GEOPHYSICAL RESEARCH LETTERS, VOL. 9, NO. 1 0 , PAGES 1207-1210, OCTOBEX 1982
PARTICLE SIZE DISTRIBUTION O F NITRATE AND SULFATE IN THE MARINE ATMOSPHERE
D.L. Savoie and J.M. Prospero
Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, Florida 33149
Abstract. Cascade impactor samples were collected
a t coastal sites on Sal Island, Barbados, and Virginia Key,
Miami during 1974 and a t two Miami coastal sites on
Virginia Key and Key Biscayne during 1981. In all of t h e
samples, t h e majority of t h e nitrate mass was found on
intermediate size particles and exhibited a mass median
diameter (MMD) of about 4 pm. The ratio of t h e MMD of
nitrate t o t h a t of sea-salt (about 7 pm) varied from 0.54
t o 0.60 which indicates t h a t t h e nitrate mass distribution
is well defined by t h e surface a r e a distribution of t h e
sea-salt aerosols.
This sharply contrasts with t h e
distribution of non-sea-salt sulfate (also produced from
gas-to-particle conversion reactions) which is present
primarily on submicron aerosols. The lower levels of
nitrate in t h e small particles a r e probably a consequence
of t h e higher volatility and photochemical reactivity of
t h e nitrate compounds expected t o exist in the acidic
aerosols.
Introduction
The nitrogen cycle in t h e atmosphere is t h e subject of
much current research. An important oxidative endmember in this nitrogen cycle is nitrate. However, t o
d a t e t h e vast majority of t h e studies of nitrate in t h e
atmosphere have been conducted in continental
(primarily urban) areas; data on nitrate in marine areas
a r e still extremely sparse especially with regard t o t h e
nitrate size distribution (Duce, 1981).
In this report, we present d a t a on t h e size distribution
of nitrate in aerosols collected over an extended time
period in maritime air masses over t h e tropical North
Atlantic. The bulk concentration data will be presented
in a l a t e r report.
The size distribution data were
obtained by analyzing samples collected with high
volume cascade impactors during 1974 and 1981. The
1974 samples were collected a t t h r e e locations a t coastal
sites on Sal Island, Cape Verde Islands, on Barbados, West
Indies, and on t h e NOAA (National Oceanic and
Atmospheric Administration) building, Virginia Key,
Miami, Florida. Site descriptions a r e presented in Savoie
and Prospero, 1977. The 1974 study was carried out a s a
part of t h e Global Atmospheric Research Program
(CARP) Atlantic Tropical Experiment (GATE). The 1981
samples were collected on t h e roof of t h e Rosenstiel
School of Marine and Atmospheric Science (RSMAS),
University of Miami, Florida and on t h e upper walkway
of t h e lighthouse a t Cape Florida S t a t e Recreation Area
(CFSRA), located on t h e southern tip of Key Biscayne,
Miami, Florida. The data from a total of 41 samples a r e
presented in this report.
Sampling and Analysis
Cascade impactor samples were collected with a
Weather Measure Corp. Model 235 High Volume Cascade
Impactor which was modified by mounting an additional
uppermost stage which extended t h e large particle range
Copyright 1982 by t h e American Geophysical Union.
Paper number 2L1194.
0094-8276/82/002L-1194$3.00
of this irnpactor by a factor of two. In t h e 1981
program, slotted, nonporous polycarbonate sheets were
used as impaction surfaces; these sheets had a "frosted"
finish which minimizes t h e blow-off loss of droplets from
t h e collection stage. A 20 x 25 c m W h a t m a n 4 1 filter
served as t h e backup to collect t h e smallest particles.
The polycarbonate inlpaction sheets were used because
of their inertness and t h e absence of artifact nitrate
production (Spicer and Schumacher, 1979). In t h e 1974
sampling program W h a t m a n 4 1 filters were used in t h e
casade impactors as both impaction substrates and
backup filters.
In the analysis procedure, w e extracted a quarter
section of t h e impactor backup filters and a half section
of the what man-41 impaction-surf a c e filters. These
were extracted in Millipore filtering-centrifuge tubes
with one 10-ml aliquot of Milli-Q water, followed by t w o
The entire polycarbonate impaction
5-ml aliquots.
sheets were extracted into 25 ml of Milli-Q water using
an ultrasonic probe t o disperse t h e particulates and
increase the r a t e of dissolution. During 1981 t h e filters
and impaction surfaces were extracted immediately upon
removal f rorn the filter holders.
