Continuous surface strain measurements on sea ice and on Erebus ...
this region of strong westerly atmospheric flow, with the weaker velocities near the coast being more variable in direction but predominantly easterly and with cyclonic curvature. Although the model produces some satisfactory results on a large scale, smaller scale features such as the observed tendency for thicker ice at the northern reaches of the Weddell Sea than at the Weddell Sea coast are generally not reproduced. Even on the large scale the simulated thicknesses and extents are dependent on a rather arbitrarily chosen oceanic heat flux employed in the model (Parkinson, 1978). Nevertheless, the current results are encouraging and reveal the feasibility of inserting sea ice into global models of the atmosphere and/or oceans. This work was supported by National Science Foundation grant ATM 76-08492 and was carried out at the National Center for Atmospheric Research with the assistance of the leader of the Global Climate Modeling Group, Warren M Washington.
Continuous surface strain measurements on sea ice and on Erebus Glacier Tongue, McMurdo Sound, Antarctica
D.J. GOODMAN Physics and Chemistry of Solids Group, Cavendish Laboratory, Cambridge, UK R. HOLDSWORTH Department of Physics, University of Waikato, Wellington, New Zealand
We measured the flexural wave energy on the sea ice close to the Erebus Glacier Tongue (a floating glacier in McMurdo Sound) and monitored surface strain changes on the glacier itself. Also, we made continuous surface strain measurements (in a frequency range between 1 hertz and d.c.) close to the annual runway on the sea ice close to McMurdo Station as LC-130 transport planes landed, and vehicles passed along the ice road to the runway. Strains were measured continuously, to an accuracy of 1 in 108 strain, over a 5-meter gage length with a geophysical wire strainmeter (King and Bilham, 1976). Such instruments have been used to measure flexure of fast ice due to swell (Goodman et al., 1975; Squire and Allan, 1977), and surface strain changes on an ice cap, and a valley glacier (Goodman,
October 1978
References
Daniel, H. C. 1957. Oceanographic atlas of the polar seas, part I, Antarctic (HO. Publication 705). U.S. Navy Hydrographic Office, Washington, D.C. Parkinson, C. L. 1978. A numerical simulation of the annual cycle of sea ice in the Arctic and Antarctic (NCAR cooperative thesis 46). National Center for Atmospheric Research, Boulder, Colorado. Untersteiner, N. 1975. Sea ice and ice sheets and their role in climatic variations. In: The physical basis of climate and climate modelling (Global Atmospheric Research Programme publication series 16). World Meteorological Organization and International Council of Scientific Unions, Stockholm. pp. 206-224. Zwally, H.J., and P. Gloersen. 1977. Passive microwave images of the polar regions and research applications. Polar Record, 18: 431-450.
1977; Evans et al., 1978). The strainmeter uses a length of INVAR wire held under constant tension as a length standard, and detects strains with a lever system and a L.V.D.T. The output, in this case, from the L.V.D.T. was amplified and recorded on a data logger, and a strip chart recorder. An automatic rezeroing system allowed the instrument to operate at a high sensitivity, for long periods, unattended. Six instruments were used in two arrays of three. To avoid temperature effects the instruments were covered with snow. Care was taken not to install the instruments across prominent cracks. Six strainmeters were installed close to the ice road near the ice runway on 14 December 1977 and removed on 19 December. LC-130s were using the runway throughout this period, and were landing approximately 1 kilometer from the site. A typical strain record, from three strainmeters in three different directions, obtained when an LC-130 weighing 105,000 pounds landed is shown in figure 1. The clear dispersion of the waves can be seen. The data collected is in a much lower frequency range to that previously collected by Press etal. (1951) or Robinson (1965) as can be seen from the time bar on the figure. The initial oscillations, which can be seen on the record before the waves from the plane landing arrived, are from swell penetrating from the open water. The fast ice was 3.44 meters thick at the site, which was 35 kilometers from the fast ice edge when the record in the figure was collected. Static, and dynamic, records of the strain also were made as vehicles moved along a road close to the array. On 23 December the first strainmeter was installed on the ice tongue, approximately 9 kilometers from the snout; data were collected more or less continously until 22 January 1978 at the initial site. Data were collected at a second site on the glacier between 11 and 26January 1978. A typical example of the data obtained is shown in figure 2a; the power spectra of 4048 s of the time series (with a five point Gaussian smoothing) is shown in figure 2b. The different peaks correspond to the different modes of oscillation of the ice tongue; the driving force for the oscillation is the swell in the sea.
