Mass size distribution study of aerosol particles

advertisement
Mass size distribution study of aerosol particles
using samples from Arctic Ocean expedition 2001
(AOE01)
Arash Gharibia,*, Keith Biggb, Erik Swietlickia, Caroline Leckb
a
Division of Nuclear Physics, Lund University, P.O. Box 118, SE-221 00 Lund, Sweden
b
Meteorological Institute of Stockholm University (S-106 91 Stockholm)
Abstract
During the last two decades, there has been an increasing interest in aerosol particles in the
Arctic. During the Arctic Ocean Expedition 2001 (AOE01) total, fine and coarse mass
concentrations of aerosol particles have been measured in the Pack Ice (PI) region. The
equivalent aerodynamic diameter (EAD < 10 m) has been used to determine total mass
concentration. Coarse fraction diameters were defined as 2m <EAD<10m and the fine
fraction as <2m. The aerosol particles were collected using stacked filter units (SFU) and
two types of cascade impactor, a Low Pressure Impactor (LPI) and a Small Deposit area
Impactor (SDI). The results from parallel samples, taken with the different devices and
analyzed by PIXE technique, were compared. The results have also been compared with those
from three previous Arctic expeditions, namely AOE-96 and IAOE-91. Total mass
concentrations of Cl and S as a function of marine transport time (Mtr) are presented. Lognormal fitting of size distributions is applied to the PIXE results of aerosol samples obtained
in the pack ice region. Modal parameters are calculated and shown for many trace elements.
1. Introduction
The Arctic Ocean is climatologically sensitive and plays a very important role in determining
the regional global radiation balance and earth climate. Over the Arctic Ocean in summer the
major factors for the radiation balance are: pack ice cover and albedo, and cloud or fog cover
and albedo. Atmospheric aerosol particles affect the climate both directly and indirectly. The
direct effect refers to back-scattering to space of short wave solar radiation. It will only be
appreciable in the absence of overlying clouds, and because of generally low concentrations
of particles in the marine boundary layer (MBL), any effects are likely to be confined to the
particles in the free troposphere. The indirect effect of aerosols is their influence on cloud
radiative properties. For a given cloud water content, the cloud albedo depends on how finely
the water is divided. This is determined by the concentration of particles on which cloud
drops form – the cloud condensation nuclei (CCN). The indirect effect remains one of the
large uncertainties in climate models and information is needed on the nature and sources of
the CCN. The ability of a particle to act as a CCN depends on the amount of soluble material
it contains and the rate of cooling of the air causing the cloud. Fully soluble particles over the
central Arctic probably need to be at least 80nm in diameter to act as CCN (Bigg E. K. and C.
Leck 2001b).
*
Corresponding author: arash.gharibi@nuclear.lu.se
1.1 Atmospherics aerosol particle sources in the Arctic Ocean
The particulate matter in the ocean consist of both living and non-living (from biological
point of view) creatures. The non-living includes both organic and inorganic component.
Most of the particulate matter in the ocean water is of biogenic origin, thus biotic processes
and transformation is responsible for the type, size distribution, abundances, and chemical
composition of particle. This particulate matter is eventually will end up in the atmosphere as
atmospheric aerosol particles. Generally, the sources of atmospherics aerosol particle, in
Arctic, are crust mineral from river runoff, sea salts, eolian dust, DMS, and biota. Rivers
transport clay minerals and debris from weathered rocks to the Arctic Ocean. The major
contribution of aerosol to the world ocean comes from river runoff (Barrie, L.A., Hoff, R.M.
1985).To make use of particulate matter transferred by river runoff to ocean by
microorganisms requires the transfer of minerals from hydrosphere to biosphere. E.g. diatoms
build up their skeletons from the product of these materials. The transfer of minerals from the
marine environment to the atmosphere initiates the biological cycle of the minerals. After
these microorganisms die the particulate matter accumulated in them dissolves. The
dissolution results particles with different sizes. The large one will sinks to the deeper water
column while small one eventually will be transported to water surface. This is the process
that involves marine biota contribution to the Arctic aerosol. It is well known that oceanic
emissions of dimethyl sulfide (DMS) to the atmosphere are a major source of atmospheric
aerosol particle (Leck, C. et al. 1996).
2. Experimental set-up
Four low pressure impactors LPIA, LPIB, SDIA and SDIB, providing highly resolved particle
sizes were used on the expedition. The LPI impactor consists of 13 impaction stages and the
experimental cut-points d50-values 0.030-10.33 m. SDI is a 12 stage multi-jet low pressure
cascade impactor which can classify the aerosol particles into 12 size fraction (d50-values
0.045-8.39 m) according to their aerodynamic diameter. Both LPI and SDI sample the
aerosol particles through inertial impaction onto filters.
