ACE-Asiacampaign (11–27 April 2001). Filters were sampledduring 24-h. Start time of each sampling was 0900local standard time (i.e., 00 UTC). All filters weresampled at ambient temperature and RH. PM10,PM2.5 and PM1.0 cut-offs were achieved usingcyclone impactors. Using these filter samples, anumber of chemical analyses were performed,including mass, major ion components, elementalcarbon (EC) and organic carbon (OC), traceelements, and water-soluble organic components(Chuang et al., 2003; Schauer et al., 2003; Yanget al., 2004; Han et al., 2004).

Size-segregated aerosol filter sampling for inor-ganic ions and water-soluble organics was per-formed by means of a five-stage Berner impactor.The cut-offs for each stage were: o0.2 mm (backupfilter); 0.2–0.5 mm (stage 1); 0.5–1.5 mm (stage 2);1.5–5.5 mm (stage 3); 5.5–10.0 mm (stage 4). Theconcentration of inorganic ions (NHþ4 , Na+, K+,Ca2+, Mg2+, Cl�, NO�3 , and SO2�

4 ) was determinedby ion chromatography (IC), by means of DionexDX4500 and DX500 systems. The accuracy of ICmeasurements was better than 5%. The water-soluble organic carbon (WSOC) content of thesamples was measured by a Shimadzu TOC-5000Aliquid analyzer. The accuracy of TOC analysisranges from 7% for 1 ppmC solution to 3% forconcentrations higher than 2 ppmC (Decesari et al.,2005). Water-insoluble OC (WISOC) concentrationin a sample was also obtained by subtracting theWSOC concentration from the OC concentration ofthe sample (Wang et al., 2005; Kleefeld et al., 2002).

An eight-stage rotating drum impactor with theequivalent aerodynamic cut-off diameters of 0.09,0.26, 0.34, 0.56, 0.75, 1.15, 2.5, 5.0, and 12.0 mm wasalso operated for determining the mass concentra-tion of trace elements (e.g., Al, Si, S, Cl, K, Ca, Ti,V, Cr, Mn, Fe, Ni, Cu, Zn, As, Se, Br, Rb, and Pb).Impaction substrates were analyzed by synchrotronX-ray fluorescence (SXRF) by the DELTA Groupat the University of California at Davis. A detaileddescription of the analytical method and uncertain-ties appears in Bench et al. (2002) and Han et al.(2004).

Super-micrometer concentrations of each compo-nent were determined by the difference between thesub-micrometer and sub-10 mm concentrations. Thedust (inorganic oxidized material) concentration foreach size fraction was constructed from the traceelement data (e.g., Al, Si, Ca, Fe, and Ti) that weredetermined by SXRF. It was assumed that eachelement was present in the aerosol in its mostcommon oxide form (Al2O3, SiO2, CaO, K2O, FeO,Fe2O3, and TiO2). The measured elemental massconcentration was multiplied by the appropriatemolar correction factor as follows (Malm et al.,1994):

½Dust� ¼ 2:2½Al� þ 2:49½Si� þ 1:63½Ca�

þ 2:42½Fe� þ 1:94½Ti�: (1)

3. Results and discussion

3.1. Temporal variation of aerosol chemical

composition

Day-to-day variation of aerosol chemical compo-sition including major inorganic ions, EC, OC, anddust components for sub-micrometer (Dpo1.0mm),super-micrometer (1.0oDpo10.0mm), and sub-10mm (Dpo10.0 mm) size is shown in Fig. 1. Asshown in Fig. 1, it is noted that there was largetemporal variability in aerosol mass concentrationand chemical composition according to air masssource region and episodic events (e.g., Asian dust,pollution, or smoke) (Kim et al., 2005a, 2006). Themean concentration and standard deviation (SD) ofsub-micrometer aerosol (Dpo1.0mm) mass concen-trations during 11–27 April 2001 were 14.476.2mgm�3. Sulfate ion ðSO2�

4 Þ and OC componentsare dominant in sub micrometer range with 17-daymeans and SDs of 3.971.6 and 3.271.7mgm�3,respectively, while the mass concentration for submicrometer dust aerosol was enhanced during 11–13April 2001 (one of the major episodic dust periods atthe site during ACE-Asia) (Chuang et al., 2003; Kimet al., 2005a, 2006). It is noteworthy that theconcentrations of major anthropogenic components(e.g., sulfate, nitrate, and ammonium) are alsoincreased during the aforementioned dust event.The sub-micrometer mass faction of Na+ and Cl�

ion components during the whole sampling periodwas less than 4% of the total sub-micrometeraerosol mass concentration.