Sulfate and nitrate concentrations were determined
by ion chromatography.
The Dionex Model-10 ion
chromatograph
was
operated
with
a
0.0024 M N a COj/0.0009 M KOH eluent a t a flow r a t e of
3 ml/min an8 sample volumes of 0.1 ml. Calibration
curves were developed from t h e areas of the peaks (as
measured with a Hewlett-Packard 3390A integrator)
produced from a series of mixed standards. Sodium was
determined by atomic absorption spectrometry using a
series of mixed standards approximating t h e composition
of diluted seawater.
Standard errors of - the
concentrations in the extracts were 2 5% for SO- and
NO3 and 2 2% for ~ a * . The volume flow rates th?ough
t h e fllters were determined froin t h e pressure drop
across a calibrated orifice plate in t h e exhaust outlet.
The flow r t e for t h e impactor was essentially constant
a t 0.86 m 9/ m .~ nfor all samples; total sample y l u m e s
were generally in t h e range of 1000 t o 2000 m
The
50% cutoff diameters for t h e cascade impactors were
16.7, 8.4, 4.2, 2.1, 1.0, 0.50 microns for s y g e s 1-6,
respectively, assuming a density of 1.2 g cm- for seasalt particles a t 80% relative humidity (Patterson g &.,
1980). During 1981, t h e pumps were controlled by a wind
direction sensor s o t h a t sampling occurred only during
periods of easterly winds (i.e., when t h e winds were
blowing directly off t h e ocean). In the case of t h e
RSMAS samples, t h e in-sector condition included winds
which would pass over Key Biscayne; these samples
conceivably could be biased by pollutant inputs. The only
land masses upwind of t h e CFSRA s i t e were t h e Bahama
Islands.
.
.
Results
Representative particle- size d i s t r i b ~ t i o n s for Na+
NO;, and non-sea-salt SO& (total SO - - 0.2517 Na'i
from samples collected a t RSMAS a n 8 CFSRA during
1981 a r e shown in Figure 1. Although t h e concentrations
varied substantially, t h e precise locations of t h e peaks
varied only slightly and t h e general trends in t h e
S a v o i e and P r o s p e r o : N i t r a t e and S u l f a t e S i z e D i s t r i b u t i o n s
CFSRA
Y
BARBADOS
A
PARTICLE DIAMETER RANGE
(MICRONS)
Figure I. Typical particle-size dsitributions of sodium,
non-sea-salt (xs) sulfate, and n i t r a t e a t several coastal
sites.
distributions were essentially t h e same for all samples.
The sodium and total sulfate distributions were typical of
those which have been previously reported in t h e
literature, both exhibiting peaks a t about 6 t o 8 microns
diameter a s a consequence of their presence in sea-salt
aerosols. The non-sea-salt sulfate, being derived from
gaseous precursors, is contained primarily on submicron
particles. Although n i t r a t e is presumed t o be derived
from gaseous precursors, its size distribution differs
substantially from t h a t of non-sea-salt sulfate and, in
f a c t , is more similar t o t h a t of sea-salt (i.e., sodium).
The geometric mean n i t r a t e concentration a t RSMAS,
0.975 pg/SCM (standard cubic meter), was about 60%
g r e a t e r than t h a t a t CFSRA, 0.623 pg/SCM (Table I).
This result is undoubtedly a consequence of t h e f a c t t h a t
a i r masses were strictly marine during t h e e n t i r e CFSRA
sampling program from 4 August t o 14 August 1981.
Although t h e RSMAS samples were collected only during
periods of on shore winds, t h e synoptic scale circulation
patterns indicated t h a t several of t h e a i r masses sampled
may have been over continental a r e a s within t h e previous
24 t o 48 hours.
Also, a s s t a t e d earlier, air parcels
sampled a t RSMAS may have passed over Key Biscayne
(population:
4,500) which has heavy tourist and
recreational traffic. Despite this difference, t h e m e a n
MMD1s (mass median diameters) for both ~ a and
+ NO,
w e r e essentially t h e s a m e a t the t w o locati_ons; t h e MMD
of ~ a was
+ ca. 7.3 pm and t h a t of NO3 was 4.0 um
(Table 1).
