Regional and seasonal variability of particle and element-concentrations from snowpits in Antarctica
M. Kriews, S. Dietrich, D. Dick, I. Stölting, A. Wegner, H. MillerAlfred Wegener Institute for Polar and Marine ResearchAm Handelshafen 12, 27570 Bremerhaven, Germany
email: [email protected]
Introduction
Mineral dust measured in Antarctic ice cores is a unique tool
to reconstruct palaeo climate variations, but even the
knowledge about present day dust in Antarctic snow is still
very poor. In this study we analyzed different proxies of
mineral dust (particle concentration, size distribrution,
elemental composition) in snowpits from Berkner Island, a
coastal influenced site, and from Kohnen Station located on
the plateau, in order to investigate differences in the transport
behaviour of dust from the source to its deposition site and
how this might change during the year.
Fig. 1: Sampling in the snowpit at Kohnen-Station
Snowpit Kohnen-Station
75°00'S, 0°04 E, 2890 m a.s.l.
2 m deep, 2 cm resolution
season 2003/2004
mean annual accumulation rate:
62 mm w.eq
Snowpit Berkner Island
79° 34' N, 45° 39' W, 880 m a.s.l.
2 m deep, 2 cm resolution
season 2003/2004
mean annual accumulation rate:
130 mm w.eq
Experimental
Sampling
The samples were taken in polypropylen tubes, which were pressed into the snowpit walls. For the investigation of elemental composition the tubes
were precleaned in 3 different acid baths (1:4 HCl, 1:4 HNO3, 1:10 HNO3) and rinsed with ultrapure water. After sampling, the tubes were packed
into two polyethylene bags and kept frozen until analysis.
11 ppt59Co
0,4 ppt238U0,3 ppt159Tb2 ppt109Ag58 ppt58Ni
1 ppt232Th1 ppt153Eu2 ppt107Ag1,5 ppb57Fe
1 ppt209Bi1 ppt151Eu1 ppt89Y1,4 ppb56Fe
13 ppt208Pb1 ppt149Sm33 ppt88Sr0,1 ppb55Mn
11 ppt207Pb1 ppt147Sm39 ppt86Sr0,1 ppb52Cr
12 ppt206Pb0,8 ppt144Nd4 ppt85Rb15 ppt51V
6 ppt205Tl1 ppt143Nd38 ppt82Se4,0 ppb44Ca
12 ppt204Pb1 ppt142Nd0,4 ppb78Se3,6 ppb43Ca
0,2 ppt175Lu0,4 ppt141Pr11 ppt75As2,8 ppb39K
0,4 ppt174Yb2 ppt140Ce13 ppt69Ga21 ppb27Al
0,2 ppt169Tm1 ppt139La2,8 ppb66Zn1,5 ppb25Mg
0,4 ppt166Er88 ppt138Ba0,2 ppb65Cu1,8 ppb24Mg
0,2 ppt165Ho0,3 ppt115In2,6 ppb64Zn8,0 ppb23Na
0,4 ppt164Dy1 ppt114Cd0,2 ppb63Cu1 ppt9Be
1 ppt160Gd1 ppt111Cd52 ppt60Ni2 ppt8Li
MDLElementMDLElementMDLElementMDLElement
Fig. 2: Laser-sensor-setup, for particle size
and -concentration measurements
Fig. 3: Setup of full acid digestion under
pressure for sample treatment and
enrichment before ICP-MS analysis
Measurements of particle distribution
For the particle concentration and size distribution
the samples were measured within 15 minutes after
melting with a laser-sensor device (λ=670 nm). A
cross-calibration with a Coulter-Counter was made
by using samples from the EPICA deep-drilling-
sites in Dronning Maud Land and on Dome C for
particles <5 µm and samples from Greenland and
spherical particles for particle >5 µm. To determine
blank-levels ultrapure water was measured before
and after the samples. The measuring range of the
laser-sensor is between 1 and 10 µm. Each sample
was measured at least three times using a sample
volume of 4-10 ml.