ARTICLE IN PRESS

Dates in April 200111 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

Mas

s C

once

ntra

tion

(Dp<

1 µm

) (µ

g/m

3 )

0

5

10

15

20

25

30

35

Na+

NH4+

Cl-NO3

-

SO42-

ECOCDust

Na+

NH4+

Cl-NO3

-

SO42-

ECOCDust

Na+

NH4+

Cl-NO3

-

SO42-

ECOCDust

Dates in April 200111 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27M

ass

Con

cent

ratio

n (1

<D

p<10

µm

) (µ

g/m

3 )

0

20

40

60

80

100

120

Dates in April 200111 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

Mas

s C

once

ntra

tion

(Dp<

10 µ

m)

(µg/

m3 )

0

20

40

60

80

100

120

140

160

(a) (b)

(c)

Fig. 1. Day-to-day variation of major inorganic ions, elemental carbon, organic carbon, and dust aerosols in the sub-micrometer

(Dpo1.0 mm (a)), super-micrometer (1.0oDpo10.0mm (b)), and sub-10mm (Dpo10.0mm (c)) size ranges at the Gosan super-site, Korea

during ACE-Asia (11–27 April 2001).

J. Kim et al. / Atmospheric Environment 40 (2006) 6718–6729 6721

The mass concentration of super-micrometer(1.0oDpo10.0 mm) aerosols was mainly composedof dust (see Fig. 1) while OC and nitrate ions werealso enhanced during the dust episodic period. Themean (7SD) of the super-micrometer aerosolmass concentration for the 17-day period was51.9731.0 mgm�3. The mean (7SD) of the dustcomponent in the super-micrometer size range was41.1726.4 mgm�3. The 17-day mean mass concen-trations of OC, nitrate and sulfate ions in the super-micrometer size range were 3.8, 2.2, and 1.8 mgm�3,respectively. The fractional sum of the super-micrometer aerosol mass contributed by Na+,Cl�, NHþ4 , and EC was only 6.2%.

Sea salt aerosol concentrations for sub-micrometer(Dpo1.0mm) and super-micrometer (1.0oDpo10.0mm) size ranges were calculated using Cl� andNa+ concentrations for each mass range as follows:

½Sea salt� ¼ ½Cl�� þ 1:47½Naþ�; (2)

where 1.47 is the seawater ratio of (Na++K++Mg2++Ca2++SO2�

4 +HCO3� )/Na+ (Holland,1978). The above Eq. (1) prevents the inclusion ofnon-sea salt components (e.g., K+, Mg2+, Ca2+,SO2�

4 , and HCO�3 ) in the sea salt mass and allowsfor the loss of Cl� mass through Cl� depletionprocesses (Quinn et al., 2004). However, the smallobserved mass fractions of sea salt components (i.e.,Na+ and Cl�), both in sub- and super-micrometersize ranges imply that the optical or radiativeimpacts contributed by sea salt aerosols are smallat the Gosan site during ACE-Asia.

Previous studies reveal that a large portion of OCis water-insoluble (Wang et al., 2005; Kleefeld et al.,2002). Mean mass fraction of WISOC of the totalOC was about 0.33. The WISOC component wasmore enriched in the super-micrometer size range(not shown in Fig. 1). Aerosol compositionalchange with time for the sub-10 mm (Dpo10.0 mm)aerosols is also shown in Fig. 1.

ARTICLE IN PRESSJ. Kim et al. / Atmospheric Environment 40 (2006) 6718–67296722

3.2. Extinction closure studies

Closure studies of light extinction due to atmo-spheric aerosols at a given wavelength usually usean overdetermined data set of the chemical, micro-physical, optical, and radiative properties of atmo-spheric aerosols. Measured aerosol extinction(scattering plus absorption) and calculated aerosolextinction from Mie theory are applied to themeasured number and chemical mass size distribu-tion. In this context, closure studies can be used toassess uncertainties in using aerosol chemical andmicrophysical properties to yield aerosol opticalproperties (Bates et al., 2006).

In order to calculate scattering and absorptioncoefficients contributed by several key aerosol types(e.g., mineral dust, water-soluble components, EC,and sea salt), the aerosol optical model (referred toas optical properties of aerosols and clouds(OPAC)) and results from previous studies wereused (Hess et al., 1998; Bates et al., 2006). Theaerosol size distribution and complex spectralrefractive indices of the water-soluble component,EC, and sea salt aerosol types prescribed by theOPAC were used to determine the aerosol radiativeproperties. The water-soluble type consists ofsulfate, nitrate, ammonium, and WSOC (Hesset al., 1998). Scattering coefficients contributed bydust aerosols in the sub- and super-micrometer sizerange were calculated from the empirical relation-ship between mass scattering efficiency and the ratioof the super- and sub-micrometer mass concentra-tion that was introduced by Bates et al. (2006).

The light scattering coefficient (ssp) and absorp-tion coefficient (sap) due to the key aerosol types ata given wavelength were calculated by multiplyingthe mass concentrations of each aerosol type by themass scattering and absorption efficiencies. Sinceprevious studies have reported that the differencesbetween external and internal models were usuallyless than about 10% (Sloane, 1986; Tang, 1996;Malm and Kreidenweis, 1997), we assumed that anestimate of ssp and sap can be achieved byconsidering that each aerosol type is externallymixed (Sciare et al., 2005). The water uptake ofhygroscopic aerosols with RH influences the particlesize and its distribution and complex refractiveindices (Hegg et al., 2002). Hence several key opticalproperties of aerosols (e.g., ssp, single scatteringalbedo, asymmetry parameter, and AOD) are alsoaffected by the aerosol hygroscopicity (Kim et al.,2006). In this study, we assumed that the water-

soluble component and sea salt aerosol arehygroscopic. The aerosol radiative properties ofthese hygroscopic aerosols (e.g., the mass scatteringand absorption efficiencies, single scattering albedo,and asymmetry parameter) calculated from theOPAC model are used to reflect the hygroscopicgrowth effect in the calculation of ssp undernephelometer RH or ambient RH conditions (Hesset al., 1998).