The samples collected a t Miami during 1974 exhibit
essentially t h e same characteristics a s thpse collected in
1981 (Table I). The t o t a l NO3 and Na concentrations
and their MMD's a r e not significantly different from
those obtained from either t h e RSMAS o r CFSRA
samples despite .the use of W h a t m a n 4 1 filter material a s
t h e impaction surface for t h e 1974 samples. It should be
noted t h a t bulk aerosol samples collected on 20 x 25 cm
Whatman-41 filters during 1981 (total number = 33)
without t h e use of a wind direction controller yielded a
geometric mean n i t r a t e concentration of 1.3 ug/SCM,
33% higher than t h a t for t h e wind controlled samples
collected a t t h e s a m e location.
The particle size distribution of sodium and n i t r a t e a t
Sal Island (Table I) during 1974 were also essentially t h e
s a m e a s those a t Miami. The only major difference was
t h a t t h e sodium concentration was a f a c t o r of 2 t o 2.5
higher at Sal than a t Miami. The mean t o t a l n i t r a t e
concentration of 0.96 pg/SCM a t Sal compares favorably
with a geometric mean of 0.99 pg/SCM determined from
10 bulk aerosol samples collected for us aboard t h e R/V
Researcher during February, 1979 and aboard the R/V
Meteor during January and June, 1979; during those
periods, t h e ships were cruising several hundred
kilgmeters from t h e African coast between latitudes
10 N and 2 5 ' ~ .
A major difference in the sodium distribution at
Barbados is t h e absence of a peak below ,the upper limit
cut-off of our impactor. This f e a t u r e is a consequence
of t h e high concentration and large size of sea-salt
particles produced by breakers along t h e reefs and within
t h e shoals immediately offshore of t h e sampling s i t e
which was located on a 20 m bluff.
The n i t r a t e
distribution, however, was not a f f e c t e d by these
processes; t h e MMD of n i t r a t e at Barbados was not
significantly different from t h a t a t any of t h e o t h e r
stations.
In contrast, t h e mean t o t a l nitrate
concentration, 0.39 ug/SCM, a t Barbados is a f a c t o r of 2
t o 2.5 lower than t h a t a t t h e other stations. The mean of
t h e 1974 impactor samples, however, is in fair agreement
with t h e mean of 0.52 ug/SCM from 114 bulk aerosol
samples collected a t t h e s a m e s i t e in 1979 and t h e mean
of 0.37 pg/SCM for 6 bulk samples collected for us
aboard t h e R/V Researcher during July and August, 1979
in t h e s a m e region. In addition, t h e mean a t Barbados is
in good agreement with t h a t of 0.28 ug/SCM reported f o r
t h e marine boundary layer over t h e c e n t r a l and southern
Pacific (Huebert, 1980; Huebert and Lazrus, 1978, 1980a,
1980b).
The above d a t a a r e indicative of a very consistent
particle size distribution of n i t r a t e in m a r i t i m e air
masses. Further insight into t h e n i t r a t e distribution is
provided by considering t h e ratio of t h e MMD of n i t r a t e
t o t h e MMD of sodium. If t h e sea-salt size distribution i s
assumed t o be log-normal, then t h e ratio of t h e surface
median diameter (SMD) t o .the MMD is given by:
SMD/MMD = e x p -(In ag)2
where og i s t h e standard geometric deviation of t h e size
distribution which is t h e s a m e for t h e mass, surface a r e a ,
and number distributions. Typical estimates of a g f o r
sea-salt aerosols range from about 1.7 t o 2.4 (see e.g.
Walsh g &., 1978), yielding SMD/MMD ratios of 0.76 t o
0.47, respectively. In our d a t a (Table I ) t h e ratios of t h e
MMD of nitrate t o t h a t of sodium range from 0.54 t o
0.60, thus indicating t h a t t h e n i t r a t e mass is roughly
distributed a s t h e surface a r e a of t h e sea-salt aerosols.
Discussions and Conclusions
Because of previously reported a r t i f a c t problems
(Appel and Tokiwa, 1981; Appel etg., 1979, 1980, 1981;
S a v o i e and P r o s p e r o : N i t r a t e and S u l f a t e S i z e D i s t r i b u t i o n s
TABLE 1.
-LOCATION
Geometric mean mass median diameters (MDD,pm) and total concentrations (pg/SCM)
of sodium and nitrate.
