Aerosol particle chemical characteristics measured from aircraft in the lower troposphere during ACE-2


Schmeling, M.; Russell, L.M.; Erlick, C.; Collins, D.R.; Jonsson, H.; Wang, Q.; Kregsamer, P.; Streli, C.

Tellus Series B Chemical & Physical Meteorology 52B(2): 185-200

2000


During the Aerosol Characterization Experiment (ACE-2), filter samples were collected aboard the Center for Interdisciplinary Remotely Piloted Aircraft Studies (CIRPAS) Pelican aircraft near Tenerife in June and July of 1997. The flights included constant altitude measurements in the boundary layer as well as profiles up to 3800 m providing detailed chemical information about the composition of the aerosol distribution in the lower troposphere. Three cases with different air mass origins - clean marine air, anthropogenically-influenced air from the European continent, and dust-laden air from the Sahara - were identified. The samples were analyzed by ion chromatography (IC) for ionic species, by combined thermal and optical analysis (TOA) for organic carbon, and by total reflection X-ray fluorescence (TXRF) for elemental composition. Particle composition and size distributions for the range of air masses encountered illustrate links in the chemical and microphysical characteristics of aerosol from different sources. Clean marine air masses were characterized by low particle number and mass concentrations with no detectable metals, while anthropogenically-influenced and dust-laden air had high number, mass, and trace metal concentrations. Anthropogenic sources were characterized by high concentrations of submicron particles and some Fe and Cu, whereas dust particle loadings included a significant mass of micron-sized particles and significant loadings of Fe, in addition to small amounts of Mn, Cu, and Ni. These results showed similar tracers for air mass origin as those found in other measurements of oceanic and continental air masses. Aerosol optical properties were estimated with a simplified model of the aerosol based on the measured compositions. The real and imaginary refractive indices and single scattering albedos differed significantly among the three types of aerosol measured, with clean marine aerosol properties showing the least absorption and dust-containing aerosols showing the most. There were only small differences in optical properties for the two different cases of clean marine aerosol, but some significant differences between the two dust cases. Since measurement uncertainties affect these calculations, we studied the type of mixing and the fraction of absorbing species and found the calculation was sensitive to these variations only for the dust-containing aerosol case, probably due to the small amount of water present. While the optical properties varied little with composition for clean marine and anthropogenically-influenced cases, they showed a strong dependence on variations in particle composition and mixing state for the dust-containing cases.

Tellus
(2000),
52B,
185-200
Copyright
©
Munksgaard,
2000
Printed
in
UK.
All
rights
reserved
TELLUS
ISSN
0280-6509
Aerosol
particle
chemical
characteristics
measured
from
aircraft
in
the
lower
troposphere
during
ACE-2
By
MARTINA
SCHMELING
1
,
LYNN
M.
RUSSELL
1
*,
CARYNELISA
ERLICK
2
,
DONALD
R.
COLLINS',
HAFLIDI
JONSSON
4
,
QING
WANG
4
,
PETER
KREGSAMER
5
and
CHRISTINA
STRELI
6
,
1
Department
of
Chemical
Engineering,
Princeton
University,
Princeton,
New
Jersey;
2
Atmospheric
and
Oceanic
Sciences
Program,
Princeton
University,
Princeton,
New
Jersey;
'Department
of
Chemical
Engineering,
California
Institute
of
Technology,
Pasadena,
California;
4
Department
of
Meteorology,
Naval
Postgraduate
School,
Monterey,
California,
USA;
5
Atominstitut
of
the
Austrian
Universities,
Vienna,
Austria
(Manuscript
received
7
January
1999;
in
final
form
13
September
1999)
ABSTRACT
During
the
Aerosol
Characterization
Experiment
(ACE-2),
filter
samples
were
collected
aboard
the
Center
for
Interdisciplinary
Remotely
Piloted
Aircraft
Studies
(CIRPAS)
Pelican
aircraft
near
Tenerife
in
June
and
July
of
1997.
The
flights
included
constant
altitude
measurements
in
the
boundary
layer
as
well
as
profiles
up
to
3800
m
providing
detailed
chemical
information
about
the
composition
of
the
aerosol
distribution
in
the
lower
troposphere.
Three
cases
with
different
air
mass
origins
clean
marine
air,
anthropogenically-influenced
air
from
the
European
continent,
and
dust-laden
air
from
the
Sahara
were
identified.
The
samples
were
analyzed
by
ion
chromatography
(IC)
for
ionic
species,
by
combined
thermal
and
optical
analysis
(TOA)
for
organic
carbon,
and
by
total
reflection
X-ray
fluorescence
(TXRF)
for
elemental
composition.
Particle
composition
and
size
distributions
for
the
range
of
air
masses
encountered
illustrate
links
in
the
chemical
and
microphysical
characteristics
of
aerosol
from
different
sources.
Clean
marine
air
masses
were
characterized
by
low
particle
number
and
mass
concentrations
with
no
detectable
metals,
while
anthropogenically-influenced
and
dust-laden
air
had
high
number,
mass,
and
trace
metal
concentrations.
Anthropogenic
sources
were
charac-
terized
by
high
concentrations
of
submicron
particles
and
some
Fe
and
Cu,
whereas
dust
particle
loadings
included
a
significant
mass
of
micron-sized
particles
and
significant
loadings
of
Fe,
in
addition
to
small
amounts
of
Mn,
Cu,
and
Ni.
These
results
showed
similar
tracers
for
air
mass
origin
as
those
found
in
other
measurements
of
oceanic
and
continental
air
masses.
Aerosol
optical
properties
were
estimated
with
a
simplified
model
of
the
aerosol
based
on
the
measured
compositions.
The
real
and
imaginary
refractive
indices
and
single
scattering
albedos
differed
significantly
among
the
three
types
of
aerosol
measured,
with
clean
marine
aerosol
properties
showing
the
least
absorption
and
dust-containing
aerosols
showing
the
most.
There
were
only
small
differences
in
optical
properties
for
the
two
different
cases
of
clean
marine
aerosol,
but
some
significant
differences
between
the
two
dust
cases.
Since
measurement
uncertainties
affect
these
calculations,
we
studied
the
type
of
mixing
and
the
fraction
of
absorbing
species
and
found
the
calculation
was
sensitive
to
these
variations
only
for
the
dust-containing
aerosol
case,
probably
due
to
the
small
amount
of
water
present.
While
the
optical
properties
varied
little
*
Corresponding
author:
A317
Engineering
Quadrangle,
Princeton
University,
Princeton,
NJ
08544,
USA.
email:
lrussell@princeton.edu.
Tellus
52B
(2000),
2
186
M.
SCHMELING
ET
AL.
with
composition
for
clean
marine
and
anthropogenically-influenced
cases,
they
showed
a
strong
dependence
on
variations
in
particle
composition
and
mixing
state
for
the
dust-containing
cases.
1.
Introduction
Aerosol
properties
in
the
North
Atlantic
region
include
a
variety
of
aerosol
types,
resulting
from
influences
from
several
different
kinds
of
air
masses
(Huebert
et
al.,
1996;
Russell
et
al.,
1996a).
Clean
marine air,
including
air
masses
from
Arctic
regions,
contains
aerosol
that
has
been
processed
under
pristine
conditions
for
several
days
(Hoppel
et
al.,
1990).
Anthropogenically-influenced
air
masses
that
originate
in
Europe
frequently
include
high
particle
concentrations
in
the
boundary
layer
from
continental
sources.
Saharan
air
masses,
originating
in
northern
Africa,
contain
dust
par-
ticles
that
have
been
lifted
above
the
boundary
layer
by
updrafts,
providing
long-lived
and
dis-
tinctive
aerosol
layers
in
the
free
troposphere
(Bergametti
et
al.,
1989;
Arimoto
et
al.,
1995).
Aerosol
chemical
composition,
particle
size
distri-
butions,
and
air
mass
back
trajectories
can
serve
as
tracers
for
the
origin
of
the
air
mass
sampled
(Zhang
et
al.,
1993;
1998).
Differing
particle
origins
result
in
mixtures
of
chemical
species
with
different
optical
properties
(d'Almeida
et
al.,
1991).
During
the
second
Aerosol
Characterization
Experiment
(ACE-2)
in
the
northeastern
Atlantic
Ocean,
a
series
of
aircraft
and
surface-based
meas-
urements
was
conducted
in
order
to
study
the
chemical
and
physical
properties
of
the
aerosol
found
in
this
region
of
the
world
(Raes
et
al.,
2000;
Verver
et
al.,
2000).
Suites
of
microphysical
instru-
mentation
on
several
platforms
measured
radiative
and
microphysical
properties
of
the
aerosol
par-
ticles
to
quantify
their
contribution
to
the
so-called
direct
effect
of
aerosol
on
radiative
transfer
in
the
atmosphere
(Russell
and
Heintzenberg,
2000).
Aerosol
filter
samples
were
collected
on
the
CIRPAS
Pelican
aircraft
as
well
as
on
several
surface
and
ship-based
sites
to
identify
chemical
species
contributing
to
optical
properties
for
the
aerosol
measured
(Putaud
et
al,
2000;
Swietlicki
et
al.,
2000;
Quinn
et
al.,
2000).
Here,
we
present
the
analysis
of
elemental
and
ionic
species,
collected
with
filter
holders
aboard
the
CIRPAS
Pelican
aircraft
near
Tenerife.
The
Pelican
aircraft
(a
modified
Cessna)
flew
at
low-
altitude
patterns
for
several
hours
during
some
flights,
offering
the
ability
to
measure
aerosol
particles
in
the
marine
boundary
layer.