Measurements of elemental composition
For determination of elemental composition the samples were processed and measured in
cleanroom laboratories at AWI. After melting the initial sample volume was approximately 20
ml. The samples were treated under pressure with a full-acid digestion (HNO3, HF, H2O2) and
enriched by a factor of 10 (fig. 3). Trace and ultrace element analyses were carried out with an
ICP-MS (ELAN 6000-Perkin-Elmer/Sciex) in combination with membran desolvatation
system MCN 6000 (Cetac) and a 100 µl PFA nebulizer. The instruments setups are shown in
table 1 and 2. As internal standard Rh was added to each sample. The method detection limits
(MDL) of the measured elements are given in table 3. The MDL depence on the method of
sample preparation and was calculated as three times the standard deviation of 91 blank
treatments.
Tab. 3: Method detection limits ( 3 σ criteria, n = 91)
Results
Particle concentration
At Berkner Island the particle concentration is low (3.75 ng/ml
without very pronounced particle peaks), we find not a clear
seasonality, but a slightly increase of particle-concentration in
winter. There are very prominent peaks at 122-136 cm depth
(year 2001), as shown in fig.4. This event is also detected in
elemental composition investigations.
At Kohnen the particle concentration is higher (6.7 ng/ml), we
find a clear seasonality in the particle-concentration with a
maximum in winter (Fig.5).
Chemical composition
Additional the concentration for Fe as a tracer for mostly
insoluble mineral dust particles is shown in fig. 4 and 5. For
Kohnen station there is a good agreement obvious in the depth
range from 16-58 and 76-96 cm.
At Berkner Island are only some few peaks for Fe which reflect
the particle distribution. The prominent peaks for particle
concentrations at 122-136 cm correspond very well with the
distribution for sea salt elements like Na, Mg, Ca (fig. 6). This
finding could not be explained yet. It could be an effect of
chemical reactions during atmospheric transport whereby
insoluble particles are produced.
1500 µsSettlingtime
DualDetector mode
10 msDwelltime
on Peakhopping
0,8 l/minAuxillary gas
15 l/minPlasma gas
onAutolens
1450 WPlasma
4,75 VLens voltage 103Rh
0,7 ml/minNebulizer gas
SetupParameter
~ 100 µl/minSample uptake rate
18 ml/minNitrogen flow
1,88 l/minArgon flow
SetupParameter
Tab. 1: MCN 6000 parameters Tab. 2: ELAN 6000 parameters
Depth[cm]
0 8 16 24 32 40 48 56 64 72 80 88 96 104 112 120 128 136 144 152 160 168 176 184
Concentr
ation [ppb]
0
10
20
30
200
Delta 1
8O
[‰
]
-40
-35
-30
-25
-20
-15
2004 2003 2002 2001 2000
Age [a]
Particle concentration [ppb]
Delta 18
O [‰]
Fe [ppb]
23Na
Ko
nz.
[ppb
]
0
400
800
1200
1600
25Mg
Konz.
[pp
b]
0
50
100
150
200
44Ca
Konz.
[pp
b]
0
20
40
140Ce
Ko
nz.
[ppb
]
0,00
0,020,100,120,14
142Nd
Ko
nz.
[ppb
]
0,000
0,005
0,010
208Pb
depth (cm)
0 8 16 24 32 40 48 56 64 72 80 88 96 104 112 120 128 136 144 152 160 168 176 184
Konz. [p
pb]
0,00
0,05
0,10
Concentration with error
Instrument detection limit (aligned)
Method detection limit (aligned)
Fig. 4: Particle and Fe concentration as well as Delta 18O
(temperature proxy) at Berkner Island
Fig. 6: Na, Mg, Ca (sea salt), Ce, Nd (mineral dust) and Pb
(anthropogenic) concentration at Berkner Island
2004 2003 2002 2001 1997 19962000 1999 1998 1995 1994
Age [a]
Depth [cm]
0 8 16 24 32 40 48 56 64 72 80 88 96 104 112 120 128 136 144 152 160 168 176 184 192 200
Con
cen
tration [
ppb]
0
5
10
15
20
25
30
100
Delta
18
O [
‰]
-60
-55
-50
-45
-40
-35
-30
Particle concentration [ppb]
Delta 18O [‰]
Fe [ppb]
24Mg
Con
c. [
ppb
]
0
5
10
15
20
25
70
23Na
Conc. [p
pb]
0
40
80
120
300
44Ca
Co
nc.
