Brine Exclusion Process from Growing Sea Ice* Masaaki W AKA TSUCHI :;s: ± le B~ The Institute of Low Temperature Science Received December PDF

Description
TitleBrine Exclusion Process from Growin Author(s) WAKATSUCHI, Masaaki Citation Contributions from the Institute of A33: Issue Date DOI Doc URLhttp://hdl.handle.net/2115/20249 Right Type

Please download to get full document.

View again

of 20
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Jobs & Career

Publish on:

Views: 20 | Pages: 20

Extension: PDF | Download: 0

Share
Transcript
TitleBrine Exclusion Process from Growin Author(s) WAKATSUCHI, Masaaki Citation Contributions from the Institute of A33: Issue Date DOI Doc URLhttp://hdl.handle.net/2115/20249 Right Type bulletin Additional Information File Information A33_p29-65.pdf Instructions for use Hokkaido University Collection of Scholarly and 29 Brine Exclusion Process from Growing Sea Ice* by Masaaki W AKA TSUCHI :;s: ± le B~ The Institute of Low Temperature Science Received December 1983 Abstract The exclusion process of brine from growing sea ice, which plays an important role in the polar ocean, has been quantitatively studied in the field and laboratory in which sea ice grew under various conditions. From oceanographic observations during the growth of sea ice in the Antarctic winter it was found that a haline convection induced by brine exclusion changed the salinity structure of shelf water from stratified to homogenized distribution, yielding the salinization of the water. The appearance of the excluded brine falling into the seawater was experimentally observed by means of a schlieren optical system, with the following findings: The excluded brine formed long vertical filaments starting at fixed positions on the water ice interface and fell into the underlying seawater without appreciable diffusion. The characteristics of the brine filaments varied with the ice growth rate. At a higher ice growth rate, a larger number of thinner brine filaments fell at a lower velocity. Direct measurements were made of the salinity and volume of the excluded brine under various conditions of ice growth. The brine salinity increases and the volume flux decreases as the ice growth rate decreases. Consequently, the salt flux of the brine decreases with decreasing ice growth rate and hence the amount of salt excluded as brine depends highly upon its volume rather than its salinity. For sea ice growth rates between 1.7 x 10-5 and 1.4 x 10-4 cm. s- and for a seawater salinity of 33.0%0, the brine salinity ranged from 42.3 to 92.7 %0, the volume flux ranged from 6.3 X 10-6 to 3.4 X 10-5 cc. cm- 2 s- and the salt flux ranged from 6.4 X 10-7 to 1.5 X 10-6 g. cm- 2 S-I. The total volume of brine excluded during one sequence of ice formation depends upon the duration of formation as well as the ice growth rate. It increases with an increase in both growth rate and duration; that is, with an increase in the mass of ice grown. The brine exclusion process thus disclosed sheds light on a change in salinity of natural sea ice under various conditions of ice growth. When it takes a longer time for a sea ice to grow to a fixed thickness, the sea ice has a lower salinity. This is due to the exclusion of a larger amount of brine with a higher salinity during the formation of sea ice. Meanwhile, a thicker sea ice grown during a fixed period of time has a higher salinity. This is due to the exclusion of a smaller amount of brine with a lower salinity per unit mass of ice grown during the period. The foregoing results suggest that the salinity and volume of brine excluded during the formation of sea ice can be estimated approximately in future by measuring the duration of formation, thickness and salinity of the sea ice. * Contribution No from the Institute of Low Temperature Science ~ti.fijoi1i:**'lifjlt*1jf:~)( 30 Masaaki W AKA TSUCHI Contents I. Introduction II. Salinization of shallow shelf water during ice growth process 31 II. 1. Pool experiment II. 2. Oceanographic observations III. 3. Cabbeling effect III. Sea ice characteristics III. 1. Structure of sea ice III. 2. Desalination of sea ice III. 3. Relation between sea ice salinity and ice growth rate IV. Optical observation of brine streamers IV. 1. Experimental apparatus and measurement technique 42 IV. 2. Observation results V. Measurements of salinity and volume of brine streamers falling into seawater V.I. Basic model V. 2. Apparatus and experimental procedures V. 3. Analytical method V. 4. Effect of ice growth rate 55 V. 5. Effect of increase in ice thickness V. 6. Brine exclusion process VI. Concluding remarks Acknowledgements References... 