Flight
patterns
included
profiles
while
descending
or
climbing
in
altitude
between
50
and
3800
m
above
sea
level
to
show
the
vertical
variation
in
the
size
distribution.
The
chemical
composition
provides
information
about
the
origin
of
the
air
mass.
Non-
sea
salt
(nss)
sulfate,
for
example,
is
an
indicator
of
anthropogenic
sources.
Trace
metal
constituents
in
particles
indicate
both
specific
source
types
and
regions
(Schutz
and
Sebert,
1987;
Arimoto
et
al.,
1995).
The
resulting
characterization
of
aerosol
sources
provides
specific
examples
of
the
aerosol
contribution
to
scattering
and
absorbing
proper-
ties
of
the
atmosphere
on
local
and
regional
scales.
2.
Chemical
analysis
Samples
were
collected
with
a
stainless
steel
filter
holder,
loaded
with
47
mm
quartz-fiber
filters
(2500QAT-UP;
Pallflex
Products
Corp.)
that
were
precleaned
by
heating
at
900
°C,
and
with
a
teflon
filter
holder,
loaded
with
47
mm
teflon
filters
(Zefluor;
Gelman
Sciences).
Several
preferable
approaches,
including
more
detailed
chemical
identification,
larger
mass
collection
or
multiple
size-cut
particle
impaction
were
precluded
by
space,
power
and
sampling
time
limitations
aboard
the
Pelican
aircraft.
Quartz-fiber
filters
were
used
for
collection
of
organic
and
elemental
carbon
and
teflon
filters
for
ionic
and
elemental
species.
Both
filter
holders
were
mounted
on
the
CIRPAS
Pelican
Aircraft
for
sampling
representative
aero-
sol
particles
during
ACE-2
(Raes
et
al.,
2000;
Vever
et
al.,
2000).
Sampling
times
varied
between
41
min
and
126
min,
and
air
volumes
filtered
lie
in
the
range
of
0.205
m
3
and
0.630
m
3
at
a
flow
rate
of
51
min
-1
for
each
filter
holder.
The
sam-
pling
inlet
has
a
50%
cut-off
at
approximately
2.5
itm
at
ambient
humidity.
Flights
included
several
patterns
to
sample
different
regions
of
the
lower
troposphere,
including
some
slowly
descending
profiles
to
characterize
a
vertically-
Tellus
52B
(2000),
2
AEROSOL
PARTICLE
CHEMICAL
CHARACTERISTICS
187
averaged
composition
in
flights
15
and
20,
constant-
altitude
box
patterns
to
provide
detailed
measure-
ments
of
the
boundary
layer
on
flights
11
and
16,
and
circles
at
several
altitudes
to
quantify
mixing
of
aerosol
in
flight
21.
Elemental
(EC)
and
organic
(OC)
carbon
were
identified
by
thermal-optical
analysis
(TOA),
described
by
Birch
and
Carey
(1996).
In
TOA,
the
aerosol-containing
quartz
fiber
filter
is
placed
in
an
oven
under
helium
atmosphere
and
is
heated
in
four
increasing
temperature
steps
to
a
final
temperature
of
725
°C.
Organic
carbon
volatilizes
completely
at
this
temperature
and
immediately
oxidizes
to
CO
2
on
MnO
2
,
and
then
reduces
to
methane.
A
flame
ionization
detector
measures
the
methane
concentration
evolved
by this
process.
For
elemental
carbon,
the
oven
is
cooled
to
850
°C
before
heating,
and
the
carrier
gas
is
changed
to
a
helium/oxygen
mixture
in
order
to
oxidize
all
remaining
carbon.
Given
sufficient
loading,
ele-
mental
carbon
can
be
determined
from
the
differ-
ence
of
total
and
organic
carbon
evolved
but
was
below
detection
by
thermal
decomposition
for
the
ACE-2
samples.
In
addition,
there
was
no
change
in
the
light
fraction
reflected
in
all
samples,
so
that
black
carbon
was
also
below
detection
by
the
simultaneous
optical
approach
used.
Ionic
species
(Na
+
,
Cl
-
,
and
SO4
-
)
were
ana-
lyzed
by
ion
chromatography
(IC)
(Dionex
2010i,
Dionex
Corp.),
after
extraction
of
the
filter
with
5
ml
of
methanol/water
(1
:10)
solution
in
an
ultrasonic
bath
for
at
least
6
hr
(Huebert
et
al.,
1996).
The
anion
analysis
was
performed
with
an
AS4
column
(Dionex
Corp.)
and
the
cation
ana-
lysis
with
a
CS12A
column
(Dionex
Corp.)
under
standard
conditions.
In
addition
to
ions,
elemental
species
were
measured
by
total
reflection
X-ray
Fluorescence
(TXRF),
after
extraction
of
the
filter
with
1
ml
hot
concentrated
nitric
acid
(Suprapure,
EM).
In
this
analysis,
10
µ1
from
each
sample
was
pipetted
onto
a
clean
Si-wafer
and
3
µI
of
a
1
µg
m1
-1
solution
of
Cr
was
added
as
internal
standard
for
quantification.
The
droplet
was
dried
under
vacuum
and
irradiated
by
vv-L(3
for
500
s
(Schmeling
and
Klockow,
1997;
Streli,
1997;
Klockenkamper,
1997).
In
addition
to
atmospheric
samples,
blank
filters
were
taken
before
and
after
each
flight
and
were
handled
with
the
same
protocol
as
the
sample
filters.
Values
of
blanks
for
each
flight
were
aver-
aged
and
subtracted
from
the
measured
aerosol
mass.
Detection
limits
for
each
species
for
the
three
methods
described
are
presented
in
Table
1.
Since
three
main
types
of
air
masses
are
typically
found
in
the
Canary
Islands
clean
marine
(flights
11
and
16
on
4
and
9
July,
respectively),
anthropogenically-influenced
(flight
21
on
18
July),
and
Saharan
dust-containing
(flights
15
and
20
on
8
and
17
July,
respectively)
we
will
focus
here
on
the
distinctive
features
of
the
size
distributions
and
chemical
compositions
of
each
of
these
types
during
ACE-2.
Section
3
describes
measured
chemical
compositions
and
particle
size
distributions
for
each
of
the
flights
studied.
Section
4
compares
our
results
with
other
meas-
urements
that
have
been
reported
in
the
literature
for
similar
and
contrasting
regions.
Chemical
com-
position
and
size
distribution
are
key
factors
in
determing
particle
optical
properties
including
refractive
indices
and
single
scatter
albedo.
Section
5
uses
the
measured
compositions
to
study
the
sensitivities
of
these
optical
properties
for
each
air
mass
type.
3.
Aerosol
characterization
In
Table
1,
chemical
concentrations
obtained
for
five
ACE-2
flights
(namely
flights
11, 15, 16,
20,
and
21)
are
shown.
Figs.
1
c,
d,
2c,
d,
and
3b
show
these
values
for
each
component
analyzed
in
each
flight.
In
order
to
compare
the
measured
composition
with
the
corresponding
particle
size
distribution,
we
have
shown
representative
size
distributions
measured
from
the
Automated
Classified
Aerosol
Detector
(ACAD,
Russell
et
al.,
1996b)
for
the
range
0.005
to
0.190
pm
dry
diameter
and
the
Passive
Cavity
Aerosol
Spectrometer
Probe
(PCASP,
Particle
Measuring
Systems,
Inc.)
for
the
range
0.1
to
3.0
µm
ambient
diameter
for
each
flight
in
Figs.
I
a,
b,
2a,
b,
and
3a
(Collins
et
al.,
2000).
To
describe
the
character-
istics
for
each
flight,
the
particle
size
distribution
has
been
classified
in
3
size
categories:
Aitken
mode
particles
comprising
those
particles
up
to
0.1
µm
dry
diameter,
accumulation
mode
particles
up
to
1.0
µm,
and
coarse
mode
particles
larger
than
1.0
pm.
In
general,
the
results
for
ionic
and
carbona-
ceous
aerosol
are
comparable
to
other
measure-
ments
made
during
ACE-2
aboard
the
R/V
Tellus
52B
(2000),
2
Table
1.
Chemical
composition
of
fine
aerosol
particles
measured
during
ACE-2
Pelican
flights
00
00
NSS
SO
4
2-
a
OC'
EC'
Cla
Na
Mg
b
Mn
b
Fe
b
Co
b
Ni
b
Cu
b
1.0
0.20
0.27
3.1
1.7a/0.03
b
0.3
0.00002
0.00005
0.00003
0.0001
0.00001
1.6
<
DL
+0.05
<
DL
1.1
<
DL
1.1
<
DL
<
DL
0.333
<
DL
<
DL
0.045
+0.06
±0.2
+0.002
+0.0008
<DL <DL
<DL
0.11b
5.3
0.180
2.95
0.045
0.372
<
DL
±0.01
±0.03
±0.001
±0.05
±0.0007
±0.001
6.5
<DL <DL
<DL
7.8a
<
DL
0.036
2.58
<
DL
<
DL
0.071
±0.4
+0.0008
+0.06
+0.0009
3.8
3.9
<
DL
0.29
b
<
DL
<
DL
0.851
<
DL
<
DL
0.108
±0.10
±0.02
±0.003
±0.001
Concentration
[µg/m
3
]
S0
4
2-
a
Detection
limits
[ig]
1.0
Clean
marine
cases
flight
11
(0.50
m
3
)
flight
16
(0.41
m
3
)
<
DL
Dust-containing
cases
flight
15
(0.34
m
3
)
flight
20
(0.38
m
3
)
8.6
+0.20
Anthropogenically-influenced
case
flight
21
(0.42
m
3
)
3.9
±0.08
-
IV
.