In order to evaluate the consistency between themeasured and modeled aerosol optical properties,the ssp, sap, and sext (extinction coefficient) calcu-lated from the aerosol chemical composition arecompared to those measured by the nephelometerand PSAP (Fig. 2). The measured and modeledcoefficients for sub-10 mm (Dpo10.0 mm) aerosolsare in good agreement with r2 ¼ 0:92 and root meansquare error (RMSE) ¼ 23.6Mm�1 for ssp, r2 ¼

0:93 and RMSE ¼ 3.0Mm�1 for sap, and r2 ¼ 0:93and RMSE ¼ 23.2Mm�1 for sext. These RMSEvalues correspond to 19.7%, 22.2%, and 17.5 %of the period mean ssp (119.4Mm�1), sap(13.6Mm�1), ssp (133.0Mm�1), respectively. Thesedifferences between measured and modeled valuesmay be attributable to diurnal variation of aerosolloading and chemical composition, the changes ofoptical properties due to complex mixing statesbetween particles, and regional (or local) differencesin the aerosol optical parameters (e.g., complexrefractive indices of dust, water-soluble compo-nents, and EC).

Fig. 3 shows the day-to-day variation of themeasured and modeled ssp, sap, and sext at 550 nmfor the size range of Dpo10.0 mm. The accumulatedvertical bars indicate the daily sum of ssp, sap, andsext that were contributed by each aerosol compo-nent. In Fig. 3, the white open circles and verticallines with whiskers in each panel indicate the dailymeans and SDs of ssp, sap, and sext that wereobtained from hourly nephelometer and PSAPmeasurement data. As presented in Fig. 3, thereare large temporal variations of the aerosol opticalproperties. The fractional contributions due to eachaerosol component are also highly variable fromday to day. On the whole the calculated ssp, sap, andsext values are within71 SD from the daily mean ofthe measured ssp, sap, and sext. The aerosolextinction at the surface contributed by each aerosolcomponent is mainly accounted for by dust(51.7%), water-soluble components (40.9%), andEC (5.9%). The light extinction by sea salt aerosoldoes not play a major role.

ARTICLE IN PRESS

σsp (measured) (Mm-1)

0 50 100 150 200 250 300 350

σ sp

(cal

cula

ted)

(M

m-1

)

0

50

100

150

200

250

300

350

Y = -13.97 + 1.14 X (r2=0.92)

σap (measured) (Mm-1)

0 5 10 15 20 25 30 35

σ ap

(cal

cula

ted)

(M

m-1

)

0

5

10

15

20

25

30

35

Y = 0.74 + 0.78 X (r2=0.93)

σext (measured) (Mm-1)

0 100 200 300 400

σ ext

(ca

lcul

ated

) (M

m-1

)

0

100

200

300

400

Y = -15.67 + 1.12 X (r2=0.93)

(a)

(b)

(c)

J. Kim et al. / Atmospheric Environment 40 (2006) 6718–6729 6723

3.3. Aerosol optical depth

AOD is a measure of aerosol loading in thevertical atmospheric column. In general, AOD isdirectly observed by sunphotometer measurements(Kim et al., 2002). Alternatively, AOD at a givenwavelength can be obtained by vertically integratingthe extinction coefficients contributed by eachaerosol component under ambient temperatureand RH conditions. By assuming that water-solubleand sea salt aerosol types are hygroscopic and theaerosol chemical composition is vertically uniform(Xu et al., 2003; Markowicz et al., 2003), thechemically apportioned AOD can be determined byvertically integrating the chemically apportionedscattering and absorption coefficients at ambientRH as follows:

t550 ¼X

i

ti;550 ¼

Z h

surface

Xi

fssp;i;550ðz;RHÞ

þ sap;i;550ðz;RHÞgdz, ð3Þ

where ti,550 is the chemically apportioned AODfrom component i at 550 nm. Based on the aboveassumptions, the height of the homogeneous aerosollayer (h) can be estimated by dividing the AODmeasured at 550 nm by the total sext (i.e., the sum ofthe extinction due to each component) at ambientRH. Although the AOD represents the verticalintegration of light extinction under ambient con-ditions contributed by all the aerosol components inthe atmospheric column, it is very difficult tomeasure the vertical compositional distribution ofeach aerosol component accurately. In order todetermine the fractional contribution of the keyaerosol components to the total AOD, we extendedthe chemically apportioned extinction coefficientsat the surface to the entire columnar atmosphereusing the AODs determined by the sunphotometersand the mass scattering and absorption efficienciesof each aerosol component at ambient RH. In fact,the vertical profile shape and chemical compositionof each aerosol layer must be important todetermine the radiative impact of aerosols. There-fore, the assumptions used in this study are more

Fig. 2. A comparison between measured and calculated scatter-

ing (ssp (a)), absorption (sap (b)), and extinction (sext (c))

coefficients of sub-10mm aerosols at Gosan during ACE-Asia.