2.1 The sampling system
The data discussed in the present paper were all obtained from aerosol collectors and
instruments that were connected to a sampling manifold that was equipped with an equivalent
aerodynamic diameter (EAD) of 10 m cut-off. The SFU sampler collected particles on two
sequential 47-mm diameter polycarbonate filters a coarse mode filter of 8- m pore size and a
fine mode filter of 0.4 m pore size. The coarse filter had a 50% collection efficiency at 2 m
EAD, and thus collected particles in the size range of 2 to 10 m EAD. The fine filter
collected all particles less than 2 m EAD. The SFU operated at a flow rate of 32 (average)
Lmin-1. Twenty two samples were collected with this device. The SDI and LPI were operated
at a flow rate of 10 L min-1. The collection foils in SDI consisted of a very thin AP1 film
whereas LPIA7 to LPIA14 consisted of a 1.5 m thick KIMFOL film and LPIA1 to LPIA6
consisted of a 1.5 m thick Nuclepore polycarbonate film. The collection time per SDI
sample was 20- 60 hours while for LPI was 10-75 hours. A total of 9 and 14 samples were
collected for SDI and LPI respectively.
Sampler
Inlet
Stages
Flow (l/min)
Backing
LPIA(1-6)
LPIA(7-14)
SDI
SFUA
10 m
"
"
"
13
"
12
2
10
"
"
32
Nuclepore
Kimfol
AP1
Nuclepore
Table 1. Summary of sampler properties.
2.2 The PIXE analyses
A proton beam of 2.55 MeV was used in PIXE analyses (Macro beam-line). The beam size
was 0.63 mm2 and 0.92 mm2 for SDI, LPI and SFU samples, respectively. The PIXE spectra
were accumulated with an HPGe X-ray detector with an absorber of 1mm hole (myler). The
spectra were collected for a proton beam preset charge of 25 C to 50 C. Finally, the
resulting X-ray spectra were evaluated with GUPIX a fitting program which converts peak
areas to elemental concentrations. Details of the Macro beam line set up and PIXE spectrum
analysis and quantification procedures can be found in [5].
3 Results
The AOE01 aerosol study program covered the following Arctic region, MIZ, OW, PI 3 Ice
Drift and PI4. Tables 2a and 2b describe effective sampling time, the overlap of sampling time
with the different collectors, wind velocity, and marine transport time (Mtr, the time since last
contact with open water). The Ice Drift, the ship was moored to an ice floe, started on 1
August near latitude 89.0o N, longitude 1.8o E and ended on 22 August 88.2o, 9.4o W. 
When using different samplers, with different degrees of overlap in time, some scatter in the
results is inevitable. Figures 1, and 2 using chlorine (Cl) determinations as an indicator, show
the extent of this scatter. The meteorological conditions often changes rapidly and drastically
in Arctic Ocean. Therefore most of the scatter must have been due to the incomplete
correspondence in sampling time. Figure 3 compares the sulfur mass size distributions
obtained with two SDI impactors operating in parallel for the same sampling time. The
agreement suggests that errors in the determination of this element were small.
Total
1000
SFU
LPIA
SDIB
M(ng/m 3)
SDIA
100
Cl
10
1
0
2
4
6
8
10
12
14
Figure 1. Atmospheric Cl concentration for overlapped sampler obtained from various
aerosol collection devices.
Total
1000
LPIA
SDIA
SDIB
M (ng/m 3)
SFU
Cl
100
10
1
0
2
4
6
8
10
12
14
W (m/s)
Figure 2. Atmospheric Cl concentration from various aerosol particles collection
devices as function of median wind speed.
400
SDIA7
350
SDIB7
3
dM/dlogDp (ng/m )
S
300
250
200
150
100
50
0
0,01
0,1
1
10
100
d(µm)
Figure 3. Size distribution of the Arctic aerosol over pack ice has been plotted to show
validity of the data. As expected S show a large peak for fine and a small peak for
coarse mode.
Sample
Station
LPIA
MIZ
"
"
"
"
OW
PI 3
DRIFT
"
"
"
"
"
"
"
"
"
"
"
"
"
"
"
"
PI 4
Sample no.