La
ONIIHNIHOS
'1,1
The
results
include
analyses
by
IC,
TOA,
and
TXRF.
Blank
values
indicate
that
a
suitable
sample
was
not
available
for
this
element's
analysis.
"<
DL"
indicates
that
no
value
is
reported
because
the
measured
value
of
the
sample
was
below
the
detection
limit
of
the
technique
used.
Detection
limits
are
reported
as
total
mass
of
the
chemical
species
required
for
an
individual
sample
for
analysis
by
(a)
ion
chromatography
(IC),
(b)
total
reflection
X-ray
fluorescence
(TXRF),
and
(c)
thermal-optical
analysis
(TOA).
The
corresponding
atmospheric
loadings
can
be
found
by
dividing
the
volume
of
standard
air
filtered
for
each
flight,
which
varied
from
0.2
m
3
to
1.9
m
3
during
ACE-2.
Z
l000
) u
s
sna
u
d)
12.00
11.00
10.00
E
9.00
T.
8.00
a
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
.4
7-
2
C
0
C.)
AEROSOL
PARTICLE
CHEMICAL
CHARACTERISTICS
189
a)
b)
_11111
1
1
1
1
11111
11
1
1
11111
,_
111111
1
1 1
1
11111
1
1
1
1
11111
1
10
4
10
3
10
4
10
3
a.
a)
10
2
10
2
10
1
10
1
--
Flight
11
Average
-
Flight
16
Average
1111
6
2
4
6
2
4
6
2
4
6
2
4
6
2
4
6
2
4
10
100
1000
10
100
1000
Dry
Diameter
(nm)
Dry
Diameter
(nm)
0
12.00
-
11.00
7
10.00
-
E
9.00
-
8.00
-
7.00
-
6.00
-
5.00
-
4.00
-
3.00
-
2.00
-
1
.
0
0
.
0.00
OC
nss
CI
Na
Mn
Fe
Co
Ni
Cu
sulfate
1'1
OC
nss
CI
Na
Mn
Fe
Co
Ni
Cu
sulfate
rn
0
47.
CO
0
Fig.
1.
Aerosol
size
distributions
and
chemical
compositions
for
the
two
clean
cases
studied,
flights
11
and
16.
Size
distributions
from
integrated
ACAD
and
PCASP
measurements
are
shown
in
(a)
and
(b)
for
flights
11
and
16.
The
solid
line
shows
the
average
value
during
the
filter
sampling
period
for
the
flight,
and
the
dotted
lines
give
the
range
of
values
measured
showing
the
minimum
and
maximum
measured
concentrations
with
size.
Chemical
composition
data
are
shown
in
(c)
and
(d)
for
flights
11
and
16,
respectively.
Filled
bars
indicate
measured
values,
and
empty
bars
are
shown
to
indicate
detection
limits.
For
species
where
no
bar
is
shown,
both
values
are
too
small
to
appear
in
the
range
of
the
graph
and
can
be
found
in
Table
1.
Vodyanitskiy
and
from
land-based
platforms
at
Sagres,
Portugal,
and
Punta
del
Hidalgo,
Tenerife
(Quinn
et
al.,
2000).
The
low
values
of
both
organic
and
black
carbon
in
the
clean
air
masses
were
below
detection
in
many
cases
for
the
air-
borne
and
ship-based sampling
and
present
below
0.1
itg
m
-3
from
ground
based
data
(Quinn
et
al.,
2000).
In
the
more
polluted
samples,
both
airborne
and
surface
samples
measured
concentrations
above
I
ptg
m
-3
,
although
the
vertically-averaged
flight
21
sample
has
the
highest
organic
carbon
concentration
at
3.9
ptg
m
-3
,
as
would
be
consist-
ent
with
higher
carbonaceous
particle
loadings
aloft.
The
Na
concentration
measured
varied
between
0.1
and
0.4
itg
m
-3
for
air,
ship
and
land
measurements
for
all
cases
except
flight
20,
where
the
masses
of
sulfates,
sodium
and
iron
were
each
above
2µg
m
-3
.
The
clean
non-seasalt
sulfate
ion
Tellus
52B
(2000),
2
d)
c)
Co
ncen
tra
t
ion
Gtg
m
Na
Mn
Fe
Co
Ni
Cu
12.00
11.00
-
10.00
9.00
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
OC
nss
CI
sulfate
12.00
11.00-
e
10.00-
E
9.00-
a)
8.00
c
7.00
6.00
1
2
5.00
a)
4.00-
o
O
3.00
0
2.00
1.00
0.00
OC
nss
CI
Na
Mn
Fe
Co
Ni
Cu
sulfate
190
M.
SCHMELING
ET
AL.
a)
10
4
0
3
r..
1
1111
1
1
1
1
1
1111
1
1
11111
E
o.
rn
0
-
Ts
10
2
-a
.
.
10
1
Flight
15
Boundary
Layer
Flight
15
Free
Troposphere
1
1111
1 1
1
11111
1
1
1
1
11111
1 1 1
I
•::
Flight
20
Boundary
Layer
--
Flight
20
Free
Troposphere
..JI
b)
1
0
4
0
3
10
1
6
10
100
2
4
6
2
4
x1000
2
4
Dry
Diameter
(nm)
6
2
4
6
2
4
6
2
4
10
100
1000
Dry
Diameter
(nm)
Fig.
2.
Aerosol
size
distributions
and
chemical
compositions
for
the
two
dust
cases
studied,
flights
15
and
20.
Size
distributions
from
integrated
ACAD
and
PCASP
measurements
are
shown
in
(a)
and
(b)
for
flights
15
and
20,
respectively,
where
thick
lines
represent
the
average
size
distribution
above
the
boundary
layer
in
the
free
troposphere
and
thin
lines
represent
the
average
size
distribution
in
the
boundary
layer.
The
solid
lines
show
the
average
values
during
the
filter
sampling
period
for
the
two
parts
of
the
flight,
and
the
dotted
lines
give
the
range
of
values
measured
showing
the
minimum
and
maximum
concentrations
with
size.
Chemical
composition
data
are
shown
in
(c)
and
(d)
for
flights
15
and
20,
respectively.
Filled
bars
indicate
measured
values,
and
empty
bars
are
shown
to
indicate
detection
limits.
For
species
where
no
bar
is
shown,
both
values
are
too
small
to
appear
in
the
range
of
the
graph
and
can
be
found
in
Table
1.
concentrations
were
too
low
to
be
detected
on
flight
16,
which
is
consistent
with
concentrations
found
on
the
other
three
platforms,
as
were
pol-
luted
non-seasalt
sulfate
concentrations
of
3.8
and
6.5
ptg
nr
3
on
flights
20
and
21
consistent
with
the
range
of
4.5
to
11
ptg
nr
3
reported
for
polluted
conditions
on
Punta
del
Hidalgo
and
Sagres
(Quinn
et
al.,
2000).
3.1.
Clean
marine
air
During
flights
11
(4
July,
JDT
185)
and
16
(9
July,
JDT
190),
the
Pelican
aircraft
sampled
clean
marine
air
that
had
spent
several
days
over
the
North
Atlantic
Ocean.
Northwesterly
winds
had
transported
these
clean
air
masses
to
the
Canary
Islands.
After
transport
times
of
several
Tellus
52B
(2000),
2
AEROSOL
PARTICLE
CHEMICAL
CHARACTERISTICS
191
a)
E
0
rn
0
10
2
z
-a
6
2
4
6
2
4
6
2
4
10
100
1000
Dry
Diameter
(nm)
b)
12.00
11.00
10.00
E
9.00
c.
8.00
c
7.00
T.
,
6.00
5.00
u
4.00
o
3.00
2.00
1.00
0.00
OC
nss
CI
Na
Mn
Fe
Co
NI
Cu
sulfate
Fig.
3.
Aerosol
size
distributions
and
chemical
composi-
tions
for
the
anthropogenically-influenced
case
studied,
flight
21.
Size
distributions
from
integrated
ACAD
and
PCASP
measurements
are
shown
in
(a)
for
flight
21.
The
solid
line
shows
the
average
value
during
the
filter
sam-
pling
period
for
the
flight,
and
the
dotted
lines
give
the
range
of
values
measured
showing
the
minimum
and
maximum
concentrations
with
size.
Chemical
composi-
tion
data
are
shown
in
(b)
for
flight
21.
Filled
bars
indicate
measured
values,
and
empty
bars
are
shown
to
indicate
detection
limits.
For
species
where
no
bar
is
shown,
both
values
are
too
small
to
appear
in
the
range
of
the
graph
and
can
be
found
in
Table
1.
days,
the
air
was
well-mixed
and
contained
mainly
sea
salt
and
marine
components.
The
sample
collected
at
a
single
altitude
in
the
boundary
layer
was
likely
to
be
representative
of
the
boundary
layer.
Figs.
la,
b
shows
average
particle
size
distri-
butions
for
each
of
the
2
clean
flights.
The
two
cases
show
clearly
different
concentra-
tions,
indicating
two
types
of
marine
air
conditions.
Flight
11
is
a
case
with
clean
marine
air
character-
ized
by
back
trajectories
that
were
consistently
over
the
ocean
for
5
days
before
the
flight.