--
(YEAR)
N
MMD
SODIUM
CONC
-
11
7.4
( 1 . 3 ) 4.0
(1.1 9
1 1 2 . 3
6
7.1
1 . 2 6
(1.2
4.0
(11.62
5
6.6
(1.8) 3.9
(1.3) .86
(1.2) 2.2
9
7.0
(1.9) 4.2
(1.2) 5.2
(1.3) .96
10
(- 1 (- ) 3.6
(1.2) .39
Numbers in parentheses a r e t h e sample standard geometric deviations for t h e
them.
RSMAS
CFSRA
Miami
Sal Island
Barbados
(1981)
(1981)
(1974)
(1974)
(1974)
NITRATE
MMD
CONC
Forrest g &., 1979, 1980, 1982; Okita g &., 1976;
Pierson g &., 1980; Spicer and Schumacher, 1977, 1979;
Spicer
&., 1982), we were very much concerned about
t h e possible generation of nitrate artifacts within t h e
impactor which might yield apparent size distributions
significantly different from those which actually occur in
t h e atmosphere.
Possible problems include:
1) t h e
reaction of vapor phase nitric acid with aerosols (e.g.
calcite and NaCI) already collected on t h e impaction
surfaces and/or t h e adsorption of t h e vapors on t h e
impaction surfaces themselves, and 2) t h e release of
nitric acid from small particles as a consequence of t h e
presence of acid sulfate species on t h e same collection
surface. To be significant, t h e first requires that a
substantial portion of t h e total nitrate exist as nitric
acid vapor. Results from a subsequent study of maritime
air masses a t Miami, using virtual impactors in which
only about 1% of t h e air flow is through t h e collected
coarse fraction substantiate our contention that on t h e
order of 80% of t h e nitrate is in t h e greater than 1 um
diameter fraction. Furthermore, less than 10% of t h e
nitrate was found to exist in t h e vapor phase.
The
absence of a significant difference between the
distributions obtained when using Whatman-41 a s
compared t o polycarbonate impaction surfaces indicates
that nitric acid absorption on the surfaces was also
negligible. In our most recent cascade impactor study
aboard ship in t h e Gulf of Mexico, our normal cascade
i m p a c t o r - W h a t m a n - backup filter system was followed
immediatley by a glycerine-K CO
impregnated
Whatman-41 filter to t r a p any v a p d p h a 2 e nitrate which
may have volatilized from t h e upstream collection
surfaces. Nitrate collection on t h e impregnated filter
was near t h e detection limit and represented a maximum
of 5% of t h e total nitrate collected in a sample. Reports
on these l a t t e r two studies a r e currently in preparation.
Our conclusion that gaseous HN03 represents a minor
fraction of t h e total nitrate mass in t h e lower marine
troposphere contrasts with that of Huebert (1980) who
reported t h a t t h e concentration of gaseous nitric acid in
t h e marine boundary layer over t h e central Pacific is
about 50% of t h e particulate nitrate concentration.
However, recent evidence (Appel g &., 1980, 1981;
Forrest g g., 1981) suggests that t h e high values
measured by Huebert may be an artifact generated by
nitrate loss from their Teflon aerosol prefilter. On t h e
other hand, we would expect to find substantially higher
percentages of t h e nitrate in t h e gaseous form in t h e
f r e e marine troposphere where the concentrations of
neutral or basic aerosols a r e usually very low.
The data presented here indicate a major difference
in t h e particle size distributions of nitrate and non-seasalt sulfate, species which a r e both produced from
gaseous precursors.
In contrast to t h e non-sea-salt
(1.3)
(1.3)
(1.9)
(1.5)
(1.4)
--
MMD(NO~-)/
MMD(N~+)
.54
.56
.59
.60
-
(1.1)
(1.1)
(1.3)
(1.3)
(-
1
parameters preceding
sulfate which is present primarily on submicron particles,
the nitrate exists almost exclusively on particles larger
than I vm diameter with its mass being distributed as t h e
surface area of t h e sea-salt aerosols. Similarly, Junge
(1956, 1957) had found that aerosols collected with a
two-stage impactor a t coastal stations in Massachusetts,
Florida, and Hawaii contained t h e majority of the nitrate
on particles greater than 1.6 um diameter. Gravenhorst
e t &. (1979) noted that t h e peak in the nitrate mass
coincided with the small particle end of t h e sea-salt
mass distribution in samples collected over t h e North
Atlantic; although they performed no confirming
calculations, they suggested that this may reflect a
similarity in t h e distributions of nitrate mass and seasalt surface area. Thus, t h e nitrate size distribution
appears to be rather uniform over t h e oceans.
