Trends in atmospheric elemental carbon concentrations from 1835 to 2005

Trends in atmospheric elemental carbon concentrations from 1835

to 2005

Liaquat Husain,1,2 A. J. Khan,1 Tanveer Ahmed,2 Kamal Swami,1 A. Bari,1

James S. Webber,1,2 and Jianjun Li2,3

Received 17 September 2007; revised 25 January 2008; accepted 28 February 2008; published 1 July 2008.

[1] Elemental carbon (EC) aerosols absorb solar radiation which results in heating of theatmosphere. Recent increases in the atmospheric burden of EC may account for �10to 15 % of global warming. Long-term EC data, however, are sparse. We report here ourmeasurements of annual mean atmospheric EC concentration, [EC]atm, from filtersamples collected daily from 1978 to 2005 at Whiteface Mountain, NY using the thermaloptical method. The [EC]atm for 1978–1986, 1987–1996, and 1997–2005 were, 550,225, and 62 ng m�3, respectively. We also collected �55 cm long sediment cores fromWest Pine Pond near Whiteface Mountain. The cores were sliced and their 210Pb agesdetermined. The first (top) five slices each represented sediment deposition over 7 yearsand the remaining 13 years each. EC was chemically separated from the sedimentsamples from four cores, and its concentration in each slice was determined using thethermal optical method. The [EC]sed followed closely that of [EC]atm from 1978 to 2005.Assuming wet and dry deposition as the only source, we can show that [EC]sed =K[EC]atm, where K (m3 g�1) is a constant for a given lake. From [EC]atm, and [EC]sed forthe 1978–2005 period, K was determined to be 10,400 ± 4,400 m3 g�1. With this valueused for K and [EC]sed, the [EC]atm values were determined from 1835 to 1978. The[EC]atm from 1835–1862 was �30 ng m�3, which may be close to the preindustrialbackground level. The [EC]atm was 65 ng m�3 for the 1863–1875 period, then increasedsharply, reaching a maximum value, 760 ng m�3, from 1917–1930. From 1931–1943through 1978–1984, the concentration decreased gradually, from 680 to 560 ng m�3. Theconcentrations for 1985–1991, 1992–1998, and 1999–2005 were 295, 195, and60 ng m�3, respectively. Model calculations for BC emissions from fossil fuelcombustion for the US by Novakov et al. (2003) qualitatively reproduce the trenddetermined experimentally in this work.

Citation: Husain, L., A. J. Khan, T. Ahmed, K. Swami, A. Bari, J. S. Webber, and J. Li (2008), Trends in atmospheric elemental

carbon concentrations from 1835 to 2005, J. Geophys. Res., 113, D13102, doi:10.1029/2007JD009398.

1. Introduction

[2] The atmospheric loading of carbonaceous aerosols,consisting of elemental carbon (EC) and organic carbon(OC), has increased substantially since preindustrial times[IPCC, 1995]. EC is a product of incomplete combustion. Itis an inert material [Gelinas et al., 2001], characterized byhigh molecular weight, nonvolatility [Burtscher et al.,2001], a high degree of aromaticity, few functional groups,and a graphitic structure. The term elemental carbon, blackcarbon, and soot are used interchangeably. The term is anoperational one, dependent on the particular method of

measurement. (We have used the term EC for our thermal-optical measurements, and EC or BC as used by theauthors of the papers cited). EC concentration, [EC], where[X] refers to the concentration of X, has been found tovary widely, from a background level of 1 ng m�3 at theSouth Pole [Hansen et al., 1988], to between 7000 and21,000 ng m�3 in Paris, France [Ruellan and Cachier,2001]. Exceedingly high concentrations, with daily averagevalues of up to �25,000 ng m�3, are now commonlyfound in urban areas of South Asia [Husain et al., 2007].[3] EC and other carbonaceous aerosols constitute as

much as 50% of the PM2.5 mass in urban air, and havebeen linked to increasing morbidity and mortality due torespiratory and cardiac stresses [e.g., Dockery et al., 1993;Dockery and Pope, 1994; Gwynn et al., 2000]. EC aerosolsalso play a key role in the Earth’s temperature regulation,because of their strong ability to absorb solar radiation.They may be the second biggest contributor to globalwarming after greenhouse gases [Jacobson, 2002]. Themagnitude of the forcing by EC is uncertain, estimated at

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 113, D13102, doi:10.1029/2007JD009398, 2008ClickHere

for

FullArticle

1Wadsworth Center, New York State Department of Health, Albany,New York, USA.

2Department of Environmental Health Sciences, School of PublicHealth, State University of New York, Albany, New York, USA.

3Ambient Air Quality Monitoring, China National EnvironmentalMonitoring Center, Beijing, China.

Copyright 2008 by the American Geophysical Union.0148-0227/08/2007JD009398$09.00

D13102 1 of 10

�0.5 W/m2 [Sato et al., 2003]. Recent studies of bothterrestrial and marine sediments suggest that EC plays avery important role in the global carbon cycle and couldinfluence atmospheric {O2} and {CO2} [Crutzen andAndreae, 1990; Kuhlbusch and Crutzen, 1995; Kuhlbusch,1998]. EC can also affect climates on a regional scale, e.g.,increased precipitation and droughts in different regions inChina and a moderate cooling in India over the last decadeappear to be related to increases in atmospheric [EC][Menon et al., 2002].[4] Owing to its atmospheric lifetime of about a week, the

[EC] at a given location is dependent not only on theemissions from local sources in the vicinity but also thosehundreds of kilometers away. Therefore radiative forcingcalculations require [EC] values representative of specificregions. Such data are relatively sparse. Hence models areused to estimate [EC] based on emissions from varioustypes of fossil fuels and biofuels. Emission of EC, however,is dependent not only on the fuel type burned but also uponthe type of technology used. Long–term field data areneeded to validate these models.[5] Efforts have been made to deduce long term annual

