The McMurdo "Pegasus Site"

Pegasus after the crash
Above, the Super Constellation (C121J) aircraft "Pegasus" immediately
after its crash on 8 October 1970. Everyone survived (the story) (photo
from Larry Lister)

[from "Airfields on Antarctic Glacier Ice" by Malcolm Mellor and Charles Swithinbank, CRREL Report 89-21, 1989 full report (PDF)]

To the south and west of McMurdo station, a lobe of the Ross Ice Shelf flows into McMurdo Sound (Fig. 63). To the east and southeast of Cape Armitage, the surface of the ice shelf is a net accumulation area, with a permanent snow cover that is of great depth. Winds are relatively light in this area, as evidenced by the name "Windless Bight" for the area between Hut Point Peninsula and Cape Mackay. To the west and southwest of Cape Armitage, the surface of the ice shelf suffers very strong net ablation, with intense summer melting (Swithinbank, 1979). This is a consequence of southerly winds which limit snow deposition and carry dark-colored dust auto the ice from Black Island and Brown Peninsula.

The Williams Field skiway, almost due east of Cape Armitage is in the accumulation area, and it receives about 0.6 m (2 ft) of new snow per year. The water-equivalent of this net accumulation is about 0.23 m of water. Traveling southwest from Williams Field, the annual accumulation gradually decreases, becoming zero at the transition from the accumulation area to the ablation area. This transition zone is encountered at about longitude 166º 35' E in latitude 78º S. Just to the west of the transition there is the wreck of an old C-121J Constellation which carried the name "Pegasus". Since this aircraft, which crashed in October 1970, is the only landmark, the area immediately southeast of the wreck is known to us as the Pegasus site.

About 1 km west of the transition, and just within the zone of net ablation, is the site of the former Outer Williams Field (OWF). During the period when OWF operated as a backup airfield for McMurdo station (from 1966-67 to 1970-71) , its nominal location was approximately 77º 57.7' S, 166º 28.5' E. According to old U.S. Navy drawings, the elevation of the ice surface was about 19 ft (5.87 m) above sea level, and the ice thickness at the site was approximately 114 ft (35 m). If accurate, these figures imply a low mean density for the ice (0.855 Mg/m3). The ice was moving slowly in the direction WNW at about 95 ft/yr (29 m/yr). The ice surface sloped down to the NNE at about 3 ft/mile (0.57 m/km); this is a gradient of only 0.06%.

When the field was first established (1966-67), the main runway was meant to be aligned with the prevailing wind, in a direction approximately 155º/335º true, There was also a crosswind runway aligned with the storm wind direction. No record of the exact orientation has been found, but it was perhaps about 040º/220º. In 1967-68 the main runway was realigned about 15º closer to north-south, i.e. to about 170º/350º true. Drawings for the 1970/71 season indicate that it was changed again, to about 004º/184º. The ablation area where OWF was located proved to have an unusual and disconcerting characteristic; subsurface melt cavities form in January, creating hummocks in the ice surface when they re-freeze during the following winter (Paige, 1968). The ice surface in this area usually remains at sub-freezing temperatures throughout the summer, partly because of the persistent flow of cold air from the south. however, solar radiation is strong on clear days in December and January (up to 19.5 J/m2) , and the albedo of the bare ice is relatively low (Paige mentions 9.48 for very blue ice). Radiation thus penetrates into the ice, where the energy is absorbed, allowing Internal melting to occur. At the time of the study by Paige (1968), melting at OWF began in mid-December, typically at a depth of 49 cm (16 in.) or more. The melting initiates, or perhaps concentrates, in scattered patches (it is not uniform in the horizontal plane). Eventually, lenticular water cavities develop and they grow, in both vertical and horizontal extent, until mid-January. Paige reported diameters up to 10-15 m, depths of 0.5 to 1.0 m (assumed to be below ice surface), and ice cover thicknesses decreasing from about 40 cm to 7 cm as the season progressed. During re-freezing the trapped water expands (about 8% by volume) , the ice cover over the cavity heaves, and a cracked ice hummock is formed. Paige gave the size of these hummocks as 2-8 m in diameter and 0.3-0.6 m high at the site of OWF.

The 1988 proposal for a new airfield at the Pegasus site (Mellor, 1988a) called for one or wore runways to be laid out immediately east of the snow/ice transition instead of using the OWF site, which is west of the transition. In other words, whereas OWF was just in the ablation area, the Pegasus site was to be just inside the accumulation area. The intent was to maintain a thin snow cover over the ice in order to limit ablation problems.

