A two-dimensional model of glacier flow is presented which includes periodical surging as a natural phenomenon for a certain class of glaciers. The input consists of the bedrock and balance profiles along the glacier, together with the ice flow properties and a frictional lubrication factor. The basal stress is determined from the condition of gross equilibrium for the whole glacier, together with the distribution of the frictional lubrication from energy dissipation along the glacier.

The difference between the basal stress and the down-slope stress of the glaciers produces longitudinal strain-rates which determine the basal sliding velocity. Since the velocity is also involved in the frictional lubrication, feed-back develops between the basal stress and sliding velocity.

For a given lubrication factor, a critical stage can be reached for which the velocity becomes sufficiently high to lower the basal stress, enough to cause very high velocities to develop. The model thus gives rise to three classes of glaciers with two modes of flow.

“Ordinary” glaciers do not have sufficient mass flux, for the given bedrock profile, to go beyond the “slow mode” in which the basal stress and velocity increase together as the glacier builds up to steady state.

“Fast” glaciers have sufficient flux to remain continuously in the “fast mode” with high velocities and relatively low basal stress.

“Surging” glaciers have sufficient flux to reach the fast mode but not sufficient to maintain it, and thus develop a periodically oscillating state between the fast and slow modes with gradual build up and rapid drainage.

Sample results are presented for models of a typical large valley surging glacier and for a very high-speed surging glacier.

Temperatures have been measured at two sites in a temperate glacier (Blue Glacier, Washington, U.S.A.) to depths of 192 and 76 m. The accuracy, which varies between about 0.002 and 0.005 deg, is about an order of magnitude better than previously obtained. Except near the surface, temperatures vary linearly with depth but are in disagreement with the simplest model of a temperate glacier, being about 0.02 deg colder near the surface and 0.04 deg colder at 192 m depth.

The surface-perpendicular component of velocity and strain-rate have been determined at one site in the ablation area of Blue Glacier, Washington, U.S.A., where the total depth is about 250 m. The strain-rate is near zero at the surface but increases with depth to about 4% a−1 at 175 m. The results were obtained with the help of a finite deformation theory from the measured stretch of cables frozen into the ice.

Compressional wave velocity gradients at 43 of 50 Antarctic traverse stations plot as sequences of straight lines on semilogarithmic graph paper. Intersections of the lines appear to correlate with depths at which the predominant metamorphic mechanism in polar snow changes. The seismie pattern supports a three-layer interpretation of snow densification. The base of the upper layer (8.4±12.3 m) corresponds to the “critical depth” of Anderson and Benson (1963) at which snow grains settle into a “random closepacked” arrangement. The base of the lower layer may correspond to the firn: ice transition depth, but more data are needed to confirm this conclusion. It is unclear what densification phenomenon is marked by the base of the middle layer (27.7±4.4 m). The distinction between the middle and lower layers tends to disappear and the velocity gradient at a fixed depth increases as mean annual accumulation decreases.

The displacement of the surface of an ice sheet and of markers set in its top layers can be measured geodetically, and also, it is expected, by radio-echo methods. The paper discusses how such measurements could be interpreted as showing long-term changes in the thickness of the ice sheet; in particular it discusses how one might design an experiment so as to avoid unwanted effects due to short-term changes in rate of accumulation. The analysis is similar to that of Federer and others (1970), but it corrects an error, so that when applied to their results for central Greenland it gives a different result for the lowering of the surface. Federer and others have already concluded that the average accumulation rates during the past 100 years have been below those needed to keep in balance with the velocity of the ice sheet as a whole. Using a particular model, it is found that this has resulted in the surface lowering at a mean rate of 0.050 m a−1 between 1871 and 1968, and a mean rate of 0.140 m a−1 between 1959 and 1968.

A high-resolution radio echo sounder operating at a frequency of 620 MHz has been developed for studies of temperate glaciers. Excellent spatial resolution is obtained through the use of a short pulse length (70 ns) and an antenna beam width of 5.2°. Large amounts of high-quality data may be rapidly collected since the sounder incorporates an automatic positioning system and an on-line computer. Real time analysis of the echoes facilitates the understanding of complex reflecting horizons observed in temperate glaciers.

Results obtained during field trials of the echo sounder on both the Wapta Icefield and Athabasca Glacier are given. Intraglacial structures which may be due to water levels within the ice have been detected.

Radio echo soundings on Rusty Glacier, a small surge-type glacier in Yukon Territory reveal that the ice is considerably thicker than previously believed. A reinterpretation of deep ice-temperature measurements made in 1969 and 1970 suggests that a large zone of temperate basal ice exists. This result supports thermal instability as the surge mechanism for Rusty Glacier.

As part of a program to study surge-type glaciers, a radar-depth survey, using a frequency of 620 MHz, has been made of Trapridge Glacier, Yukon Territory. Soundings were taken at 26 locations on the glacier surface and a maximum ice thickness of 143 m was measured. A rapid change in surface slope in the lower ablation region marks the boundary between active and stagnant ice and is suggestive of an “ice dam” or the water “collection zone” postulated by Robin and Weertman for surging glaciers.

Steady-state gravity flow of air (inversion wind) on sloping snow-covered ice sheets is analyzed for sensitivity to local topography. Topographic features of the order of a few kilometres or less in length are too small to affect the direction and speed of this air flow. Air flow on a longer scale should however conform cosely to topography. Surface roughness on ice sheets is consistent with these results. Features of length shorter than a few kilometers (drifts and sastrugi) are transient, but longer features (surface undulations) remain essentially unaltered for many years. On the longer scale, inversion wind speed and therefore the amount of drifting and blowing snow should vary with the surface slope even where slope changes by as little as 1/10%. Observed variations in surface mass balance (aceumulated snow) in upper Marie Byrd Land, Antarctica, support this hypothesis.

Snow drift and inversion winds thus constitute a feed-back mechanism on the form of ice sheets and some of the topographic detail, formerly attributed to ice-flow character alone, may be in large part due to this mechanism.

During the summers of 1969 and 1970, we recorded in the ablation zone of the Glacier de St-Sorlin (Massif des Grandes Rousses, France) temperature, air moisture, and wind profiles, as well as the radiation balance and the daily ablation. Numerous profiles characterize a katabatic flow following the line of greatest slope, and there appears to be a correlation between the speed of the “glacier wind” and the corresponding temperature gradients. Computed according to Prandtl’s theory of turbulent transfers, the flux of sensible and latent heat added to the radiation flux leads to theoretical values for the daily melting in good agreement with the measured values. The relative importance of the radiation balance on the melting of the snow is 57%; that of the sensible heat flux is 43%; the latent heat flux is very weak and negative.

It is shown that anisotropic scattering with a strong forward peak can give reasonable agreement with angular reflectance data for snow. As a result of the forward peak, solar radiation penetrates deeper into the medium, when measured in terms of photon mean free paths, than it does for isotropic scattering. The radiation transmitted directly through finite slabs can be seen to an optical depth of seven, and decreases much more rapidly with optical depth than does the diffusely transmitted (scattered) radiation.

Holes drilled into thin areas of the Brunt Ice Shelf encounter a layer of liquid brine less than 1 m thick approximately at sea-level. Assuming the brine to be moving horizontally, analysis of its effects on thermal equilibrium gives an estimate of steady-state annual brine flow that is in good agreement with the value deduced from a percolation model. The effect of firn density on percolation rates is such that the slope of an active brine layer increases rapidly as ice thickness increases. However, the heat transport model predicts that brine layers are unlikely to be active in both very thick and very thin ice shelves.

Laboratory experiments on the growth of sea ice in a very thin plastic tank filled with salt water, cooled from above and insulated with thermopane, clearly show the formation and development of brine drainage channels. The sea-water freezing cell is 0.3 cm thick by 35 cm wide by 50 cm deep; the thermopane insulation permits the ice interior to be photographed. Experimentally, we observe that vertical channels with diameters of 1 to 3 mm and associated smaller feeder channels extend throughout the ice sheet. Close examination of the brine channels show that their diameter at the ice-water interface is much narrower than higher up in the ice, so that the channel has a “neck” at the interface. Further, oscillations occur in the brine channels, in that brine flows out of the channel followed by a flow of sea-water up into the channel. Theoretically, a qualitative theory based on the difference in pressure head between the brine inside the ice and the sea-water provides a consistent explanation for the formation of the channels, and the onset of a convective instability explains the existence of the neck. Finally, an analysis based on the presence of the brine-channel neck provides an explanation for the observed oscillations.

The extension of the Quaternary glaciations has been studied in the semi-arid Andes of north-central Chile, where the glacial modelling is striking. In the Elqui valley (lat. 30°S.), two glacial advances were identified reaching down to 3 100 m (Laguna glaciation) and 2 500 m (Tapado glaciation). In the Aconcagua valley (lat. 33°S.), moraines from three major glacial advances were found, at 2 800 m (Portillo glaciation), 1 600 m (Guardia Vieja glaciation) and 1 300 m (Salto del Soldado glaciation).

The Quaternary glaciations were linked with a decrease of temperature, but more significantly with a marked increase of precipitation probably related to an equatorward shift of 5–6 degrees of the austral polar front. The results obtained in the semi-arid Chilean Andes are correlated with those recently reported from other sectors of the southern Andes.

The study of temperate glaciers has led to the development of three drilling processes. The steam drill used for ablation stakes is easily portable and its speed is 30–40 m/h for the first 10 m. With the hot-water drill we can drill a fairly large but irregular hole; and reach 100 m depth in less than 5 h. The electro-thermal drill, usable to great depths, is particularly suitable for rapid drilling (15m/h and more) of small diameter holes (25 min) with easily portable equipment.

Studies of Norwegian glaciers indicate that foliation is frequently derived from sedimentary stratification. Shearing and accompanying recrystallization of the ice is most likely to occur in the planes of the sedimentary layers, particularly where they are steeply dipping and have a longitudinal trend, Foliation not related to pre-existing layering is uncommon. Steeply dipping transverse layers, often referred to as foliation, are believed to be simply the traces of former crevasses.

Dirty snow avalanches have been observed to carry considerable amounts of rock debris on to lake ice at the foot of scree slopes. As ice breaks up in the spring thaw, this material is carried back and forth on ice floes and is gradually deposited in the lake. In some areas this produces typical drop stones of rock debris in predominantly fine-grained deposits. Most avalanche debris is very angular which enables avalanche drop stones to be differentiated from those of glacial or other drift-ice origins. However, where avalanches incorporate glacial debris, such deposits may be indistinguishable from those formed by floating glacier ice.