[pp
b]
0
20
40
60
140Ce
Co
nc. [p
pb]
0,00
0,01
0,02
0,03
0,04
0,05
142Nd
Co
nc. [p
pb]
0,00
0,01
0,02
0,04
208Pb
Depth (cm)
0 8 16 24 32 40 48 56 64 72 80 88 96 104 112 120 128 136 144 152 160 168 176 184 192 200
Con
c. [
ppb
]
0,0
0,1
0,2
0,6
Concentration with error
Instrument detection limit (aligned)
Method detection limit (aligned)
Fig. 5: Particle and Fe concentration as well as Delta 18O
(temperature proxy) at Kohnen station
Fig. 7: Na, Mg, Ca (sea salt), Ce, Nd (mineral dust) and Pb
(anthropogenic) concentration at Kohnen station
Tab. 4: Mean element concentrations with standard deviation as well as min. and max.
values for snowpits as well as the deposition rate at Kohnen station and Berkner Island
From tab. 4 it is obvious that the mean con-
centrations for sea salt elements are higher at
Berkner Island, while the mineral dust tracer
show higher values at Kohnen. Anthropogenic
tracer elements show slightly higher concentrations at Kohnen. All elements show a high variability.
This strong variabilty is also
shown for some selected ele-
ments in fig. 4-7. There is a
clear seasonality in the sea-salt
elements at Berkner Island.
At Kohnen sea salt-aerosol is
not the dominating source,
but shows also a seasonality with
maxima in winter.
Mineral dust aerosol is the dominat
source at Kohnen and shows a clear
seasonality with maxima in winter.
Taking into account the yearly
accumulation rate it can be seen
from tab. 4 that the flux for sea salt
aerosol is by a factor of 8-15 higher
at Berkner Island than on the plateau
at Kohnen. For mineral dust as well
as for anthropogenic elements the
deposition rate at both stations is
nearly similar whereas there is a
slightly higher amount at Kohnen.
0.0780,2 ppt15 ppt0,01 ppt0,6 ppt0.0500,2 ppt19 pptn.d.0,8 ppt238U
0.210,2 ppt15 ppt0,09 ppt1,6 ppt0.0810,2 ppt27 pptn.d.1,3 ppt232Th
0.0260,2 ppt11 pptn.d.0,2 ppt0.130,3 ppt27 pptn.d.2,1 ppt209Bi
2.10,9 ppt0,2 ppb0,1 ppt16 ppt3.21,2 ppt0,6 ppbn.d.51 ppt208Pb
0.290,3 ppt10 pptn.d.2,2 ppt0.440,8 ppt30 pptn.d.7,1 ppt205Tl
0.0260,2 ppt2,8 pptn.d.0,2 ppt0.0430,1 ppt18 pptn.d.0,7 ppt175Lu
0.0390,2 ppt2,9 ppt0,04 ppt0,3 ppt0.0560,2 ppt19 pptn.d.0,9 ppt174Yb
0.0260,2 ppt2,8 pptn.d.0,2 ppt0.0370,1 ppt17 pptn.d.0,6 ppt169Tm
0.0390,2 ppt2,9 pptn.d.0,3 ppt0.0560,1 ppt19 pptn.d.0,9 ppt166Er
0.0260,2 ppt2,8 pptn.d.0,2 ppt0.0430,1 ppt17 pptn.d.0,7 ppt165Ho