Mtr Median (h)
LPIA SDIA SDIB SFUA
1
2
3
1
4
2
5
"
6
3
"
4
7
5
8
2
6
"
"
7
9
3
8
"
"
9
10
4
10
"
"
11
11
5
12
"
"
13
12
6
14
"
"
15
13
7
7
16
"
"
"
17
14
8
8
18
"
"
"
19
"
9
9
20
"
"
"
21
22
LPIA
0
0
>120
>120
>120
62,8
"
72
84
"
44
"
72
"
55,5
"
>120
"
47,6
"
36
"
"
"
Sampling period
LPIA
SDIA
SDIB
(Julian day)
(Julian day)
(Julian day)
186,53
187
187,56 187,9
190,25 190,8
192,38 192,9
192,93 193,4
196,58 197,6
"
"
207,96
209
214,42 215,8
214,42 215,8
"
"
"
"
216,72 218,7
216,8 218,7
"
"
"
"
218,82 220,8
218,84 220,7
"
"
"
"
221,54 223,6
221,54 223,6
"
"
"
"
224,48 227,5
224,48 227,5
"
"
"
"
227,58 229,5 227,63 229,5 227,63 229,5
"
"
"
"
"
"
229,57 233,4 229,63 231,5 229,63 231,5
"
"
"
"
"
"
"
"
231,63 233,4 231,63 233,4
"
"
"
"
"
"
Eff. sam. time (min)
SDIA SDIB SFUA LPIA
676
455
>120 747
>120 724
>120 675
13,8 1440
72
"
72 1436
84
84 1994
96,2
"
"
44 113 2860
"
44
"
48
48 2788
84
"
"
55,5 49,7 2456
70,5
"
"
>120 >120 3610
"
>120
"
47,6 47,6 >120 2780
47,6
"
"
"
36
36 28,6 4512
"
"
40
"
45,3 45,3 94,8
"
34,5
"
"
"
25,8
SDIA SDIB SFUA
1994
"
2755
"
2715
"
2456
"
3610
"
2703 2703
"
"
2262 2262
"
"
2091 2091
"
"
766
1477
"
725
673
1436
1378
588
1420
1366
1359
1367
1013
1365
2264
1265
1338
1370
1030
1325
1124
1018
1200
SFUA
(Julian day)
190,25
192,38
"
196,58
197,12
207,96
214,42
215,44
216,72
217,76
218,82
219,8
221,54
222,64
224,48
226,56
227,58
228,56
229,57
230,59
231,57
232,69
236,76
190,78
193,4
"
197,09
197,58
208,96
215,42
215,85
217,71
218,71
219,76
220,75
222,58
223,59
226,5
227,5
228,51
229,51
230,29
231,51
232,5
233,43
237,59
Table 2.a
% time average of SFUA by
*=LPIA
*=SDIA
*=SDIB
%=SFU / *
%=SFU / *
%=SFU / *
Wind Speed (m/s)
LPIA
min
median
SDIA
max
3
7,2
10,9
min
median
SDIB
max
min
median
SFUA
max
min
median
max
3,1
4,9
6,1
(3-1) 98
1,8
4,6
7,3
1,8
4,6
7,3
(4-2) 49
2,6
5,1
6,9
2,6
5,9
9,1
(5-2) 46
4,4
7,8
9,1
"
"
"
(6-3) 50
8,7
11,3
13,5
8,7
11
13,3
13,5
(6-4) 47
"
"
"
9,5
11,6
(7-5)100
3,1
6,3
9,6
3,1
6,3
9,6
0,7
9,9
15,2
15,2
7,6
12,8
15,2
(8-6) 69
(2-6) 69
(8-7) 29
(2-7) 29
"
"
"
"
"
"
0,7
3,1
7,9
(9-8) 51
(3-8) 52
2,7
7,1
10,5
2,7
7,1
10,5
5,1
8,1
10,5
0,7
9,9
(9-9) 48
(3-9) 50
"
"
"
"
"
"
2,7
5,8
9
(10-10) 49
(4-10) 50
1,2
7,3
13
1,7
7,5
13
5
8,7
13
(10-11) 49
(4-11) 50
"
"
"
"
"
"
1,2
4,4
10,8
(11-12) 41
(5-12) 41
0,3
5,2
9,6
0,3
5,2
9,6
0,3
4
8,5
(11-13) 56
(5-13) 55
"
"
"
"
"
"
2,3
6,5
9,6
(12-14) 63
(6-14) 63
1,1
3,3
9,1
1,1
3,3
9,1
1,1
3,4
9,1
(12-15) 35
(6-15) 35
"
"
"
"
"
"
1,4
3,1
5,6
4,7
6,6
11,9
6,7
11,9
4,7
8,05
11,9
4,1
5,9
10,1
5,2
9,4
0,4
4,8
6,1
1,7
6,4
9,4
(13-16) 48
(7-16) 50
(7-16) 50
(13-17) 49
(7-17) 51
(7-17) 51
"
"
"
"
(14-18) 23
(8-18) 45
(8-18) 45
0,2
5
10,2
0,2
(14-19) 30
(8-19) 58
(8-19) 58
"
"
"
"
(14-20) 25
(9-20) 54
(9-20) 54
"
"
"
0,9
4,8
10,2
0,9
4,8
10,2
0,9
3,3
5,8
(14-21) 23
(9-21) 49
(9-21) 49
"
"
"
"
"
"
"
"
"
5,4
7,5
10,2
0,5
3
6,3
5
6,7
11,9
5
5,2
9,4
0,2
"
"
Table 2.b.The number inside brackets is a sum of % time average of two or more impactor.
For example (8-6) 69 and (8-7) 29 means the running time of SFUA6 and SFUA7 together
cover 98 % of the running time of the LPIA8.
3.1 Source profile
To identify the probable source of atmospheric aerosols in the Arctic, we calculate the ratio of
the concentrations relative to Si both for our result and a crust material taken from a reference
(Stuart Ross Taylor, Scott M. McLennan 1985). From figure 4, this suggests the source
profile all studied elements matches with crust material.