[EC] in the United Kingdom based on black smoke datafrom a network of stations and SO2 emissions [Novakov andHansen, 2004], in the Canadian Arctic by aerosol absorp-tion measurements [Sharma et al., 2004], and from a ruralnetwork by the IMPROVE program [Malm et al., 2004].Lake sediments provide an archive of environmental changeover long periods of time. For remote mountain lakes,atmospheric transport is a major pathway for the loadingof contaminants to the system. Pollutants are primarilyremoved from the atmosphere by precipitation and depos-ited in lake sediments. Thus undisturbed, dated sedimentmay provide a record of atmospheric deposition of thesepollutants and allow temporal trends in their levels to beassessed over many decades or even centuries [Kralovecet al., 2002; Muri et al., 2006, 2002; Lavanchy et al., 1999;Jenk et al., 2006]. Muri et al. [2002] studied the depositionof EC in five remote alpine lakes in Slovenia. Owing to alack of information of the transfer rates for EC aerosolsfrom the atmosphere to lake sediments, these data could notbe converted into atmospheric concentrations with anyreasonable certainty. We propose to address this issue bycalibrating atmospheric deposition to the sediments of alake in the Adirondack Mountains of New York State,through comparison by direct measurements of the atmo-spheric aerosols over nearly three decades to values deter-mined in the lake sediments. We have been collectingaerosols daily or every 48 h at Whiteface Mountain, NY,from July 1978 to the present. This site is rural, and devoidof large population and emission sources. We present herethe measurements of [EC]atm in monthly composites and usethe data to determine annual mean [EC]atm for the 1978 to2005 period. We extend the [EC] measurements by aboutanother 100 years by a new approach. As we show later inthis paper, for a given lake, [EC]atm in the atmosphere canbe related to [EC] in sediments by [EC]sed = K[EC]atmwhere K is a constant dependent on wet and dry depositionand sedimentation rate of EC. We determined K from themeasured [EC]atm and [EC]sed for the period �1978 to�2005, and then applied it to the measurements of [EC]sedfor the �1835 to 1980 period, to obtain the historical

[EC]atm. The [EC]atm concentrations so deduced are thencompared with the model estimates based on emissionsfrom fossil fuel and biofuel combustion.

2. Experimental Methods

2.1. Aerosol Collection

[6] Whiteface Mountain (44.37�N, 73.90�W) is locatedin the Adirondack Mountains of New York State. Ourobservatory is located at the summit, 1.5 km above meansea level. The nearest urban centers are Albany, 200 km tothe south; Montreal, Canada, 130 km to the north; andSyracuse, 220 km to the southwest. Samples have beencollected continuously from July 1978, with data missingonly for 1980 and 1982. Daily aerosol samples werecollected on Whatman 41 filters using hi-vol samplesequipped with mass flow controllers (airflow controlled to± 5%) and back flow protection. Samples were collectedevery 24 h from 1978 to 1988 and every 48 h from 1989 to2005. In addition, many sampling campaigns were con-ducted in which sampling frequency was reduced to a fewhours. These campaigns lasted for months at a time. Thedata have been used to study composition, sources, trans-formation, and transport of aerosols (see, e.g., Husain et al.[1998, 2004], and references therein).2.1.1. Determination of EC in Aerosols[7] Owing to the high carbon content of Whatman 41

filters, this material is not suitable for a direct determinationof the concentration of EC collected on it. Therefore wedeveloped a technique to transfer EC from Whatman 41 toquartz filters with little loss [Li et al., 2002]. Briefly,cellulose substrate is dissolved in a high-purity 70% ZnCl2solution at 50�C, and EC is transferred to a quartz filter. Thelatter has very low background for EC and is suitable for ECdetermination using the thermal-optical method. Recoveriesof EC using this analytical technique were determinedthrough deposition of known amounts of EC standard(carbon black from acetylene, Alfa Aesar, Ward Hill, MA)on Whatman 41 filters and processing of the filters asdescribed elsewhere [Li et al., 2002]. The recoveries were93 ± 4%.[8] Commonly, two methods are in use to measure the EC

and OC by the thermal-optical method. These are (1) theIMPROVE method [Chow et al., 1993], known as TotalOptical Reflectance (TOR) and (2) the NIOSH method[NIOSH, 1996; Birch and Cary, 1996] referred as TotalOptical Transmission (TOT). We used the TOT method.Details of the experimental method used were given byKhan et al. [2006].2.1.2. Aerosol Collection Efficiency of Whatman41 Filters[9] To assess the collection efficiency of Whatman 41

filters for EC aerosols, we determined [EC] in sampleswhere two Whatman 41 filters in tandem were used tocollect aerosols from January to April, 1983 at WhitefaceMountain, NY. The [EC] were determined in the monthlycomposites of the front and back filters. The results (Table 1)show that �85 to 90% (mean, 88.4%) of the EC wascollected in the front one. This finding is in very goodagreement with our SO4

2- measurements in the aliquot of thesame samples, which showed a collection efficiency of 95%[Husain and Dutkiewicz, 1990]. A similar conclusion was

D13102 HUSAIN ET AL.: ATMOSPHERIC EC CONCENTRATIONS, 1835–2005

2 of 10

D13102

reached when we analyzed individually front and back of30 daily filters.[10] Since quartz filters are commonly used for EC

analysis by the thermal-optical method, we compared thecollection efficiency of the Whatman 41 filters relative toquartz by collecting simultaneously, side by side, aerosolson quartz and Whatman 41 filters in Albany. The datayielded a linear relationship with a slope of 0.92 and r2 of0.82. The collection efficiency of 92% was in excellentagreement with the results obtained above using a stack oftwo filters. From these data and the ones cited above wededuced a mean collection efficiency of 90% and appliedit to all measurements.2.1.3. Field Blanks[11] Over 28 years of continuous sampling at Whiteface

Mountain, we have collected a large number of 24-h fieldblanks. The sampling system has been protected against thecollection of exhaust or emissions from pumps by installinga check valve to eliminate any possibility of back-flowthrough the motor. For the past 10 years, motors withcarbonless brushes have been used. Approximately 40 fieldblanks collected over the duration of the sampling periodwere analyzed for EC. The EC concentrations were invari-

ably at or below the detection limit for the entire period. Noblank correction was needed.2.1.4. Method Detection Limit[12] The method detection limit (MDL) was established

by analyzing the samples in which the EC loadings werenear background levels. We took �81 cm2 of Whatman 41filter and followed the procedure as given by Li et al.[2002]. The MDL for our EC measurement depends uponthe filter loading and the amount of filter paper used fordissolution. If the filter loading is low, the amount of filteranalyzed can be increased, to obtain EC signals muchhigher than the instrument detection limit of 20 ng cm�2

(IDL). For an air volume of �1500 m3, the MDL of our ECmeasurement is 10 ng m�3.2.1.5. Quality Control[13] For quality assurance, three standard sucrose samples

with varying but known (OC) concentrations were analyzeddaily before analysis of the samples. Sucrose pyrolysis takesplace during analysis and can be tested for the charringcorrections. No EC was detected in these samples, suggest-ing that the instrument properly corrected the charring effectof organic material. For EC, Carbon Black (acetylene, 100%compressed from Alfa Aesar, Ward Hill, MA) was used asstandard. One standard sample was analyzed with each setof the EC samples. The recoveries of EC varied from 98.5to 100.5%.