0.0390,2 ppt3,6 pptn.d.0,3 ppt0.0560,1 ppt20 pptn.d.0,9 ppt164Dy
0.0390,2 ppt3,1 pptn.d.0,3 ppt0.0560,1 ppt19 pptn.d.0,9 ppt160Gd
0.0260,2 ppt2,9 pptn.d.0,2 ppt0.0370,1 ppt17 pptn.d.0,6 ppt159Tb
0.0390,2 ppt3,1 pptn.d.0,3 ppt0.0370,1 ppt17 pptn.d.0,6 ppt151Eu
0.0650,2 ppt3,2 pptn.d.0,5 ppt0.0560,2 ppt18 pptn.d.0,9 ppt147Sm
0.180,3 ppt18 ppt0,2 ppt1,4 ppt0.180,2 ppt39 ppt0,1 ppt2,9 ppt143Nd
0.0650,2 ppt5,5 ppt0,02 ppt0,5 ppt0.0810,1 ppt20 pptn.d.1,3 ppt141Pr
0.830,3 ppt0,1 ppb0,6 ppt6,4 ppt0.930,3 ppt0,6 ppb0,5 ppt15 ppt140Ce
0.390,3 ppt63 ppt0,2 ppt3,0 ppt0.420,2 ppt0,3 ppb0,2 ppt6,8 ppt139La
26.04,8 ppt5,8 ppbn.d.0,2 ppb24.86,6 ppt14 ppb3,7 ppt0,4 ppb138Ba
0.0390,2 ppt15 pptn.d.0,3 ppt0.0500,1 ppt17 pptn.d.0,8 ppt115In
0.520,6 ppt20 pptn.d.4,0 ppt0.811,8 ppt0,1 ppbn.d.13 ppt114Cd
0.360,2 ppt13 pptn.d.2,8 ppt0.510,2 ppt0,1 ppb0,5 ppt8,2 ppt109Ag
0.210,2 ppt25 ppt0,1 ppt1,6 ppt0.180,1 ppt29 ppt0,05 ppt2,9 ppt89Y
265,4 ppt1,3 ppb14 ppt0,2 ppb3.11,2 ppt0,3 ppb0,4 ppt50 ppt86Sr
1.20,7 ppt35 ppt1,8 ppt9,1 ppt5.20,3 ppt48 ppt1,3 ppt84 ppt85Rb
2.50,7 ppt0,5 ppbn.d.19 ppt2.30,6 ppt1,2 ppbn.d.37 ppt69Ga
6.84,3 ppt1,3 ppbn.d.52 ppt6.25,4 ppt1,4 ppbn.d.0,1 ppb65Cu
5330,1 ppb12 ppb1,2 ppb4,1 ppb86810 ppb105 ppb2,7 ppb14 ppb64Zn
0.950,5 ppt85 pptn.d.7,3 ppt0.870,3 ppt0,1 ppbn.d.14 ppt59Co
8.21,6 ppt0,8 ppbn.d.63 ppt6.22,5 ppt1,4 ppb2,8 ppt0,1 ppb58Ni
37757 ppt50 ppbn.d.2,9 ppb2540,1 ppb52 ppb0,6 ppb4,1 ppb56Fe
3.41,8 ppt0,4 ppbn.d.26 ppt6.24,3 ppt1,7 ppbn.d.0,1 ppb55Mn
8.54,5 ppt0,8 ppbn.d.65 ppt12.48,2 ppt1,9 ppbn.d.0,2 ppb52Cr
3.33,9 ppt0,9 ppbn.d.25 ppt2.52,8 ppt0,8 ppbn.d.41 ppt51V
14300,3 ppb53 ppb1,0 ppb11 ppb4840,2 ppb94 ppb0,7 ppb7,8 ppb44Ca
14300,3 ppb66 ppb0,6 ppb11 ppb17470 ppt10 ppbn.d.2,8 ppb39K
7410,2 ppb459 ppbn.d.5,7 ppb26048,8 ppb1015 ppbn.d.42 ppb27Al
37701,0 ppb176 ppb1,1 ppb29 ppb4900,2 ppb65 ppb1,4 ppb7,9 ppb24Mg
276907,6 ppb1491 ppb8,2 ppb213 ppb16740,7 ppb327 ppb68 ppt27 ppb23Na
0.0330,2 ppt9,5 pptn.d.0,25 ppt0.0430,2 ppt18 pptn.d.0,7 ppt9Be
0.730,5 ppt63 pptn.d.5,6 ppt0.400,4 ppt44 ppt0,5 ppt6,4 ppt8Li
Flux (µg m-2 a-1)Std. dev.Max.Min.AverageFlux (µg m-2 a-1)Std. dev.Max.Min.AverageElement
Berkner IslandKohnen
In contrast to Berkner Island we found at Kohnen
only a good correlation for Na and Mg. For all
shown elements the seawater ratio could not be
found. The reason might be fractionation effects
during atmospheric transport or there are other
sources.