Relative contribution of species to source profile
Source profile (ultrafine+fine)
100
SDIB02
SDIB03
10
SDIB04
SDIB05
1
SDIB06
SDIB07
0,1
SDIB08
SDIB09
0,01
SDIA07
0,001
SDIA08
SDIA09
0,0001
Soil
0,00001
Si
S
Cl
K
Ca
Ti
Fe
Zn
Br
Figure 4. Source profile for aerosol particle compared to soil.
3.2 Log-normal fitting (results)
Log-normal fitting of size distributions is a major tool for structural analyses because it
indicates the appearance of relative maxima while reducing the number of distribution
parameters (Jost Heintzenberg et al.). The results of best-fit log-normal distribution as
obtained by minimising 2 are shown in table 3. Previous studies have shown that in addition
to “nucleation” and “Aitken” modes < 50nm, there is always an “accumulation mode” at
around 120nm. Modes present in mass distributions >0.1m obtained by the SDI impactors,
and their numerical parameters are shown for all detected elements in Table 3. The
accumulation mode and a coarse mode were present in most. The estimated mass
concentration of the aerosol particles was comparable to that obtained by PIXE analysis.
Figure 5 shows these diagrammatically for selected elements in one sample.
1000
SDIB02
Si
S
Cl
K
Ca
Fe
M (ng/m3)
100
10
1
0,1
0,01
0,1
1
10
100
Dg (m)
Figure 5.The best-fitting, log-normal distribution on SDIB02 data for elements Si, S,
Cl, K, Ca, and Fe. Two modes Dg corresponding to 0.1-0.3 m and 0.3-1.0 m
particles is evident.
3.3 Marine Transport Time (Mtr)
The time since the last contact with open water is defined as Mtr. The open waters of the
Arctic Ocean south of the ice edge are biologically very productive in summer, leading to
much DMS production and sulfate in aerosols. Sea salt is also produced by wave action. Over
the pack ice, both DMS and sea salt production are greatly reduced, (Leck and Persson, 1996)
while the frequent fog and low cloud of the marginal ice zone and pack ice rapidly reduce the
concentrations of all soluble particles. Transport time from open water is important since it is
a rough measure of the balance between loss and production of particles. Figures 6a and 6b
show the way in which chlorine and sulfur mass change with transport time. An initial steep
decline is seen during the first 40 hours of transport, followed by an increase and levelling
out. Similar trends were found in the concentration of cloud condensation nuclei by Bigg and
Leck (2001). During summer, there is infrequent atmospheric transport of air within the MBL
to the central Arctic Ocean, so that concentrations of anthropogenic pollutants are usually
low. However, there is more frequent transport of air to the central Arctic from lower latitudes
in the free troposphere, much higher number concentrations of particles >0.3m than in the
MBL having often been recorded from a helicopter during AO-01 (M. Tjernström et al.).
Total
1000
LPIA
SDIA
M (ng/m 3)
SDIB
SFU
S
100
10
1
0
20
40
60
Mtr (h)
80
100
120
140
Total
1000
LPIA
SDIA
M (ng/m 3)
SDIB
SFU
Cl
100
10
1
0
20
40
60
80
100
120
140
Mtr (h)
Figure 6a and 6b.Total mass concentration (ng/m3) of Cl and S for all samplers, SDI,
LPI and SFU, as a function of time since contact with open water (Mtr).
10000
4
Enrichment Factor VS Cl
SDI
5
6
1000
S
100
10
1
0,01
0,1
1
10
100
d(µm)
Figure 7.Enrichment factor for S versus Cl in sea salt for three SDI impactor
samples taken over pack ice. At sizes greater than 2 m, coarse mode, the particles
are almost composed of primary sea salt aerosol. The reference value for Cl in sea salt
is taken from J.P. Rileyand G. Skirrow 1975.