65 Brine Exclusion Process from Growing Sea Ice 31 I. Introduction In the polar ocean, Arctic and Antarctic, atmospheric cooling freezes the water surface and forms sea ice on it. The rate of sea ice growth, which changes with such cooling conditions as air temperature, wind speed and radiation, determines the structure of the sea ice thus formed. It is well known that growing sea ice excludes brine which is at low temperature and has high salinity and hence the formed sea ice has salinity less than that of the original seawater. The sea ice salinity also depends upon the ice growth rate. Since the excluded brine is denser than the underlying seawater, the brine exclusion induces a haline convection [Foster, 1969,1972 ; Farhadieh and Tankin, 1972 ; Wakatsuchi, 1977].With progress in ice growth the haline convection is further invigorated, yielding an appreciable salinization of shallow shelf water in the Antarctic winter [Brennecke, 1921 ; Wakatsuchi, 1982J. Gill [1973J points out theoretically that the salinity increased shelf water sinks along the slope of the continental shelf mixing with deeper water, and that Antarctic Bottom Water may be formed through the convective-mixing process. Though the formation process of Antarctic Bottom Water is not yet completely understood, it is now widely believed from the works of Brennecke [1921J and Mosby [1934J that the primary influence upon the formation process is the exclusion of brine from growing sea ice. The present study aims at a quantitative understanding of the brine exclusion process, which has a marked influence upon various phenomena in the polar ocean, as mentioned above. In this connection, the next section II is devoted to the results of observations concerning the salinization, with progress in ice growth, of shallow shelf water in the Antarctic region during the winter. Then, explained in section III are sea ice characteristics, structure and salinity, as a function of ice growth rate. The amount of salt excluded during the ice formation has been estimated through the measurements of salinities of the formed sea ice and the original seawater. Little has been understood, however, about the process of salt exclusion. The observation results about how the excluded brine falls into the underlying seawater, which were obtained with a schlieren optical system, are described in section IV. Finally the results of direct measurements of the salinity and volume of the excluded brine under various ice growing conditions are presented in section V. II. Salinization of shallow shelf water during ice growth process Oceanographic observations under the growing sea ice sheet were carried out near Syowa Station (69 S and 39.5 E) in Ltitzow-Holm Bay, Antarctica, from March 1976 to January 1977, to examine how the salinity of shallow shelf water increases as a result of brine 32 Masaaki W AKA TSUCHI 37'E 200 j.. a 10 I km I 'E Fig. 1 Bathymetry of Ltitzow-Holm Bay. exclusion accompanied with growth of sea ice. Figure 1 shows a bathymetry in Liitzow Holm Bay. This station is located on Ongul Island in the bay. The continental shelf spreads here with the mean width of 60 km and the mean depth of m. Several glacial troughs and narrow channels deeper than 600 m are found near the S6ya Coast [Fujiwara, 1971J. Since the sea surface in this area was already covered with fast ice sheets thicker than 1 m, oceanographic observations during the initial process of freezing were performed in an artificially made pool; it was a square hole made by removing ice blocks of 5 m x 5 m in area from the ice sheet. The opened seawater surface in the pool was exposed to cold air and sea ice newly grew there. In addition to the above observations, seasonal variations in water structure under the fast ice sheet were observed at several stations located over glacial troughs and narrow channels near the S6ya Coast. II. 1. Pool experiment The pool was located over a small basin (about 55 m deep) within an extremely shallow Brine Exclusion Process from Growing Sea Ice 33 MARCH APRIL SEA ICE SALINITY (., ) o 0 0 E E u u ~10 0 ~10 Jl f) f) Jl W 0 W Z z :.:: u 15 0 :.:: u 15 0 I I t- t- Original Seawater W 20 0 W 20 Salinity u 0 0 u 32.2 (a) (b) Fig. 2 lee thickness vs. date of measurement (a) ; Bulk ice salinity vs. thickness (b). continental shelf with a mean depth of about m just near Syowa Station. Sea ice grew to about 25 cm thick in the pool during the observation period from March 18 to April 3 in 1976 and its bulk salinity decreased 10 with an increase in thickness of growing sea ice, as shown in Figure 2. The water salinity E 20 profiles under the sea ice were continuously I I obtained by lowering a portable salinometer a. w o (Type MC5/2, made in England). 30 With an increase in thickness of ice and a decrease in its salinity, the salinity of the underlying seawater increased appreciably, as shown in Figure 3. Although the seawater temperature profiles were also obtained at the same time, they changed little 50 during the observation period and the temper ature of the entire water column was near the MAR.18 -APR o r r-~'-~~al~,-' r- MARCH 18 (start) o 19 (3.5cm).. 20 (6.4) x 21 (10.0) 40 e 24 (13.0) 25 (15.4) 26 (16.6) 27 (20.0) o 28 (21.2) ~ 31 (21.8) APRIL 3 (2~POOL ICE THICKNESS equilibrium freezing point. Because of the Fig. 3 Change in water salinity profiles under strong salinity dependence of density at low growing sea ice in the pool. temperature, salinity profiles, which are shown in Figure 3, can also be regarded as density profiles. To follow simply the sequence of salinization of seawater, the salinity profiles were extracted from Figure 3 at three typical 34 Masaaki W AKA TSUCHI SALINITY ('I,,) ,-----, ,---, o SALINITY ('100) E 20 :r: t- ~ 30 o... under fast ice under pool ice MAR.18 E :r: I- tl. UJ under fast ice under pool ice MAR (a) 50 (b) SALINITY ('I.. ) o ~--~--_ ~---.