This article was downloaded by: [North Carolina State University]On: 18 November 2014, At: 10:59Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: MortimerHouse, 37-41 Mortimer Street, London W1T 3JH, UK

Air & WastePublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/uawm19

Chemical Dynamics of Clouds at Mt. Mitchell, NorthCarolinaViney P. Aneja a & Dueg-Soo Kim aa North Carolina State University , Raleigh , North Carolina , USAPublished online: 06 Mar 2012.

To cite this article: Viney P. Aneja & Dueg-Soo Kim (1993) Chemical Dynamics of Clouds at Mt. Mitchell, North Carolina,Air & Waste, 43:8, 1074-1083, DOI: 10.1080/1073161X.1993.10467185

To link to this article: http://dx.doi.org/10.1080/1073161X.1993.10467185

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purposeof the Content. Any opinions and views expressed in this publication are the opinions and views of theauthors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should notbe relied upon and should be independently verified with primary sources of information. Taylor and Francisshall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, andother liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relationto or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

FEATURE

Chemical Dynamics of Clouds atML Mitchell, North Carolina

Viney P. Aneja and Dueg-Soo KimNorth Carolina State University

Raleigh, North Carolina

The role of clouds as the primary pathway for deposition of airpollutants into ecosystems has recently acquired much attention.Moreover, the acidity of clouds is highly variable over short periodsof time. Cloud water collections were made at Mt. Mitchell StatePark, North Carolina, using a real-time cloud and rain acidity/conductivity (CRAC) analyzer during May to September 1987,1988and 1989 in an effort to explore extremes of chemical exposure. Onthe average, the mountain peak was exposed to cloud episodesabout 70 percent of experimental days. The lowest pH of cloudwater in nearly real-time (~10 min.) samples was 2.4, while that inhourly integrated samples was 2.6. The cloud pH during shortcloud events (mean pH 3.1), whjch results from the orographiclifting mechanism, was lower than that during long cloud events(mean pH 3.5), which are associated with mesoscale or synopticatmospheric disturbances. On the average, the pH values in non-precipitating cloud events were about 0.4 pH unit lower than thosein precipitating cloud events. Sulfate, nitrate, ammonium andhydrogen ions were found to be the major constituents of cloudwater, and these accounted for -90 percent of the ionic concentra-tion. Total ionic concentrations were found to be much higher innon-precipitating clouds (670-3,010 [ieq/L) than those in precipi-tating clouds (220-370 (jeq/L). At low acidity, ionic balance issometimes not obtained. It is suggested that organic acids mayprovide this balance.

The profile of cloud water ionic concentration versus time wasfrequently observed to show decrease at the beginning and risingtoward the end during short cloud events. Before the dissipation ofclouds, a decrease in cloud water pH and an increase in ionicconcentration were found. At the same time, temperature and solarradiation increased, and relative humidity and microphysical pa-rameters (liquid water content, average droplet size, and dropletconcentration) decreased. These observations suggest that evapo-rative dissipation of cloud droplets leads to acidification of cloudwater. Mean pH of cloud water was 3.4 when the prevailing windwas from the northwest direction, and it was 3.9 when the wind wasfrom the west direction. The effects of variations in cloud liquidwater content have been separated from variations in pre-cloudpollutant concentrations to determine the relationship betweensource intensity and cloud water concentrations.

In 1986, the Mountain Cloud Chemistry Program (MCCP),funded by the U.S. Environmental Protection Agency, began aroutine program of one hour integrated cloud water collection,chemical analysis, gas monitoring and meteorological data sam-pling at Mt. Gibbs [~2,006 m mean sea level (MSL)] located inMt. Mitchell State Park (35°44' N, 82° 17' W). This program wasdesigned to evaluate the hypothesis that acidic and other airbornechemicals contribute to the observed decline in spruce-fir forestsin the eastern United States. A real-time cloud and rain acidity/conductivity (CRAC) analyzer was developed in 1987 as part of

SCALE (km)

60 m ContoursI — 12 m Contour*

KEY MAP

Figure 1 . Location and topographical map of Mt. Mitchell Research Site (Mt. Gibbs, NC).

a research program aimed at determining the chemical dynamicsof montane clouds, to complement cloud chemistry informationfrom MCCP at the same site.

It is now recognized that deposition of cloud water maysignificantly contribute to the water, nutrient and health status ofhigh-elevation forests where immersion in clouds occurs fre-quently.1 This recognition arises largely from the recent studieswhich address the frequency of cloud interception, cloud waterdeposition and chemical composition of the intercepted cloudwater at a number of high-elevation locations in the U.S. and otherplaces.212 Ithas now been shown13 that clouds provide theprimarypathway (~60 percent) for pollutant deposition on high elevationforest ecosystems when compared to rainfall (~20 percent), anddry deposition (~20 percent).

The majority of the rain monitors in operation today aredesigned to collect rain either on an event basis or weekly basis.There have been fewer attempts to design remote monitors thatcan collect rain events on a near real-time basis, chemicallyanalyze the samples in real-time and also preserve the samples formore thorough laboratory analysis to be conducted later.14 In thesame way, most of the studies done so far on chemical character-ization of clouds were also based on integrated water sampling.Only recently9-11 have dynamic pH characterization of cloudsbeen performed. However, here we describe and present results ofan automated CRAC real-time sampling system, which storesnearly real-time samples for more thorough chemical analysis

Copyright 1993 - Air & Waste Management Association

1074 • August 1993 • Vol. 43 • AIR & WASTE

Dow

nloa

ded

by [

Nor

th C

arol

ina

Stat

e U

nive

rsity

] at

10:

59 1

8 N

ovem

ber

2014

so2

NO, mr uoxCloud detector

* Throughfalt

later in the laboratory. This instrumentwas field tested during 1987, 1988 and1989 in a stand-alone configuration at aremote location at Mt. Gibbs in MtMitchell State Park, NC.

The analysis of the real-time cloudsamples collected by using the CRACsampler has provided more fundamentalunderstanding of cloud chemistry duringan entire event than the currently availableintegrated water chemical characteriza-tion. In particular, real-time chemical char-acterization would be useful both in thestudy of acidic deposition, and in the studyof the effects of deposition on biologicalsystems.6

In this study, cloud water was collectedby a CRAC real-time sampler as part ofthe Dynamic Chemical Characterizationof Montane Clouds Project. The objec-tives of this paper are to: (1) characterizethe extremes in chemical exposure to theseecosystems due to cloud water by the useof the CRAC real-time analyzer; (2) char-acterize the extremes of physical expo-sure; (3) examine the temporal variabili-ties of cloud water acidity for short (oro-graphic effect) and long (frontal effect)cloud events; (4) investigate the depen-dency of cloud water acidity on meteoro-logical parameters; and (5) examine theeffects of microphysical processes on cloudwater acidity. These data will provide uswith insight on the extremes of chemicalexposure and their significance in contrib-uting to the water, nutrient and healthstatus of mountain forests.