2.2. Lake Sediment Core Collection

[14] Sediment cores were collected from four lakes within�50 km of Whiteface Mountain, NY, which is located in the5.9 million acres Adirondack Park (Figure 1). There areapproximately three thousand lakes, many at high altitudeand some hundreds of meters away from roads. It is a truewilderness. We selected lakes with minimal human activity(no motor boats, no picnicking, no camping, and locatedaway from major roads), low watershed to surface area ratio(<10), low flushing rates (�1 per year), high lake clarity,

Table 1. Collection Efficiency of Whatman 41 Filters for EC-

Bearing Aerosols

1983

Number ofDays

Sampled

EC inFront Filter,ng m�3

EC inBack Filter,ng m�3

CollectionEfficiency,

%

January 26 712 ± 38 128 ± 10 84.8February 28 1430 ± 76 168 ± 11 89.5March 31 671 ± 37 80 ± 7 89.3April 19 385 ± 24 43 ± 6 90.0Mean 88.4

Figure 1. Map of New York State showing the locations of the Whiteface Mountain and lakes sampled.The cores used for [EC] determinations were from West Pine Pond. See text.

D13102 HUSAIN ET AL.: ATMOSPHERIC EC CONCENTRATIONS, 1835–2005

3 of 10

D13102

and low deciduous organic matter, to avoid effects of localcontamination. After visiting a large number of lakes, weselected four lakes which met our criteria. These lakes wereClear Pond, West Pine Pond, Bear Pond, and Deer Pond(Figure 1). The location and other relevant information aregiven in Table 2. We collected six cores from differentlocations from the deepest part of each lake. Coring wascarried out with a gravity-driven coring device attached toremovable plastic tubes with 5.12 cm inner diameter. Thecores were �55 cm long and sectioned in the field imme-diately after collection to avoid perturbation. Because thesedimentation rate was not known a priori the cores werehorizontally sectioned in 1.25 to 4.5 cm thick slices. Thesections were transferred to 5-cm diameter plastic jars in thefield, and transported to the laboratory at 4�C. Individualsamples were weighed, freeze dried, and weighed again, fordetermination of moisture content. The freeze-dried sampleswere ground to a fine powder, homogenized, and stored foranalysis.2.2.1. Chemical Separation of EC From Sediment[15] Determination of EC in complex organic and mineral

matrices such as lake sediments is quite challenging. Pro-cedures to chemically separate EC from sediment have beendescribed by various authors [Lim and Cachier, 1996;Gelinas et al., 2001; Song et al., 2002]. We have modifiedthe method to meet the requirements of the thermal opticaltechnique to measure EC. Briefly, the following steps wereused. (1) 0.5–2.0 g of dried sediment sample + 3M HCl at60�C were left overnight for removal of carbonates.(2) Next the sample was heated with 22M HF and 6M HClin an evaporating dish on a steam bath until dryness, toremove silicates. (3) The sample was heated at 60�C over-night in 10 M HCl, to remove any remaining carbonates.(4) It was heated at 60�C overnight in 0.1M NaOH toremove humic acids. (5) It was heated in 0.1M K2Cr2O7 +2M H2SO4 at 60�C for 72 h to remove organic carbon;and (6) The residue was treated with 70% ZnCl2 and thesolution was filtered through a quartz filter. Recoverieswere determined using NIST SRM 1649A (urban partic-ulate matter), Alfa Aesar carbon black standard, and NISTSRM 8704 (Buffalo River sediment). An average recoveryof 92% was obtained for EC.2.2.2. Age Determination of Sediment Core Via 210PbDating[16] Naturally occurring 210Pb (half-life, 22.3 years) is

widely used in radioactive dating of lake sediments [e.g.,Appleby, 2001]. It is a decay product of 222Rn, formed by226Ra decay in soils. 222Rn diffuses through the soil into theatmosphere, where it decays through a sequence of short-

lived radionuclides to 210Pb, which has a residence time ofabout a week, and is removed primarily by precipitation. Itis efficiently adsorbed on sediment particles in the watercolumn and these sink to the bottom of the lake. If there isno perturbation of the deposited sediments, 210Pb activity inthe sediment will decrease exponentially, with a half-life of22.3 years. However, the lake sediments also contain 226Ra(half-life, 1600 years), which contributes to 210Pb via in situdecay. This contribution, known as supported 210Pb, isusually very small compared to that due to deposition fromthe atmosphere; nevertheless, it must be subtracted from thetotal observed 210Pb activity. This excess 210Pb, known asunsupported 210Pb {[210Pb]un]}, over and above that inequilibrium with the in situ 226Ra present in the soilparticles, decays with a half-life of 22.3 years in anundisturbed sediment. This excess 210Pb is also used todetermine the time or the age of the deposit and thesedimentation rate.[17] An independent verification of the 210Pb technique is

often provided by 137Cs activity, which is produced innuclear fission and was introduced into the atmosphere bynuclear weapons detonations. The nuclear fallout began in1945 and reached its maximum level in 1964–65. Themaximum 137Cs activity in undisturbed sediments occursaround 1964 and can be used as a marker to verify 210Pbage.[18] Freeze dried sediment samples of about 0.4 to 3.0 g

from each section were accurately weighed and counted forgamma activity using low-background, well-type high-purity germanium detector systems. The efficiencies ofthe spectrometers were determined using NIST-traceable210Pb, and 137Cs standards. The counting times variedfrom 1000 to 4000 min. The age of each section of thecore was calculated using the equation:

210Pb� �

un

� �t¼ 210Pb

� �un

� �0e�lt ð1Þ

where l is the decay constant of 210Pb (0.03114 a�1), and tis the age.[19] The average sedimentation rates were calculated

using the cumulative depth of the core sections and themeasured age of the core sections.

3. Theory of Atmospheric EC DeterminationFrom EC in Sediment

[20] Dry and wet deposition in term of EC flux into thesediment can be expressed as [Seinfeld and Pandis, 1998]follows.