To identify possible other sources e.g. for Na and Ca the non sea salt Na and Ca-
concentration (nss-Na, Ca) was calculated. Fig. 8 c shows the results. At Kohnen
most of the Ca is not from a marine source, while for Na except some few depths
more than 90% is from sea salt. At Berkner Island except some few depths more
than 90% of the total Ca and nearly 100% of Na is from an oceanic source.
Fig. 8a: Correlation of Na versus , Mg, Ca,
Sr at Berkner Island
Na/K
Concentration K [ppb]
0 10 20 30 40 50 60 70
Co
nce
ntr
ati
on N
a [
pp
b]
0
200
400
600
800
1000
1200
1400
1600
r ²=0,9843
Na/Sr
Concentration Sr [ppb]
0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4
Concen
tration
Na [p
pb
]
0
200
400
600
800
1000
1200
1400
1600
r ²=0,9547
Na/Mg
Concentration Mg [ppb]
0 50 100 150 200
Concen
tration
Na [p
pb
]
0
200
400
600
800
1000
1200
1400
1600
r ²=0,9537
Na/Ca
Concentration Ca [ppb]
0 10 20 30 40 50 60
Co
ncen
tration N
a [p
pb]
0
200
400
600
800
1000
1200
1400
1600
r ²=0,7874
0 25 50 75 100 125 150 175
Co
nc
. [
ppb
]
0
10
20
30
40 nss-Ca
total-Ca
depth [cm] depth [cm]
0 50 100 150 200
Con
c. [p
pb]
0
20
40
80
total-Ca
nss-Ca
depth [cm]
Berkner Island
0 25 50 75 100 125 150 175
Co
nc.
[p
pb
]
0
200
400
600
800
1000
1200
1400
nss-Na
total-Na
Kohnen
depth [cm]
0 50 100 150 200
Conc. [p
pb
]
0
50
100
350
total-Na
nss-Na
Na/K
Concentration K [ppb]
0 2 4 6 8 10 12
Con
centr
atio
n N
a [p
pb
]
0
50
100
150
200
250
300
350Na/Sr
Concentration Sr [ppb]
0,00 0,05 0,10 0,15 0,20 0,25 0,30
Conce
ntr
atio
n N
a [p
pb]
0
50
100
150
200
250
300
350
Na/Mg
Concentration Mg [ppb]
0 20 40 60
Co
ncentr
ation
Na [
pp
b]
0
100
200
300
Na/Ca
Concentration Ca [ppb]
0 10 20 30 40 50
Con
centr
ation N
a [
ppb
]
0
50
100
150
200
250
Ocean water ratioOcean water ratio
Fig. 8b: Correlation of Na versus
Mg, Ca, Sr at Kohnen station
From fig. 6 it can bee seen that at Berkner Island is a good
agreement in the depth profile for the sea salt elements Na, Mg
and Ca. At Kohnen station (fig. 7) the profile for Ca is often
different. In fig. 8 a and b the Na concentration is plotted
versus typical other sea salt elements (Mg, Sr, Ca). For
Berkner Island it can be seen that there is a good correlation
between these elements and Na and the element ratios (marked
by a red cross) in open ocean water agree very well.
Fig. 8c: Total and non sea salt (nss) Na and Ca
Conclusions and Outlook
The plateau position at Kohnen
Station has a clear seasonality in
dust concentration. The chemical
composition shows that mineral dust
is dominating. At Berkner Island the
particle concentration and size
distribution is dominated by single
events, that might have a local
source. A strong influence from the
ocean could be observed when
looking to typical sea salt tracer
elements.
For future investigations of
deep ice cores it is necessary
to take into account at which
position the drilling will be
carried out. In ice core
studies up to now Ca is used
as a tracer for mineral dust in
contnious flow analyses. This
contribution shows that
depending on the sampling
location Ca has at least two
different sources.
In further steps analyses will be carried out for samples from potential sources in the southern
hemisphere to identify sources and possible changes during the year. For this analyses the Rare
earth element distribution patterns will be used.
Acknowledgement
This work is a contribution to the “European Project for Ice
Coring in Antarctica” (EPICA), a joint European Science
Foundation/European Commission (EC) scientific program,
funded by the EC under the Environment and Climate Program.