67
L
:1: 1
I 11411urdo Souzd Antarctjc Stxiin Site I ku from to. runway. Isoorda frm three atrujo..ters it 120° to each other after C-130trenspor piai* weIghing 105,000 The ( I 8,000 kg) laMed. I t :U IWS'
I j I
46
I -
I
Figure 1. The strain changes observed by three strainmeters at 120 0 to each other when an LC-130 weighing 105,000 pounds landed on the sea Ice runway about 1 kilometer away. The initial waves are due to swell penetrating through 35 kilometers of fast Ice. 055 0-1281 1811 I0JRN78 Dl BROWN 19(C1) STRAIN DI 7 2.585
065 0-1281 15:11 I0JRN78 GI BROWN BId) POWER SPECTRUM, FREO. STEP .2441E-03 1/SEC
SEC
rn" 60
1701 50
40
11301
ix
4
30
41 1000-000 2050 380000 4000 000 TIME (SECS)
Figure 2a. Part of the time series recorded on the array of strainmeters on the surface of Erebus Glacier Tongue.
While strainmeters were running on the glacier, at different times to the north and the south of the glacier, an array of three strainmeters was installed on the sea ice about 400 meters from the ice front. The array measured the wave energy incident on the glacier, and will enable the transfer of energy between the sea and the glacier to be determined. A typical record of the swell induced flexural waves is shown in figure 3a (the wave amplitude is greater than the waves in figure 1 because the fast ice edge had advanced to a line between the end of the ice tongue and Cape Evans). The power spectra (also with a five point Gaussian smoothing) of part of the time series (4048 s) is shown in figure 3b. This is typical
68
900
500
20
10-100
I
100 FREQUENCY (Hz)
Figure 2b. The log 10 (power) spectra of 4048 a of the time series shown in figure 2a. The peaks are the normal modes of oscillation of the ice tongue.
of a sea swell spectra with the high frequency components removed by attenuation in the fast ice (Wadhams, 1973). Further data analysis is under way to generate frequency spectra to follow how the wave energy in the ice tongue changes with the progress of the season, and the approach of the fast ice edge. This will later be combined with the other work out this season on the ice tongue (Holdsworth and Holdsworth, 1978, this issue), the theoretical model for oscillations of the ice tongue (Holdsworth and Glynn, 1978), and fracture studies of ice (Goodman and Tabor, 1978) to predict the conditions for the ice tongue to calve. The ice tongue has now grown to the same length as it was when it
ANTARCTIC JOURNAL
CBS CHRRT BLUE 25JAN SITE S3 OT .5000 SEC
STRBIN
MV
go.
70
50
30
Figure 3e. Part of the time series recorded on the sea Ice close to the Ice tongue. The waves are swell induced flexural waves propagating into the fast Ice from the open water.
10
-.10
CBS CHART BLUE 25JRN SITE 53 POWER SPECTRUM FREO. STEP .6333E-03 1/SEC 3000.900 17 SECONDS
T
SEA
2øøO.@Oø >. 4 Cr
.1
co 4
ideal for the testing of present theories of the creep of a large floating ice mass. G. Holdsworth, who was the third member of our party, describes elsewhere in this volume the extent of information collected about the glacier. We are grateful to C. S. Neal, who designed and built part of the strainmeter recording system. Our project would not have been successful without the close support of the helicopter crews, and members of the field centre in McMurdo. A grant from the Royal Society and the Science Research Council of the U.K. paid for part of the equipment and the freight charges; the Scott Polar Research Institute, Sea Ice Group, loaned the remainder of the equipment. The National Science Foundation provided complete logistics support.
H
1geo.oeel.