Experimental Site and InstrumentationThe experimental site is located in Mt. Mitchell State Park. Mt.

Mitchell is the highest peak (35°44'03" N, 82°17' 15" W; -2,038m MSL) in the eastern United States, and is located about 600 kmwest of the Atlantic coast. A topographical map of Mt. Mitchellarea is shown in Figure 1. Mt. Mitchell State Park has a forestecosystem composed primarily of red spruce (Picea rubenssarg.), and Fraser fir {Abiesfraseri (Pursh.)Poir.}.

According to the survey results of the Mt. Mitchell forest from2,024 m elevations to above 1,575 m:ls (1) red spruce growing atelevation of 1,935 m or higher are in a severe state of decline; (2)defoliation and significant annual increment suppression was notobserved for spruce below 1,935 m MSL; (3) the growth suppres-sion cannot be attributed to lack of moisture; (4) red spruce's shortroots at high elevation would be affected at pH <4.0; and (5) nofir or spruce trees are thought to reproduce successfully ataltitudes above 1,935 m MSL.

All measurements to characterize the chemical climatologywere taken on or near a 16.5 m tall aluminum walk-up towerinstalled near Mt. Gibbs (~2,006 m MSL) which is locatedapproximately 2.0 km southwest of Mt. Mitchell. The tower isequipped with the standard meteorological instrumentations (tem-perature, relative humidity, wind speed and direction, solar radia-tion, and ambient pressure), a passive and an active cloud watercollector, cloud liquid water content (LWC),16 cloud detector,17

and gas pollutant sensors ((),, SO2, and NOX). The instrumentsused are described in Table 1. The maintenance and calibrationfor these instruments were performed on a routine basis under the

Table I. Measured parameters and instruments used at the Mt. Mitchell Research Station during theobservational periods (1987 -1989).

Freijueney

ContinuousContinuousContinuousContinuousContinuousContinuous

Thermistor

ContinuousContinuousContinuousContinuous

(FSSP)

HourlyEvent

Hourly

Electrode probeElectrode probe

* Ammonium ion

instructions of MCCP protocol and Quality Assurance/QualityControl plans.18

The meteorological sensors were placed at least 10 m above theforest canopy which mostly consists of 6-7 m tall Fraser fir trees.There are no known anthropogenic emission sources in the vicinityof the tower.19 The investigation of the dynamic chemical character-ization of clouds was performed by coupling a Caltech (CaliforniaInstitute of Technology) active cloud water collector, installed at thetower-top, to an automated precipitation monitoring (APM) system,which is located on the ground near the tower. The combined systemof the APM and the Caltech cloud water collector is defined as theCRAC (cloud and rain acidity/conductivity) automated sampler inthis research. A detailed description of the CRAC automatedsampling system is presented in the next section.

Description of Cloud and Rain Acidity/Conductivity (CRAC)Automated Sampling System

The monitoring system is based on a prototype AutomatedPrecipitation Monitoring (APM) system which was developed byKronmiller et al.u APM was modified by interfacing a Caltechactive cloud collector to automatically collect the cloud samplesdelivered from a cloud water collector.9 The modified deviceincludes a switching mechanism which can direct the sample flowfrom either a rain or a cloud collector. The CRAC analyzer is arobotic monitoring system to collect cloud and/or rain samples on areal-time basis. Figure 2 gives an information flow diagram for theoperation of the CRAC analyzer during a cloud and/or rain event.

AIR & WASTE • Vol. 43 • August 1993 • 1075

Dow

nloa

ded

by [

Nor

th C

arol

ina

Stat

e U

nive

rsity

] at

10:

59 1

8 N

ovem

ber

2014

FEATURE

No

Rain Cover Open

Collect RainWater i y Rain

MonitorRain Cover

Close

Data Analysis ForCollecting Water

Operate CalteehCloud Water

CollectorNot Operate

Beth Of Them

Collect CloudWater By

Calteeh CollectorNo Sampling

Figure 2. CRAC real time sampler information flow diagram.

The CRAC analyzer has three major units: (1) a cloud or raincollector unit, (2) a microprocessor-based electronics unit for thepH and conductivity measurements, and (3) a sample storage unit.The major components of the CRAC analyzer include the cloudcollector, rain sensor, rain cover, microprocessor electronicsmodules, pH meter, conductivity meter, accumulation chambersand printer. The main sensor and the rain cover mechanism areboth mounted on top of the monitor.

In the case of a rain event, once the event is detected by thesensor, the microprocessor sends a signal to a series of pneumaticcylinders that control the opening and closing of the rain cover.The cover remains open until the rain event is over. During a cloudevent, cloud water samples are collected with the Calteeh activecloud water collector.20 The Calteeh cloud water collector con-sists of a sampler duct, e.g., in the form of a square box open atboth ends with a fan at the back of the duct. Droplets in a cloudsample are drawn into the sampler from the front of the duct by thefan. The velocity of flow is about 9 m s1. The droplets areimpacted on a series of vertical Teflon strings (diameter 5 lOum)which are arranged in the form of a screen on a frame. The 50percent collection efficiency cutoff predicted from impactiontheory corresponds to a droplet diameter of about 3.5nm8. Once~50 ml of cloud water is collected in the accumulator, it is divertedto a sample vial which is equipped in a storage unit of CRACmonitoring system, and collected cloud water is then analyzedchemically. The conductivity is measured with a Cole-Parmerconductivity meter in conjunction with a conductivity probe. AnOrion Model 611 pH meter, in which the temperature is automati-cally compensated, and an Orion 91-62 electrode has been usedbecause it responds quickly at 5°C. A special quality assuranceprocedure has been included in the microprocessor programs sothat the electrode response is automatically checked at pro-grammed intervals.

A modified commercial refrigerator is used to store the cloudwater samples. The collected samples are stored in the refrigeratorslightly above freezing temperatures to minimize chemical changes

in the samples. For the convenience of the storage and lateranalysis in a laboratory, an Eldex carousel is placed in the bottomof the refrigerator. The carousel holds 50 polycarbonate samplevials. The carousel is controlled by the microprocessor andadvances after each sample is collected. The samples are keptrefrigerated until collection.

Other components include an ALPS AGS 1100 printer. Datareports produced on the roll-paper printout include a stationidentification code, time of sample collection, sample vial posi-tion in the carousel, final pH, and conductivity readings andquality control solution results.