~--r_--_r 10 E 20 :r: t- tl. UJ under 'fast ice un der pool ice APR (C) Fig.4 Water-salinity profiles under growing sea ice in the pool (0) and under the fast ice sheet (e) at (a) start of surface freezing, (b) midway time when the sea ice grew to about 10 em thick and (c) finish of the observation (ice thickness of about 25 em) in the pool. Brine Exclusion Process from Growing Sea Ice 35 times : the start of surface freezing, a midway point in ice growth and the finish of the observation. They are indicated in Figure 4. It was observed that the salinity structure of the water column in the pool changed from stratified to homogenized distribution and that the depth of layers increasing in salinity lowered. This sea area was observed to have no current and tidal change between the depths 30 and 80 cm during the observation period. Even in case the tide is considered to cause a change in the salinity profile in the pool, the major change in it should be due to convective mixing induced by the exclusion of brine from the sea ice which has been formed there. For reference, the salinity profiles were also observed in a water column under the fast ice sheet located about 10 m apart from the pool. The observation results are plotted in Figure 4 against the data in the pool obtained simultaneously. Since the mature fast ice (about 2 m thick) grew only by about 2 cm in thickness during the observation period, the salinization of the underlying seawater was found little, as expected. II. 2. Oceanographic observations Oceanographic observations were carried out at least once a month from May to December in 1976, using standard Nansen bottles and paired reversing thermometers in a low temperature range (-2 to 10 C). The sea surface at the start of observation had already been covered with an old ice as thick as 80 cm and the ice grew to the thickness of about 2 m by the end of the winter. An ice hole of about 30 cm in diameter was bored with an electric core drill and the obtained ice cores were saved for measuring ice salinity profiles. Water temperature and salinity profiles were measured through the ice hole. The observation results obtained in the Ongul Strait are shown as an example in Figures 5 and 6. Since the old ice kept relatively low salinities (less than 2%0 ) throughout the ob servation period, a distinct boundary is shown between the parts of old ice and newly '976 MAY JUNE JULV AU G. SE P. OCT. NOV. so+---~--~----~--~--~--~--~.... o -~---~ ~ ~---~ ' Fig.;; Seasonal variations in sea ice salinity profiles in the Ongul Strait. The broken line represents the sea level, and the dots indicate locations of ice samples collected. accreted ice in the salinity profiles, as shown in Figure 5. In the latter a relatively high salinity layer produced initially was kept in the original level changing little in salinity, whereas salinities in the underlying layers gradually decreased during the growth process. 36 Masaaki W AKA TSUCHI Since the sea ice originates from seawater with a salinity of about 33.5%0, the salinity decrease would lead to a salinization in the underlying water column, if no lateral advection exists there. In May when the observation started, the seawater column of the Ongul Strait was found to have had a stratified structure concerning salinity already, as shown in Figure 6. In this figure surface water less than 33.5%0 in salinity seems to be a remnant of the layer diluted by ice melting in summer. The figure provides the following interpretation ; from May to July the salinity of the water column above a depth of 200 m gradually increased and the water temperature lowered with progress in ice growth. The stratified structure composed of superimposed multi layers disappeared on July 11 and two homo geneous water masses with the salinities of and 33.93%0 were observed over and under a relatively strong halocline at a depth of 50 m. In particular the latter extended from 50 to 300 m in depth. The salinities of both homogeneous water masses increased further during the period from late July to early September when the ice growth rate was relatively high, as shown in Figure 5. M J J A 5 0 o to 2, i :~!-, :-L-!~l ~: :~:-1'~8:~.... ~ E ;/ L~j~- : -'_~( H_ ~1.4\ 600 \. \.. W C-1j '~d \5 \.