In
contrast, t h e major fraction of the nitrate in continental
areas is concentrated on submicron aerosols (Appel g
al., 1978). The results suggest that sea-salt particles
may play an important role in t h e atmosphere-to-ocean
flux of odd nitrogen and, further, that t h e modes and
rates of deposition of nitrate and sulfate t o t h e s e a
surface a r e likely t o be quite different.
If aerosol nitrate is produced by t h e absorption and
subsequent reaction of NOi or by disso?ution of gaseous
n described
HNO on sea-salt aerosols, he d ~ s t r l b u t ~ ojust
is prc?cisely what one would expect. The extremely low
concentration of nitrate on submicron particles is
probably due t o the relative instability of HN03 and
NH NO a s compared t o H SO4, NH4HS04, (NH ) SO
wh$h dould be formed in d o s e particles. ~ l t h o t ~ ! thg
'?
sulfate compounds all have significant vapor pressures,
they a r e all much lower than those of t h e nitrates. In
addition, sulfuric acid is essentially stable with respect
t o photochemical degradation. In contrast, nitric acid
undergoes degradation t o NO and/or N 0 (see e.g.,
Ore1 and Sienfeld, 1977, and &vis, 1980):
is, in f a c t ,
t h e NO produced from these reactions which results in
the disc%loration of nitric acid in the presence of light.
There appears t o be a small but consistent shoulder or
peak in t h e non-sea-salt sulfate mass distribution in t h e
same size range as the nitrate peak (Figure I). Due t o
the combined analytical errors in t h e sodium and total
sulfate on a given impactor stage and the fairly low
percentage of non-sea-salt sulfate in this size range, our
present data, while certainly suggestive, a r e not
completely definitive in this respect.
Although t h e
heterogeneous reaction of SO on sea-salt particles could
conceivably contribute t o s u c i a phenomenon, t h e major
causative process is probably the agglomeration of small
sulfate aerosols with t h e sea-salt particles; i t is notable,
in this regard, .that a small percentage of particulate
ammonium has also been found in this size fraction
(Gravenhorst g &., 1979).
14
1210
S a v o i e and P r o s p e r o : N i t r a t e and S u l f a t e S i z e D i s t r i b u t i o n s
Acknowledgments. We sincerely thank Major James
Stevenson (Chief Naturalist, Division of Recreation and
Parks, Florida Department of Natural Resources) and
Captain C a r l Nielson (Park Superintendent, Bill Baggs
C a p e Florida S t a t e Recreation Area) for their interest
and support in permitting us t o sample from t h e CFSRA
lighthouse. Thanks a r e also due Mr. Tom Snowdon and
Ms. Ruby Nees for their technical assistance and B.J.
Huebert for helpful discussions. This material is based
upon work supported by t h e Division of Atmospheric
Sciences, National Science Foundation Grants ATM7812246 and ATM-8016127 and by t h e National Oceanic
and Atmospheric Administration. Contribution from t h e
Rosenstiel School of Marine and Atmospheric Science,
University of Miami, Miami, Florida.
References
Appe1,B.R. and Y. Tokiwa.
1981.
Atmospheric
n nerrors due t o reactions
particulate nitrate s a m ~ l i "
k i t h particulate and gaseous strong acids. Atmos.
Environ. 15:1087-1089.
-A.P.E ~ , B.R.. E.L. Kothnv, E.M. Hoffer. G.M. Hidv. and
J.J. ~ e s k l o w s k i . 1978: Sulfate and Aitrate d a t a f r o m
t h e California aerosol characterization experiment
(ACHEX). Environ. Sci. Technol. 12(4): 418-425.
Appel,
. . B.R., S.M. Wall, Y. Tokiwa, and M. Haik. 1979.
interference effect; in samplini particulate nitrate in
ambient air. Atmos. ~ n v i r o n .fii 319-325.
Appel,
. . . B.R., S.M. Wall, Y. Tokiwa, and M. Haik. 1980.
Simultaneous nitric acid,
nitrate and
acidity measurements in ambient air.
Atmos.
Environ.
14: 549-554.
Davis, D.D. 1980. Project Gametag: An overview. 3.
Geophys. Res. E ( C 12):7285-7292.
Duce, R.A. 1981. Biogeochemical cycles and t h e air/sea
exchange of aerosols. Paper presented a t t h e SCOPE
Workshop on t h e Interaction of Biogeochemical
Cycles, Orsundsbro, Sweden. 25-30 May.
1979.
Forrest, J., R. Garber, and L. Newman.
Formation of sulfate, ammonium and n i t r a t e in an
oil-fired power plant plume.
Atmos Environ.
13:1287-1297.
orr rest, J., R.L. Tanner, D. Spandau, T. DtOttavio, and
L. Newman. 1980. Determination of t o t a l inorganic
n i t r a t e utilizing collection of nitric acid on NaCIimpregnated filters. Atmos. Environ. fi:137-144.
Forrest, J., D.J. Spandau, R.L. Tanner, and L. Newman.
1982. Determination of atmospheric nitrate and
nitric acid e m ~ l o v i n na diffusion denuder with a filter
16(6): 1473-1485.
pack. Atmos. '~n;ir;n.
Gravenhorst, G., K.P. Muller, and H. Franken. 1979.
inorganic nitrogen in marine aerosols. Gesellschaft
fur Aerosolforsch 2:182-187.
~ u e E r t B.J.
.
1980. Nitric acid and aerosol n i t r a t e
.
measurements in t h e equatorial Pacific region.
Geophys. Res. Lett. 7(5):325-328.
Huebert, B.J. and A.L.
Lanus.
1978.
Global
tropospheric measurements of nitric acid vapor and
particulate nitrate. Geophys.
Lett. (7):577580.
Huebert, B.J. and A.L. L a n u s . 1980a. Tropospheric gasphase and particulate n i t r a t e neasurements.
-J.
Geophys. R&. g ( C 12):7322-7328.
Huebert, B.J. and A.L. L a n u s . 1980b. Bulk composition
of aerosols in t h e r e m o t e troposphere. J. deophys.
Res. g(C12):7337-7344.
~ u n x ~ . ~1956.
.
Recent investigations
in air
chemistry. Tellus 8(2): 127-139.
Junne. C.E. 1957. Chemical analysis of aerosol particles
i n d of gas traces on the island of ~ a w a i i Tellus
~
9(4):528-537.
~ k i t a T.,
, S. Morimoto, M. Izawa, and S. Komno. 1976.
Measurements of gaseous and particulate n i t r a t e in
t h e atmosphere. Atmos. Environ. g:1085-1089.
Orel, A.E. and J.H. Sienfeld. 1977. Nitrate formation in
Technol.
atmospheric aerosols.
Environ.
1 l(10):lOOO-1007.
Patterson, E.M., C.S. Kiang, A.C. Delany, A.F. Wartburg,
A.C.D. Leslie and B.J.
Huebert.
1980. Global
measurements of aerosols in r e m o t e continental and
marine regions: concentrations size distributions, and
optical properties. J. Geophys.
g ( C 12):73617376.
Pierson, W.R.,
W.W.
Brachaczek,
T.J.
Korniski,
T.J. Truex, and J.W. Butler.
1980.
Artifact
formation of sulfate, nitrate, and hydrogen ion on
backup filters: Allenhenv Mountain experiment. 3.
Air ~ o l l u t .Contr. ~ s c o c .*30(1): 30-34.
Savoie. D.L. and J.M. P r o s p e r ~ .
1977.
Aerosol
concentration statistics for t h e northern tropical
Atlantic. 3. Geophys. Res. 82(37): 5954-5964.
Spicer, C.W. and P. Schumacher. 1977. Interference in
sampling atmospheric particulate nitrate. Atmos.
Environ.
- 11:873-876.
Spicer, C.W. and P.M. Schumacher. 1979. Particulate
nitrate:
Laboratory and field studies of major
sampling interferences. Atmos. Environ. 13(4):543552.
Spicer, C.W., J.E. Howes Jr., T.A. Bishop, and L.H.
Arnold. 1982. Nitric acid measurement methods: a n
intercomparison. Atmos. Environ. 16(6): 1487-1500.
Walsh, P.R., K.A. Rahn, and R.A. Duce.
1978.
~ r r o n e o u selemental mass-size functions from a highvolume cascade impactor.
Atmos. Environ.
12(8):1793-1795.
-
&.
s.
e.
A
(Received O c t o b e r 2 7 , 1981;
a c c e p t e d J u l y 29, 1 9 8 2 . )