The linear regression equations and r2 values between the

measured and calculated properties are exhibited in the bottom

right of each panel. The RMSEs for ssp, sap, and sext were 23.6,

3.0, and 23.2Mm�1, respectively.

ARTICLE IN PRESS

Dates in April 200111 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27Sc

atte

ring

Coe

ffic

ient

(D

p<10

µm

) (M

m-1

)

0

100

200

300

400

Water-solubleSea salt (Dp<1µm)Sea salt (1<Dp<10 µm)Dust (Dp<1 µm)

Dust (1<Dp<10 µm)

Elemental carbon

σsp (Dp<10 µm)-observed

Water-solubleSea salt (Dp<1µm)Sea salt (1<Dp<10 µm)Dust (Dp<1 µm)

Dust (1<Dp<10 µm)

Elemental carbon

σsp (Dp<10 µm)-observed

Water-solubleSea salt (Dp<1µm)Sea salt (1<Dp<10 µm)Dust (Dp<1 µm)

Dust (1<Dp<10 µm)

Elemental carbon

σsp (Dp<10 µm)-observed

Dates in April 200111 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27A

bsor

ptio

n C

oeff

icie

nt (

Dp<

10 µ

m)

(Mm

-1)

0

10

20

30

40

Dates in April 200111 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27E

xtin

ctio

n C

oeff

icie

nt (

Dp<

10 µ

m)

(Mm

-1)

0

100

200

300

400

(a) (b)

(c)

Fig. 3. Light scattering (a), absorption (b), and extinction (c) budgets contributed by several key aerosol types (e.g., water-soluble

components, sea salt, dust, and elemental carbon) measured at Gosan, Korea during ACE-Asia. Daily mean and SD of the measured

scattering, absorption, and extinction coefficients are exhibited by open circles and vertical lines with whiskers, respectively. See text for

more discussion.

J. Kim et al. / Atmospheric Environment 40 (2006) 6718–67296724

appropriate when aerosols are enriched near surfaceor boundary layer.

Fig. 4a shows the temporal variation of AODat 550 nm and the Angstrom exponent (a��log {AOD(368)/AOD(500)}/log {368/500}; one in-dicator representing relative aerosol size distributionin the columnar atmosphere). a was calculated usingthe AODs measured at 368 and 500nm. Largetemporal variation was observed in AOD and a. Thehighest AOD was on 13 April 2001 (0.727 at 550 nm)due to a large input of dust and pollution aerosolstransported from the Asian continent. The lowestAOD at the same wavelength was on 14 April 2001(0.17) affected by a post-frontal clean air mass. The17-day mean (7SD) of AOD at 550 nm was0.4070.14 (Kim et al., 2002). Chemically appor-tioned AOD on four clear days (13, 15, 22, and 26April 2001) during the study period is exhibited inFig. 4b. According to Kim et al. (2006), the air

masses which arrived at Gosan during the ACE-Asia2001 intensive field campaign can be categorized intofour classes: dust: 11–13 April, pollution from theChinese sector: 14–19 April, pollution from theKorean sector: 20–24 April, and smoke and dust:25–27 April. Four days representing each category inFig. 4b are selected for case studies of chemicalapportioning of AOD and aerosol radiative forcing(the forcing estimation will be discussed in the nextsection). In the two cases affected by dust events, 13and 26 April 2001, the AOD contributed by the dustcomponent was dominant and occupied 62.8% and52.9% of the total AOD, respectively. On the otherhand, on the two other days affected by pollutionaerosols (15 and 22 April), water-soluble componentsaccounted for 50.6% and 52.2% of the total AOD.The AOD fraction due to EC ranged from 4.7% to6.9%. Sea salt contribution to the total AOD was nolarger than 3%.

ARTICLE IN PRESS

Dates in April 200111 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

Aer

osol

Opt

ical

Dep

th a

t 550

nm

0.0

0.2

0.4

0.6

0.8

1.0

Ang

stro

m E

xpon

ent (

368/

500

nm)

0.0

0.5

1.0

1.5AOD@550 nmAngstrom Exponent

0.727

Month/Date4/13 4/15 4/22 4/26Fr

actio

n of

Aer

osol

Opt

ical

Dep

th a

t 550

nm (

%)

0

20

40

60

80

1000.231 0.401 0.390

Water-solubleSea salt (Dp<1 µm)Sea salt (1<Dp<10 µm)Dust (Dp<1 µm)Dust (1<Dp<10 µm)Elemental carbon

(a)

(b)

Fig. 4. (a) Aerosol optical depth (AOD) at 550 nm (vertical cyan

bar: daily mean AOD, vertical black line: its SD and Angstrom

exponent (red circle: daily mean, vertical red line and whisker: its

SD) measured at the Gosan super-site, Korea during ACE-Asia.