References
0.0 I g ioe g o+ee 8.20000'2 0.3000D.ee g.øOOD+OO _I FREQUENCY (Hz)
Figure 3b. The 1og 10(power) spectra of 1200 s of the time series shown in figure 3a. The spectrum is typical of a wind Induced swell. was first observed to calve in 1911 (Debenham, 1965), and was thought to have calved in the 1940s (Holdsworth, 1974). The end result will have important implications for the calving of other ice tongues, and icebergs. The strain data collected can also yield long term strain rate data, which is of interest in the modelling of the dynamics of the ice tongue. This is particularly important here because the size and location of the ice tongue make it October 1978
Debenham, F. 1965. The glacier tongues of McMurdo Sound. Geographical Journal, 10(2): 369-371. Evans, K., D. J . Goodman, and G. Holdsworth. In press. The installation of three continuously recording wire strainmeters on the Barnes Ice Cap, Baffin Island (Journal of Glaciology). Goodman, D. J . 1977. Creep and fracture of ice, and surface strain measurements on glaciers and sea ice. Unpublished doctoral dissertation. Cavendish Laboratory, University of Cambridge. Goodman, D.J., A. Allan, and R. C. Bilham. 1975. Wire strainmeters on ice. Nature, 255(5503): 45-46. Goodman, D. J . , and D. Tabor. In press. Fracture toughness of ice, some preliminary results from a new experiment (Journal of Glaciology). Holdsworth, C. 1974. Erebus Glacier Tongue, McMurdo Sound, Antarctica. journal of Glaciology, 13(67): 27-35. Holdsworth, G., and J . Glynn. 1978. Iceberg calving from floating glaciers by a vibrating mechanism. Nature, 274(5670): 464-466. 69
Holdsworth, G., and R. Holdsworth. 1978. Erebus Glacier Tongue movement. Antarctic Journal of the U.S., 13(4): 61-63, King, C. C. P., and R. C. Bilham. 1973. Strain measurements, instrumentation and technique. Philosophical Transactions of the Royal Society, London, A(274): 209-217. Press, F., A. P. Crary,J. Oliver, and S. Katz. 1951. Air-coupled flexural waves in floating ice. Transactions of the American Geophysical Union, 32: 673-678. Robinson, E. S. 1965. Seismic surface wave dispersion on the antarctic ice cap and adjacent floating ice—A preliminary study. Department of Geophysics, University of Utah, Utah. Special Report 1-34, Wadhams, P. 1973. Attenuation of swell by sea ice.JournalofGeophysical Research, 78(18): 3552-3563.
Sea ice and ice algae relationships in the Weddell Sea S. F. ACKLEY U.
Cold Regions Research and Engineering Laboratory Hanover, New Hampshire 03755 S. TAGUCHI
Hawaii Institute of Marine Biology Kaneohe, Hawaii 96744 K. R BUCK
Department of Oceanography Texas A&M University College Station, Texas 77843
Preliminary findings on ice properties and ice-associated algae were given using data obtained during a 1977 cruise in the Weddell Sea (Ackley, 1977; El-Sayed and Taguchi, 1977). Further analysis of these data indicates that the ice algal community found during that cruise is distinct from others that have been described (for example, the bottom epontic community in the land-fast ice in McMurdo Sound, the surface communities off East Antarctica, and the bottom communities in Arctic pack ice). The surface communities off East Antarctica (Meguro, 1962) depend on a thick snow cover, in excess of one-fourth the ice thickness. These snow loads depress the ice surface to below sea level, causing seawater to infiltrate the bottom few centimeters of snow. Increased light levels and high nutrient concentrations in the snow-seawater mixture enable the growth of ice algae (Meguro, 1962). Bunt (1963) has described extensively the bottom epontic community formation which appears to occur primarily in fast ice regions and depends for its existence on a relatively low level of mechanical disturbance at the bottom ice surface. In moving pack ice, shear between the ice and water probably would be strong enough to disturb this fragile layer. 70