The
large
number
of
particles
near
20
nm
dry
diameter
indi-
cates
a
source
of
recently-produced
particles,
although
we
do
not
have
sufficiently
size-resolved
chemical
composition
measurements
to
distinguish
among
marine,
biogenic,
and
anthropogenic
poten-
tial
sources.
However,
in
the
case
of
flight
16,
the
chemical
composition
indicates
that
other
species
were
present,
including
Fe
and
Cu,
which
are
not
typically
found
in
clean
marine
air
(Allen
et
al.,
1997;
Brechtel
et
al.,
1998).
Back
trajectories
for
flight
16
support
this
assumption
as
back
trajector-
ies
at
3
altitudes
from
different
altitudes
in
the
lower
troposphere
lead
to
North
Africa
and
one
to
Europe.
Consequently,
we
believe
that
the
air
mass
sampled
in
flight
16
included
some
dust-containing
or
anthropogenically-influenced
air
masses.
Size
distributions
show
that
high
concentrations
(20
cm
-3
)
are
present
in
the
supermicron
part
of
the
measured
distribution.
These
particles
may
be
derived
from
oceanic
(sea
salt)
or
dust
sources.
Chemical
analysis
shows
that
there
are
small
amounts
of
dust
constituents
including
Fe
and
Cu
present,
but
these
species
may
also
be
of
anthropo-
genic
origin.
Since
this
sample
was
collected
entirely
in
the
boundary
layer
and
dust
particles
are
usually
found
predominantly
in
the
free
tropo-
sphere,
these
tracers
are
likely
to
indicate
anthro-
pogenically-influenced
air
from
Europe.
These
characteristics
are
consistent
with
an
air
mass
of
mixed
origin
with
some
recent
sources
of
anthro-
pogenic
particles
influencing
a
predominantly
marine
particle
size
distribution.
3.2.
Dust-laden
air
The
back
trajectories
indicate
that
part
of
the
air
mass
sampled
on
flights
15
and
20
(7
and
17
July,
or
JDT
189
and
199,
respectively)
originated
from
North
Africa.
In
addition
to
oceanic
sources,
we
expect
the
aerosol
particles
to
consist
mainly
of
10
4
10
3
1
I
1
1
11111
1
1 1
1
11111
1
-
-
Flight
21
Average
-
111
Tellus
52B
(2000),
2
192
M.
SCHMELING
ET
AL.
mineral
dust
derived
from
the
Sahara
Desert.
In
order
to
obtain
more
detailed
information
about
the
vertical
distribution
of
particulates
in
this
air
mass,
the
aircraft
performed
flights
in
spirals
down
from
3000
m
altitude
(in
the
free
troposphere)
to
60
m
altitude
(in
the
boundary
layer).
In
Fig.
2a,
b
the
average
size
distributions
of
particles
in
the
part
of
the
free
troposphere
sampled
and
in
the
boundary
layer
are
shown.
During
aircraft
profiles,
relative
humidity
above
1500
m
altitude
was
less
than
10%
and
increased
to
70%
in
the
boundary
layer.
The
submicron
size
distribution
above
the
boundary
layer
is
dominated
by
a
single
mode,
typical
of
aerosol
size
distributions
observed
in
the
free
tropo-
sphere
during
ACE-2
(Collins et
al.,
2000).
For
the
particle
sizes
between
0.9
gm
and
1.8
gm
dry
dia-
meter,
where
dust
particles
have
been
identified
previously
(Li-Jones
et
al.,
1998),
the
particle
con-
centrations
in
these
two
cases
are
higher
by
a
factor
of
5
or
more
at
concentrations
of
10
to
30
cm
-3
than
in
the
non-dust
cases.
Once
in
the
boundary
layer
the
particle
size
distribution
changes
and
the
particle
number
increases
in
the
fine
mode
by
about
an
order
of
magnitude.
This
change
is
particularly
significant
for
flight
20,
in
which
some
of
the
bound-
ary
layer
size
distributions
may
have
been
influenced
by
anthropogenic
particles
from
Tenerife.
In
the
boundary
layer,
the
particle
source
may
be
the
ocean
surface,
producing
sea
salt
particles
with
a
peak
in
concentration
at
about
2.0
gm
due
to
bubbles
bursting
at
the
sea
surface.
Chemical
com-
positions
in
flights
15
and
20
are
very
different
from
the
clean
cases
and
show
high
amounts
of
trace
metals
(Mn,
Fe,
Co,
Ni,
Cu).
However,
in
flight
20
Cu
has
a
concentration
of
0.071
gg
m
-3
.
This
high
amount
of
Cu
is
more
indicative
of
anthropogenic
sources
than
of
dust,
which
is
consistent
with
the
high
nss
sulfate
concentration
of
6.5
gg
m
-3
during
this
flight.
Back
trajectories
also
indicate
an
air
mass
of
mixed
origin
with
some
back
trajectories
in
this
case
originating
over
the
European
continent.
The
chemical
composition
for
flight
15
shows
high
concentrations
of
trace
metals
and
Na.
High
amounts
of
Fe
(2.95
gg
m
-3
)
and
Mg
(5.4
gg
m
-3
)
provide
strong
indicators
for
crustal
material
that
has
formed
dust.
Al
and
Zn
were
measured
as
well
but
showed
high
blank
values
resulting
from
hand-
ling
and
sampling
artifacts
(the
aircraft
inlet
was
made
of
Al),
such
that
no
concentrations
for
these
elements
are
reported
here.
3.3.
Anthropogenically-influenced
air
Some
air
masses
are
transported
by
north-
easterly
or
easterly
winds
from
the
European
continent
to
the
Canary
Islands.
Their
particle
size
distributions
and
compositions
are
expected
to
be
strongly
influenced
by
anthropogenic
emis-
sions.
Back
trajectories
for
flight
21
on
18
July
(JDT
200)
indicate
that
the
observed
air
mass
originated
in
southwestern
Europe.
The
flight
pattern
included
two
circles
at
two
different
alti-
tudes
(180
m
and
60
m)
during
the
aerosol
sampling.
In
Fig.
3,
the
average
size
distribution
is
shown.
Particle
size
distributions
with
high
concentrations
typical
of
anthropogenically-influenced
particles
were
found
including
concentrations
of
more
than
5000
cm
-3
and
a
peak
dry
diameter
of
0.06
gm,
as
shown
in
Fig.
3a.
Many
of
the
particles
are
present
in
the
accumulation
mode,
which
is
typical
of
anthropogenically-influenced
air
masses
(Hoppel
et
al.,
1990;
Russell
et
al.
1996a).
There
are
also
some
larger
particles,
which
may
originate
from
either
dust
or
sea
salt.
Since
these
measure-
ments
were
made
at
low
altitudes
in
the
boundary
layer,
coarse
particles
in
these
size
distributions
are
dominated
by
sea
salt.
The
chemical
composition
measured
on
this
flight
is
also
consistent
with
an
anthropogenically-
influenced
air
mass
with
a
small
amount
of
Na
(0.29
gg
m
-3
)
and
a
substantial
mass
of
organic
carbon
(3.89
gg
m
-3
),
but
only
minor
concentra-
tions
of
Fe
(less
than
1
gg
m
-3
)
and
Cu
(less
than
0.2
gg
m
-3
).
Additional
evidence
for
the
anthropo-
genic
origin
of
this
air
mass
is
provided
by
the
high
nss
sulfate
concentration
(3.85
gg
m
-3
).
4.
Comparison
with
previous
atmospheric
measurements
Table
2
shows
chemical
compositions
of
aerosol
particles
in
several
previous
studies
of
ionic
and
trace
metal
compositions.
Table
2
indicates
that
our
results
are
in
good
agreement
with
results
from
other
field
measurements,
especially
with
those
performed
in
the
northern
Atlantic
Ocean.
Na
concentration
varies
significantly
with
wind
conditions
in
the
boundary
layer,
with
our
results
falling
slightly
below
other
reported
values.
Most
Na
concentrations
in
studies
with
reported
surface
wind
speeds
were
measured
on
ship
or
land-based
Tellus
52B
(2000),
2
AEROSOL
PARTICLE
CHEMICAL
CHARACTERISTICS
193
Table
2.
Concentrations
of
chemical
compounds
in
fine
aerosol
particles
at
a
series
of
locations
illustrating
clean,
dust-containing
and
anthropogenically-influenced
conditions
Concentration
[µ,g/m1
Na
Nss
sulfate
OC
Mn
Fe
Co
Ni
Cu
ACE-2
Flight
15
0.11
<
DL
<
DL
0.18
2.95
0.045
0.372
<
DL
ACE-2
Flight
16
0.752
<
DL
1.1
<DL
0.333
<
DL
<
DL
0.045
ACE-2
Flight
20
7.8
6.5
<
DL
0.036
2.58
<
DL
<
DL
0.071
ACE-2
Flight
21
0.29
3.8
3.9
<DL
0.851
<
DL
<
DL
0.108
Bermuda
3.97a
2.65
h
Cape
Grim
3.3a
Hawaii
4.76a
Heimaey
7.52a
Mace
Head
6.12a
1.94
h
Barbados
1.15`
S.
Ocean
0.033
d
NE.
Atlantic
5.16`
Brisbane
40
f
0.092
f
2.6f
Boston
0.088g
33g
Long
Beach
39g
0.148g
5.1g
Kashima
0.02"
4.1h
Aberg
0.002'
0.037'
0.0037' 0.0002'
Xiang
0.74'
43i
0.52'
0.24' 0.39'
Mexico
City
0.07
k
1.126k
0.0018
k
0.084
k
Fuertaventura
0.31'
2.77'
aGong
et
al.