AOD at 550 nm was obtained from the interpolation of multi-

channel AODs with wavelength. The Angstrom exponent was

calculated by using AODs at 368 and 500 nm. (b) Chemical

fraction of AOD at 550 nm on 13, 15, 22, and 26 April 2001 at

Gosan. Daily mean of AOD at 550 nm of each day is exhibited at

the top of the figure.

J. Kim et al. / Atmospheric Environment 40 (2006) 6718–6729 6725

3.4. Aerosol radiative forcing at the surface

The aerosol radiative forcing at the surface wasestimated by using a modified Fu–Liou RTM (Fuand Liou, 1993), which is currently employed by theSurface and Atmospheric Radiation Budget(SARB) (Charlock et al., 1997). The chemicallyapportioned aerosol optical depth (CAOD) of thekey aerosol types at 550 nm was used as input datato calculate the direct and diffuse irradiances forestimating aerosol radiative forcing in the solarspectral region. Based on the International Global

Biosphere Program (IGBP), a category 17-oceantype was set as the surface albedo (Jin et al., 2002).In addition, the vertical profile of pressure, tem-perature, and water vapor mixing ratio, and ozonemixing ratio were used as input data for running theRTM. The ozone mixing ratio was obtained fromozonesonde measurements (launched on 19 April2001) performed by NOAA/CMDL while the otherparameters were obtained from routine radiosondemeasurements performed twice a day (at 00 and 12UTC) at the Gosan station of the Korea Meteor-ological Administration.

If the CAOD estimated in this study reflects wellthe aerosol optical properties and loading, both themeasured and modeled values of the direct anddiffuse irradiances should be in good agreement. Tovalidate the assumption and the performance of themodel used in this study, the measured and modeledirradiances on 15 and 22 April 2001 were comparedin Fig. 5. As shown in Fig. 5, there was goodagreement between the measured and modeledirradiances. The diurnal variations of aerosolradiative forcing at the surface contributed by eachaerosol component are shown in the lower part ofeach panel of Fig. 5. On 15 April 2001, the diurnalaveraged aerosol radiative forcing (DDF) at thesurface was �27.6Wm�2. This value is comparableto those obtained on the same day in previousstudies (e.g., Bush and Valero, 2003: �34.0Wm�2;Won et al., 2004: �28.7Wm�2). The differences inthese DDF values may be a result of the differencesin the AODs determined by the sunphotometers(Kim et al., 2005b) and RTMs (Boucher et al.,1998). Consequently, the small difference of esti-mated forcing among different approaches confirmsthat the approach and assumptions of this study arevalid. As shown in Fig. 5, aerosol radiative forcingon 2 days shows a large diurnal variability with thesolar zenith angle. The difference in the morning on22 April between measured and modeled radiativefluxes might result from the time variation of theaerosol loading and properties during the daytime.DDF values on 13, 22, and 26 April 2001 were�63.1, �37.6, and �36.2Wm�2, respectively.

The chemical fraction of the surface aerosolradiative forcing on the selected 4 days appears inFig. 6. DDF values for each day are shown at thetop of the figure. On 13 April DDF values due tosuper-micrometer (1.0oDpo10.0 mm) and sub-mi-crometer (Dpo1.0 mm) dust particles were �32.6and �4.9Wm�2, respectively. These correspond toabout 60% of the total diurnal forcing. And DDF

ARTICLE IN PRESS

Local Standard Time6 8 10 12 14 16 18 20

Rad

iati

ve F

orci

ng a

t the

Sur

face

-100

-80

-60

-40

-20

0

Water-solubleSea salt (Dp<1 µm) Sea salt (1<Dp<10 µm)

Dust (Dp<1 µm) Dust (1<Dp<10 µm)Elemental carbon

Irra

dian

ce (

W/m

2 )

200

400

600

800

1000

1200Total (cal.)Dir. (cal.)Diff. (cal.)Total (obs.)Dir. (obs.)Diff. (obs.)

April 15, 2001

Local Standard Time6 8 10 12 14 16 18 20

Rad

iati

ve F

orci

ng a

t the

Sur

face

-140

-120

-100

-80

-60

-40

-20

0

Water-solubleSea salt (Dp<1 µm)Sea salt (1<Dp<10 µm)

Dust (Dp<1 µm) Dust (1<Dp<10 µm)

Elemental carbon

Irra

dian

ce (

W/m

2 )

200

400

600

800

1000

1200Total (cal.)Dir. (cal.)Diff. (cal.)Total (o bs.)Dir. (obs.)Diff. (obs.)

April 22, 2001

(a) (b)

Fig. 5. Direct and diffuse irradiances determined by radiative flux measurements (obs.) and the RTM calculations (cal.) at Gosan in Korea

on the representative cloud-free days of 15 April 2001 (a) and 22 April 2001 (b). The diurnal variations of the chemically apportioned

shortwave radiative forcing for each aerosol component are also shown.

J. Kim et al. / Atmospheric Environment 40 (2006) 6718–67296726

values contributed by water-soluble componentsand EC were �14.1 and �11.1Wm�2, respectively.On 22 April the fractions due to water-soluble andEC were �13.7 and �9.3Wm�2, respectively. Thesecomprise about 61.1% of the total diurnal forcingso a high contribution to the surface forcing wasmade by anthropogenic sub-micrometer particles.In the case of a dust and smoke episode day (26April 2001), DDF due to the dust componentwas �18.8Wm�2 while DDF values due to ECand water-soluble components were �6.7 and�10.7Wm�2, respectively. The above results showthat the mineral dust component, especially for thesuper-micrometer (1.0oDpo10.0 mm) dust, greatlycontributed to the DARF at the surface. The finewater-soluble components (mixture of sulfate,nitrate, ammonium and WSOC) and EC alsosignificantly contributed to aerosol radiative for-cing. The contribution due to sea salt aerosolswas minor. Consequently, the 17-day mean(�38.3Wm�2) of the aerosol radiative forcing atthe surface is attributable to dust (�17.5Wm�2),water-soluble components (�10.3Wm�2), and EC(�10.1Wm�2). In this study, WISOC componentwas not considered for estimating extinction, AOD,

and the aerosol forcing because of the limitedinformation on its optical properties and relativesmall mass fraction in sub-micrometer size range.However, a small amount of WISOC componentmay significantly affect aerosol extinction andforcing because of large extinction and forcingefficiencies of fine aerosols. In order to betterunderstand the seasonality and long-term trends ofchemically apportioned aerosol radiative forcingover East Asia, long-term network measurementsthat employ size-resolved measurements of aerosolchemical, microphysical, and optical properties aswell as Sun photometer and radiative flux measure-ments, are needed.

4. Conclusion

Chemical apportionment of light extinction,optical depth, and surface radiative forcing due toseveral key aerosol types (mineral dust, water-soluble, EC, and sea salt) was investigated for theGosan super-site, South Korea during the ACE-Asia 2001 intensive field campaign. The size-resolved chemical, and optical measurements ofatmospheric aerosols were used to estimate the

ARTICLE IN PRESS

-63.1

Month/Date4/13 4/15 4/22 4/26

Frac

tion

of A

eros

ol R

adia

tive

Forc

ing

at th

e Su

rfac

e (%

)

0

20

40

60

80

100-27.6 -37.6 -36.2 W/m2

Water- soluble

Sea salt (Dp<1 µm)

Sea salt (1<Dp<10 µm)Dust (Dp<1 µm)

Dust (1<Dp<10 µm)Elemental carbon

Fig. 6. Chemical fraction of shortwave direct aerosol radiative

forcing at the surface estimated on 13, 15, 22, and 26 April 2001

at Gosan, Korea. Diurnal averaged forcing of each day is

exhibited at the top of the figure in Wm�2.

J. Kim et al. / Atmospheric Environment 40 (2006) 6718–6729 6727

aerosol-light extinction coefficients at the surface.The calculated scattering, absorption, and extinc-tion coefficients for Dpo10 mm aerosols agreed wellwith measured scattering, absorption, and extinc-tion coefficients with r2 ¼ 0:92, 0.93, and 0.93,respectively, and RMSEs of 23.6, 3.0, 23.2Mm�1,respectively. Most of the calculated scattering,absorption, and extinction coefficients were distrib-uted within 71 SD from the diurnal means of thescattering, absorption, and extinction coefficients,respectively, obtained from hourly measurementdata. The chemical apportionment of measured

AOD was achieved under the assumption of avertically uniform distribution of aerosol chemicalcomposition. Vertical distribution of light extinctiondue to aerosol with its chemical composition is animportant factor to determine the radiative andclimatic forcing effects of aerosols, including anestimation of radiative heating/cooling rates in theatmosphere. Therefore, the approaches used in thisstudy are more realistic when aerosols are enrichedin the well-mixed boundary layer. Thus situationmay reduce the uncertainty of the aerosol radiativeforcing estimated in this study.

A RTM was employed to estimate the diurnalvariation of aerosol radiative forcing at the surface.The calculated direct and diffuse fluxes at thesurface generally coincided well with the measureddirect and diffuse fluxes. A large fraction of the 24-haveraged aerosol radiative forcing at the surface isattributable to mineral dust, water-soluble compo-nents, and EC. The surface aerosol radiative forcingcontributed by sea salt aerosol was minor. Theapproach used in this study could be readily appliedto monitor aerosol radiative forcing with chemicallyspecified information, especially for long-termground-based measurement sites.

Acknowledgments

The authors are deeply grateful to Drs. ThomasP. Charlock and Fred Rose (NASA/LaRC) forproviding the RTM. This research was supported bythe BK 21, the Korea Meteorological Administra-tion Research and Development Program underGrant CATER_2006-4104. FBJ was supported bythe US National Science Foundation AtmosphericChemistry Program under Award ATM-0333817.This research is a contribution to the InternationalGlobal Atmospheric Chemistry (IGAC) Core Pro-ject of the International Geosphere BiosphereProgram (IGBP) and is part of the IGAC AerosolCharacterization Experiments (ACE). Authors aregrateful to two anonymous reviewers for theirthorough comments and insightful suggestions.