The other bottom ice communities, observed in the Arctic, depend on thermal processes leading to brine migration to the bottom of the sea ice. Coupled with summer light levels, algae growth is enhanced in this nutrient-rich region at the bottom (Meguro et al., 1967). Unlike these other communities, the Weddell pack ice algae is dominantly an interior one, existing not at the surface or bottom but at mid-depth (.65 to 2.15 meters) within the ice. The formation of this community is dependent on the unique thermal and physical setting for Weddell sea pack ice. Brine drainage processes are initiated by summer warming, but are not carried through to completion as in the Arctic. This process causes a redistribution of salinity, maximizing in the middepth regions of the ice and apparently leading to algae production because of the relatively higher nutrient levels at these mid-depths. A qualitative model indicating the relationship between the thermally induced brine migration and subsequent algae growth is given in the figure. In assessing how this primary production contributes to the food chain, we examine how the algae enters the ocean through pack ice breakup processes. The disintegration of pack ice in the Weddell region seems to be dominated by mechanical processes (divergence and wave action) occurring primarily at the ice edge region. "Ice edge region" refers to the pack ice-open ocean boundary and is different from "the edges of the ice," [implying the edges of particular ice floes]. The broken-up ice mixes with the sun-warmed ocean water until completely melted. These processes affect the time and location of release of ice-associated algae into the pelagic system and differ considerably from the thermal disintegration and consequent pulse input of ice algae to the water column observed in the Arctic (Homer, 1976). In the table we use satellite-derived estimates of the ice extent together with the field measurements of sea ice chlorophyll a to estimate the organic carbon input to the water column from the annual ice retreat in the Weddell Sea. As seen in the table, organic material is supplied to the water column over a prolonged period of roughly half the year. Although these are not high values for carbon and chlorophyll typically found in the Antarctic oceanic region, the constant supply of biomass may provide supplemental nutrition to the pelagic grazers. This algae may also be the major contributor for areas where the water column in situ Chlorophyll content and particulate organic carbon (POC) contributions to the water column by retreating sea ice (Weddell Sea, south of 55°S., between 60°W. and 30°E.). Mass of POC input to Sea ice chlorophyll a water column Date extent (million in sea ice' from retreating sq km) (million kg) ice (million kg)a 15 Sept 15 Oct 15 Nov 15 Dec 15Jan 15 Feb 15 Mar 15 Apr
7.75 7.44 825b 7.92
7.15 5.85 2.25 1.25 1.15 1.65
6.86 5.62 2.16 1.20 1.10 1.58
46 128 36 -
asumes area shown in column 2 is 80 percent ice covered and 20 percent open ocean at any given time. bMaximum area of sea ice coverage.
ANTARCTIC JOURNAL
Continuous surface strain measurements on sea ice and on Erebus Glacier Tongue, McMurdo Sound, Antarctica
D.J. GOODMAN Physics and Chemistry of Solids Group, Cavendish Laboratory, Cambridge, UK R. HOLDSWORTH Department of Physics, University of Waikato, Wellington, New Zealand
We measured the flexural wave energy on the sea ice close to the Erebus Glacier Tongue (a floating glacier in McMurdo Sound) and monitored surface strain changes on the glacier itself. Also, we made continuous surface strain measurements (in a frequency range between 1 hertz and d.c.) close to the annual runway on the sea ice close to McMurdo Station as LC-130 transport planes landed, and vehicles passed along the ice road to the runway. Strains were measured continuously, to an accuracy of 1 in 108 strain, over a 5-meter gage length with a geophysical wire strainmeter (King and Bilham, 1976). Such instruments have been used to measure flexure of fast ice due to swell (Goodman et al., 1975; Squire and Allan, 1977), and surface strain changes on an ice cap, and a valley glacier (Goodman,
October 1978
References
Daniel, H. C. 1957. Oceanographic atlas of the polar seas, part I, Antarctic (HO. Publication 705). U.S. Navy Hydrographic Office, Washington, D.C. Parkinson, C. L. 1978. A numerical simulation of the annual cycle of sea ice in the Arctic and Antarctic (NCAR cooperative thesis 46). National Center for Atmospheric Research, Boulder, Colorado. Untersteiner, N. 1975. Sea ice and ice sheets and their role in climatic variations. In: The physical basis of climate and climate modelling (Global Atmospheric Research Programme publication series 16). World Meteorological Organization and International Council of Scientific Unions, Stockholm. pp. 206-224. Zwally, H.J., and P. Gloersen. 1977. Passive microwave images of the polar regions and research applications. Polar Record, 18: 431-450.