Table 2. Description and Sedimentation Rates of Four Adirondack Lakes

Lake LocationAltitude,

mSize,ha

CoringDepth, m

SedimentationRate, cm�a�1

West Pine Pond Core#1 44�20.2830N, 74�25.5830W 484 25.5 13.1 0.19West Pine Pond Core#2 44�20.2730N, 74�25.5830W 484 25.5 13.0 0.18West Pine Pond Core#3 44�20.2790N, 74�25.5790W 484 25.5 13.3 0.20West Pine Pond Core#5 44�20.2810N, 74�25.5790W 484 25.5 13.1 0.18Clear Pond CP(II)-3 43�59.720N, 73�49.220W 583 70.4 25.9 0.09Deer Pond Core # 1 44�210N, 74�380W 459 11.3 13.7 0.14Deer Pond Core # 6 44�210N, 74�380W 459 11.3 14.0 0.12Bear Pond Core #2 44�23.9560N, 74�17.2490W 499 21.9 18.3 �0.08

D13102 HUSAIN ET AL.: ATMOSPHERIC EC CONCENTRATIONS, 1835–2005

4 of 10

D13102

[21] For dry EC deposition:

Fd ¼ �Vd EC½ atm ð2Þ

where Fd is the vertical dry deposition flux, Vd is the drydeposition velocity in length per unit of time at a certainreference height, and [EC]atm is the [EC] at that height.[22] For wet deposition:

FW ¼ EC½ precip x; y; 0; tð Þpo ð3Þ

where [EC]precip (x, y, 0, t) is the [EC] in precipitation at alocation x,y and zero height at a given time and po is theprecipitation intensity (mm/h or m/y).[23] The wash out ratio wr relates [EC] in the atmosphere

and in precipitation:

wr ¼ EC½ precip x; y; 0; tð Þ= EC½ atm x; y; 0; tð Þ ð4Þ

[24] This can be substituted in equation (3):

Fw ¼ EC½ atm x; y; 0; tð Þwrpo ð5Þ

[25] Equations (2) and (5) can be combined to define thetotal flux FT:

FT ¼ �Vd EC½ atmþ EC½ atm x; y; 0; tð Þwrpo ¼ �Vd þ wrpo½ EC½ atmð6Þ

[26] Equation (6) assumes homogeneous mixing in thelower troposphere. For a long-term integrated sample [�Vd

+ wr po] can be taken as a constant, k1, then for a givenlocation equation (6) can be written as

FT ¼ k1 EC½ atm ð7Þ

[27] Since EC is insoluble, chemically inert and thusresistant to further biological and chemical degradation[Forbes et al., 2006]. It will be deposited in the lakesediment. A term, k2, should be included to account forthe effect of lake focusing and any contributions from thelake catchments. Hence equation (7) can be written as

FT ¼ EC½ sed¼ k1 þ k2ð Þ EC½ atm¼ K EC½ atm ð8Þ

[28] In addition to lake focusing and catchment contribu-tions, K (m3 g�1) is dependent on the amounts of precip-itation. The slow rate of sedimentation requires thatmeasurements be made on samples integrated over a num-ber of years, say, �5 to 10 years, thereby minimizing anyeffects due to variations in precipitation. The precipitationdata going back to �1890 is available for Lake Placid, onlyabout 10 miles from Whiteface Mountain. [http://www.erh.noaa.gov/er/okx/climate/records/monthannualpcpn.html].The decadal mean for the entire period was 112.5 cm. Themaximum deviation from the mean was observed for 1970–1979(+18.1%). The maximum deficiency occurred in1950–1959 (�10.7%). For the years 1980–1989, and1990–1999, decades that are covered by our filter samples,

the deviations from the mean were +12.8, and –6.7%,respectively. The uncertainties caused by these variationsin the determination of [EC] should be relatively minor,<10%. Kada and Heit [1992] have shown that 210Pb and137Cs can be used to account for the depositional focusingof the anthropogenic trace elements Pb, Zn, As, and Cd.Since the size distribution of EC is similar to that of Pb-bearing aerosols [Harley et al., 2005], the use of theseisotopes should also be equally applicable to EC-bearingaerosols. Several investigators have addressed the questionof catchment contributions [Lewis, 1977; Scott et al., 1985;Dominik et al., 1987]. These investigations suggest that just1 to 2 % of the annual 210Pb fallout on the catchments isremoved to the lake. This is further supported by a recentstudy by Appleby et al. [1999] who suggested that just over1% of fallout per year onto the catchment is delivered to thelake.[29] Aerosol samples collected at Whiteface Mt. provide

an excellent opportunity for us to determine [EC]atm. The[EC]sed can be determined in a lake in the vicinity of thefilter sampling site. The concentration of SO4

2- and traceelements such as As, Se, and Sb in precipitation at Wilsboroand Moss Lakes, �100 km apart, were found to be the samewithin analytical uncertainties [Huang et al., 2001]. There-fore long-term data from the Whiteface Mountain summitand a lake located within �50 km of Whiteface Mountaincan be used in equation (8). The lake should have had nosignificant local sources of EC that are not reflected in thechemical composition of the aerosols collected at thesampling sites. Once [EC]sed is known for slices of alake-bed core, we can use these data to estimate [EC]atmfor the same period. For the period since 1978, [EC]atm willbe determined from our archived Whatman 41 filters. Therewill be roughly four determinations of [EC]sed from �1978to 2005, each integrated over �7 years, so that K can befitted to give the best correlation between [EC]sed and[EC]atm. We can then estimate historical [EC]atm from�1835 to 1980. Once a better understanding of K isachieved it may be possible to recreate historical [EC]atmtrends with this technique elsewhere.

4. Results and Discussion

4.1. Determination of [EC]atm in Monthly FilterComposites

[30] Monthly composites were prepared by cutting 2.5 to10.1 cm2 out of the daily filter sample. If samples weremissing for more than 10 d in a given month, that monthwas not included in the study. There were no samples for1980 and 1982. For 1978, samples were available for6 months, and for 11 months in 1979. Samples wereavailable for 8 months for 1981 and 9 months in 1994. Inthe remaining 21 years we missed the samples only for11 months. The composites were chemically processed forEC analysis by the method given of Li et al. [2002]. Afterthe cellulose from the Whatman 41 filter had been removed,EC was transferred onto a 4.7-cm diameter quartz filter. A1.0 cm � 1.5 cm aliquot of the quartz filter was analyzed bythe thermal-optical method for the determination of [EC]atm.Three aliquots were analyzed from each quartz filter. Theuncertainty in the monthly data was nominally ±10% orless. For 18 of the 28 years we also prepared one duplicate

D13102 HUSAIN ET AL.: ATMOSPHERIC EC CONCENTRATIONS, 1835–2005

5 of 10

D13102

monthly composite for each year. The duplicates generallyagreed within ±10%. The data are reported for 280 out of324 months, or about 87% coverage for the period. Thequarterly mean [EC]atm for January–March, April–June,July–September, and October–December were calculatedfor the entire study period. However, the differences werenot statistically significant. We conclude that [EC]atmshowed no seasonal differences at the Whiteface Mountainsite. Apparently, the EC aerosols at this site are primarilyderived from long range transport. Any contribution fromwood burning during excessively cold winter in the Adir-ondack must be relatively small.[31] Annual mean [EC]atm were calculated from the