Cl
K
Ca
Ti
Fe
Zn
Br
SDIB03
m(ng/m3)
Dg (m)
sg(m)
SDIB02
223,09 151,06
0,14
0,11
1,12
1,31
Si
S
10,39
0,11
1,37
1,15
0,14
1,14
2,45
0,14
1,14
0,65
0,11
1,32
0,72
0,11
1,21
0,45
0,21
3,49
0,03
0,13
1,20
m(ng/m3)
Dg (m)
sg(m)
118,63 1200,0 19,18
0,27
0,3
0,25
2,34
1,3
1,14
m(ng/m3)
Dg (m)
sg(m)
241,02 114,21 142,78
1,18
0,67
0,85
1,71
1,43
1,30
13,48
1,33
1,59
14,84
1,31
1,57
1,69
0,86
1,17
3,43
1,23
2,60
4,71
0,81
1,10
0,00
0,47
1,02
m(ng/m3)
Dg (m)
sg(m)
28,50
0,95
1,31
300,0
1,4
1,3
m(ng/m3)
Dg (m)
sg(m)
264,40 450,16
4,31
1,82
1,44
1,34
0,43
3,46
5,00
0,00
0,00
0,00
0,56
3,19
1,75
0,12
1,42
1,75
m(ng/m3)
Dg (m)
sg(m)
86,45
2,57
2,25
160,0
3,5
2,6
3
mj tot(ng/m ) 464,12 529,67 603,33
mi tot(ng/m3) 416,41 540,04 625,12
SDIB04
Si
S
Cl
14,62
17,30
2,77
4,15
5,72
0,15
17,95
K
21,75
Ca
2,80
Ti
4,83
Fe
4,02
Zn
0,15
Br
Si
S
40,55
0,10
1,53
6,91
0,18
1,70
2,47
0,14
1,70
1,60
0,13
1,52
1,52
0,13
1,32
0,40
0,17
1,42
0,16
0,14
1,71
0,00
0,18
1,01
m(ng/m3)
Dg (m)
sg(m)
255,34 657,36
0,21
0,28
4,11
1,92
m(ng/m3)
Dg (m)
sg(m)
65,09
0,76
1,55
39,03
1,07
2,46
60,28
1,58
1,50
2,13
1,58
1,50
3,50
1,54
1,83
0,30
1,54
1,83
1,00
1,38
2,11
0,25
1,18
1,73
0,01
2,79
1,23
m(ng/m3)
Dg (m)
sg(m)
29,86
0,80
1,44
m(ng/m3)
Dg (m)
sg(m)
77,05
3,24
2,18
m(ng/m3)
Dg (m)
sg(m)
47,71
3,68
1,77
67,19
4,60
5,10
1,82
1,40
0,41
0,01
74,13
Cl
4,03
K
4,52
Ca
2,04
Ti
1,41
Fe
0,51
Zn
0,02
Br
0,39
0,78
4,85
UDL
m(ng/m3)
Dg (m)
sg(m)
24,82 117,45
0,11
0,30
1,55
1,81
6,76
0,19
5,00
2,47
0,69
3,75
2,38
0,15
3,46
1,64
0,25
1,76
0,41
0,06
5,00
m(ng/m3)
Dg (m)
sg(m)
157,53 41,89
0,90
2,15
5,00
1,80
51,55
2,08
1,72
1,59
2,70
1,48
4,80
2,23
1,58
0,25
1,17
1,11
0,95
1,08
5,00
mj tot(ng/m ) 182,35 159,34
mi tot(ng/m3) 167,47 181,12
SDIB08
Si
S
58,32
4,06
7,17
1,89
1,36
0,39
60,85
Cl
4,26
K
7,88
Ca
1,98
Ti
1,17
Fe
0,39
Zn
3
3
m(ng/m )
Dg (m)
sg(m)
6,20
0,61
5,00
m(ng/m3)
Dg (m)
sg(m)
9,62
0,40
1,52
0,16
0,17
4,80
0,56
1,01
1,88
3,22
1,86
1,63
0,29
0,93
4,94
0,42
2,15
2,76
UDL
1,75
0,78
4,91
0,42
0,17
1,23
Br
0,02
0,93
4,92
Fe
Zn
Br
0,68
0,15
2,64
0,10
0,15
1,75
0,01
0,14
1,24
2,00
2,12
1,80
0,79
0,63
1,10
0,47
0,77
1,41
0,11
0,70
1,20
0,07
1,33
1,33
0,29
1,52
1,15
2,02
5,11
1,62
0,41
1,58
5,00
0,06
4,49
1,14
6,00
1,33
3,17
0,62
0,14
5,58
Ca
1,34
Ti
3,04
Fe
0,53
Zn
0,11
Br
1,43
0,10
1,64
3,82
0,37
1,52
0,34
0,16
1,23
0,27
0,15
2,92
0,04
0,31
1,56
1,71
1,22
5,00
0,40
1,35
1,52
3,82
2,05
0,67
0,04
3,95
Ti
1,63
Fe
0,73
Zn
0,04
Br
98,50
1,90
1,81
7,47
0,14
2,22
12,07
1,28
2,20
27,54 236,05
1,91
1,64
1,85
1,65
10,60
1,22
2,41
3
m(ng/m3)
Dg (m)
sg(m)
83,946
0,74
5,00
191,1 2,909 4,215 2,095
0,40
0,27 1,27 2,15
1,86
4,80 2,22 1,76
UDL
27,908 9,9083 0,343
0,78
0,78
1,31
4,91
4,91