Malcolm Mellor at the Pegasus site
Malcolm Mellor working at the Pegasus site
(photo from Krist Hammargren)
A reconnaissance of the Pegasus site was made by Mellor in November 1988 and again in January 1989. Stakes were drilled into the ice to mark the snow/ice transition in November (brown and white checkered flags) and January (black flags). In November, the ice hummocks described by Paige for OWF could be seen on the bare ice at the transition. They were typically up to 0.4 m (16 in.) high, and some collapsed under the bulldozer tracks, leaving fragment-filled depressions up to 0.15 m (6 in.) deep. Long swaths were bulldozed through the thin snow cover, parallel to the transition inside the snow-covered area and in zig-zags across the transition (Fig. 64, 65). More hummocks were found under the thin snow cover (0.1 m, or 4 in.) and it was assumed that this was winter extension of the accumulation area. In January, the transition appeared to have retreated eastward, by a few hundred metres, but in some places there was hardly any change. Bulldozing tests were repeated, and subsurface melt cavities were found under bare ice and under the thin (0.1 m, or 4 in.) snow cover near the transition (Figs. 66, 67). The 35 ton (32 tonne) LGP D-8 dozer was a reliable detector of the cavities, which were typically up to 10 m (33 ft) in diameter, with tile deep part about 4 to 6 m (13-20 ft) in diameter. The ice cover over the water was 0.1 to 0.15 m (4-6 in.) thick. The air bubbles in the ice were quite large, some several millimetres in diameter, suggesting local melting around the original small bubbles. the total depth, from general ice surface to the deep part of the cavity, was 0.25 to 0.46 m (10-18 in.), which is less than the depths reported by Paige for OWF. As the snow cover became thicker with increasing distance from the transition, the melt cavities become smaller and fewer. In areas where there was 0.25 to 0.3 m (10-12 in.) of snow over the ice the bulldozer ceased to reveal evidence of meltwater. Because it is hard to imagine signi- ficant penetration of solar radiation through more than 9.1 m (4 in.) of snow cover, it is suspected that the snow over the melt cavities was of recent origin.

Since the albedo and the extinction coefficient for snow are both high, it would be surprising if subsurface melt cavities could form under a stable snow cover that is 0.1 m (4 in.) or more thick. Paige (1968) indicated that a layer of ice chips with an albedo of 0.76 and a thickness of 3 cm (1.2 in.) or more was sufficient to prevent the formation of melt cavities. For now, we are assuming that the melt cavities formed in bare ice that was covered by new snow just before the January reconnaissance (there was a lot of snow at McMurdo during the 1988-89 summer).

Further investigations of the Pegasus site are scheduled for 1989-90. The intention is to lay out a runway just inside the accumulation area, where a snow cover can be maintained throughout the Summer. The runway will have to be plowed in late winter for the start of the flying season, and snow will probably have to be blown or pushed back onto the ice from time to time during summer. Any hummocks that are encountered will have to be planed flat. One of the primary goals is to develop procedures for preventing the formation of melt cavities.

At OWF, melt cavities were detected by towing a heavy load cart up and down the runway. To detect water pockets at the Pegasus site, we propose to acquire a dual antenna radar unit of the type used earlier in this general area by Kovacs et al. (1982). The equipment will be towed along the runway in a series of parallel sweeps until the full area of the runway is covered. One design goal is to achieve a wide search path and to travel as fast as is reasonably possible. Preliminary discussions with a CRREL consultant (S. Arcone, personal communication) suggests that an appropriate system would be an impulse-type subsurface radar, such as is manufactured by tile GSSI Company in Hudson, NH. The control unit of the standard radar can operate two antennas at once, which helps in covering a wide search path. Each antenna unit contains separate transmit and receive antennas. The recommended units are GSSI Model 3102; when placed on ice, these usually radiate a short 2 to 3 cycle wavelet with a frequency spectrum centered near 400 MHz. The estimated beam width in ice is about 70º, so that each antenna covers a 1-m swath at a depth of 0.7 m and a 0.5-m swath at a depth of 0.35 m. With a transverse separation of 1 m between the antennas, therm would be continuous coverage across a 2-m swath at 0.7 m depth and two swaths of 0.5 m with a 1-m gap between them at 0.35 m depth. A realistic goal might be detection of water pockets when they reach a diameter of 1-2 m at 0.35 m depth. At a towing speed of 2 m/s (4.5 mph), data could be collected at a density of 13 echoscans for every metre. The antennas can be carried by a dielectric sled (metal-free) and need not be shielded from the weather. The control unit and the tape recorder need an operating environment where the temperature is no lower than 2ºC (35ºF). One or two operators would be needed.

The ice shelf calves occasionally, producing major charges in the position of the ice front. In 1947 the ice front to the southwest of Cape Armitage was much further south than the present position, almost as far back as the current location of Pegasus. In siting a new airfield and camp facilities, periodic calving has to be taken into account.