monthly data from 1978 to 2005 (Figure 2). In 1978, themean [EC]atm was 566 ng m�3. However, the data wereavailable only for July through December. During 1979, the[EC]atm decreased to 440 ng m�3. The data were availablefor all months except April. The highest annual mean[EC]atm was found for 1981, �780 ng m�3. Exceptionallyhigh [EC]atm values of 1065, 1100, and 1495 ng m�3 weremeasured for January, February, and December 1981, re-spectively. Such high concentrations were never foundagain except for February 1983 (1575 ng m�3), and January1986 (1040 ng m�3). Beginning with 1983, [EC]atm valueswere usually available for all 12 months, with more thantwo months worth data rarely missed. The annual meanconcentrations for 1983, 1984, 1985, 1986 and 1987 were535, 415, 430, 600, and 300 ng m�3, respectively. Subse-quently, the [EC]atm fluctuated between 150 and 200 ngm�3. The [EC]atm increased to 250, 330, and 310 ng m�3

for 1994, 1995, and 1996, respectively. A very sharp declinewas observed beginning in 1997. From 1997 through 2005,the mean annual [EC]atm varied between 40 and �80 ngm�3, with a mean value of 66 ng m�3. Mean concentrationsfor the 1978–1986 and 1987–1996 periods were 550 and225 ng m�3, respectively. A 59% decrease in mean [EC]atmoccurred between 1978–1986 and 1987–1996 periods,while an 88% decrease was found between the 1978–

1986 and 1997–2005 periods. The decrease in [EC]atm isperhaps due to the reduction in coal combustion in indus-trial, residential, commercial and coke sectors [EnergyInformation Administration (EIA), 2005] and improveddiesel combustion technology in motor vehicles [Novakovet al., 2003]. The observed decreases are very large but, aswe shall see later in this paper, were mimicked in sedimentsfrom an Adirondack lake.[32] Not many long-term EC data are available for the

Northeastern US. However, there are data available from theIMPROVE network for Acadia National Park (ANP) in MEfrom 1990 to 2004 and for Lye Brook, VT from 1993 to2003. The [EC] at ANP showed a 55% decrease from 1990and 2004, while at Lye Brook, the [EC] remained nearlyunchanged from 1993 to 2003. Measurements of aerosolblack carbon (BC) at a Canadian arctic location from 1989to 2002 show a decrease of 60% in [BC] over that period[Sharma et al., 2004]. Husain et al. [2004] showed that[SO4

2�], a major aerosol component in the Northeastern USemitted from fossil fuel sources, decreased by 59% from1979 to 2002. The reduction in SO4

2� concentrations wasattributed to the reduction in SO2 emissions from coal�firedpower plants in the Midwest under acid rain program. Thecoal-fired power plants are not believed to be a major sourceof EC.

4.2. 210Pb Ages and Sedimentation Rates

[33] We analyzed two cores from Deer Pond, and oneeach from Clear Pond and Bear Pond. The sedimentationrates for Dear Pond, Clear Pond, and Bear Pond were 0.13,0.09, and 0.08 cm a�1, respectively (Table 2). The coreswere sliced into 4.5 cm sections, owing to the low sedi-mentation rates each section represented accumulation over�20 to �30 years and were not suitable to meet the goals ofthis study. Six cores were collected from the deepest part ofWest Pine Pond (WP). On the basis of coring experience wesectioned the West Pine Pond cores differently; the top5 sections into 1.27-cm thick slices; and the remaining into2.54-cm thick slices. The lengths of the cores varied from54 to 58 cm. Each core was sliced into 23 or 24 sections.The top 5 sections, contained �90 to 97% water by weight,and weighed 0.37 to 0.75 gdw (gdw, gram dry weight). Themoisture content of the remaining 19 sections varied from80 to 90% and the mass from �1 to 4.5 gdw. The four coresanalyzed yielded 210Pb ages and sedimentation rates thatwere in excellent agreement (Table 2). For brevity, the dataare shown for two cores in Figure 3. Least-squares fits,through the unsupported 210Pb (Figure 3), yielded half livesof about 22.25 years, in excellent agreement with the known23.3 years half-life of 210Pb. Measurement of 137Cs activityin the West Pine Pond sediment core confirmed the result of210Pb dating. The 210Pb activity for this slice yielded a timeof deposition of 1959–1971, consistent with the maximumdeposition of fallout 137Cs around 1964. Each of the firstfive sections of the core represented deposition over 7 years,and the following sections each represented �13 years asthey were twice as thick. This time resolution is quiteappropriate as it allows determination of four K valuesusing equation (8). Hence the sediment samples from theappropriately sectioned West Pine Pond cores were highlysuitable for the primary goal of this study, and wereanalyzed for EC in an effort to test our hypothesis.

Figure 2. Annual atmospheric elemental carbon concen-trations measured at Whiteface Mountain from 1978 to2005. Inset shows the concentrations for the 1997–2005period on an expanded scale.

D13102 HUSAIN ET AL.: ATMOSPHERIC EC CONCENTRATIONS, 1835–2005

6 of 10

D13102

4.3. EC Concentrations in West Pine Pond SedimentCores

[34] An approximately 200 mg aliquot of the driedsediment from each of the upper top 15 slices of the WP-1(core #1) was weighed and chemically processed using theprocedure described in section 2.2.1. The chemicallyseparated EC was transferred to a 4.7-cm diameter quartzfiber filter. The uppermost two samples, representing thedeposition for 1999–2005 and for 1992–1998, and thelowermost samples (�1835–1848) contained the lowestamounts of EC. Five replicate measurements of [EC]sedvalues were made for each sample. In order to increase theEC signal, amount of sediment was almost doubled whenwe analyzed the WP-2, WP-3, and WP-5 cores. We alsoanalyzed �80% of the quartz filter, the maximum our set upwould permit. In all [EC]sed values were determined in fourof the six cores collected for West Pine Pond. The [EC]sedvalues determined in the individual slices from four WestPine Pond cores, along with the mean [EC]sed for the entiredata set, are shown in Figure 4. The [EC]sed values variedfrom �0.6 to �8 mg gdw

�1. The [EC]sed values in the fourcores were generally in good agreement, except that thedata for WP-1 show large deviations relative to others at adepth of �16 and 18.5 cm. The [EC]sed values at a depthof �27-cm were very low and were comparable to thevalues for the most recent deposition.