References

Bates, T., et al., 2006. Aerosol direct radiative effects over the

northwest Atlantic, northwest Pacific, and North Indian

Oceans: estimates based on in situ chemical and optical

measurements and chemical transport modeling. Atmospheric

Chemistry and Physics 6, 1657–1732.

ARTICLE IN PRESSJ. Kim et al. / Atmospheric Environment 40 (2006) 6718–67296728

Bench, G., Grant, P.G., Ueda, D., Cliff, S.S., Perry, K.D., Cahill,

T.A., 2002. The use of STIM and PESA to measure profiles of

aerosol mass and hydrogen content, respectively, across

Mylar rotating drums impactor samples. Aerosol Science &

Technology 36, 642–651.

Boucher, O., et al., 1998. Intercomparison of models representing

direct shortwave radiative forcing by sulfate aerosols. Journal

of Geophysical Research 103 (D14), 16979–16998.

Bush, B.C., Valero, F.P.J., 2003. Surface aerosol radiative forcing

at Gosan during the ACE-Asia campaign. Journal of

Geophysical Research 108 (D23), 8660.

Charlock, T.P., et al., 1997. Compute Surface and Atmospheric

Fluxes (System 5.0). CERES Algorithm Theoretical Basis

Document, 84pp.

Charlson, R.J., et al., 1992. Climate forcing by anthropogenic

aerosols. Science 255, 423–430.

Christopher, S.A., Zhang, J., 2002. Shortwave aerosol radiative

forcing from MODIS and CERES observations over the

oceans. Geophysical Research Letters 29 (18).

Chuang, P.Y., Duvall, R.M., Bae, M.S., Jefferson, A., Schauer,

J.J., Yang, H., Yu, J.Z., Kim, J., 2003. Observations of

elemental carbon and absorption during ACE-Asia and

implications for aerosol radiative properties and climate

forcing. Journal of Geophysical Research 108 (D23), 8634.

Conant, W.C., et al., 2003. A model for the radiative forcing

during ACE-Asia derived from CIRPAS Twin Otter and R/V

Ronald H. Brown data and comparison with observations.

Journal of Geophysical Research 108 (D23), 8661.

Decesari, S., Facchini, M.C., Fuzzi, S., McFiggans, G.B., Coe,

H., Bower, K.N., 2005. The water-soluble organic component

of size-segregated aerosol, cloud water and wet depositions

from Jeju Island during ACE-Asia. Atmospheric Environ-

ment 39, 211–222.

Dutton, E.G., Bodhaine, B.A., 2001. Solar irradiance anomalies

caused by clear-sky transmission variations above Mauna

Loa: 1958–99. Journal of Climate 14, 3255–3262.

Fu, Q., Liou, K.-N., 1993. Parameterization of the radiative

properties of cirrus clouds. Journal of the Atmospheric

Sciences 50, 2008–2025.

Han, J.S., et al., 2004. Characteristics of ion components and

trace elements of fine particles at Gosan, Korea in spring time

from 2001 to 2002. Environmental Monitoring and Assess-

ment 92, 73–93.

Hegg, D.A., Covert, D.S., Crahan, K., Jonssen, H., 2002. The

dependence of aerosol light scattering on RH over the Pacific

Ocean. Geophysical Research Letters 29 (8), 1219.

Hess, M., Koepke, P., Schult, I., 1998. Optical properties of

aerosols and clouds: The software package OPAC. Bulletin of

the American Meteorological Society 79, 831–844.

Holland, H.D., 1978. The Chemistry of the Atmosphere and

Oceans. Wiley, Hoboken, NJ, 154pp.

Huebert, B.J., et al., 2003. An overview of ACE-Asia: strategies

for quantifying the relationships between Asian aerosols and

their climatic impacts. Journal of Geophysical Research 108

(D23), 8633.

Intergovernmental Panel on Climate Change (IPCC), 2001.

Climate Change 2001. In: Houghton, J.T., et al. (Eds.).

Cambridge University Press, New York.

Jin, Z., Charlock, T.P., Rutledge, K., 2002. Analysis of broad-

band solar radiation and albedo over the ocean surface at

COVE. Journal of Atmospheric and Oceanic Technology 19,

1585–1601.

Kim, J., Choi, B.-C., Oh, S.-N., 2002. Spectral aerosol optical

depth retrieved from ground-based sunphotometer measure-

ments at Kosan during ACE-Asia. Journal of the Korean

Meteorological Society 38 (4), 355–367.

Kim, J., Yoon, S.-C., Jefferson, A., Zahorowski, W., Kang,

C.-H., 2005a. Air mass characterization and source region

analysis for the Gosan super-site, Korea during ACE-Asia

2001 field campaign. Atmospheric Environment 39,

6513–6523.

Kim, S.-W., Jefferson, A., Yoon, S.-C., Dutton, E.G., Ogren,

J.A., Valero, F.P.J., Kim, J., Holben, B.N., 2005b. Compar-

isons of aerosol optical depth and surface shortwave

irradiance and their effect on the aerosol surface forcing

estimation. Journal of Geophysical Research 110, D07204.