1977; Evans et al., 1978). The strainmeter uses a length of INVAR wire held under constant tension as a length standard, and detects strains with a lever system and a L.V.D.T. The output, in this case, from the L.V.D.T. was amplified and recorded on a data logger, and a strip chart recorder. An automatic rezeroing system allowed the instrument to operate at a high sensitivity, for long periods, unattended. Six instruments were used in two arrays of three. To avoid temperature effects the instruments were covered with snow. Care was taken not to install the instruments across prominent cracks. Six strainmeters were installed close to the ice road near the ice runway on 14 December 1977 and removed on 19 December. LC-130s were using the runway throughout this period, and were landing approximately 1 kilometer from the site. A typical strain record, from three strainmeters in three different directions, obtained when an LC-130 weighing 105,000 pounds landed is shown in figure 1. The clear dispersion of the waves can be seen. The data collected is in a much lower frequency range to that previously collected by Press etal. (1951) or Robinson (1965) as can be seen from the time bar on the figure. The initial oscillations, which can be seen on the record before the waves from the plane landing arrived, are from swell penetrating from the open water. The fast ice was 3.44 meters thick at the site, which was 35 kilometers from the fast ice edge when the record in the figure was collected. Static, and dynamic, records of the strain also were made as vehicles moved along a road close to the array. On 23 December the first strainmeter was installed on the ice tongue, approximately 9 kilometers from the snout; data were collected more or less continously until 22 January 1978 at the initial site. Data were collected at a second site on the glacier between 11 and 26January 1978. A typical example of the data obtained is shown in figure 2a; the power spectra of 4048 s of the time series (with a five point Gaussian smoothing) is shown in figure 2b. The different peaks correspond to the different modes of oscillation of the ice tongue; the driving force for the oscillation is the swell in the sea.
67
L
:1: 1
I 11411urdo Souzd Antarctjc Stxiin Site I ku from to. runway. Isoorda frm three atrujo..ters it 120° to each other after C-130trenspor piai* weIghing 105,000 The ( I 8,000 kg) laMed. I t :U IWS'
I j I
46
I -
I
Figure 1. The strain changes observed by three strainmeters at 120 0 to each other when an LC-130 weighing 105,000 pounds landed on the sea Ice runway about 1 kilometer away. The initial waves are due to swell penetrating through 35 kilometers of fast Ice. 055 0-1281 1811 I0JRN78 Dl BROWN 19(C1) STRAIN DI 7 2.585
065 0-1281 15:11 I0JRN78 GI BROWN BId) POWER SPECTRUM, FREO. STEP .2441E-03 1/SEC
SEC
rn" 60
1701 50
40
11301
ix
4
30
41 1000-000 2050 380000 4000 000 TIME (SECS)
Figure 2a. Part of the time series recorded on the array of strainmeters on the surface of Erebus Glacier Tongue.
While strainmeters were running on the glacier, at different times to the north and the south of the glacier, an array of three strainmeters was installed on the sea ice about 400 meters from the ice front. The array measured the wave energy incident on the glacier, and will enable the transfer of energy between the sea and the glacier to be determined. A typical record of the swell induced flexural waves is shown in figure 3a (the wave amplitude is greater than the waves in figure 1 because the fast ice edge had advanced to a line between the end of the ice tongue and Cape Evans). The power spectra (also with a five point Gaussian smoothing) of part of the time series (4048 s) is shown in figure 3b. This is typical
68
900
500
20
10-100
I
100 FREQUENCY (Hz)
Figure 2b. The log 10 (power) spectra of 4048 a of the time series shown in figure 2a. The peaks are the normal modes of oscillation of the ice tongue.
of a sea swell spectra with the high frequency components removed by attenuation in the fast ice (Wadhams, 1973). Further data analysis is under way to generate frequency spectra to follow how the wave energy in the ice tongue changes with the progress of the season, and the approach of the fast ice edge. This will later be combined with the other work out this season on the ice tongue (Holdsworth and Holdsworth, 1978, this issue), the theoretical model for oscillations of the ice tongue (Holdsworth and Glynn, 1978), and fracture studies of ice (Goodman and Tabor, 1978) to predict the conditions for the ice tongue to calve. The ice tongue has now grown to the same length as it was when it
ANTARCTIC JOURNAL
CBS CHRRT BLUE 25JAN SITE S3 OT .5000 SEC
STRBIN
MV
go.
