Introduction
Experiments leading to the collection of data herein were conducted in 1959 and 1960 on the sea ice of Drifting Station “Charley” in the Arctic Ocean; at the Arctic Research Laboratory, Point Barrow, Alaska and in the Cascade Mountains of Washington State.
Weston photoelectric cells, horizontally embedded, were used at all locations. On ice floe station “Charley” they were installed at the following depths: (1) ice surface, (2) 70 cm. and (3) 350 cm. The latter was just beneath the ice–water interface. Three cells were unfiltered and three (each) provided with red and yellow special-work filters. It was planned to use three blue filters in addition, but these were not available in special-work quality before the experiment began. To keep millivoltage within recorder-scale limits resistors were hooked in parallel in each circuit.
On the shore-fast ice at Point Barrow the cells were planted at the following places: (1) snow surface; (2) snow–ice interface; (3) 100 cm. in the ice and (4) immediately above the ice–water interface, 180 cm. beneath the ice surface. Because the study was carried out during a period of the year (May and June) when ice was deteriorating, intensity of light was measured manually to avoid risking the loss of a recorder.
For investigating the properties of snow as a medium of visible energy exchange, densities were measured and grain-size was estimated with a magnifying glass, as microscopic equipment was not available. In general, the technique of Reference LiljequistLiljequist (1956) was used in the collection of data from snow. Blue, green and red special-work filters were used and in some cases the polychromatic band was measured. Care was taken at each station to insure the homogeneous composition of the snow and that the incident intensity did not vary appreciably during the readings. Resistors were used on a switch to keep the millivoltmeter from “pegging”.
Discussion
In computing the absorption of radiation in homogeneous ice and snow it is assumed that the Lambert law is valid. This may be written
where I z is the intensity of radiation at depth z, I 0 is the non-reflected radiation available for transfer into the medium, and k is the coefficient of absorption.
To compute coefficients of absorption in a downward flux, the equation is solved for −k and written
At Point Barrow albedo values were available, but at other places they were computed by determining I 0 from (1) and (2) and deducing the albedo A from the equation
where I s is the incident intensity.
Porosity of snow and its permeability to air was investigated by Reference BaderBader (1939) who related porosity n to density by the equation
where γ e is the density of ice (0.917 g. cm.−3), and γ s is the density of snow.
The results of the investigation of the transfer of visible radiation in sea ice and snow are shown in Table I. Because concentration of sediments in shore-fast ice at Point Barrow is high, the coefficient of absorption is nearly double that of pack ice in the central Arctic Ocean.
Table I Coefficients of Absorption and Albedo of Sea Ice and Snow as a Function of Density and Porosity
Table II compares some data in Table I with representative observations of other authors, among whom it shows a lack of accord. Moreover, Reference LiljequistLiljequist (1956) observed a marked selective extinction of light while Reference SaubererSauberer (1938) remarked that the extinction of different colors within the band from 0.38μ to 0.78μ is, in the main, the same. Some minor variation in results of measurements in common media is to be expected due to multiple light reflection at depth and instrumentation factors. However, the magnitude of disagreement in Table II makes it apparent that absorption of light cannot be solely a function of wave-length and the porosity, density, grain-size or humidity of the medium. Discounting such foreign elements as brine pockets, interstices and inclusions, orientation of the c-axes of ice crystals (which are ordinarily uniaxial) must significantly affect transparency of a medium for refracted light. This may well account for incongruities in the value of −k for common media. In fresh snow the shape of the crystals, and in old snow that of the grains, surely also influences transparency.
Table II Values of Coefficients of Absorption as determined by Various Authors for Visible Radiation in Certain Types of Snow
For example, dendritic, plate, cup and needle crystal types should have different optical properties even though the density and porosity be identical. Furthermore, in fresh snow, crystals—particularly dendritic ones—may become unevenly rimed. In old snow, the formation of depth-hoar layers can also vary the absorption of energy.
Conclusions
The above hypothesis suggests a need for investigation of absorption properties of ice and snow as a function of crystal size and orientation.
Notwithstanding the incongruity of results shown in Table II, the conclusions of other investigators support the data presented in Table I in making apparent the following fact: Absorption coefficients and albedos vary inversely with density of the medium and frequency of the energy.
When sufficient data are collected and tabulated it will be possible to determine by inspection, any one of the several variables governing the transfer of visible radiation. For example, with space as a variable, the size and shape of a medium may be delimited. In areas of pack ice where surface observations are available, thickness and perhaps other characteristics may be inferred by a submarine on a traverse beneath the ice-water interface. Or the ice seaman (who is concerned with crystal orientation) may improve his breaking and navigating techniques.
Acknowledgement
The help of Robert Ditzler, Daniel Hale and Richard Surnmerfèld in the collection of data presented herein is gratefully acknowledged.