Quality Control/Quality AssuranceThe CRAC system was serviced on a weekly basis though it

was usually also checked daily. In addition, it was serviced after eachevent. The routine weekly service included once-a-week pH elec-trode calibration checks with pH buffers of 3.0 and 7.0. Thecalibration of the conductivity meter was also evaluated on a weeklybasis with 36 and 186 ̂ mhos/cm potassium chloride solutions. Atthe start of a cloud event, the cloud water collector and Teflon tubingto the automated precipitation monitor are thoroughly cleaned bydeionized water until the conductivity of the rinse water is within 5(imhos/cm of that of the deionized water.

The system has been designed to have an automatic qualitycontrol solution processed at programmable intervals. The pH ofthe quality control solution, ignoring any temperature or activityeffects, was 4.00. The pH measurements of the QC solution areplotted in Figure 3 for the 1987,1988 and 1989 field seasons at Mt.Mitchell. The average and standard deviation values for the pH ofQC solution during 1987,1988 and 1989 field seasons are pH 4.02± 0.03, pH 4.01 ± 0.03, and pH 4.03 ± 0.04, respectively. Theaverage value for all three years is pH 4.02 ± 0.03. The responseto the QC solution is a good indicator of the state of the electrodefor pH field measurement during the field seasons, and it func-tioned satisfactorily.

Cloud Chemistry at Mt. Mitchell as Revealed by the CRACAnalyzer

The field season for cloud events observation for each experi-mental year are as follows:

(1) 1987: July 29 - October 12 (22 events monitored, and 492total samples)

(2) 1988: June3 - September 16(23 events monitored, and 978total samples)

(3) 1989: May 30 - August 1 (20 events monitored, and 402total samples)

1989—*

40 60 80 100 120 140 160 180

Number of Measurements

Figure 3. Cloud pH monitor response to pH 4.00 QC solution at Mt. Mitchell Research Siteduring the field season (1987 -1989).

1076 • August 1993 • Vol. 43 • AIR & WASTE

Dow

nloa

ded

by [

Nor

th C

arol

ina

Stat

e U

nive

rsity

] at

10:

59 1

8 N

ovem

ber

2014

The pH and conductivity of the cloudsamples were measured immediately aftercollection, and aliquots were separated forlater chemical analyses. The major an-ions, Cl\ SO4

2- and NO3', were measuredin the laboratory using a Dionex 201 Oi ionchromatograph. The concentrations ofmajor metal cations, Na+, K+, Ca2+ andMg2+, were determined by atomic absorp-tion spectrophotometry. NH4

+ concentra-tion was determined by a standard colon-metric method. Detailed chemical analy-sis in the laboratory was performed for

4 • S . J J 4 J , « V"?

only selected cloud events each year. Fourhundred and sixteen cloud water sampleswere analyzed for the major cations andanions to explore the chemical dynamicsof the clouds at the site and their associa-tion to meteorology and microphysical

palamcicra.The average and/or extreme values ofmeteorological parameters observed dur-

ing 1987 through 1989 field seasons aresummarized in Table 2 to aid in the inter-pretation of the cloud water chemical data,

ana 10 uescriDc uic climatic conuinon aithe site. Summary of the sampled eventsfor different types of clouds, accumulatedhours, mean pH of cloud water, and aver-age LWC during 1987, 1988 and 1989field seasons are presented in Table 3. Theliquid water content was performed by agravimetric sampler, as described in detailby Valente et al.16 Based on Saxena andLin's12 terminology, cloud events werecategorized into long cloud events (dura-tion >8 hours), and short cloud events

(duration s8 hours). Generally, the formerare caused by a frontal passage or associ-ated with large-scale disturbances, and thelatter are caused by the orographic lifting

Table II. Summary of meteorological data during the field seasons at Mt. Mitchell Research Site(1987-1989).

Year

Average Tempi198719661989MaximumMinimumMean

Average netsa198719881989MeanTotal Pmeinifai198719881989Mean

Mumtmmm Ofind inwvfmffw wtitv <s

19871988 ,1989Max. hourlyMean_ _ _ — . . . „ . . , » _

Awerage WfmfL198719881989Mean

mechanisms. Cloudevents accompanied with precipitation are defined as precipitat-ing cloud events (or mixed events). Otherwise, they are defined asnon-precipitating cloud events.

May

VWttlfS12.810.810.920.$2.0

11.5

WmUR90.271479.230.3

fan/tinnnt f fit*

79.226.932.246.1

ntnmtiipevuft

425.68.2

216.0

mctfo267285247266

June

m13,313.613.322.10.4

13.4

tiiUu iO/. ltatty (%/

84.473.389.682.4

ml145.336.8

130.6104.2

mM6.55.36.7

19.26.3

n (degrees)290293227270

The cloud pH

July

15.114.814.623.66.1

14.8

84.484.294.837.8

51.396.6

173.2107.1

— — —

5.35.55.9

21.6$.6

293271229264

data shows seasonal

August

14.615.213.623.5-0.614.5

8888.190.188.7

47.2105.2110.887.7

————-

6.55.65.6

205.9

35024$234276

September

10.31212.719.92.2

11.7

90.888.595.891.7

306.6145,3

4.4152.1

6.66.50.0

18.96.2

83266193181

variation in cloud wateracidity. The pH value in the summer season (pH 3.15~pH3.82)was significantly lower than those in spring and fall seasons (pH3.48 ~pH 4.02). This reflects the influence of the seasonal shift

Table III. Summary of CRAC sampled events, time-averaged pH, pH ranges, and average LWC values for different event types (1987 -1989).

Year Type Events Samples A<

198? Short 14 226(July- Long 3 222Oct.) Mixed 5 47

Total 22 495

1988 Short 9 149(June- Long 2 287Sept) Mixed 12 572

Total 23 978

1989 Short 8 77(May- Long 5 211Aug.) Mixed 7 125

Total 20 413

$JI1H* hf t . A w . fci&»

67.99 §.2336.79 12.288.09 1.62

41.22 4.5836.70 18.36

102.35 8.53

31.14 $MS4.67 10.9330,62 4.36

Ave.pH

i3.23(0.17)3.40(0.02)4.05(0.02)3.34(0.07)

3.03(0.29)3.55(0.21)357(0.38)3.44(0.29)

1,09(0.18)3.59(0.20)4.03(0.34)3.57(0.24)

pHrangwCRAC 1

Continuous) (1hrl

2.07-4.82 3.0

2.44 • §,64 2.S

2.79-4.91 2.9-

LSRC

0*4,55

3*5.27

4-4.50

LWC(g/m3)

0.15(0.11)0.54(0,20)0.71(0.24)

0J4(0,17)0.19(0.24)0.30(0.21)

0.20(0.13)0.22(0.14)0.32(0.15)

AIR & WASTE • Vol. 43 • August 1993 • 1077

Dow

nloa

ded

by [

Nor

th C

arol

ina

Stat

e U

nive

rsity

] at

10:

59 1

8 N

ovem

ber

2014

FEATURE

Table IV. Non-precipitating and precipitating cloud events summary sampled during observationalperiods of 1987 through 1989.

Types ofEvents

Non-precipitating{Short and Long Events)

Precipitating(Mixed Events)

Number ofEvents

41

24

Total PHRanges

Time-averagedmeans pH(*SD)

268.51 2.44-4.88 3.32(+0.29)

140.96 2.71-5.64 3.69(±0.45)

(pHnonJ and precipitating cloud events(pH J ) . A one-sided hypothesis was usedin this statistic test because it was sus-pected that pHprec is greater than pHnonp inthe sample mean pH. The null hypothesis,Ho, stated as follows:

The alternate hypothesis, HA, is:

accompanied by significant change in large scale weather systemse.g., changes in typical pressure systems and predominant winddirection. In addition, the variability of pH and chemical compo-sitions in cloud water depends on the origin of air masses. Thesedynamic effects on variability of pH and chemical composition incloud water are discussed later.

Table 3 also shows that mean pH values for short cloud events(pH 3.23 ± 0.17 for 1987; pH 3.03 ± 0.29 for 1988; andpH 3.09± 0.20 for 1989) were much lower than those values for long cloudevents (pH 3.40 ± 0.02 for 1987; pH 3.55 ± 0.21 for 1988; andpH3.59 ± 0.20 for 1989). The pH values for mixed cloud (cloud andrain together) events (pH 4.05 ± 0.02 for 1987; pH 3.57 ± 0.38 for1988; and pH 4.03 ± 0.34 for 1989) were much higher than thosevalues for both short and long cloud events.

During the field seasons, average LWC values ranged from0.15 to 0.24 g nr3 for the short cloud events, 0.22 to 0.54 g nr3 forthe long cloud events', and from 0.32 to 0.71 g nr3 for the mixedcloud events. The results of cloud water acidity data and LWCdata analyses revealed an increase of cloud acidity from mixedcloud events (pH 3.57 to pH ~4.05) to short cloud events (pH 3.03to pH ~3.23) as the LWC increases from short cloud events (0.15-0.24 g nr3) to mixed cloud events (0.32 -0.71 g nr3). Thisincrease in LWC may be explained as follows: in general, theshort cloud events are associated with stratus (St) and stratocumu-lus (Sc) cloud types, in which the cloud droplets are a veryconcentrated solution of soluble cloud condensation nuclei.21

According to these observations, it appears that these short cloudevents can be attributed to the orographic lifting mechanism alongthe slope of Mt. Mitchell site, and thereby bring some near-surfaceor soil-derived aerosols and scavenging of pollutants (aerosol andgases) which have already accumulated in the valley.22 On the otherhand, long cloud events are attributed to the frontal clouds ofnimbostratus (Ns) and altostratus (As) type which have larger clouddroplets as well as abundant moisture.12 The highest pH of cloudwater for the mixed events suggests that the rain drops are consid-erably more dilute man the non-precipitating cloud droplets.

Table 4 illustrates the time-averagedmean pH values and pH range (from mini-mum pH to maximum pH) for non-pre-cipitating clouds (short and long cloudevents), and precipitating clouds (mixedcloud events). The time-averaged meanpH values show that the cloud watersamples collected in non-precipitatingevents (mean pH 3.32) are more acidicthan those collected in precipitating events(mean pH 3.69).

A hypothesis-test (t - statistic test) hasbeen used to evaluate the statistical sig-nificance of differences in sample meanpHs for non-precipitating cloud events

HA:pHprec-pHnonp>0

where pHprec is a time-averaged mean pHvalue for precipitating cloud events, and

pHnon is a time-averaged mean pH for non-precipitating cloudevents. A 5 percent level of significance has been used for thehypothesis test. According to the results oft-statistic by using thestatistic information listed in Table 4, we reject the null hypothesis,Ho, at the 5 percent level of significance (t - statistic value=4.035,163 05 = 1.670). It indicates that there is enough evidence in thesesamples to suggest thatthecloud water acidities of non-precipitatingcloud events were significantly higher than those of precipitatingcloud events. According to confidence interval analysis, a 95percent confidence interval for (pH -pH ) was [0.187,0.553]pH units. Therefore, we are 95 percent confident that our confidenceinterval covers the true value of (pH - pHnonp), and this intervaldoes not contain zero.

Figure 4 shows the cloud water pH variation with time andprecipitation during a long mixed cloud event which occurred onSeptember 16 through 17, 1988. A trend in cloud water pHvariation with time was observed. For the first few hours (about4 hours) of the event there was no precipitation. In this non-precipitating period, the cloud water pH value gradually rose atthe beginning of the event (pH -3.11) for about two hours (pH-3.50) and decreased to its initial pH value (pH -3.10) just beforeprecipitation began. This kind of time trend of cloud pH inabsence of precipitation was generally observed in non-precipi-tating short cloud events sampled at the Mt. Mitchell site. On theother hand, in the presence of precipitation, the cloud pH gradu-ally rose as the amount of precipitation increased, and it reachedthe maximum pH value (pH -5.4) when the maximum amounts ofprecipitation were observed. After the precipitation had ceasedthe cloud pH value decreased gradually as the cloud dissipated.This is in agreement with results reported by Aneja et al.9 Thereduced acidity of precipitating cloud water during the timeprecipitation is present may be explained by dilution of airpollutants (aerosol and gases) with rain water.

Based on 416 cloud water samples collected during the fieldseasons of 1987 through 1989, the duration of the cloud event, themean acidity and ion concentration of cloud water are summa-

4 -

3 -

17:00

5 -?cEm

23:00 05:00Time (EST)

11:00 17:00

Figure 4. Cloud water pH variation with time and amount of precipitation for a mixed cloud event on September 16,1988.

1078*August 1993 -Vol. 4 3 - AIR & WASTE

Dow

nloa

ded

by [

Nor

th C

arol

ina

Stat

e U

nive

rsity

] at

10:

59 1

8 N

ovem

ber

2014

Table V. Statistical summary of cloud chemical analysis for the events sampled during 1987, 1988 and 1989. Concentration units are in jxeq/L.

Sate D

Si30-Jul

6-Aug

17-AUS

19-AltO

12«0Ct

30-Jun

30-Jun

18-Sept

— _

20-May

6-Jun

19-Jul

23-Jul

27-Jul

28-Jul

31-Jul

iinrtiofl(

7,7

4.1

4.7

3.7

4.1

5

3.1

4.8

13.3

13.4

7.8

2.6

3.54

9,27

2.92

hit)

— '

meanSD

% of summean

SD% of sum

meanSD

% of summean

SD% of sum

meanSD

% of sum

meanSD

% of summean

SD% of sum

meanSD

% of sum

mean$0

% of summean

SD% of sum

meanSD

% of summean

SD% of sum

meanSD

% of summean

SD% of sum

meanSD

% of sum

PH

3.070.15

3.140.20

3.240.21

3.170.07

4.020.12

2,940.15

3,020.49

3,250.18

3.440,18

3,640.21

3.670.12

4.310.09

3,170.08

3.280.16

3,370,1

848.0230.9$3.4

726.3343,136.6

577,1231028.4

679,8104.034.596.026.828.2

1137.5354.637.7

970.7695.142,5

564.0236.647.5

364.0150.825.4

228.982.433.6

212.757.831.748.29.8

22.2677.3140.742?

