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 2m <EAD<10m and the fine fraction as <2m. 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 >0.1m 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 >0.3m 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 (<1m), surface microlayer (<1000m) 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