[35] The mean [EC]sed in the topmost slices representingdeposition over 1999–2005, and 1992–1998 from the fourcores were 0.60 ± 0.15 (standard error of the mean =standard deviation/

pn), and 0.60 ± 0.25 mg gdw

�1, respective-ly. The third slice representing deposition during 1985–1991 showed a sharp increase in [EC]sed to 2.97 ± 0.50mg gdw

�1. The [EC]sed more than doubled to 6.40 ± 1.10 mggdw�1 in the fourth slice covering the period 1984–1978.Subsequently, the [EC]sed values varied by 15 to �20%until �1902, then it sharply decreased to 4.4 ± 0.9 mggdw�1. The [EC]sed continued to decrease reaching the valueof �0.3 mg gdw

�1 in the slices dating 1835–1848. The[EC]sed values determined in the West Pine Pond cores arecomparable to those observed in elsewhere. For example,Kralovec et al. [2002] found [EC] values in a Lake Eriesediment core to range from 2.5 to 7.4 mg of C g�1 ofsediment, from 1850–1998. Karls and Christensen [1998]reported [EC] varying from 0 and 10 mg gdw

�1 in sedimentcores from Lake Michigan. BC concentrations in NewEngland Harbors and suburban lake sediments rangedbetween 2.0 and 7.0 mg gdw

�1 [Gustafsson and Gschwend,1998]. Muri et al. [2002] determined that the [BC] inSlovenian alpine lake sediments ranged from 1 to 11 mggdw�1. However, much higher [EC] values of up to 30 mgggw�1 have been measured in an urban lake in Central Parkin New York City, due apparently to very high vehicleemissions [Louchouarn et al., 2007].[36] To properly account for differences in moisture con-

tents and density, we have converted the [EC]sed into ECflux, and have plotted it as a function of time in Figure 5. Thedata showed an increase in EC deposition from 1863 to�1930. EC fluxes ranged from 0.08 to 0.23 g m�2 a�1, whilemodern EC flux ranged from 0.53 to 0.77, and finallydecreased to 0.026 g m�2 a�1 during 1998 to 2005. Thistrend is in agreement with that observed for several Adir-ondack lakes [Lorey and Driscoll, 1999].[37] The [EC] values determined in the filters at White-

face Mountain and in the lake sediments for the 1999–2005, 1992–1998, 1985–1991, and 1978–1984 are shownin Figure 6. The aerosol data and the sediment data show

Figure 3. Lead isotopic analysis of two cores from WestPine Pond for age determination: (a) activity of 210Pb as afunction of depth in samples from core WP-5, and (b)activities of 210Pb, and 137Cs in WP-1. The peak in137Csactivity occurred in the fifth sample. The 210Pb age of thissample is 1959–1971, which is in good agreement with theactivity of 137Cs peak at �1964.

Figure 4. [EC]sed in the four analyzed West Pine Pondcores, as a function of depth.

D13102 HUSAIN ET AL.: ATMOSPHERIC EC CONCENTRATIONS, 1835–2005

7 of 10

D13102

the same trend. The ratio [EC]sed/[EC]atm for the 1999–2005, 1985–1991, and 1978–1984 periods is essentiallyconstant. However, the data for 1992–1998 show consid-erable deviation. From the trend it appears that [EC]atm maybe slightly elevated. This could be a result of local forestfires in the Adirondacks during the period. Despite this theratio for this period is within the uncertainties of the ratiodetermined for the other three periods. All data points takentogether are highly correlated, R2 = 0.93, and support thehypothesis presented in section 3. Using the [EC]atm and[EC]sed, we calculated the constant K from equation (8) tobe 10,400 ± 4000 m3 g�1. The large uncertainty of �40% in

the value for is due primarily to the uncertainty in the[EC]sed measurements in the most recent sediment sample.We are not aware of any measurements of [EC]atm anywhere over the period 1835 to 1970. Therefore data evenwith 40% uncertainty represent an important advance. Thevalue of K was then used to reconstruct the atmospheric[EC]atm values going back to �1835 (Figure 7). The[EC]atm varied from 28 ng m �3 for 1835–1848, to760 ng m�3 for the 1917–1930 period. The directlymeasured [EC]atm at Whiteface Mountain for the 1999–2005, 1992–1998, 1985–1991, and 1978–1984 periodswere 62, 196, 297, and 560 ng m�3. The [EC]atm for1835–1848 was 28 ng m�3, and that for 1849–1863 was37 ng m�3, respectively. These concentrations are essentiallythe same and could be attributed to the natural backgroundfor the region. From 1863 on, the concentrations showedlarge increases. During the 1890–1902 period, the [EC]atm(425 ng m�3) more than doubled from the preceding 1876–1889 period (190 ng m�3). The [EC]atm continued toincrease, reaching a maximum of 760 ng m�3 in 1917–1930.The [EC]atm decreased to 680 ng m�3 in 1931–1943and continued to decrease very slowly until 1978–1984.Sharp decreases occurred subsequently, as indicated inFigure 7.[38] There are limited model estimates of BC emissions

for the US, although not specifically for the NortheasternU.S. Novakov et al. [2003] estimated the atmosphericburden of BC based only on fossil fuel combustion in USfrom 1875 to 1999. Five-year averages of their BC esti-mates are presented in Figure 7. These authors suggestedthat the emissions from fossil-fuel consumption in UScontinuously increased and reached a peak in 1920–1930,then decreased during the great depression in US. Subse-quently, more petroleum was used instead of coal, and BCemissions increased slightly after 1980. Nearly a four-folddecrease occurred in BC burden in the 1980s from the peakvalues in the 1920s. The calculations of Novakov et al.

Figure 5. Mean EC flux for the West Pine Pond, N.Y.core.

Figure 6. Relationship between [EC] determined inaerosols and sediment. Except for the data point for1992–1998, the ratio of [EC]atm/[EC]sed is essentiallyconstant.

Figure 7. Atmospheric [EC] determined in this work (seetext) for the Adirondack region from 1835 to 2005. Themeasurements are compared with US BC emissions[Novakov et al., 2003].

D13102 HUSAIN ET AL.: ATMOSPHERIC EC CONCENTRATIONS, 1835–2005

8 of 10

D13102

[2003] do not include emissions from forest fires. Thesediment data showed a 46% increase in [EC]atm in1903–1916, compared to 1890–1902. At least a part ofthis increase could be attributed to forest fires. From 1900 to1910 there were huge forest fires in the Adirondacks; thelargest occurred in1903. During the fire of 1903, 292,000acres of timber and 172,000 acres of brush land burned[Middleton, 1904]. This is perhaps reflected in the enhancedvalue of EC for the lake in 1890–1902. The maximum in[EC]sed values coincided with the peak in energy consump-tion from coal in the U.S. Energy consumption from coal inthe industrial, commercial and residential sectors decreasedsubstantially from 1950 to present and it is reflected both inthe [EC] in atmosphere as well as in sediment.