Kim, J., Yoon, S.-C., Jefferson, A., Kim, S.-W., 2006. Aerosol

hygroscopic properties during Asian dust, pollution, and

biomass burning episodes at Gosan, Korea in April 2001.

Atmospheric Environment.

Kleefeld, S., et al., 2002. Importance of organic and black carbon

in atmospheric aerosols at Mace Head, on the West Coast of

Ireland. Atmospheric Environment 36, 4479–4490.

Malm, W.C., Kreidenweis, S.M., 1997. The effects of models of

aerosol hygroscopicity on the apportionment of extinction.

Atmospheric Environment 31, 1965–1976.

Malm, W.C., et al., 1994. Spatial and seasonal trends in particle

concentration and optical extinction in the United States.

Journal of Geophysical Research 99, 1347–1370.

Markowicz, K.M., et al., 2003. Influence of relative humidity

on aerosol radiative forcing: an ACE-Asia experiment

perspective. Journal of Geophysical Research 108 (D23),

8662.

McInnes, L., et al., 1998. Apportionment of light scattering and

hygroscopic growth to aerosol composition. Geophysical

Research Letters 25 (4), 513–516.

Nemesure, S., Wagener, R., Schwartz, S.E., 1995. Direct short-

wave forcing of climate by the anthropogenic sulfate

aerosol: sensitivity to particle size, composition, and relative

humidity. Journal of Geophysical Research 100 (D12),

26105–26116.

Penner, J.E., et al., 1994. Quantifying and minimizing uncertainty

of climate forcing by anthropogenic aerosols. Bulletin of the

American Meteorological Society 75, 375–400.

Quinn, P.K., et al., 2004. Aerosol optical properties measured on

board the Ronald H. Brown during ACE-Asia as a function of

aerosol chemical composition and source region. Journal of

Geophysical Research 109, D19S01.

Schauer, J.J., et al., 2003. ACE-Asia intercomparison of a

thermal–optical method for the determination of particle-

phase organic and elemental carbon. Environmental Science

and Technology 37, 993–1001.

Sciare, J., et al., 2005. Aerosol mass closure and reconstruction of

the light scattering coefficient over the Eastern Mediterranean

Sea during the MINOS campaign. Atmospheric Chemistry

and Physics 5, 2253–2265.

Sloane, C.S., 1986. Effect of composition on aerosol

light scattering efficiencies. Atmospheric Environment 20,

1025–1037.

Takemura, T., Nakajima, T., Higurashi, A., Ohta, S., Sugimoto,

N., 2003. Aerosol distributions and radiative forcing over the

Asian Pacific region simulated by Spectral Radiation-Trans-

port Model for Aerosol Species (SPRINTARS). Journal of

Geophysical Research 108 (D23), 8659.

ARTICLE IN PRESSJ. Kim et al. / Atmospheric Environment 40 (2006) 6718–6729 6729

Tang, I.N., 1996. Chemical and size effects of hygroscopic

aerosols on light scattering coefficients. Journal of Geophy-

sical Research 101 (D14), 19245–19250.

Wang, H., Kawamura, K., Shooter, D., 2005. Carbonaceous and

ionic components in wintertime atmospheric aerosols from

two New Zealand cities: implications for solid fuel combus-

tion. Atmospheric Environment 39, 5865–5875.

Won, J.-G., et al., 2004. Estimation of direct radiative forcing of

Asian dust aerosols with sun/sky radiometer and lidar

measurements at Gosan, Korea. Journal of the Meteorologi-

cal Society of Japan 82, 115–130.

Xu, J., Bergin, M.H., Greenwald, R., Russell, P.B., 2003. Direct

aerosol radiative forcing in the Yangtze delta region of China:

observation and model estimation. Journal of Geophysical

Research 108 (D2), 4060.

Yang, H., Xu, J., Wu, W.-S., Wan, C.-H., Yu, J.Z., 2004.

Chemical characterization of water-soluble organic aerosols

at Jeju Island collected during ACE-Asia. Environmental

Chemistry 1, 1–5.

Yoon, S.-C., Kim, J., 2006. Influences of relative humidity on

aerosol optical properties and aerosol radiative forcing during

ACE-Asia. Atmospheric Environment 40 (23), 4328–4338.

Yoon, S.-C., Won, J.-G., Omar, A.H., Kim, S.-W., Sohn, B.-J.,

2005. Estimation of the radiative forcing by key aerosol types

in worldwide locations using a column model and AERO-

NET data. Atmospheric Environment 39, 6620–6630.

Yu, H., et al., 2004. Direct radiative effect of aerosols as

determined from a combination of MODIS retrievals and

GOCART simulations. Journal of Geophysical Research 109,

D03206.

Zhang, J., Christopher, S.A., Remer, L.A., Kaufman, Y.J., 2005.

Shortwave aerosol radiative forcing over cloud-free oceans

from Terra: 2. Seasonal and global distributions. Journal of

Geophysical Research 110, D10S24.