GREGORY H. L EONARD and HAYLEY H. SHEN, Department of Civil and Environmental Engineering, Clarkson University, Potsdam, New York 13699
STEVEN F. ACKLEY, Snow and Ice Division, U.S. Army Cold Regions Research and Engineering Laboratory, Hanover, New Hampshire 03755
Waves are ubiquitous in the southern oceans. Their presence has been linked to the predominant pancake-ice formation in the antarctic marginal ice zone. The morphology of an ice cover formed from a pancake ice field is determined by the initial pancake floe's size and thickness. The incorporation of any biological material into an ice cover also begins with this initial formation. No systematic field observations have been made of this formation process. This article describes a laboratory study in which various wave fields, with or without wind and current, were created in a refrigerated recirculating flume. The evolution of the resulting ice covers clearly depends on these hydrodynamic conditions.
Six ice-growth tests were conducted in the Hamburg Ship Model Basin in Hamburg, Germany. The parameters are shown in the table. The instruments used for these tests included a pressure transducer, an infrared videocamera, and a regular videocamera. The pressure transducer was placed in the middle section of the tank. This pressure transducer traversed a length of the tank along at least 10 locations, and the separation between locations was 15 centimeters. Time series were taken at these locations, from the paddle side down, one at a time. Each time series consists of 1,024 readings at 50-hertz data rate. A 30-second infrared video image and regular video image were recorded after each session of pressure-data acquisition. All three recordings were done at roughly a 30-minute interval.
In these tests, test B had very high wave amplitude in which ice crystals were swept down from the paddle to the extent that most of the test section was open water throughout the test duration. The remaining tests are separated into two: C, D, and E were under wave-only conditions and A and F had wind (with or without current) in addition. The presence of wind greatly changes the morphology of the ice cover as will be discussed below.
The results are summarized in terms of time evolution of wave amplitude under the ice cover and the observation of the ice-cover evolution. An example of these results is given in the figure. The wave energy is attenuated by the ice cover as it grows, but the amount of attenuation depends on the type of the ice cover, which changes with time.
Although the number of tests conducted is low compared with the number of variables observed, some general trends can be observed in these three tests.
Wind has a definitive effect on the formation of ice cover because it creates frazil at a much higher rate so that grease-ice production is prolonged and pancake-ice formation is hindered. In tests A and F, the ice cover remained slushy. Only toward the end of the tests were some sparsely distributed pancakes able to form above a very thick layer of frazil.
The microstructure of the ice cover was also observed. The initial crystals before a well-defined grease-ice layer formed were flakelike. Their size was about 1 centimeter in diameter in all these tests. Such crystals went through a rapid kinetic growth under super-cooled conditions. Well-defined pancakes consist of a frozen top and a slushy bottom. The slush is an agglomerate of platelets of several millimeters in diameter stacked up horizontally. These platelets are formed from the initial frazils rubbing against and freezing onto each other due to the oscillatory motion in the wave field. The rubbing causes the initial crystals to lose their dendritic offshoots and to reduce their sizes. The horizontal orientation indicates the possibility of highly anisotropic permeability in the ice cover that is subsequently formed by these pancakes.
From these tests, it is suggested that the dynamic formation of an ice cover in a wave field may be described by four stages.
Provided that the above scenario is correct, it is clear that the morphology of an ice cover is a product of the thermodyamic and hydrodynamic conditions prescribed by air temperature and by the wind, wave, and current conditions. Of the latter three, wind produces waves, especially the high-frequency spectrum part. Wind also promotes heat loss and frazil production rate. Waves shape the platelets formed from the initial crystals, promote vertical growth of these platelets by rafting and compaction, and create lateral growth of mature pancakes by fusion through differential motion of pancakes. On top of these, air temperature is a key factor that highly influences the initial frazil production and the late fusion rate. Finally, current may determine the frazil distribution in the water through the induced turbulence and, thus, the surface accumulation efficiency.
This research was supported by National Science Foundation grants OPP 93-15934 and OPP 92-19165 as well as by the National Science Foundation's International Program. Technical support from the Hamburg Ship Model Basin, Hamburg, Germany, is deeply appreciated.