5. Summary

[39] The [EC]atm were determined in monthly compositesof daily filters collected at Whiteface Mountain (1.5 kmabove mean sea level), New York from 1978 to 2005. Themean [EC]atm for 1978–1986, 1987–1996, and 1997–2005periods were 550, 225, and 62 ng m�3, respectively. Theconcentrations have, thus, decreased nearly 10-fold over aperiod of �28 years.[40] Approximately 55-cm long sediment cores were

collected from the bottom (�13 m) of West Pine Pond,about 40 km southwest of Whiteface Mountain. Each corewas sliced in 15 sections representing deposition from�1835 to 2005. The EC was chemically isolated and itsconcentration determined in each slice using the thermal-optical method.[41] The [EC]sed for the 1978 to 2005 period mimicked

the [EC]atm directly determined in the filters. The [EC]atmvalues for the past 28 years were used to calibrate thedeposition of EC in the West Pine Pond using equation[EC]sed = K [EC]atm. A value of 10,400 ± 4000 (m3 g�1)was determined for K from the EC data in filters and lakecores for 2005–1978 period. From K and [EC]sed, the[EC]atm were determined from 1835 to 1978. The [EC]atmfrom 1835–1862 was �30 ng m�3, and may be close to thepreindustrial background for the region. Subsequently, theconcentrations increased sharply, reaching a maximumvalue, 760 ng m�3, for the 1917–1930 period. From 1931through 1970, the [EC]atm decreased gradually from 680 to610 ng m�3, and then to the current level of �60 ng m�3.[42] Model calculations by Novakov et al. [2003] for BC

emissions from fossil fuel combustion in the U.S. qualita-tively reproduced the trend obtained experimentally in thiswork.

[43] Acknowledgments. The authors are grateful to Karen Roy, JedDukett, Susan Capone, and Paul Casson for advice on lake selection andsediment coring, and V. A. Dutkiewicz for many discussions on all aspectsof the work. This work was partly funded by an NSF grant ATM 0503850.

ReferencesAppleby, P. G. (2001), Chronostratiographic techniques in recent sedi-ments, in Tracking Environmental Change Using Lake Sediments,vol. 11, edited by W. M. Last and J. P. Smol, pp. 171–203, KluwerAcad. Publishers, Netherlands.

Appleby, P. G., E. Y. Haworth, G. Barci, H. Michel, D. B. Short, G. Laptev,J. Merino, B. M. Simon, and A. J. Lawler (1999), Internal project reportto the European Commission: ‘‘The large scale and long term environ-

mental behaviour of transuranic element as modelled through Europeansurface water system’’, final report, contract F14PCT960046, DGXII,81 pp., European Commission.

Birch, M. E., and R. A. Cary (1996), Elemental carbon based for monitor-ing occupational exposures to particulate diesel exhaust, Aerosol Sci.Technol., 25, 221–241.

Burtscher, H., et al. (2001), Separation of volatile and non-volatile aerosolfractions by thermodesorption: Instrumental development and applica-tions, Aerosol Sci., 32, 427–442.

Chow, J. C., J. G. Watson, L. C. Pritchett, W. R. Pierson, C. A. Frazier, andR. G. Purcell (1993), The DRI thermal optical reflectance carbon analysissystem-description, evaluation and applications in United States air qual-ity studies, Atmos. Environ., 27, 1185–1201.

Crutzen, P. J., and M. O. Andreae (1990), Biomass burning in the tropics -Impact on atmospheric chemistry and biogeochemical cycles, Science,250, 1669–1678.

Dockery, D. W., and C. A. Pope III (1994), Acute respiratory effects ofparticulate air pollution, Ann. Rev. Public Health, 15, 107–132.

Dockery, D. W., C. A. Pope III, X. Xu, J. D. Spenglar, J. H. Ware,E. Martha, M. E. Fay, B. G. Ferris Jr., and F. E. Speitzer (1993), Anassociation between air pollution and mortality in six U.S. cities, NewEngl. J. Medicine, 329, 1753–1759.

Dominik, J., D. Burns, and J. P. Wernet (1987), Transport of environmentalradionuclides in an alpine watershed, Earth Planet. Sci. Lett., 84, 165–180.

Energy Information Administration (EIA) (2005), Annual Energy Review,DOE/EIA-0384 (2005), July 2006.

Forbes, M. S., R. J. Raison, and J. O. Skjemstd (2006), Formation, trans-formation and transport of black carbon (charcoal) in terrestrial and aqua-tic ecosystem, Sci. Total Environ., 370, 190–206.

Gelinas, Y., K. M. Prentice, J. A. Baldock, and J. I. Hedges (2001), Animproved thermal oxidation method for the quantification of soot/graphi-tic black carbon in sediment and soils, Environ. Sci. Technol., 35, 3519–3525.

Gustafsson, O., and P. M. Gschwend (1998), The flux of black carbon tosurface sediments on the New England continental shelf, Geochim. Cos-mochim. Acta, 62, 465–472.

Gwynn, C. R., R. T. Burnett, and G. D. Thurston (2000), A time seriesanalysis of acidic particulate matter and daily mortality and morbidity inthe Buffalo, New York, region, Environ. Health Perspect., 108, 125–133.

Hansen, A. D. A., B. A. Bohaine, and B. G. Dutton (1988), Aerosol blackcarbon measurements at the South Pole: Initial results, Geophys. Res.Lett., 15, 1193–1196.

Harley, N. H., P. Chittaporn, M. S. A. Heikkinen, R. Medora, and R. Merrill(2005), Airborne particle size distribution measurements at USDOEFernald, J. Am. Chem. Soc. Sym. Ser., 904, 342–350.

Huang, X., S. S. Keskin, and I. Olmez (2001), Wet deposition flux of traceelements to the Adirondack region, J. Radioanal. Nucl. Chem., 247,317–323.

Husain, L., and V. A. Dutkiewicz (1990), A long-term (1975–1988) studyof atmospheric SO4

2: Regional contributions and concentration trends,Atmos. Environ., 24A, 1175–1187.

Husain, L., V. A. Dutkiewicz, and M. Das (1998), Evidence for decrease inatmospheric sulfur burden in the eastern United States caused by reduc-tion in SO2 emissions, Geophys. Res. Lett., 25, 967–970.

Husain, L., P. P. Parekh, V. A. Dutkiewicz, A. R. Khan, C. Yang, andK. Swami (2004), Long-term trends in atmospheric concentrations ofsulfate, total sulfur, and trace elements in the Northeastern United States,J. Geophys. Res., 109, D18305, doi:10.1029/2004JD004877.

Husain, L., V. A. Dutkiewicz, A. J. Khan, and B. M. Ghauri (2007),Characterization of carbonaceous aerosols in urban air, Atmos. Environ.,41, 6872–6883.

IPCC (1995), Climate Change 1995: A Report of the IntergovernmentalPanel on Climate Change, edited by J. T. Houghton et al., CambridgeUniv. Press, New York.

Jacobson, M. Z. (2002), Control of fossil-fuel particulate black carbonand organic matter, possibly the most effective method of slowingglobal warming, J. Geophys. Res., 107(D19), 4410, doi:10.1029/2001JD001376.