(1997);
b
Husar
et
al.
(1997);
cProspero
et
al.
(1993);
d
Huebert
et
al.
(1998);
`Hegg
et
al.
(1993);
1
Chan
et
al.
(1997);
gThursdon
and
Spengler
(1985);
h
Okamoto
et
al.
(1986);
'Schmeling
(1997);
'Zhang
et
al.
(1993);
k
Miranda
et
al.
(1994);
'Bergametti
et
al.
(1989).
stations,
located
between
5
and
20
m
above
the
sea
surface,
causing
them
to
reflect
higher
concen-
trations
than
the
aircraft
measurements
(collected
above
60
m)
due
to
the
rapid
settling
of
coarse
particles
which
results
in
steep
Na
concentration
gradients
above
the
ocean
surface.
In
addition,
Na
(and
the
associated
Cl)
may
be
truncated
by
the
upper
size
cut-off
of
our
inlet
since
significant
sea
salt
mass
can
occur
above
2.5
gm
ambient
dia-
meter
(Huebert
et
al.,
1998).
Fig.
4
shows
trace
metal
concentrations
for
flights
15,
16,
20,
and
21
with
results
from
several
of
the
studies
noted
in
Table
2.
The
Shapoutou
samples
were
collected
in
China
during
dust
storm
events
and
non-dust
storm
events
from
the
Gobi
Desert
(Zhang
et
al.
1993;
Zhang
et
al.,
1998).
The
Fuerteventura
sample
shows
an
average
value
of
samples
taken
with
back
trajectories
from
the
Sahara
Desert
(Bergametti
et
al.
1989).
The
Aberg
study
sampled
in
a
non-urban
area
in
western
Germany
(Schmeling,
1997).
Concentrations
in
these
two
desert
regions
have
different
elemental
compositions,
but
variability
within
each
region
makes
it
difficult
to
establish
definitive
character-
istics
for
each
without
more
measurements.
Shapoutou
is
located
in
central
China
and
is
exposed
directly
to
desert-derived
aerosols,
whereas
air
masses
reaching
the
Canary
Islands
from
North
Africa
have
already
traveled
over
the
ocean
and
have
mixed
with
marine
air.
There
are
few
industries
in
the
Canary
Islands,
so
the
anthro-
pogenic
contribution to
the
aerosol
is
limited
compared
to
central
China.
Given
these
limita-
tions
we
note
that
dust
particles
from
the
Sahara
Desert
contain
more
Ni
and
Zn,
but
less
Fe,
Co,
Cu
and
Mn
than
samples
from
China.
These
species
are
also
present
in
the
absence
of
dust
storms
in
this
region,
but
most
of
the
elemental
concentrations
(with
the
exception
of
Zn)
are
much
lower
than
during
the
dust
storm
periods.
5.
Chemical
effects
on
aerosol
optical
properties
For
each
type
of
aerosol
studied,
we
want
to
assess
how
much
the
chemical
composition
of
aerosol
affects
the
aerosol
optical
properties.
To
Tellus
52B
(2000),
2
194
M.
SCHMELING
ET
AL.
10
0.001
Flight
15
Ea
Flight
20
Aberg
Shapoutou
(dust)
ono
Shapoutou
(non
dust)
Fuerteventura
Mn
Fe
Co
Ni
CU
rn
0
0.1
0
0.01
Fig.
4.
Trace
metal
concentrations
illustrating
concentration
differences
in
anthropogenically
and
dust-influenced
aerosol
samples
from
ACE-2
and
other
measurement
regions.
Flights
15
(black
bars)
and
20
(diagonally-striped
bars)
are
shown
from
the
ACE-2
dust
cases.
A
continental
background
sample
in
a
non-urban
area
of
a
very
industrialized
region
is
illustrated
by
the
Aberg
sample
(horizontally-striped bars)
from
Schmeling
(1997).
Data
from
Zhang
et
al.
(1993;
1998)
illustrate
concentrations
measured
in
Shapoutou
during
dust
storms
(dotted
bars)
and
at
other
times
(vertically-striped
bars).
The
Fuerteventura
sample
(Bergametti
et
al.,
1989)
was
taken
during
a
dust
event
traced
back
to
the
Sahara
Desert
(brick-patterned
bars).
answer
this
question
of
the
role
of
chemistry
in
determining
the
scattering
and
absorption
of
light
by
particles
in
these
five
cases
studied,
we
compare
the
sensitivity
of
modelled
optical
properties
to
differences
in
particle
composition
and
mixing
state.
In
order
to
describe
aerosol
optical
properties,
we
need
to
estimate
the
mass
of
volatile
species
associated
with
the
mass
collected
on
the
filter,
the
fraction
of
mass
not
identified,
and
the
degree
of
mixing
in
the
aerosol
population.
Since
the
mass
of
species
identified
does
not
account
for
the
total
mass
present,
our
approach
relies
on
a
model
for
the
aerosol
composition
which
is
simplified
by
assumptions
about
the
unmeasured
species.
We
use
the
resulting
inferred
ambient
composition
to
calculate
the
particle
refractive
index.
From
meas-
ured
values
of
the
ambient
relative
humidity,
we
have
estimated
the
fraction
of
water
present
in
the
condensed
phase
(Howell
and
Huebert,
1998).
We
have
used
the
parameterization
proposed
by
Howell
and
Huebert
(1998)
for
the
general
categories
of
marine
and
anthropogenically-influ-
enced
aerosols
to
estimate
water
uptake
by
the
measured
dry
particle
size
distribution.
Their
cor-
relations
predict
diameter
changes
of
factors
from
1.4
to
1.9
corresponding
to
water
mass
fractions
varying
from
30%
to
90%
at
the
boundary
layer
humidity
of
80%.
In
dust
cases,
filter
samples
were
dominated
by
the
high
aerosol
accumulation-
mode
mass
in
the
dust
layers
sampled.
For
this
reason,
optical
properties
for
particle
size
distribu-
tions
at
low
relative
humidity
that
were
character-
istic
of
free
tropospheric
dust
layers
have
been
assumed
to
take
up
a
negligible
mass
of
water
(d'Almeida
and
Schutz,
1983).
Using
measured
compositions
of
chemical
species
as
indicators
of
the
fractions
of
sea
salt,
dust,
organic
carbon,
elemental
carbon,
and
sulfate,
the
remaining
frac-
tion
of
unidentified
dry
mass
has
been
assumed
to
be
other
soluble
ions
and
has
been
modeled
using
the
refractive
index
estimated
for
dissolved
nitrate
(d'Almeida
et
al.,
1991).
To
calculate
the
aerosol
optical
properties,
we
have
converted
the
chemical
compositions
in
Table
1
into
volume
fractions
of
several
aerosol
types:
SO4
-
are
categorized
as
"sulfate,"
OC
and
EC
are
each
separate
types,
Cl
and
Na
are
grouped
together
as
"sea
salt,"
and
Mg,
Mn,
Fe,
Co,
Ni,
and
Cu
are
grouped
together
as
"dust."
For
each
Tellus
52B
(2000),
2
AEROSOL
PARTICLE
CHEMICAL
CHARACTERISTICS
195
flight,
the
sum
of
the
mass
of
all
species
in
each
aerosol
type
gives
the
mass
of
that
type.
For
a
given
flight,
the
amount
by
which
the
total
mass
calculated
from
the
physical
size
distribution
exceeds
the
measured
dry
aerosol
mass
is
assigned
to
the
"other"
soluble
species
category,
since
the
refractive
indices
for
several
organic
and
inorganic
(such
as
nitrate)
species
are
similar.
Since
the
scattering
properties
predicted
for
these
soluble
species
are
similar
to
water
and
sulfate
and
their
predicted
absorption
is
negligible,
the
results
are
relatively
insensitive
to
the
fraction
of
this
"neut-
ral"
species
present
in
particles
for
these
calcula-
tions.
For
the
non-dust
cases,
the
difference
between
the
total
dry
aerosol
mass
and
the
wet
mass
derived
from
the
measured
ambient
size
distributions
(Collins
et
al.,
2000;
Howell
and
Huebert,
1998)
(particles
less
than
2.5
gm
dia-
meter)
is
assigned
to
water.
In
the
absence
of
detailed
information
on
the
vapor
concentration
of
HC1,
we
do
not
estimate
here
the
potential
contribution
of
this
species
to
the
particles.
For
density,
the
following
approximate
values
were
assumed
in
the
calculation
in
order
to
convert
the
measured
mass
fraction
into
a
volume
fraction
for
each
aerosol
type:
sulfate
1.78
g
cm
-3
,
OC
1.9
g
cm
-3
,
EC
2.3
g
cm
-3
,
sea
salt
2.2
g
cm
-3
,
dust
2.5
g
cm
-3
,
other
soluble
species
1.78
g
cm
-3
,
and
water
1.0
g
cm
-3
.
The
resulting
approximate
models
for
the
different
aerosols
measured
in
the
five
cases
studied
here
are
shown
in
Table
3.
Using
the
volume
percent
of
each
species
in
each
flight
given
in
Table
3,
we
have
calculated
internally-mixed
average
refractive
indices
for
fine
particles
with
spectrally-resolved
refractive
indices
for
each
component
and
assuming
perfect
internal
mixing
(d'Almeida
et
al.,
1991;
Weast
et
al.,
1985-1986;
Hale
and
Querry,
1973;
Toon
et
al.,
1976).
Refractive
indices
of
non-absorbing
species
are
mixed
assuming
volume-weighted
mixing
(Hanel,
1976)
and
absorbing
species
(OC,
EC,
and
dust)
are
treated
with
Maxwell-Garnett
theory
(Bohren
and
Huffman,
1983).