70
50
30
Figure 3e. Part of the time series recorded on the sea Ice close to the Ice tongue. The waves are swell induced flexural waves propagating into the fast Ice from the open water.
10
-.10
CBS CHART BLUE 25JRN SITE 53 POWER SPECTRUM FREO. STEP .6333E-03 1/SEC 3000.900 17 SECONDS
T
SEA
2øøO.@Oø >. 4 Cr
.1
co 4
ideal for the testing of present theories of the creep of a large floating ice mass. G. Holdsworth, who was the third member of our party, describes elsewhere in this volume the extent of information collected about the glacier. We are grateful to C. S. Neal, who designed and built part of the strainmeter recording system. Our project would not have been successful without the close support of the helicopter crews, and members of the field centre in McMurdo. A grant from the Royal Society and the Science Research Council of the U.K. paid for part of the equipment and the freight charges; the Scott Polar Research Institute, Sea Ice Group, loaned the remainder of the equipment. The National Science Foundation provided complete logistics support.
H
1geo.oeel.
References
0.0 I g ioe g o+ee 8.20000'2 0.3000D.ee g.øOOD+OO _I FREQUENCY (Hz)
Figure 3b. The 1og 10(power) spectra of 1200 s of the time series shown in figure 3a. The spectrum is typical of a wind Induced swell. was first observed to calve in 1911 (Debenham, 1965), and was thought to have calved in the 1940s (Holdsworth, 1974). The end result will have important implications for the calving of other ice tongues, and icebergs. The strain data collected can also yield long term strain rate data, which is of interest in the modelling of the dynamics of the ice tongue. This is particularly important here because the size and location of the ice tongue make it October 1978
Debenham, F. 1965. The glacier tongues of McMurdo Sound. Geographical Journal, 10(2): 369-371. Evans, K., D. J . Goodman, and G. Holdsworth. In press. The installation of three continuously recording wire strainmeters on the Barnes Ice Cap, Baffin Island (Journal of Glaciology). Goodman, D. J . 1977. Creep and fracture of ice, and surface strain measurements on glaciers and sea ice. Unpublished doctoral dissertation. Cavendish Laboratory, University of Cambridge. Goodman, D.J., A. Allan, and R. C. Bilham. 1975. Wire strainmeters on ice. Nature, 255(5503): 45-46. Goodman, D. J . , and D. Tabor. In press. Fracture toughness of ice, some preliminary results from a new experiment (Journal of Glaciology). Holdsworth, C. 1974. Erebus Glacier Tongue, McMurdo Sound, Antarctica. journal of Glaciology, 13(67): 27-35. Holdsworth, G., and J . Glynn. 1978. Iceberg calving from floating glaciers by a vibrating mechanism. Nature, 274(5670): 464-466. 69
Holdsworth, G., and R. Holdsworth. 1978. Erebus Glacier Tongue movement. Antarctic Journal of the U.S., 13(4): 61-63, King, C. C. P., and R. C. Bilham. 1973. Strain measurements, instrumentation and technique. Philosophical Transactions of the Royal Society, London, A(274): 209-217. Press, F., A. P. Crary,J. Oliver, and S. Katz. 1951. Air-coupled flexural waves in floating ice. Transactions of the American Geophysical Union, 32: 673-678. Robinson, E. S. 1965. Seismic surface wave dispersion on the antarctic ice cap and adjacent floating ice—A preliminary study. Department of Geophysics, University of Utah, Utah. Special Report 1-34, Wadhams, P. 1973. Attenuation of swell by sea ice.JournalofGeophysical Research, 78(18): 3552-3563.