527.1138,437.0

429.093.334.7

NH4+

" ""*353.0133.123.9

218.7104.6110

243.5$7.212.0

198,876.510.150.024.415.8

273.587.09.1

132.808.35.8

90.030.77.6

274.0145.519.165.120.79.6

86.225.812.919.78.19,1

132.838.68.3

169.050.6118

170.952.413.8

_

35.017.71.4

12,3S.20.6

44.017.42.2

24.410.512

22.018.56,0

56,034.01.9

3t,718.814

15.05.913

33.0I IS2.3

18.312.42.7

10.94.2169,32.64.3

15.94.610

22.65.916

12.32.410

Mg++

9.05.20.33.22.80,2

29.413.214

10.24.80,59.0

1192.5

7,75.60.33.32.00,11,40.90.1

7JU0.53.20,70.52,4120.44.1171,92.9100.23,4120,22.70.70.2

Na+

""5.03.00,23,73.50.2

54.92S.52,7

1105.50.6

13.02103.6

§.92,80,29,47.80.46.93.90,6

9.04.40.63,94.10.62.1180,3

17,88.48.23.3180,22.32.40.23.6190,3

K+

4.01,80.114170,14,6150.22.6120,12,0190.5

7,53.10.27,54.60.33,53.50,3

5,53.60.43.02.10.5130.60.23.54,118190.90.1150.80.10.70.60.1

804-."™~

930.0318.336.6

832,7449,042.0

7513315.637.0

673.7186,934.2

108.047.329,5

1194.0177.439.5

893,0467.639,1

357.087.130.1

515.0203.635.9

229,2101433.7

270,478.140,340.57.1

22.8544.943.534,0

49S.668,934.7

464,855.137,0

N03-

305.093.712,0

160,085.08.1

254.895.112,5

340.960.617.347,026,812.8

205.196.29.4

194.7140.7

8.5130.072.3110

213.081414.8

111341.116.474.926.011.244.110.220.3

206.468,012.9

188,345.213,2

136.633.9111

eh

53.024.92.1

26,318,31.3

73.219.03.6

26.21161.3

11.08.33.0

77.030.12.6

40.527.31,8

19.012.51.6

15.05.01.0

17.13152.59,83.61.4

21.56,79,9

19.76.41.2

16.54.81,2

14.82.612

^cations

1253,0

49.3905.7

48.79514

46.9926.8

47.1200.0

54.6

1488.0

49.31156.0

50.6681.0

57.4

692.0

48.2323.0

47.5315.8

47.1102.2

47.0033,9

52.0725,0

50,9619.2

50.1

Cantons* * * * * * " W " H - ' H * * - • — ' •

1288.0

50.71019.0

5131079.3

53.11040.8

62.9160.0

45.4

1531.0

S0.71128.2

49.4506.0

42.6

743.0

510357.0

52.5355.0

52.9115.2

53,0770.9

48.0700,4

49.1010.2

49.9

Ratio, . - 1 1 1 • • ! • . . »

,97

.95

0.88

0.89

120

0.97

102

135

0.93

0.90

0.89

0.89

1,08

104

100

rized (Table 5) with respect to cloud events for 1987, 1988 and1989. From these analysis, the ionic balance between anions andcations for cloud water from each event are presented in Figure 5.In general, for most cloud water samples, a good balance betweenanions and cations was obtained (slope =1.04 and coefficient ofdeterminations, R2 = 0.98). The slight imbalance between mea-sured anions and cations can only mean either or both of thefollowing: (1) some anions or cations were not measured; we didnot measure organic acid (e.g., HCOOH) in this study. Formicacid may be produced because of the reaction of HCHO (aq) withOH (aq) in cloud water23 and (2) there were errors in one or morespecies' measurements.3

From Table 5, it appears that H+, NH4+, SO4

2*, andNOj" ions arethe principal ions of cloud water at Mt. Mitchell as also reportedin recent studies by other investigators.4>12>H25 On average, forcloud events, the fraction of equivalents of all ionic constituentsexpressed as a percent (hereafter referred to as mass contribution)found these four principal ions in cloud water to range from 90percent to 98 percent during the field seasons. For individualchemical components of cloud water samples, mean mass contri-bution ranged from ~23 percent to ~42 percent for SO4

2-; from -22percent to ~47 percent for H+; from ~8 percent to ~20 percent forNO3'; and from ~6 percent to ~ 19 percent for NH.+. According toSaxena and Lin,12 in general, ammonium is attributed mostly to

AIR & WASTE • Vol. 43 • August 1993 • 1079

Dow

nloa

ded

by [

Nor

th C

arol

ina

Stat

e U

nive

rsity

] at

10:

59 1

8 N

ovem

ber

2014

FEATURE

4)

C

cent

raco

n<A

nio

ns

1500-

1000-

5 0 0 -

C

Slope-1.04RA2-0.98

) 500 1000

Cations concentration (f.ieq/1)

1:1

198719881989

1500

y

2000

Figure S. Ion balance for the continuously collected cloud water samples.

bacterial production from agricultural activities and hydrolysis ofurea, and partly to man-made pollutants, and is thought to neutral-ize the cloud water acidity. Sulfate and nitrate are mainly derivedfrom anthropogenic sources, and are mainly thought to acidify thecloud water. In addition, Cl", Na+ and Ca2+ were other minorconstituents. The first two, which in general are derived from theocean, have to balance each other, if no other sources exist.Nevertheless, our results show that more Cl' than Na+ was mea-sured during cloud events. This high con-centration of Cl" may be the result of HC1gas uptake.26 Ca2+ was also found to be lessthan Cl" in our cloud water samples. Pri-marily, Ca2+ is a component of carbonateand silicate soil minerals blown into theatmosphere after cultivation.27 Secondly,it may be derived from dirt roads andconstruction sites.

Figure 6 shows contribution of chemi-cal composition to total fraction of equiva-lents of all ionic constituents expressed asa percent with respect to cloud event cat-egories for 1987,1988 and 1989. Contri-bution diagrams for long events during1987 and 1988 field seasons are missingbecause no chemical analysis were per-formed on the samples during those years.Figure 6 shows that the mass contributionof the four principal ionic species (SO4

2",NO3, H+, and NH4

+) in long and shortevents are above 95 percent for 1989 fieldseasons. However, those chemical speciesin mixed events contribute less to totalconcentration (~74 percent), while the restof chemical species in mixed events aremuch more (five to ten times) than incloud only events. It also shows that themass contribution of SO4

2" (~35 percent)and H+ (~35 percent) in non-precipitatingcloud events (both long and short events)was much greater than in precipitatingcloud events (mixed events) (~23 percentfor SO4

2 and ~22 percent for H+). On theother hand, mass contribution of Cl" andother metal cations (Ca2+, Mg2+, Na+ andK+) was about ten times greater in non-precipitating cloud events (~1. percent for

Cl" and ~2 percent for metal cations), than in precipitating cloudevents (~10 percent for Cl" and ~23 percent for metal cations).

Scatter plots of the principal anions' concentrations versuscations'concentrations were performed for all cloud water samplescollected, during 1987,1988 and 1989 field seasons, which wereanalyzed chemically (total number of samples = 416). A scatterplot of sulfate ion concentration versus hydrogen ion concentra-tion is shown in Figure 7. The plot shows that, based on statisticalanalysis (slope of regression line = 1.03, R2=0.70), the sulfate ionconcentrations were slightly higher than hydrogen ion concentra-tions. The nitrate ion concentrations were slightly lower thanammonium ion concentrations. The sum of sulfate and nitrate ionconcentrations is almost equivalent to the sum of hydrogen andammonium ion concentration (Figure 7). However, Saxena andLin12 observed that sulfate and nitrate ion concentrations were 5to 10 percent more than hydrogen and ammonium ion concentra-tions in cloud water, based on one hour integrated sample chemi-cal analysis during 1986 and 1987 field seasons. Figure 7 showsthat for most cloud water samples, the ammonium ion concentra-tion has a linear relation with sulfate and nitrate ([SO4

2] + [NCy]) ion concentration on the basis of our chemical data results (R2

= 0.70). This could be due to neutralization of acidic species bygaseous ammonia, or it could be due to the dissolution of ammo-nium aerosols in cloud water.

[SO42] and [NCy] are negatively correlated with the pH of

cloud water with a correlation coefficient of 0.84 (R2 = 0.70) forsulfate, and 0.71 (R2 = 0.50) for nitrate. These negative correla-tions between the anions and pH of cloud water were also found

2 2 1 % Short '87 3 0 0 % Mixed '87

33.00%

2.91%

2.31%

12.44%

Short '88

39.02%

15.82%

12.61%

Mixed'88

H+NH4+Ca++,Mg++,Na+,K+SO4-NO3-Cl-

12.14%1'2O% Short '89 14.24% 1 4 0 % L o n g ' 8 9

57%

9.83%

35.41%

39.32%

35.71%

Mixed '89

22.17%30.99%

20.26%

14.44%

1.50%10-43% 3.21%22.77%

15.95%

Figure 6. The major ionic contribution to the total mass for different cloud event categories of 1987,1988 and 1989.

1080 • August 1993 • Vol. 43 • AIR & WASTE

Dow

nloa

ded

by [

Nor

th C

arol

ina

Stat

e U

nive

rsity

] at

10:

59 1

8 N

ovem

ber

2014

R"2-0.70, n=416

in earlier research of one hour integratedcloud water samples at Mt. Mitchell.12

Their correlation coefficients were about -0.70 for both ions. This is about the sameas our results. On the other hand, thecorrelation coefficient between sulfate ionconcentration and pH of cloud water overthe Los Angeles Basin was 0.42.28 This isattributed to the effect of alkaline sea saltin on-shore air during the sampling pe-riod. Mt. Mitchell is about 600 km west ofthe North Carolina coast line, and even theon-shore air flow does not bring as manymarine aerosols into clouds as in the LosAngeles Basin, which is on the coast.

The initial chemical composition oforographic clouds has been shown to bedetermined largely by the composition ofaerosol that serves as condensation nu-clei,22 and to some extent by the ambientconcentration of soluble gases.4 The com-position of aerosols largely depends onthe transport routes and chemical reac-tions in cloud water during the samplingperiod. Figure 8 shows the mean pH ofcloud water with respect to sixteen pre-dominant wind direction sectors for thecloud events collected during 1987,1988and 1989 field seasons. The most acidiccloud water samples (mean pH 3.37) werefound in the northwest (NW) directionsector which has large industrial areas(i.e., midwestern U.S.), and also in the south-southeast sectorwhich has relatively high acidity. On the other hand, the lowercloud water acidity was usually observed in unindustrializedcontinental areas (southwest and west direction: pH 3.85 ~3.90)and oceanic regions (east-northeast direction: pH 3.84).

To investigate the variation of cloud water acidity as a functionof different source regions, sixteen different sectors are defined."Using the 1985 NAPAP Emissions Inventory of SOx, NOx andtotal suspended paniculate (TSP) for eastern and middle UnitedStates provided by the U.S. Environmental Protection Agency29

source regions were categorized into sixteen predominant windsectors (Table 6). The emissions of sulfur, nitrogen and totalsuspended participate in each sector corresponds with the varia-tion of cloud water acidity (at a correlation coefficient of ~0.5).This suggests the relationship betweensource intensity and cloud water acidity.

To separate the effects of variations incloud liquid water content from variationsin pre-cloud pollutant concentrations inorder to further examine the relationshipbetween pollutant source intensity andcloud water concentrations, cloud waterhydrogen ion concentrations were con-verted to equivalent gas phase concentra-tions by using the cloud liquid water con-tent. Figure 9 and Table 6 show the con-centrations in the cloud water of the hy-drogen ion in the aqueous phase, the liquidwater content, and the equivalent gas phasehydrogen ion concentration. The concen-trations of the hydrogen ions in the aque-ous phase correspond well (at a correla-tion coefficient of '-0.5) in the same orderas a pollutant source intensity of the wind

RA2 - 0.71

• \

. n=416 •

. ' • • • • / •

/&•

/

/«.''

/

+

(c)

1987

188S

1989

n*2-0.82, n»416

500 1000 1500 2000 2500 3000[S04--MNO3-] ftieq/l)

500 1000 1500 2000 2500 3000ISO4-MNO3-] (n»q/l)

Figure 7. (a) Sulfate versus hydrogen, (b) nitrate versus hydrogen, (c) ammonium versus sulfate plus nitrate, and (d)hydrogen plusammonium versus sulfate plus nitrate ion concentration for cloud water samples collected during 1987,1988and 1989.

sectors. This suggests the relationship between source intensityand cloud water concentrations. The equivalent gas phase hydro-gen ion concentration and the emissions of SO2, NOX and TSP areplotted for various wind direction sectors (Figure 10). Again, thecalculated hydrogen ion concentration levels in the gas phase(i.e., when the role of LWC is normalized) corresponds well (ata correlation coefficient ~0.7) to the emissions of the variouspollutants from the different wind sectors. This further suggeststhe long range transport of air pollutants to the site, which mayhave a significant effect on the cloud water chemistry.

Summary and ConclusionsThis study addresses the chemical dynamics of mountain

clouds by using the real-time CRAC analyzer. The principal

Figure 8. Mean pH of cloud water as a function of wind direction during the field seasons of 1987,1988 and 1989.

AIR & WASTE • Vol. 43 • August 1993 • 1081

Dow

nloa

ded

by [

Nor

th C

arol

ina

Stat

e U

nive

rsity

] at

10:

59 1

8 N

ovem

ber

2014

FEATURE

Table VI. S02, NOX, and total suspended paniculate emissions in a 500 mile radius from Mt. Mitchell, NC (units: 105 tons/yr); and wind directionfrequency, liquid water content, pH of cloud water in short events, and equivalent gas phase [H+] for the corresponding wind sectors.

Wind Sector SE SSE S 8W8 SW W WNW NW NNW

NOxTSP

9.6

2,5

4,2$.02.S

Wind Freq.(%) <0.1 <0.1LWC(gAn3) -pH -

3,72.71.7

<0.1

3.6 3.53.1 1.82.5 0.70,2 19.6

5.41.00,9

16,1

3.50.90.52.4

3.0090.50.8

0.23 0.093.S1 3.57

3,41.00.62.1

4.01,00.63.7

4.3IS0.85.6

0.33 0.33 0.623.79 3.75 3.90

4.41.80.85.00.253,72

0.077 0.071 0.024 0.054 0.059 0.078 0.048

5.21.20.77.10.20

4.3181.2

29,7

8.7 9,12.3 2.91.4 1.77.5 <0.1

0.35 0.27 -3.58 3.37 -0.092 0.115 -

results of this study are summarized as follows: The lowest pHvalue in real-time samples was 2.44 during the field seasons,while that in hourly integrated samples was 2.58 in the same event.Seasonal variation of cloud water acidity was also observed. Ingeneral, the cloud pH in summer seasons (pH = 3.15 -3.82) waslower than those in spring and fall (pH=3.48 ~4.02); suggesting thatthe amount of moisture in the atmosphere may be important indetermining the acidity of cloud water. Cloud water acidity in wetseason is low due to dilution effects of air pollutant into abundanceof moisture in the atmosphere. Mean pH of cloud water in 1987,1988 and 1989 were 3.34,3.44 and 3.57, respectively.

The pH found in short cloud events (mean pH 3.12 for all threeyears combined) was substantially lower than that in long cloudevents (mean pH 3.52). Based on the statistical analysis, it was

2. o>

(a)

0.1E B £ SE SSE S SWS SW WSW W WNW NW

Wind Sector

(C)

BSE SE SSE S SWS SW WSW W WNW NW

Wind Sector

•I. 0.12

0.02E BSE SE SSE S SWS SW WSW W WNW NW

Wind Sector

Figure 9. H* Concentration for short cloud events for the field seasons of 1987,1988 and1989 (a) liquid phase; (b) liquid water content (LWC); and (c) calculated correspondinggas phase.

0.12

0.02

T 3

E ESE SE S SWS SW WSW W WNW NW

Wind Sector

2 -

C- 1 "

0.12

0.02E ESE SE SSE S SWS SW WSW W WNW NW

Wind Sector

0.12

E ESE SE SSE S SWS SW WSW W WNW NW

Wind Sector

Figure lO.Comparison of pollutant emission source strength and calculated correspondinggas phase [H*] versus wind sector.

shown that the acidity in non-precipitating cloud water (mean pH3.32) was higher than that in precipitating cloud water (mean pH3.69). The initiation of precipitation was demonstrated to dilutethe cloud water acidity.

Sulfate, nitrate, ammonium and hydrogen ions were found tobe the major constituents of the cloud water. Their total equivalentconcentration was above 90 percent of the total analyzed fractionof equivalents of all ionic constituents expressed as a percent inshort and long cloud events. However, these species in mixedevents contributed about 74 percent to the total fraction ofequivalents of all ionic constituents. Sulfate and nitrate ions aremainly thought to acidify the cloud water, whereas ammonium isthought to neutralize the cloud water acidity.

Sulfate ion concentrations were generally higher than hydro-gen ion concentrations for most cloud water samples. Nitrate ionconcentrations were slightly lower than ammonium ion concen-trations. The sum of sulfate and nitrate ion concentrations isalmost equivalent with the sum of hydrogen and ammonium ion

1082 • August 1993 • Vol. 43 • AIR & WASTE

Dow

nloa

ded

by [

Nor

th C

arol

ina

Stat

e U

nive

rsity

] at

10:

59 1

8 N

ovem

ber

2014

concentration. Consequently, the sum of other minor cations(Mg2+, Ca2+, Na+, and K+) have to be responsible for neutralizingthe other anion (Cl) of cloud water samples. It was found thatsulfate concentration rather than nitrate correlated better withhydrogen ion concentration. This suggests that the cloud wateracidity may be coming predominantly from sulfate aerosol andless from nitric acid.

The variation in cloud water acidity as a function of sixteendifferent source regions was explored. Mean acidities measuredwhen wind direction is from the northwest sector (most pollutedregion) are the highest (pH 3.37). Thus the results showed that, ingeneral, the highest pollutant concentration in cloud water wasfound in the air masses coming from the northwest (midwestemU.S., i.e., Ohio River Valley region). On the other hand, the lowestconcentration was found in the air mass coming from southwestor west (i.e., central U.S.) and east-northeast direction (oceanicregion). Separating the effects of variations in cloud liquid watercontent from variations in pre-cloud pollutant concentrations, therelationship between source intensity and cloud water concentra-tion was ascertained, which suggests that long-range transport hasa significant effect on determining the chemical characterizationin cloud water at Mt. Mitchell, North Carolina.

AcknowledgmentsThis research was supported by the United States Environ-

mental Protection Agency Cooperative Agreement No. CR-815289-02, but does not necessarily reflect the views of theAgency, and no official endorsement should be inferred. Weacknowledge Prof. Vinod Saxena, Prof. Ellis Cowling, Dr. RalphBaumgardner, and Mr. Keith Kronmiller for helpful discussions.Thanks to Mrs. P. Aneja, Ms. B. Batts and Ms. M. DeFeo in thepreparation of the manuscript.

References1. S.B. McLaughlin, "Effects of air pollution on forests: A critical

review;'JAPCA 35:512 (1985).2. G.M. Lovett, W.A. Reiners, R.K. Olson, "Cloud droplet deposition in

sub-alpine balsam fir forest: hydrological and chemical inputs," Sci-ence 218: 1303 (1982).

3. R.A. Castillo,J. Kadlecek, S. McLaren, "SelectedWhiteface Mountaincloud water concentrations during summers 1981 and 1982," Water,Air, SoilPollut. 24:323 (1985).

4. R. J. Vong, P. Guttorp, "Co-Occurrence of ozone and acidic cloudwaterin high-elevation forests," Environ. Sci. Technol. 25:1325 (1991).

5. K.C. Weathers, G.E. Likens, F.H. Bormann, J.S. Eaton, W.B. Bowden,J.L. Anderson, D.A. Cass, J.N. Galloway, W.C. Keene, K.D. Kimball,P. Huth, D. Smiley, "A regional acidic cloud/fog water event in theEastern United States," Nature 319:657 (1986).

6. R.I. Brack, W.P. Robarge, A. McDaniel, "Forest decline in the borealmontane ecosystems of the Southern Appalachian Mountains," Water,Air, SoilPollut. 98:161 (1989).

7. J. Collett, B. Daube, J.W. Munger, M.R. Hoffmann, "The chemicalcomposition of intercepted cloudwater in the Sierra Nevada," Atmos.Environ. 24A:959 (1990).

8. D.L. Sisterson, V.C. Bowerson, A:R. Olsen, T.P. Meyers, R.J. Vong,J.C. Simpson, V.A. Mohnen, "Deposition, Monitoring, Methods andresults," NAPAP State-of-Science Technology Report 6, NationalAcidic Precipitation Assessment Program, Washington, DC. 1991.

9. V.P. Aneja, C.S. Claibom, R.L. Bradow, R. Paur, R. Baumgardner,"Dynamic chemical characterization of montane clouds," Atmos.Environ. 24A:563 (1990).

10. V.P. Aneja, W.P. Robarge, C.S. Claibom, A. Murthy, D.S. Kim, Z. Li,E.B. Cowling, "Chemical climatology of high elevation spruce-firforests in the Southern Appalachian Mountains," Environ. Pollut.75:89(1992).

11. D.S.Kim,V.P.Aneja,"ChemicalcompositionofcloudsatMt.Mitchell,North Carolina, USA," Tellus 44B:41 (1992).

12. V.K. Saxena, N.-H. Lin, "Cloud chemistry measurements and esti-mates of acidic deposition on an above cloudbase coniferous forest,"Atmos. Environ. 24A:329 (1990).

13. A. Murthy, V.P. Aneja, "Deposition and interaction of nitrogen con-taining pollutants to a high elevation forest canopy," paper no. 90-100.4presented at 83rd Annual Meeting of A&WMA, Pittsburgh, PA, 1990.

14. K.G. Kronmiller, W.D. EUenson, R.E. Baumgardner, R.K. Stevens,R.J. Paur, "Development and evaluation of an acid precipitation moni-tor for fractional event sampling with capability for real-time pH andconductivity measurement," Atmos. Environ. 24A:525 (1990).

15. R.I. Brack, "Decline of Mountain Boreal Ecosystems in Central Europeand the southern Appalachian Mountains," in Proceedings, Researchand Development Conferences, Atlanta, Georgia, 1984.

16. R.J. Valente, R.K.A. Mallant, S.E. McLaren, R.S. Schemenaner, "Fieldintercomparisons of ground-based cloud physics instruments at WhitetopMountain, Virginia," J. Atmos. and Oceanic Tech. 6:396 (1989).

17. R.K.A.M. Mallant, G.P.A. Kos, "An optical device for the detection ofclouds and fog," Aerosol Sci. Technol. 13:196(1990).

18. MCCP QA office, "Mountain Cloud Chemistry Project 1989 QualityAssurance Report," prepared by MCCP QA office, W.S. Fleming &Assoc., December 1989.

19. V.K. Saxena, R.E. Stogner, A.H. Hendler, T.P. DeFelice, R.J.-Y. Yeh,N.-H. Lin, "Monitoring die chemical climate of the Mt. Mitchell StatePark for evaluation of its impacts on forest decline," Tellus 41B:92(1989).

20. B.C. Daube, R.C. Flagan, M.R. Hoffmann, "Active cloudwater collec-tor," U.S. Patent Number 4,697,462,1987.

21. V.K. Saxena, G.F. Fisher, "Water solubility of cloud active aerosols,"Aerosol Sci. Technol. 3:335 (1984).

22. R.L. Bradow, V.P. Aneja, "Aerosol composition effects on mountainclouds: Attributing sources of acid deposition," in Proceeding of the12th International Conference andA.A. & N., P.E. Wagner, Ed., Vienna,1988,pp.40-43.

23. S.N. Pandis, J.H. Seinfeld, "Sensitivity analysis of a chemical mecha-nism for aqueous-phase atmospheric chemistry," J. Geophys. Res.94:1105(1989).

24. P.H. Daum, T.J. Kelly, S.E. Schwartz, L. Newman, "Measurement ofthe chemical composition of stratiform clouds," Atmos, Environ.18(12):2671 (1984).

25. G.A. Issac, P.H. Daum, "A winter study of air, cloud and precipitationchemistry in Ontario, Canada," Atmos. Environ. 21:1587 (1987).

26. C.A. Johnson, L. Sigg, J. Zobrist, "Case studies on the chemicalcomposition of fogwater: The influence of local gaseous emissions,"Atmos. Environ. 21:2365 (1984).

27. E. Gorham, F.B. Martin, J.T. Litzau, "Acid rain: ionic correlations inthe eastern United States, 1980 • 1981," Science 225:407 (1984).

28. D.A. Hegg, P.V. Hobbs, "Cloud water chemistry and the production ofsulfates in clouds," Atmos. Environ. 15:1597 (1981).

29. M. Saeger, J. Langstaff, R. Walters, L. Modica, D. Zimmerman, D.Fratt, D. Dulleba, R. Ryan, J. Demsy, W. Tan, D. Spraque, D. Mudgett,A.S. Werner, "The 1985 NAPAP Emissions inventory (Version 2):Development of the annual data and modelers' tapes," U.S. Environ-mental Protection Agency Report Number EPA-600/7-89-012a, 1989.

About the AuthorsDr. Aneja is a research associate professor and Dr. Kim is a

graduate research assistant with the Department of Marine,Earth and Atmospheric Sciences, North Carolina State Univer-sity, Raleigh, NC 27695-8208. This manuscript was submittedfor peer review on January 9,1992. The revised manuscript wasreceived on June 8,1992.

AIR & WASTE • Vol. 43 • August 1993 • 1083

Dow

nloa

ded

by [

Nor

th C

arol

ina

Stat

e U

nive

rsity

] at

10:

59 1

8 N

ovem

ber

2014