1,92
11,83 50,81
1,86
1,86
1,49
1,60
mj tot(ng/m3)
83,95 202,88 53,72
4,21
2,09
UDL
27,91
9,91
0,34
mi tot(ng/m3)
4,01
Si
12,08
S
2,91
Cl
0,22
K
0,12
Ca
0,01
Ti
1,29
Fe
0,52
Zn
0,02
Br
5,88
0,69
5,00
5,25
0,45
1,50
0,18
0,18
4,80
0,20
1,35
5,00
0,11
0,28
1,41
UDL
1,85
0,78
5,00
0,55
0,78
4,91
0,02
1,31
1,92
0,28
2,31
1,61
3,32
3,46
1,60
0,02
SDIB09
3
m(ng/m )
Dg (m)
sg(m)
m(ng/m3)
Dg (m)
sg(m)
0,74
2,52
5,00
Ti
0,25
0,32
1,14
mj tot(ng/m ) 332,90 684,91 243,52 12,07 12,03
mi tot(ng/m3) 293,94 726,40 237,69 14,87 11,96
SDIB07
Si
S
Cl
K
Ca
m(ng/m3)
Dg (m)
sg(m)
UDL
Ca
4,00
0,46
1,35
3
88,66
0,14
1,13
3
K
4,34
2,02
2,57
mj tot(ng/m ) 233,57 1660,0 117,67 4,34
mi tot(ng/m3) 233,11 1321,2 130,38 5,40
SDIB05
Si
S
Cl
K
m(ng/m3)
Dg (m)
sg(m)
mj tot(ng/m ) 230,81 79,58
mi tot(ng/m3) 187,12 80,05
SDIB06
Si
S
Cl
0,08
1,56
2,83
mj tot(ng/m3)
6,20
10,17
3,39
0,29
0,42
UDL
1,75
1,15
0,02
mj tot(ng/m3)
5,88
5,53
3,50
0,20
0,19
UDL
1,85
0,55
mi tot(ng/m3)
5,76
Si
10,87
S
3,69
Cl
0,29
K
0,39
Ca
0,93
Zn
0,01
Br
mi tot(ng/m3)
Ti
1,69
Fe
SDIA08
5,70
Si
5,68
S
3,63
Cl
0,19
K
0,19
Ca
0,01
Ti
1,83
Fe
0,67
Zn
Br
m(ng/m3)
Dg (m)
sg(m)
72,47 224,45
0,11
0,32
1,49
1,79
8,88
0,16
1,49
6,47
0,51
4,64
2,97
0,13
1,32
0,26
0,13
1,32
1,99
0,33
5,00
0,79
0,46
5,00
0,04
0,21
1,72
m(ng/m3)
Dg (m)
sg(m)
2,25
0,14
2,74
9,44
0,39
1,47
0,10
0,13
2,00
0,17
0,59
5,00
0,25
0,96
2,34
UDL
1,88
0,58
5,00
0,65
0,72
5,00
0,02
0,45
5,00
m(ng/m3)
Dg (m)
sg(m)
80,22
0,60
1,87
51,42
1,80
1,66
2,10
1,59
2,36
0,20
1,59
4,86
0,00
2,03
5,00
m(ng/m3)
Dg (m)
sg(m)
0,58
0,86
1,32
1,15
1,24
1,27
2,97
1,95
1,70
m(ng/m3)
Dg (m)
sg(m)
32,94
3,45
1,57
m(ng/m3)
Dg (m)
sg(m)
3,40
2,79
4,67
SDIA07
16,81
1,95
1,89
1,26
4,10
1,16
mj tot(ng/m3) 185,62 241,26
60,30
6,47
5,07
0,46
1,99
0,79
0,04
mj tot(ng/m3)
6,23
10,60
3,07
0,17
0,25
UDL
3,14
0,65
0,02
mi tot(ng/m3) 184,95 252,68
SDIA09
Si
S
61,25
Cl
6,66
K
5,25
Ca
0,39
Ti
1,84
Fe
0,83
Zn
0,04
Br
mi tot(ng/m3)
5,27
10,85
3,07
0,22
0,25
0,02
3,45
0,68
0,02
0,44
1,72
2,65
0,41
3,48
3,41
0,08
1,49
5,00
1,98
0,75
5,00
0,76
0,88
5,00
0,04
0,50
5,00
3
m(ng/m )
Dg (m)
sg(m)
9,56
0,62
5,00
4,94
0,29
1,47
0,37
0,03
5,00
m(ng/m3)
Dg (m)
sg(m)
1,81
2,99
1,06
1,51
0,48
2,77
4,06
1,84
1,78
mj tot(ng/m3)
11,37
6,44
4,43
0,44
0,98
0,08
5,61
0,76
0,04
mi tot(ng/m3)
9,14
6,42
4,23
0,51
1,19
0,09
4,14
1,34
0,04
0,57
2,81
1,11
3,63
2,74
1,15
Table 3.Log-normal Size Distribution Parameters for Si, S, Cl, K, Ca, Ti, Fe, Zn, and Br.
The mass concentration mj (ng/m3) and mi (ng/m3) obtained from the fitting and experiments
are shown in last two rows. The Dg (Geometric mean diameter, in m) and sg (standard
deviation) of each mode are present in row two and three.