\ 'Cal Fig. 6 Seasonal variations in (a) water temperature ee) and (b) salinity (%0) profiles in the Ongul strait. Since the upper water mass increased its salinity at a relatively high rate, the seasonal halocline disappeared in early August and one homogeneous water mass was newly observed on September 9. This homogeneous water mass which ranged between 33.9 and 34.12%0 in salinity reached a depth of 400 m, as shown in Figure 6b. The salinity of the surface water above 400 m increased from 33.87%0 in May to 34.05%0 in September when the ice growth was virtually over with the end of the winter. Meanwhile, the salinity of bottom water under this depth also increased during the ice growth, its maximum attaining to 34.83%0 in mid August. In the Ongul Strait, when a stainless cable with N ansen bottles was lowered to the deepest level (600 m), the cable was at a standstill without any inclination. Current measurements were made within the upper layer over the depth of 50 m during the ice growth but no current was observed throughout the observation period. Therefore, the results of the above observation support that the primary cause of the formation of the homogeneous water Brine Exclusion Process from Growing Sea Ice 37 above the depth of 400 m is the exclusion of brine from growing sea ice ; that the brine exclusion induces a haline convection; and that the homogeneous water is produced mainly through the convection process, increasing the salinity of the upper water mass. It follows then that the maximum thickness of the convection layer should be about 400 m in early September. As for the bottom water it remains unknown as yet whether the salinity increase is caused by the brine excluded from the ice which penetrated through the con vection layer and reached the bottom layer or by the advection of the surrounding saline water. II. 3. Cabbeling effect It is considered from both the results of observations in the pool and the Ongul Strait that the major cause in salinization of ~ ; shallow shelf water is the haline convection UJ induced by the exclusion of brine from ~ «growing sea ice. The following is a qual ita- ~ tive proposal for a possible mechanism of ~ salinization using a T-S diagram indicated in -1 Figure 7. On the basis of Brennecke's observation data [1921], the surface waters in Weddell Sea at the beginning of winter are characterized by a relatively cold, low salinity surface water (referred to as A) overlying a warmer, more saline water (B). Suppose that the Water Mass ~-L ~ ~.~A ~ ~~~c -2 L SALINITY ('/ ) Fig. 7 Temperature salinity diagram. F. T., equi librium line of freezing temperature. De tails in the text. surface water A changes in quality to a water (A') with the same density as that of the underlying water B by lowering its temperature and increasing its salinity on account of the brine exclusion with progress in ice growth. If A' mixes partially with B, then the mixed water should become denser than B regardless of the mixing ratio, and it should sink through B into the deeper layer. Therefore, the depth of the convective mixing layer gradually descends through the above process and the entire shelf water column increases the salinity. III. Sea ice characteristics III. 1. Structure of sea ice To understand the brine exclusion process, we first must study the internal structure of sea ice. Figure 8 shows schematically the structure of sea ice. According to the previous studies [Tabata and Ono, 1957 ; Weeks, 1958J, a normal sea-ice sheet consists of long, vertical 38 Masaaki W AKA TSUCHI AXIS I Z =CONST. PLANE BRINE POCKETS 0 BRINE CHANNEL (a) ICE SURFACE (J r/ SKELTON SEA WATER (b) Fig. 8 Schematic diagrams of sea-ice structure. (a) Microscopic drawing of ice crystals. brine pockets and a brine drainage channel; (b) macroscopic vertical section of treeshaped channels. Ta and Tw representing air and seawater temperatures, respectively. where Ta Tw. Vl Vl UJ 4 z u io UJ U TIME (hours) Fig. 9 Change in sea ice growth rate with air temperature (Ta). Fig. 10 Horizontal sections of sea ices at the depth of 2.2 cm formed at growth rates: (a) cm. h \ (b) 0.33 cm. h-'! and (c) 0.49 cm. h-!. Brine Exclusion Process from Growing Sea Ice 39 ICE THICKNESS ( 2.2 em) z a: 1 ' w ' ~ 0.5 w ool----o~.~1--~o~.2----~o.~3----~----ol.5--~o.6 ICE GROWTH RATE (em.hr-') Fig. 11 Average grain size vs. ice growth rate. Fig. 12 Horizontal section of brine drainage channels, each resembling the shape of a bright star. ice crystals, each usually several centimetres square in horizontal cross section, frequently also much smaller. In addition, each ice crystal consists of parallel pure ice plates in the direction of growth and the thicknesses of ice plates are mm. The crystal structure of sea ice changes markedly with ice growth rate. Sea ice increases its thickness linearly with time during the initial growing process, at least up to 5 cm thick, as shown in Figure 9. The ice growth rate increases with lowering air temperature. With an increase in ice growth rate the grain size becomes smaller (see both Figures 10 and 11). Brine pockets are contained between the ice plates and between the ice crystals. Brine channels distribute frequently in boundary areas where two or more ice crystals intersect with one another [Saito and Ono, 1980; Wakatsuchi et ai., 1982]. Generally a brine channel takes the form of a tree, consisting of a large vertical tubular structure attended by smaller tributary tubes, as shown in Figure 8b. The diameters of them change from less than 1 to several mm. The spatial density of brine channels has been examined by several investi
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks