General

Martin River Glacier is located in south-central Alaska, approximately 96 km. east of Cordova, the nearest community (Fig. 1). The glacier is an expanded valley glacier originating in the Bagley Icefields approximately 64 km. farther to the east. The terminal 6.2 km. of this glacier are covered with ablation till composed largely of cobbles, pebbles and occasional boulders up to 10 m. in diameter. In this terminal zone the ablation till averages about 0.3 m. thick and is at least 9 m. thick along part of the glacier margin. The most recent end moraine complex (Little Ice Age and younger) is still largely cored with dead ice.

Fig. 1. Map of the lower part of Martin River Glacier, showing the location of “Miller Lake”

The study of Martin River Glacier was carried out by faculty and student members of the Department of Geology, The University of North Dakota, under sponsorship of a grant from the National Science Foundation (G-22016, Dr. W. M. Laird, Principal Investigator). Although the main purpose of the investigation was to correlate dead ice features with related fauna and flora, it was only natural that the degree of activity of this glacier be considered important. Much of the information presented in this paper came from observations at the “Miller Lake” camp in 1963.

“Miller Lake”

A large lake 2.8 km. long, 800 m. wide and up to 70 m. deep has formed between the active ice front and the dead ice-cored moraine complex at the terminus of Martin River Glacier (Fig. 1). This lake has been unofficially named “Miller Lake” after the late Dr. Don Miller of the U.S. Geological Survey. “Miller Lake” is an ice-walled lake formed by the coalescence of several ice sink holes which are common features in the terminal zone of Martin River Glacier (Reference Reid and ClaytonReid and Clayton, 1963; Reference ClaytonClayton, 1964). A large part of the lake basin is underlain by ice located only a short distance beneath the surface. This is inferred from thermal data which show very cold water (0.1° C.) near the bottom of the lake. The bottom of “Miller Lake” is quite irregular and is characterized by steep-sided depressions and hills which in turn contain smaller similar negative and positive features. The northern two-thirds of the basin is characterized by several elongate depressions, many of which contain steep-sided hills commonly rising 20 m. above the basin floor. The south-western part of the basin exhibits a fairly extensive gently sloping littoral shelf. The present sub-circular outline of the lake has evolved from a more irregular outline characteristic of compound sink holes. The shoreline has been partly modified by wave action and calving of ice.

Thermally, “Miller Lake” is a sub-polar lake according to the classification proposed by Yoshimura (Reference HutchinsonHutchinson, 1957, p. 437). Although the mean summer surface temperature is nearly 6° C., the deeper waters are much colder, dropping to 2° C. at −10 m. and 0.1° C. at −30 m. The thermal gradient is small and the thermocline is poorly developed; temporary cooling at various times throughout the summer generally allows fairly frequent mixing of the lake waters, thereby destroying any thermal stratification.

General

Perhaps the most striking feature of “Miller Lake” is the icebergs. Most of these icebergs are typically veneered with ablation till of varying lithology, particle size and abundance. Observation of the active ice cliff along the north-eastern, eastern and south-eastern margin of the lake, however, reveals that the characteristic tumbling action of calving ice allows only a trace of debris to be retained on the resultant floating ice masses. Furthermore, floating ice masses formed this way are normally only of growler size and are less than one-tenth the size of the largest icebergs on the lake which are the typical till-veneered icebergs. This mode of formation for the debris-covered icebergs was therefore deemed insufficient to account for the abundance and size of these ice masses in “Miller Lake”. Another explanation was therefore sought.

Grounded iceberg origin

The suggestion that clean icebergs are blown against the ablating ice cliff, thereby receiving the ablation debris from above by mass wasting, was rejected because it assumed too much of a coincidence to justify the abundance of this type of iceberg. Furthermore, the strong storm winds were always observed to blow the icebergs away from the active cliffs, not towards them.

Differential ablation origin

Another explanation is that promontories of the irregular ice cliff become detached from the rest of the glacier through differential ablation, perhaps along incipient crevasses (Fig. 2). Eventually, the mass becomes a separate unit and is buoyed up as a “dirty” iceberg, the till veneer being the original ablation till of the terminal zone of the glacier. Examination of aerial photographs taken in 1950, 1957, 1959 and 1963 (Fig. 3) shows numerous such projections that have since disappeared, presumably as icebergs of this type.

Fig. 2. Crevasse pattern and condition of the ice surrounding “Miller Lake”

Talus apron origin

A third possibility is directly related to ablation till deposited along the base of the ice cliff as talus aprons or “cones” which then inhibit further ablation of the underlying ice. Throughout the 1963 summer season several such features formed, expanded and finally became ice-cored islands when the rest of the ice cliff melted back far enough (Fig. 4). Because these debris-covered islands became submerged during rising lake levels, they were concluded to be still attached to the bottom. Since most of the ablation debris from the active ice is ultimately deposited on the bottom of the lake, it was surmised that complete detachment of sub-lake ice masses and/or debris-covered islands from the active ice front might be responsible for the large ablation till-covered icebergs.

Fig. 4. idealized cross-section showing successive stage in the formation of“Gull Island”and its relationship to the shear planes. The derpression represents the site from which the iceberg emerged. The vertical scale is three times the horizontal scale

Lake bottom origin

On 14. July a new iceberg estimated to weigh between 170 and 200 tons suddenly emerged from the bottom of the lake approximately 250 m. from the active ice front. The iceberg was 30 m. wide and 350 m. long; it floated with its surface as much as 8 m. above water level. One entire side of the iceberg was covered with assorted lake sediments and glacial debris, indicating that it had rotated only slightly upon emerging from the bottom. It was surprising that so much sediment remained on the surface as washing during emergence undoubtedly removed much of what had covered the ice. Subsequent fathometer traverses revealed a hole 20 m. deep, 360 m. long, 60 m. wide, and near the site where the iceberg first emerged (Fig. 4). The violent wave action produced by the emergence of the iceberg undoubtedly forced it away from the cliff because the hole is about 100 m. closer to the ice cliff than the place where the iceberg was first measured. Although this mechanism of formation of debris-covered icebergs in “Miller Lake” is probably not as important as that of differential ablation, there is little doubt that other icebergs in the lake originated in this manner. Similar observations were made by Reference RussellRussell (1891, p. 101–02), who postulated an ice toe projecting 1,000 ft. (305 m.) in front of the ice cliffs. The resulting icebergs from beneath the water were sometimes 200 and 300 ft. (61 and 91 m.) in diameter and, as in the case of the “Miller Lake” icebergs, they were much larger than those produced by calving of the ice cliff. Reference TarrTarr (1909, p. 31) supported the existence of such an ice projection and listed three conditions that favored the formation of the projection: “(1) The warmer water at the surface… causing undercutting of the ice; (2) the crevassed, weathered, and thereby weakened upper portion of the glacier; (3) the attack of waves, especially those generated by the iceberg falls.” Nowhere in the literature, however, have icebergs been reported as emerging from the bottom of an ice-walled lake.

The reason for the emergence of the “Miller Lake” iceberg was puzzling. The easiest explanation would be that part of the terminus was floating and that it was, in fact, a projection which broke free and rose to the surface. However, the fathometer traverses discounted the presence of any floating projection. The second suggestion was that the ice mass became an isolated remnant as a result of differential ablation along the bottom and finally broke free. But the fathograms showed that the ice mass came from a hole. Had it been a remnant, the chances are that it would have been a subdued projection on the bottom prior to its separation from the bottom. This hypothesis, however, was not completely rejected.

A third suggestion was that the ice mass became separated along crevasses that were present in the ice on the bottom of the lake. It was believed that the bottom ice was not sufficiently brittle to contain crevasses. The explanation, therefore, was not immediately known. The explanation took on an added importance when a rising island was observed in the same lake.

General

As has been previously stated, when debris from the surface of the glacier falls into the lake, a cone or apron of talus frequently accumulates near the water line. This deposit inhibits the ablation of the underlying ice and, as the rest of the ice cliff retreats through continued ablation, the deposit may be left as an ablation till-veneered island (Fig. 4). One such island, informally called “Gull Island”, had already formed by the beginning of the 1963 summer season. It had an area of about 60 m.² above water level during this early part of the season and it attracted frequent attention because of gulls nesting there.

On 4 July 1963 the lake suddenly rose 1.35 m. This rise was presumed to be due to the draining of one of the numerous supraglacial lakes above the terminus (Reference Reid and ClaytonReid and Clayton, 1963; Reference ClaytonClayton, 1964). The rise of the water almost inundated “Gull Island” and definitely revealed that the island was not floating. When the large iceberg, discussed above, emerged from the bottom ten days later, “Gull Island” appeared to have risen. Although no direct measurements had been made, further rise substantiated this conclusion. The rise appeared to be steady despite additional rises of the lake level during and subsequent to periods of heavy precipitation. By the end of the summer season the island had risen approximately 4–5 m. above the level it had been 10 weeks earlier. It had also increased its surface area by as much as 200 m.² even after taking into consideration lake-level changes since the beginning of the season.

Character of the bottom ice

Because it appeared that consideration of the character of the bottom ice would be of great importance in evaluating the cause of the rising island and the emerging iceberg, several caculations were made. With an average depth of about 40 m. (the bottom of the hole from which the iceberg emerged) and an average water temperature of 1° C., the hydrostatic stress (σ) on the bottom was equal to 56.9 lb./in.² (3.87×10⁶ dyne/cm.²). By dividing by √3 (Reference NyeNye, 1953), the resulting shear stress (τ) there was equal to 2.24×10⁶ dyne/cm.². This value of shear stress is considerably more than the 1×10⁴ dyne/cm.² used by Reference Butkovich and LandauerButkovich and Landauer (1960, fig. 3), and with which they measured appreciable rates of creep. Furthermore, their results were based on lower temperatures than the “Miller Lake” ice which, because of an ice-water interface, was at the melting point and therefore capable of even greater rates of creep.

Fig. 3. Outline of “Miller Lake” in 1930, 1957, 1959 and 1963. The location of the bathymetric profile in Figure 4 is shown on the 1963 outline

Another aspect of ice that is brought from depths is the sudden change in the relationship of the temperature to the pressure-melting point. Since ice at a depth of 40 m. is under a pressure of 3.87 atmospheres, the melting point is lowered by approximately 0.04° C. Furthermore, because the ice at the bottom is in direct contact with water, it must be at the pressure-melting point. If this ice were to be suddenly brought to the surface, as occurred when the iceberg was formed, the melting point would be raised above the temperature of the ice and it would therefore freeze harder due to the release of water pressure. When the iceberg was investigated the morning following its emergence, the canvas raft that was used to reach the iceberg froze to the surface of the ice even though the air temperature had never dropped below 7.2° C. that night. This freezing confirmed the supposition.