Sea ice and ice algae relationships in the Weddell Sea S. F. ACKLEY U.
Cold Regions Research and Engineering Laboratory Hanover, New Hampshire 03755 S. TAGUCHI
Hawaii Institute of Marine Biology Kaneohe, Hawaii 96744 K. R BUCK
Department of Oceanography Texas A&M University College Station, Texas 77843
Preliminary findings on ice properties and ice-associated algae were given using data obtained during a 1977 cruise in the Weddell Sea (Ackley, 1977; El-Sayed and Taguchi, 1977). Further analysis of these data indicates that the ice algal community found during that cruise is distinct from others that have been described (for example, the bottom epontic community in the land-fast ice in McMurdo Sound, the surface communities off East Antarctica, and the bottom communities in Arctic pack ice). The surface communities off East Antarctica (Meguro, 1962) depend on a thick snow cover, in excess of one-fourth the ice thickness. These snow loads depress the ice surface to below sea level, causing seawater to infiltrate the bottom few centimeters of snow. Increased light levels and high nutrient concentrations in the snow-seawater mixture enable the growth of ice algae (Meguro, 1962). Bunt (1963) has described extensively the bottom epontic community formation which appears to occur primarily in fast ice regions and depends for its existence on a relatively low level of mechanical disturbance at the bottom ice surface. In moving pack ice, shear between the ice and water probably would be strong enough to disturb this fragile layer. 70
The other bottom ice communities, observed in the Arctic, depend on thermal processes leading to brine migration to the bottom of the sea ice. Coupled with summer light levels, algae growth is enhanced in this nutrient-rich region at the bottom (Meguro et al., 1967). Unlike these other communities, the Weddell pack ice algae is dominantly an interior one, existing not at the surface or bottom but at mid-depth (.65 to 2.15 meters) within the ice. The formation of this community is dependent on the unique thermal and physical setting for Weddell sea pack ice. Brine drainage processes are initiated by summer warming, but are not carried through to completion as in the Arctic. This process causes a redistribution of salinity, maximizing in the middepth regions of the ice and apparently leading to algae production because of the relatively higher nutrient levels at these mid-depths. A qualitative model indicating the relationship between the thermally induced brine migration and subsequent algae growth is given in the figure. In assessing how this primary production contributes to the food chain, we examine how the algae enters the ocean through pack ice breakup processes. The disintegration of pack ice in the Weddell region seems to be dominated by mechanical processes (divergence and wave action) occurring primarily at the ice edge region. "Ice edge region" refers to the pack ice-open ocean boundary and is different from "the edges of the ice," [implying the edges of particular ice floes]. The broken-up ice mixes with the sun-warmed ocean water until completely melted. These processes affect the time and location of release of ice-associated algae into the pelagic system and differ considerably from the thermal disintegration and consequent pulse input of ice algae to the water column observed in the Arctic (Homer, 1976). In the table we use satellite-derived estimates of the ice extent together with the field measurements of sea ice chlorophyll a to estimate the organic carbon input to the water column from the annual ice retreat in the Weddell Sea. As seen in the table, organic material is supplied to the water column over a prolonged period of roughly half the year. Although these are not high values for carbon and chlorophyll typically found in the Antarctic oceanic region, the constant supply of biomass may provide supplemental nutrition to the pelagic grazers. This algae may also be the major contributor for areas where the water column in situ Chlorophyll content and particulate organic carbon (POC) contributions to the water column by retreating sea ice (Weddell Sea, south of 55°S., between 60°W. and 30°E.). Mass of POC input to Sea ice chlorophyll a water column Date extent (million in sea ice' from retreating sq km) (million kg) ice (million kg)a 15 Sept 15 Oct 15 Nov 15 Dec 15Jan 15 Feb 15 Mar 15 Apr
7.75 7.44 825b 7.92
7.15 5.85 2.25 1.25 1.15 1.65
6.86 5.62 2.16 1.20 1.10 1.58
46 128 36 -
asumes area shown in column 2 is 80 percent ice covered and 20 percent open ocean at any given time. bMaximum area of sea ice coverage.
ANTARCTIC JOURNAL