ALL SDI
ALL MODE
Enrichment Factor VS Si
Min
Max
K
10(-1)
10(1)
Ca
10(0)
10(-2)
Ti
10(-2)
10(0)
Mn
10(-1)
10(-1)
Fe
10(-3)
10(0)
Ni
10(-1)
10(1)
Zn
10(0)
10(1)
Table 4.Enrichment factors calculated for K, Ca, Ti, Mn, Fe, Ni and Zn versus Si in soil dusts
for all SDI samples in pack ice region of the Arctic Ocean.
3.4 Sea Surface Micro-layer (SML)
The sea surface microlayer is a structured interface between the atmosphere and the ocean. It
can serves as both a sink and a source of anthropogenic compounds, including chlorinated
hydrocarbons, organic compounds, polycyclic aromatic hydrocarbons (PAH) and heavy
metals due to its unique chemical composition (Oliver Wurl 2004). Gradients of temperature
and gases and enrichments in organic matter and organisms occur in the SML. The SML
comprises a series of sub layer namely, surface nanolayer (<1m), surface microlayer
(<1000m) and surface millilayer (<10mm) (Marianne Grammatika et al.). The thickness of
SML depends on both the amount of organic matter and the wind speed (Liu, K. & Dickhut,
R. 1998)
3.5 The marine silica source and cycle
Silicon is an important component of the marine biogenic matter. Planktonic organisms
include diatoms, radiolarians, and silicoflagellates. Planktonic organisms use silicon to build
up their skeletal shell (skeleton). The net delivery of silicic acid, Si (OH)4, to the ocean has
thee pathways:
1. Chemical weathering of sedimentary and rocks (River runoff)
2. Eolian erosion of the land surface
3. Seafloor weathering
The major input of Si to the world ocean is through dissolved silicate carried by rivers from
continental weathering (Paul Treguer1995).
3.5 Comparison with other Arctic data sets
Median atmospheric concentrations (ng/m3 STP) during AOE01; comparison with IAOE96
and IAOE91. The compared elements are Si, S, Cl, K, Ca, Ti, Fe, Zn, and Br. The data were
derived from the SFU data where correspond to pack ice in the Arctic Ocean. The IAOE96
and IAOE91 data used in the table below were taken from Maenhaut W. et al. [3].
Our results are, in most elements, higher in elemental concentration than the results from
previous expedition. The differences may reflect the seasonal variation of concentration for
these elements in the Arctic region.
AOE01
AOE96
IAOE91
AOE01
AOE96
IAOE91
Coarse
Fine
Coarse
Fine
Coarse
Fine
Coarse
Fine
Coarse
Fine
Coarse
Fine
Si
50,59
77,71
1,9
1,6
17,4
4,6
Ti
0,31
0,75
0,18
0,44
0,2
S
31,95
37,79
17
26
6
26
Fe
1,89
1,26
1,63
0,42
10,03
0,57
Cl
18,12
17,27
46
60
13,6
4
Zn
0,50
0,36
1,03
0,08
0,11
0,07
K
1,92
2,49
2,5
2,3
0,98
1,12
Br
2,85
2,99
Ca
1,54
3,79
2,1
2,4
2,24
0,76
0,43
0,144
0,106
Table5.
4. Discussion
In open water whitecaps are responsible for bubble formation and breaking waves are
responsible for bursting them. This is the process behind particle production open seas. In the
small open leads e.g. pack ice region in Arctic Ocean whitecaps are produced but breaking
waves are not strong enough to break them because of short space between the leads. Our
results show the total mass concentration of Cl is increasing with increasing of wind speed
(see figure 2) in pack ice region (Leck, Norman Bigg and Hillamo 2001) thus particle
production does occur in the vicinity of the open leads. It is well known that total particle
concentration logM is linearly related to wind speed. Therefore in open leads particle
production will take pace if the wind speed is high enough. In Arctic region whitecaps
production occur and breaking them also take place, now the question is what material this
process will injects to the atmosphere. When bubbles burst two clearly different sizes of
particles will be produced. There are jets and film drops and they form large and small
particles respectively. Jet and film drops contain the material of the SML and the attached
particulate matter to the bubbles while they were moving towards the surface. A size of the
bubble from a breaking wave varies from 100 m (detection limit) to a few mm (Deane
1997). The size of the film droplets is approximately between 0.1 to 0.3 m while the jet
droplets have size between 0.3 to 1.0 m. Generally the larger the original bubble, the larger
the particle. The results from enrichment factors and source profile calculation show elements
as K, Ca, Ti, Mn, Fe, Ni and Zn seem to come from crusted material (See table no. 4 and
figure no 4). The explanation is crusted material which is attached to ice can be transport to
Arctic via Arctic current (for example river runoffs in Siberia). In summer, when the ice melts
the particulate matter from the crusted material will be release either into the bulk water in the
sea or in the SML. Then it is taken up by planktonic organisms to further natural processing.
It is not conceivable that this material comes from wind blown soil since the contribution of
particulate matter from Eolian erosion to world ocean is very small (Paul Treguer et al. 1995)
If the composition of the biota in Arctic Ocean were know it can be also interpreted as the
trace elements come from biota. Since there is no data available on the concentration of the
trace elements in the ocean biota therefore our hypothesis is that these elements may derive
from biota. The materials in the SML can be enriched by up to 10000 times relatives to
concentration occurring in the bulk water emphasise this conclusion.
Acknowledgements
We would like to thank all the staff of PIXE group at Lund University.
We owe thanks to Swedish Polar Research Secretariat, which provide us a fantastic
environment to collect our data on icebreaker ODEN during Arctic Ocean 2001 expedition.
References
[1] www.fysik.lu.se/eriksw
[2] www.polar.se
[3] Maenhaut W et al. Tellus (1996), 48B, 300-321.
[4] Hillamo R et al. J. Geophys. Res. Vol. 106, No. D21, 27555-27571, Nov. 16, 2001.
[5] Asad Shariff “Development of New Experimental Facilities at the Lund Nuclear
Microprobe Laboratory”, Ph-D thesis (2004).
[6] Leck, C. and Bigg, E. K. 1999. Aerosol production over remote marine areas(a new route).
Geophys. Res. Lett. 23, 3577-3581
[7] Jost Heinyzenberg et al. “Structure, variability and persistence of the sub-micrometer
marine aerosol”, Tellus (2004), 56B, Issue 4 Page 357.
[8] Baker, A. R., S. D. Kelly, K. F. Biswas, M. Witt, and T. D. Jickells (2003), Atmospheric
deposition of nutrients to the Atlantic Ocean, Geophys. Res. Lett., 30(24), 2296.
[9] Leck, C. et al., Overview of the atmospheric research program during the
International Arctic Ocean Expedition of 1991 (IAOE-91) and its scientific results .
Tellus (1996) 48B, 137-155.
[10] Leck, C. and C. Persson, Seasonal and short-term variability in dimethyl sulfide, sulfur
dioxide and biogenic sulfur and sea salt aerosol particles in the arctic marine boundary
layer, during summer and autumn. Tellus (1996) 48B, 272-299.
[11] Leck, C., and E.K. Bigg, 1999, Aerosol production over remote marine areas - A new
route, Geophys. Res. Lett., 23, 3577-3581.
[12] Bigg, E.K., C.Leck and L.Tranvik, Particulates of the surface microlayer of open water
in the central Arctic Ocean in summer, 2004. Marine Chemistry (in press).
[13] Bigg E. K. and C. Leck, Properties of the aerosol over the central Arctic Ocean 2001a. J.
Geophys. Res. 106 (D23) 32101-32110.
[14] Bigg E. K. and C. Leck, Cloud-active particles over the central Arctic Ocean, 2001b. J.
Geophys. Res. 106 (D23) 32155-32166.
[15] Wells M.L. and E. D. Goldberg, 1991. Occurrence of small colloids in sea water.Nature
353:342-344.
[16] Sattler B, Puxbaum H, Psenner R., Bacterial growth in super-cooled cloud droplets,
Geophys. Res. Lett., 28, 239-242, 2001.
[17] Bauer H, Giebl H, Hitzenberger R, Kasper-Giebl A, Reischl G, Zibuschka F, Puxbaum
H., Airborne bacteria as cloud condensation nuclei, J. Geophys. Res., 108, (D21): Art.
No. 4658 NOV 4 2003.
[18] Giorgos Papaspiropoulos et al. J. Geophys. Res., Vol. 107, No. D23, 4671, 2002.
[19] Oliver Wurl et al., A review of pollutants in the sea-surface microlayer (SML): a unique
habitat for marine organisms, Marine Pollution Bulletin 48 (2004) 1016-1030.
[20] Marianne Grammatika et al., Micohydrodynamics of flotation processes in the sea
surface layer, Dynamics of Atmospheres and Oceans 34 (2001) 327-348.
[21] Robert A. Duce, Peter S. Liss, The surface ocean-lower atmosphere study (SOLAS),
Atmospheric Environment 36 (2002) 5119-5120.
[22] Liu, k.& Dickhut, R., Effects of wind speed and particulate matter sources on surface
microlayer characteristics and enrichment of organic matter in southern Chesapeake Bay,
Journal of Geophysical Research 103 (1998), 10571-10577.
[23] Paul Treguer et al., The Silica Balance in the World Ocean, Science Vol. 268, 21 April,
1995
[24] Svein Kristiansen et al., Production of Biogenic Silica by Spring Diatoms,
(2000) Limnology and Oceanography Vol. 45 472-478)
[25] Barrie, L.A., Hoff, R.M., Fine years of air chemistry observations in the Canadian Arctic,
Atmospheric Environment (1985) 19, 1995-2010.
[26] Barrie, L.A. et al., Arctic air pollution: an overview of current knowledge, Atmospheric
Environment (1986) 20, 643-663.
[27] Stuart Ross Taylor, Scott M. McLennan, The Continental Crust: its Composition and
Evolution, Blackwell Scientific Publications (1985)
[28] J.P. Rileyand G. Skirrow, Chemical Oceanography; Academic Press 1975
Download