For
comparison,
we
have also
computed
the
optical
properties
assuming
the
aerosol
types
were
externally
mixed
rather
than
internally
mixed,
where
the
number
concentration
for
each
aerosol
type
with
respect
to
the
total
size
distribution
was
assumed
to
be
proportional
to
its
volume
fraction.
Fig.
5a,
c
shows
that
refractive
indices
for
clean
and
anthropogenically-influenced
air
are
practic-
ally
identical,
but
for
the
dust
cases
refractive
indices
vary
significantly.
The
similarity
in
clean
and
anthropogenically-influenced
cases
results
from
the
high
water
content
and
the
lack
of
elemental
carbon
(a
strong
absorber)
in
all
samples.
The
real
part
of
the
refractive
index
is
generally
higher
for
dust
cases
than
for
other
cases,
particularly
for
wavelengths
below
2.0
gm,
where
dust
values
exceed
1.5
compared
to
non-
dust
cases
with
values
near
1.3
(the
value
for
liquid
water).
The
imaginary
part
of
the
refractive
index
shows
the
most
significant
differences
at
wave-
lengths
above
2.5
gm,
where
it
rises
for
the
clean
and
anthropogenically-influenced
cases
to
above
0.2
for
the
liquid
water
peak.
Refractive
indices
are
then
used
with
particle
size
distributions
to
calculate
single
scatter
albedo
for
the
particles
below
2.5
gm
with
a
Mie
scat-
tering
algorithm
(Bohren
and
Huffman,
1983).
Single
scatter
albedos
for
clean
and
anthropo-
genically-influenced
cases
show
similar
trends
in
Fig.
6,
corresponding
to
values
near
unity
and
almost
no
wavelength
dependence
below
2.5
gm.
Table
3.
Fraction
of
particle
volume
(given
as
a
%
of
total
submicron
ambient
particle
volume)
of
each
species
used
to
provide
representative
models
of
the
particle
composition
for
each
flight
Flight
Sulfate
OC
EC
Seasalt
Dust
Other
Water
Clean
marine
cases
11
0.0
4.1
0.0 0.0
0.0
8.9
87.0
16
1.8
2.7
0.0
3.9
0.7
20.0
70.8
Dust-containing
cases
15
0.0
0.0
0.0
0.8
79.2
20.0
0.0
20
42.7
0.0
0.0
17.9
29.3
10.1
0.0
Anthropogenically-influenced
case
21
22.2
10.4
0.0
0.7
2.0
0.3
64.5
Tellus
52B
(2000),
2
Re
frac
t
iv
e
In
dex
(rea
l)
1.2
1.1
1.6
1.5
1.4
1.3
00
I
'
I
-
Flight
11
- -
Flight
15
Flight
16
-
-
Flight
20
-
-
Flight
21
I
1.0
2.0
1.6
1.5
1.4
1.3
1.2
3.0
Re
frac
t
ive
In
dex
(
rea
l)
1.1
0 0
1.0
I I
-
Flight
11
EC
- -
Flight
15
EC
Flight
16
EC
- -
Flight
20
EC
-
• •
Flight
21
EC
1,1
2.0
3.0
0.25
0.00
0
0
-
-
1.0
2.0
3.0
-
Flight
11
EC
- -
Flight
15
EC
Flight
16
EC
- -
Flight
20
EC
-
• •
Flight
21
EC
Re
frac
t
ive
In
dex
(
imag
inary
)
0.20
0.15
0.10
0.05
196
M.
SCHMELING
ET
AL.
a)
b)
Wavelength
(µm)
Wavelength
(µm)
c)
d)
0.25
I I
I
-
Flight
11
-
-
Flight
15
0.20
Flight
16
- -
Flight
20
-
Flight
21
0.15
0.10
0.05
0.00
L
0
1.0
2.0
3.0
0
Wavelength
(Rni)
Wavelength
(gm)
Fig.
5.
Refractive
indices
are
shown
for
the
5
case
studies
for
conditions
of
clean,
dust-laden,
and
anthropogenically-
influenced
aerosol
flights.
Real
(a,
b)
and
imaginary
(c,
d)
refractive
indices
are
shown
for
clean
cases
-
flights
11
(solid
line)
and
16
(short-dashed
line),
dust-laden
cases
-
flights
15(dotted
line)
and
20
(medium-dashed
line),
and
an
anthropogenically-influenced
case
-
flight
21
(long-dashed
line).
Panels
(a)
and
(c)
show
the
refractive
indices
for
the
measured
volume
composition,
and
panels
(b)
and
(d)
show
the
results
calculated
by
assuming
elemental
carbon
("EC")
to
be
present
at
the
detection
limit.
Re
frac
tive
In
de
x
(
imag
inary
)
Because
of
the
absorbing
properties
of
dust,
dust
cases
show
lower
single
scatter
albedos
than
non-
dust
cases
below
0.5
gm,
reaching
as
low
as
0.8.
However,
above
2.5
gm
where
the
imaginary
part
of
the
refractive
index
is
high
in
non-dust
cases
and
water
begins
to
absorb
strongly,
the
single
scatter
albedo
decreases
drastically
to
0.2
and
below.
Similarities
in
the
scattering
properties
of
clean
and
anthropogenically-influenced
cases
are
an
expected
consequence
of
the
similarities
in
their
components,
especially
for
the
water-soluble
species
whose
optical
properties
are
quite
similar.
Differences
in
composition
obtained
from
the
limited
number
of
samples
here
suggest
few
composition-dependent
variations
in
aerosol
Tellus
52B
(2000),
2
Sing
le
Sca
tte
r
Albe
do
0.8
0.6
0.4
0.2
0.0
0
0
1.0
2.0
3.0
1.0
'
I
'
1
'
I
-
Flight
11
- -
Flight
16
Flight
11
EC
-
-
Flight
16
EC
I I
Sing
le
Sc
a
tte
r
Albe
do
0.8
0.6
0.4
0.2
0.0
0 0
1.0
2.0
3.0
1.0
1
'
1
-
Flight
11
- -
Flight
16
Flight
11
EC
-
-
Flight
16
EC
ill
1
Sing
le
Sca
tte
r
Albe
do
1.0
0.8
0.6
0.4
0.2
0.0
'
1
'
1 1
-
Flight15
- -
Flight
20
-
Flight
15
EC
- -
Flight
20
EC
I
I
e)
S
ing
le
Sc
a
tter
Albe
do
1'
I
.............
-
-
Flight
21
• •
Flight
21
EC
1
1
AEROSOL
PARTICLE
CHEMICAL
CHARACTERISTICS
197
a)
b)
S
ing
le
Sca
tte
r
Albe
do
Wavelength
(pm)
c)
d)
' I
r
••••
-
-
Flight
15
--
Flight
20
-
Flight
15
EC
-
-
Flight
20
EC
1
I
00
1.0
2.0
3.0
Wavelength
(p.m)
Wavelength
(pm)
1.0
0.8
0.6
0.4
0.2
0.0
0 0
1.0
2.0
3.0
Wavelength
(pm)
Sing
le
Sca
t
te
r
Albe
do
fl
1.0
0.8
0.6
0.4
0.2
0.0
0 0
1.0
2.0
3.0
Wavelength
(pm)
1.0
0.8
-
0.6
-
0.4
-
0.2
--
Flight
21
- -
Hight
21
EC
0.0
1
I
0
0
1.0
2.0
3.0
Wavelength
(pm)
Fig.
6.
Single
scatter
albedo
values
are
shown
for
the
5
case
studies
for
conditions
of
clean,
dust-laden,
and
anthropo-
genically-influenced
aerosol.
Clean
cases
during
flights
11
and
16
are
shown
assuming
an
internal
mixture
in
panel
(a)
and
assuming
an
external
mixture
in
panel
(b).
The
dust
cases
for
flights
15
and
20
are
shown
assuming
an
internal
mixture
in
panel
(c)
and
assuming
an
external
mixture
in
panel
(d).
The
anthropogenically-influenced
case
during
flight
21
is
shown
assuming
an
internal
mixture
in
(e)
and
assuming
an
external
mixture
in
(f
).
For
each
case,
the
annotation
"EC"
denotes
the
single
scatter
albedo
with
the
volume
fraction
of
the
particle
composition
modified
to
include
an
elemental
carbon
concentration
at
the
detection
limit.
light-absorbing
properties
for
the
cases
that
were
not
dust-influenced.
However,
since
sampling
con-
straints
resulted
in
detection
limits
of
between
1%
and
3%
elemental
carbon,
this
apparent
similarity
may
have
resulted
from
measurement
uncertainty.
To
study
the
sensitivity
of
the
calculated
aerosol
properties
to
this
uncertainty,
we
have
used
an
absorbing
carbon
content
of
0.2%
for
clean
and
Tellus
52B
(2000),
2
198
M.
SCHMELING
ET
AL.
2%
for
anthropogenically-influenced
air
masses,
with
the
resulting
calculated
optical
properties
shown
in
the
curves
marked
"EC"
in
Figs.
5,
6.
These
results
illustrate
that
there
is
only
a
modest
difference
in
imaginary
refractive
index
and
single
scatter
albedo
in
the
clean
marine
and
anthropo-
genically-influenced
cases.
The
imaginary
refract-
ive
index
does
increase
by
an
order
of
magnitude
due
to
the
addition
of
the
absorbing
EC,
and
the
resulting
impact
on
the
single
scatter
albedo
is
a
decrease
from
0.99
to
0.95.
Aerosol
optical
proper-
ties
in
dust
cases
differ
greatly
with
the
amount
of
absorbing
species
present,
varying
with
measured
fractions
of
sulfate-to-mineral
dust
and
with
the
shape
of
the
particle
size
distributions.
Particle
number
concentration
for
dust
cases
also
changes
markedly
at
the
boundary
layer
temperature
inver-
sion,
although
aircraft-based
data
on
the
associ-
ated
composition
changes
were
not
available.
Direct
measurements
of
in
situ
optical
proper-
ties
were
measured
aboard
the
Pelican
with
a
three-wavelength
nephelometer
for
the
two
cases
showing
the
largest
variation
in
single
scatter
albedo,
namely
the
dust-containing
flights
15
and
20
(Ostrom
and
Noone,
2000).
Our
estimates
of
the
550-nm
single
scattering
albedo
using
intern-
ally-mixed,
measured-composition
(given
in
Table
3)
calculations
tend
to
exceed
their
values
for
the
dry
single
scattering
albedo
that
were
derived
from
scattering
coefficients
measured
at
550-nm
from
the
nephelometer
and
absorption
coefficients
measured
at
565-nm
from
a
particle/soot
absorption
photometer
(Ostrom
and
Noone,
2000).
For
flight
15,
our
estimates
of
0.935
(in
the
free
troposphere)
to
0.954
(in
the
boundary
layer)
for
a
profile
vertically-averaged
from
3881
to
40
m
during
flight
15
exceeded
their
measured
values
of
0.732
(at
3881
m),
0.751
(at
980
m),
and
0.806
(at
40
m).
However,
when
we
increased
the
amount
of
EC
to
the
detection
limit
for
elemental
carbon
to
obtain
a
lower
limit
for
the
single
scatter
albedo,
our
estimates
were
quite
similar
to
those
of
Ostrom
and
Noone
(2000)
with
0.727
(in
the
free
troposphere)
and
0.782
(in
the
boundary
layer).
Some
of
the
nephelometer
measurements
from
flight
20
show
better
agreement
with
our
calculations
than
those
from
flight
15.
A
single
scatter
albedo
of
0.961
was
measured
at
30
m,
as
compared
to
our
calculated
value
for
the
boundary
layer
of
0.983.
Single
scatter
albedos
of
0.959
and
0.988
were
measured
at
1918
m
and
1916
m,
respectively,
as
compared
to
our
calculated
value
for
the
free
troposphere
of
0.976.
Flight
20
meas-
urements
show
a
large
variability
in
both
the
free
troposphere
where
they
vary
from
0.869
at
3885
m
to
0.988
at
1968
m
and
the
boundary
layer
where
the
measured
values
vary
from
0.805
to
0.961
(Ostrom
and
Noone,
2000).
Some
of
the
discrepan-
cies
in
these
values
are
expected,
since
the
filter
samples
do
not
identify
all
components
and
were
collected
over
time
segments
different
from
those
used
to
calculate
the
average
single
scatter
albedos
from
the
nephelometers.
However,
the
consistent
overprediction
of
our
modelled
single
scatter
albedos
from
the
measured
calculation
suggests
that
absorbing
species
(metal
or
elemental
carbon)
were
present
but
were
not
identified
in
our
ana-
lysis.
The
good
agreement
of
the
range
we
pre-
dicted
by
varying
the
elemental
carbon
and
the
mixing
state
is
quite
consistent
with
the
measured
range,
suggesting
that
within
the
uncertainties
of
our
calculation,
we
are
in
good
agreement
with
the
measured
values.
6.
Conclusions
During
ACE-2,
filter
samples
of
aerosol
particles
were
collected
from
the
Pelican
aircraft
and
were
analyzed
by
three
techniques
(thermal-optical
ana-
lysis,
ion
chromatography,
and
total
reflection
X-ray
fluorescence),
providing
detailed
aerosol
chemical
characterizations.
Reported
concentra-
tions
are
consistent
with
those
in
similar
condi-
tions
in
other
studies
and
illustrate
several
aerosol
types,
including
clean
marine,
anthropogenically-
influenced,
and
dust-containing
aerosols.
We
used
measured
aerosol
chemical
composi-
tion
to
assess
the
importance
of
aerosol
chemistry
in
determining
optical
properties
among
different
types
of
aerosol
and
for
variations
within
those
types.
Calculations
of
associated
aerosol
optical
properties
show
that
the
dry
dust-laden
aerosol
are
dominated
by
absorbing
mineral
components,
but
missing
liquid
water
absorption.
Few
differ-
ences
were
predicted
between
clean
marine
and
anthropogenically-influenced
cases
despite
differ-
ences
in
sulfate
and
organic
carbon
fractions,
since
both
were
dominated
by
liquid
water
properties
at
high
marine
boundary
layer
humidities.
These
results
illustrate
the
importance
of
measuring
detailed
particle
composition
for
dust-containing
Tellus
52B
(2000),
2
AEROSOL
PARTICLE
CHEMICAL
CHARACTERISTICS
199
cases
where
the
optical
properties
are
especially
sensitive
to
the
masses
of
absorbing
species
present.
7.
Acknowledgements
This
work
was
supported
by
ONR
grant
N00014-97-1-0673.
The
Pelican
aircraft
flight
time
and
instrumentation
were
supported
by
NSF
grant
ATM-9614105.
The
authors
are
indebted
to
Scott
Shoemaker
for
his
help
in
collecting
the
filter
samples
during
the
Pelican
flights
and
to
the
CIRPAS
crew
and
Pelican
science
team
for
their
assistance
in
the
field.
The
authors
are
also
grateful
to
Patricia
Quinn
and
Kevin
Noone
who
shared
the
results
of
their
ACE-2
data
in
order
to
make
the
comparisons
referenced
in
this
paper
possible.
This
research
is
a
contribution
to
the
International
Global
Atmospheric
Chemistry
(IGAC)
Core
Project
of
the
International
Geosphere-Biosphere
Programme
(IGBP)
and
is
part
of
the
IGAC
Aerosol
Characterization
Experiments
(ACE).
REFERENCES
Allen,
A.
G.,
Dick,
A.
L.
and
Davison,
B.
M.
1997.
Sources
of
atmospheric
methanesulphonate,
non-sea-
salt
sulphate,
nitrate
and
related
species
over
the
tem-
perate
south
Pacific.
Atmos.
Environ.
31,
191-205.
Arimoto,
R.,
Duce,
R.
A.,
Ray,
B.
J.,
Ellis
Jr.,
W.
G.,
Cullen,
J.
D.
and
Merrill,
J.
T.
1995.
Trace
elements
in
the
atmosphere
over
the
North
Atlantic.
J.
Geophys.
Res.
100,
1199-1213.
Bergametti,
G.,
Gomes,
L.,
Coudegaussen,
G.,
Rognon,
P.
and
Lecoustumer,
M.
N.
1989.
African
dust
observed
over
the
Canary
Islands-source-regions
I
dentification
and
transport
pattern
for
some
summer
situations.
J.
Geophys.
Res.
94,
14855-14864.
Birch,
M.
E.
and
Cary,
R.
A.
1996.
Elemental
carbon-
based
method
for
monitoring
occupational
exposures
to
particulate
diesel
exhaust.
Aerosol
Sci.
Tech.
25,
221-241.
Bohren,
C.
F.
and
Huffman,
D.
R.
1983
Absorption
and
scattering
of
light
by
small
particles.
John
Wiley,
New
York.
Brechtel,
F.
J.,
Kreidenweis,
S.
M.
and
Swan,
H.
B.
1998.
Air
mass
characterisitcs,
aerosol
particle
number
con-
centrations,
and
number
size
distributions
at
Macqua-
rie
Island
during
the
first
Aerosol
Characterization
Experiment
(ACE-1).
J.
Geophys.
Res.
103,
16351-16367.
Chan,
Y.
C.,
Simpson,
R.
W.,
Mctainsh,
G.
H.,
Vowles,
P.
D.,
Cohen,
D. D.
and
Bailey,
G.
M.
1997.
Charac-
terisation
of
chemical
species
in
PM2.5
and
PM10
aerosols
in
Brisbane
Australia.
Atmos.
Environ.
31,
3773-3785.
Collins,
D.
R.,
Jonsson,
H.
H.,
Seinfeld,
J.
H.,
Flagan,
R.
C.,
Gasso,
S.,
Hegg,
D.,
Russell,
P.
B.,
Schmid,
B.,
Livingston,
J.
M.,
Ostrom,
E.,
Noone,
K.
J.,
Russell,
L.
M.
and
Putaud,
J.
P.
2000.
In
situ
aerosol
size
distributions
and
clear
column
radiative
closure
during
ACE-2.
Tellus
52B,
498-525.
d'Almeida,
G.
A.,
Koepke,
P.
and
Shettle,
E.
P.
1991.
Atmospheric
aerosols:
global
climatology
and
radiative
characteristics,
A.
Deepak,
Hampton,
Va.
d'Almeida,
G.
A.
and
Schatz,
L.
1983.
Number,
mass,
and
volume
distribution
of
mineral
aerosol
and
soils
of
the
Sahara.
J.
Clim.
Appl.
Meteorol.
22,
233-243.
Hale,
G.
M.
and
Querry,
M.
R.
1973.
Optical
constants
of
water
in
the
200-nm
to
200
gm
wavelength
region.
Applied
Optics
12,
555-563.
Hanel,
G.
1976.
The
properties
of
atmospheric
aerosol
particles
as
functions
of
the
relative
humidity
at
ther-
modynamic
equilibrium
with
the
surrounding
moist
air.
Adv.
Geophys.
19,
73-188.
Hegg,
D.
A.,
Ferek,
R.
J.
and
Hobbs,
P.
V.
1993.
Light
scattering
and
cloud
condensation
nucleus
activity
of
sulfate
aerosols
measured
over
the
Northeast
Atlantic
ocean.
J.
Geophys.
Res.
98,
14887-14894.
Hoppel,
W.
A.,
Fitzgerald,
J.
W.,
Frick,
G.
M.,
Larson,
R.
E.
and
Mack,
E.
J.
1990.
Aerosol
size
distributions
and
optical
properties
found
in
the
marine
boundary
layer
over
the
Atlantic
Ocean.
J.
Geophys.
Res.
95,
3659-3686.
Howell,
S.
G.
and
Huebert,
B.
J.
1998.
Determining
marine
aerosol
scattering
characteristics
at
ambient
humidity
from
size-resolved
chemical
composition.
J.
Geophys.
Res.
103,
1391-1404.
Huebert,
B.
J.,
Zhuang,
L.,
Howell,
S.,
Noone,
K.
and
Noone,
B.
1996.
Sulfate,
nitrate,
methanesulfonate,
chloride,
ammonium,
and
sodium
measurements
from
ship,
island,
and
aircraft
during
the
Atlantic
Stratocu-
mulus
Transition
Experiment/Marine
Aerosol
Gas
Exchange.
J.
Geophys.
Res.
101,
4413-4423.
Huebert,
B.
J.,
Howell,
S.
G.,
Zhuang,
L.,
Heath,
J.
A.,
Litchy,
M.
R.,
Wylie,
D.
J.,
Kreidler-Moss,
J.
L.,
Coppicus,
S.
and
Pfeiffer,
J.
E.
1998.
Filter
and
impactor
measurements
of
anions
and
cations
during
the
first
Aerosol
Characterization
Experiment
(ACE
1).
J.
Geophys.
Res.
103,
16493-16509.
Husar,
R.
B.,
Prospero,
J.
M.
and
Stowe,
L. L.
1997.
Characterization
of
tropospheric
aerosols
over
the
oceans
with
the
NOAA
advanced
very
high
resolution
radiometer
optical
thickness
operational
product.
J.
Geophys.
Res.
102,
16889-16909.
Klockenkamper,
R.
1997.
Total
reflection
X-ray
fluores-
cence
analysis,
ed.
J.
D.
Winefordner.
John
Wiley
and
Sons,
New
York,
245
pp.
Tellus
52B
(2000),
2
200
M.
SCHMELING
ET
AL.
Li-Jones,
X.
and
Prospero,
J.
M.
1998.
Variations
in
the
size
distribution
of
non-sea-salt
sulfate
aerosols
in
the
marine
boundary
layer
at
Barbados:
impact
of
African
dust.
J.
Geophys.
Res.
103,
16073-26084.
Miranda,
J.,
Cahill,
T.
A.,
Morales,
R.
J.,
Adalpe,
F.,
Flores,
M.
J.
and
Diaz,
R.
V.
1994.
Determination
of
elemental
concentrations
in
atmospheric
aerosols
in
Mexico
City
using
proton
induced
X-ray
emission,
proton
elastic
scattering
and
laser
absorption.
Atmos.
Environ.
28,
2299-2306.
Okamoto,
S.,
Kobayashi,
K.
and
Yamada,
T.
1986.
Char-
acterisation
of
aerosols
in
the
Kashima
area
and
a
source
apportionment
study.
In:
Sydney
Clean
Air
Congress
1986,
pp.
253-262.
Ostrom,
E.
and
Noone,
K.
J.
2000.
Vertical
profiles
of
aerosol
scattering
and
absorption
measured
in
situ
during
the
North
Atlantic
Aerosol
Characterization
Experiment.
Tellus
52B,
526-545.
Prospero,
J.
M.,
Savoie,
D.
L.,
Arimoto,
R.
and
Huang,
F.
1993.
Long-term
trends
in
mineral
dust
concentra-
tions
over
the
western
North
Atlantic:
Relationship
to
North
African
rainfall.
Eos
Trans.
AGU
74,
Fall
Meet-
ing
Suppl.
Putaud,
J.
P.,
Van
Dingenen,
R.,
Mangoni,
M.,
Virkkula,
A.,
Raes,
F.,
Maring,
H.,
Prospero,
J.
M.,
Swietlicki,
E.,
Berg,
0.
H.,
Hillamo,
R.
and
Makela,
T.
Chemical
mass
closure
and
origin
assessment
of
the
submicron
aerosol
in
the
marine
boundary
layer
and
the
free
troposphere
at
Tenerife
during
ACE-2.
Tellus
52B,
141-168.
Quinn,
P.K.,
Coffman,
DJ.,
Bates,
T.S.,
and
Covert,
D.S.
2000.
A
comparison
of
aerosol
chemical
and
optical
properties
from
the
first
and
second
Aerosol
Charac-
terization
Experiments.
Tellus
52B,
239-257.
Raes,
F.,
Bates,
T.
S.,
McGovern,
F.
M.
and
Van
Liedek-
erke,
M.
2000.
The
second
Aerosol
Characterization
Experiment
(ACE-2):
General
overview
and
main
results.
Tellus
52B,
111-125.
Russel,
P.
A.
and
Heintzenberg,
J.
2000.
Clear
Column
Experiments
in
ACE-2.
Tellus
52B,
463-483.
Russell,
L.
M.,
Huebert,
B.
J.,
Flagan,
R.
C.
and
Seinfeld,
J.
H.
1996a.
Characterization
of
submicron
aerosol
size
distributions
from
time-resolved
measurements
in
the
Atlantic
stratocumulus
transition
experiment/
marine
aerosol
and
gas
exchange.
J.
Geophys.
Res.
101,
4469-4478.
Russell,
L.
M.,
Stolzenburg,
M.
R,
Caldow,
R.,
Zhang,
S.
H.,
Flagan,
R.
C.
and
Seinfeld,
J.
H.
1996b.
Radially-
classified
aerosol
detector
for
aircraft-based
submicron
aerosol
measurements.
J.
Atmos.
Ocean.
Tech.
13,
598-609.
Schmeling,
M.
1997.
Anwendung
der
Totalreflexions-
Rntgenfluoreszenz
zur
Analyse
von
luftgetragenen
Partikeln.
PhD-thesis,
Dortmund,
77
pp.
Schmeling,
M.
and
Klockow,
D.
1997.
Sample
collection
and
preparation
for
analysis
of
airborne
particulate
matter
by
total
reflection
X-ray
fluorescence
spectro-
metry.
Anal.
Chico.
Acta
346,
121-126.
Schutz,
L.
and
Sebert,
M.
1987.
Mineral
aerosols
and
source
identification.
J.
Aerosol
Sci.
18,
1-10.
Seinfeld,
J.
H.
1986.
Atmospheric
chemistry
and
physics
of
air
pollution.
John
Wiley
and
Sons,
New
York,
738
pp.
Streli,
C.
1997.
Total
reflection
X-ray
fluorescence
ana-
lysis
of
light
elements.
Spectrochim.
Acta
52B,
281-293.
Swietlicki,
E.,
Zhou,
J.,
Berg,
0.
H.,
Hameri,
K.,
Vakeva,
M.,
Makela,
J.,
Covert,
D.
S.,
Dusek,
U.,
Busch,
B.,
Wiedensohler,
A.
and
Stratmann,
F.
2000.
Hygro-
scopic
properties
of
aerosol
particles
in
the
North-
eastern
Atlantic
during
ACE-2.
Tellus
52B,
201-227.
Thursdon,
G.
D.
and
Spengler,
J.
D.
1985.
A
quantitative
assessment
of
source
contributions
to
inhalable
matter
pollution
in
metropolitan
Boston.
Atmos.
Environ.
19,
9-25.
Toon,
0.
B.,
Pollack,
J.
B.
and
Khare,
B.
N.
1976.
The
optical
constants
of
several
atmospheric
aerosol
species:
ammonium
sulfate,
aluminum
oxide,
and
sodium
chloride.
J.
Geophys.
Res.
81,
5733-5742.
Verver,
G.,
Raes,
F.,
Vogelezang,
D.
and
Johnson,
D.
2000.
The
second
Aerosol
Characterization
Experi-
ment
(ACE-2).
Meteorological
and
chemical
context.
Tellus
52B,
126-140.
Weast,
R.
C.,
Astle,
M.
J.
and
Beyer,
W.
H.
(Eds.)
1985-1986.
CRC
Handbook
of
chemistry
and
physics,
66th
edition.
CRC
Press,
Boca
Raton,
Fla.,
p.
D-233.
Zhang,
X,
Arimoto,
R.
An,
Z.,
Chen,
T.,
Zhang,
G.,
Zhu,
G.
and
Wang,
X
1993.
Atmospheric
trace
ele-
ments
over
source
regions
for
chinese
dust:
Concentra-
tions,
sources
and
atmospheric
deposition
on
the
Loess
Plateau.
Atmos.
Environ.
27,
2051-2067.
Zhang,
X
Y.,
Arimoto,
R.,
Zhu,
G.
H.,
Chen,
T.
and
Zhang,
G.
Y.
1998.
Concentration,
size
distribution
and
deposition
of
mineral
aerosol
over
Chinese
desert
regions.
Tellus
50B,
317-330.
Tellus
52B
(2000),
2