Jenk, T. M., S. Szidat, M. Schwikowski, H. W. Gaggeler, S. Brutsch,L. Wacker, H.-A. Synal, and M. Saurer (2006), Radiocarbon analysis inan alpine ice core: Record of anthropogenic and biogenic contributionsto carbonaceous aerosols in the past (1650–1940), Atmos. Chem. Phys.Disc., 6, 5905–5931.

Kada, J., and M. Heit (1992), The inventories of anthropogenic Pb, Zn, As,Cd, and the radionuclides 137Cs and excess 210Pb in lake sediments of theAdirondack region, USA, Hydrobiologia, 246, 231–241.

D13102 HUSAIN ET AL.: ATMOSPHERIC EC CONCENTRATIONS, 1835–2005

9 of 10

D13102

Karls, J. F., and E. R. Christensen (1998), Carbon particles in dated sedi-ments from lake Michigan, Green Bay, and Tributaries, Environ. Sci.Technol., 32, 225–231.

Khan, A. J., J. Li, and L. Husain (2006), Atmospheric transport of elemen-tal carbon, J. Geophys. Res., 111, D04303, doi:10.1029/2005JD006505.

Kralovec, A. C., E. R. Christensen, and R. P. Van Camp (2002), Fossil andwood combustion as recorded by carbon particles in lake Erie sediments1850–1998, Environ. Sci. Technol., 36, 1405–1413.

Kuhlbusch, T. A. J. (1998), Black carbon and the carbon cycle, Science,280, 1903–1904.

Kuhlbusch, T. A. J., and P. J. Crutzen (1995), Toward a global estimate ofblack carbon in residues of vegetation fires representing a sink of atmo-spheric CO2 and a source of O2, Global Biogeochem. Cycles, 9, 491–501.

Lavanchy, V. H. M., H. W. Gaggeler, U. Schotterer, M. Schwikowski, andU. Baltensperger (1999), Historical record of carbonaceous particle con-centrations from a European high-alpine glacier (Colle Gnifetti, Switzer-land), J. Geophys. Res., 104, 21,227–21,236.

Lewis, D. M. (1977), The use of 210Pb as a heavy metal tracer in Susque-hanna River system, Geochim. Cosmochim. Acta, 41, 1557–1564.

Li, J., A. J. Khan, and L. Husain (2002), A technique for determination ofblack carbon in cellulose filters, Atmos. Environ., 36, 4699–4704.

Lim, B., and H. Cachier (1996), Determination of black carbon by chemicaloxidation and thermal treatment in recent marine and lake sediments andcretaceous-tertiary clays, Chem. Geol., 131, 143–154.

Lorey, P., and C. T. Driscoll (1999), Historical trends of mercury depositionin Adirondack Lakes, Environ. Sci. Technol., 33, 718–722.

Louchouarn, P., S. N. Chillrud, S. Houel, B. Yan, D. Chaky, C. Rumpel,C. Largeau, G. Bardoux, D. Walsh, and R. E. Bopp (2007), Elementaland molecular evidence of soot- and char-derived black carbon inputs toNew York City’s atmosphere during the 20th century, Environ. Sci.Technol., 41, 82–87.

Malm, W. C., B. A. Schichtel, M. L. Pitchford, L. L. Ashbaugh, and R. A.Eldred (2004), Spatial and monthly trends in speciated fine particle con-centration in United States, J. Geophys. Res., 109, D03306, doi:10.1029/2003JD003739.

Menon, S., J. Hansen, L. Nazarenko, and Y. Luo (2002), Climate effects ofblack carbon aerosols in China and India, Science, 297, 2250–2253.

Middleton, D. C. (1904), Ninth Annual Report of the Forest, Fish, andGame Composition, 456 pp., J.B. Lyon, Albany, New York.

Muri, G., B. Cermelj, J. Faganeli, and A. Brancelj (2002), Black carbon inSlovenian alpine lacustrine sediments, Chemosphere, 46, 1225–1234.

Muri, G., S. G. Wakeham, and N. L. Rose (2006), Records of atmosphericdelivery of pyrolysis-derived pollutants in recent mountain lake sedi-ments of Julian Alps (NW Slovenia), Environ. Pollut., 139, 461–468.

NIOSH (1996), Elemental carbon (diesel particulate) method 5040, inNIOSH Manual of Analytical Methods, 4th ed. (1st Suppl.), NationalInstitute for Occupational Safety and Health, Cincinnati, Ohio.

Novakov, T., and J. E. Hansen (2004), Black carbon emissions in the UnitedKingdom during the past four decades: an empirical analysis, Atmos.Environ., 38, 4155–4163.

Novakov, T., V. Ramanathan, J. E. Hansen, T. W. Kirchstetter, M. Sato,J. E. Sinton, and J. A. Satheye (2003), Large historical changes of fossil-fuel black carbon aerosols, Geophys. Res. Lett., 30(6), 1324, doi:10.1029/2002GL016345.

Ruellan, S., and H. Cachier (2001), Characterization of fresh particulatevehicular exhausts near a Paris high flow road, Atmos. Environ., 35,453–488.

Sato, M., J. Hansen, D. Koch, A. Lacis, A. Ruedy, O. Dubovik, B. Holben,M. Chin, and T. Novakov (2003), Global atmospheric black carbon in-ferred from AERONET, Proc. Natl. Acad. Sci., 100, 6319–6324.

Scott, M. R., R. J. Rotter, and P. F. Salter (1985), Transport of fallout Pu tothe ocean by the Mississippi river, Earth Planet. Sci. Lett., 75, 321–326.

Seinfeld, J. H., and S. N. Pandis (1998), Atmospheric Chemistry and Phy-sics, John Wiley , Hoboken, N. J.

Sharma, S., D. Lavoue, H. Cachier, L. A. Barrie, and S. L. Gong (2004),Long term trends of the black carbon concentrations in the CanadianArtic, J. Geophys. Res., 109, D15203, doi:10.1029/2003JD004331.

Song, J., P. Peng, and W. Huang (2002), Black carbon and kerogens in soilsand sediments. Part I: Quantification and characterization, Environ. Sci.Technol, 36, 3960–3967.

�����������������������T. Ahmed and J. Li, Department of Environmental Health Sciences,

School of Public Health, State University of New York, Empire State Plaza,Albany, NY 12201-0509, USA.A. Bari, L. Husain, A. J. Khan, K. Swami, and J. S. Webber, Wadsworth

Center, New York State Department of Health, Empire State Plaza, RoomD224, Albany, NY 12201-0509, USA. ([email protected])

D13102 HUSAIN ET AL.: ATMOSPHERIC EC CONCENTRATIONS, 1835–2005

10 of 10

D13102