Firn-densification modeling based on hot isostatic pressing with power-law creep is investigated using depth–density data from 38 sites that collectively have mean annual temperatures ranging from 216 to 256 K and accumulation rates ranging from 0.022 to 1.2 m w.e. a−1. We use an inverse technique to obtain free parameters in a simple physical model for different stages of time-dependent firn densification. Our model works as well as or slightly better than previous models interpolating within the data range, but extrapolating would require additional physics.

Three episodes of strong basal motion occurred at Trapridge Glacier, Yukon Territory, Canada, on 11 June 1995 following the establishment of a connected subglacial drainage system. Responses to these “spring events” are noted in the records for 42 instruments and were recorded throughout the ∼60 000 m2 study area. Strong basal motion during the events is indicated by ploughmeter, load-bolt and vertical-strain records, and abrupt pressure changes in several transducer records denote damage caused by extreme pressure pulses. These pressure pulses, generated by the abrupt basal motion, also resulted in the failure of seven pressure sensors. Records for pressure, turbidity and conductivity sensors indicate that basal drainage patterns did not change significantly during the events. Geophone records suggest that the episodes of basal motion were precipitated by the gradual failure of a “sticky spot” following hydraulic connection of part of the study area. This failure resulted in the transfer of basal stress to the unconnected region of the bed during the course of the events. No evidence for strong basal motion is seen in the instrument records for several weeks following the events, suggesting that the mechanical adjustments resulted in a stable configuration of basal stresses. This event illustrates how unstable situations can be quickly accommodated by mechanical adjustments at the glacier bed.

Modelling the heat flow in small, cold high-altitude glaciers is important for the interpretation of paleoclimatic data from ice cores. Coupled glacier-flow and heat-flow models are presented that incorporate the densification, heat advection and possible phase transitions at the permafrost boundaries within the bedrock. Marked bends observed in the temperature profiles from two recent boreholes on Colle Gnifetti, Swiss Alps, are interpreted with the help of a transient heat-flow model, driven with a temperature history. The conclusion is that substantial warming of the mean firn temperature at shallow depths has taken place over the last few decades. This has not been observed before in cold-firn regions of the Alps. Modelled heat fluxes in the Monte Rosa massif are strongly influenced by the mountain topography. This leads to a spatial variability of the temperature gradient near the glacier base which has been observed in boreholes to the bedrock. In order to match the measured temperature profiles in the glacier, the vertical heat flux at great depth must be set to an extremely low value. It is shown with the help of the transient heat-flow model that this is a paleoclimatic effect, possibly enhanced by a degrading permafrost base.

A glacier at the summit of Ushkovsky volcano, Kamchatka peninsula, Russia, was studied in order to obtain information about the physical characteristics of a glacier that fills a volcanic crater. The glacier has a gentle surface and a concave basal profile with a maximum measured depth of 240 m at site K2. The annual accumulation rate was 0.54 m a−1 w.e., and the 10 m depth temperature was −15.8°C. A 211.70 m long ice core drilled at K2 indicates that (1) the site is categorized as a percolation zone, (2) the stress field in the glacier changes at 180 m depth from vertical and longitudinal compression with transversal extension, which is divergent flow, to a shear-dominated stress field, and (3) the frequent occurrence of ash layers can be a good tool for dating the ice core. The borehole temperature profiles were considered to be non-stationary, but the linear profile made it possible to estimate the basal temperature and the geothermal heat flux at K2. Assuming constant surface and the basal boundary conditions, we constructed two depth–age relationships at K2. These predicted that the bottom ages of the ice core were about 511 or 603 years.

We have mapped Antarctic blue-ice areas using the U.S. National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) Antarctica cloud-free image mosaic established by the United States Geological Survey. The mosaic consists of 38 scenes acquired from 1980 to 1994. Our results show that approximately 60 000 km2 of blue ice exist for each of the two main types of blue ice: “melt-induced” and “wind-induced”. Normally, the former type is located on slopes in coastal areas where climate conditions (i.e. persistent winds and temperature), together with favourable surface orientation, sustain conditions for surface and near surface melt. The latter blue-ice category occurs near mountains or on outlet glaciers, often at higher elevations, where persistent winds erode snow away year-round, and combined with sublimation creates areas of net ablation. Furthermore, we have identified an additional area of 121 000 km2 as having potential for blue ice. However, in these areas features such as mixed pixels, glazed snow surfaces, crevasses and/or shadows make interpretation more uncertain. In conclusion, a conservative estimate of Antarctic blue-ice area coverage by this method is found to be 120 000 km2 (∼0.8% of the Antarctic continent), with a potential maximum of 241 000 km2 (∼1.6% of the Antarctic continent).

Sortebræ is a surge-type tidewater glacier complex draining southeastward from the Geikie Plateau, East Greenland. Sortebræ’s main flow unit surged around 1950 and again between 1992 and 1995. The 1990s surge affected the lower 50 km of Sortebræ over an area of approximately 335 km2. Over a period of <1 year the tidewater front advanced >5 km. Surge velocities in the order of kilometres per annum are about 100-fold the quiescent velocities. Multi-model photogrammetric analysis shows a thinning of the reservoir zone of up to 219 m and thickening of the receiving zone of up to 74 m. The surge transported approximately 18.6 km3 of ice down-glacier. The total calving volume as a result of the surge amounted to 11.7 km3, equivalent to a calving flux of 3.9–7.3 km3 a−1. The surge characteristics and environmental setting suggest that the surge mechanism involves a switch in the subglacial drainage. This surge of Sortebræ is more similar to the fast, short Alaskan-type surges than to the sluggish, long Svalbard-type surges.

Laboratory experiments that simulate natural ice-formation processes and sediment entrainment in shallow water are presented. A 10–30 cm s−1 current was forced with impellers in a 20 m long, 1 m deep indoor tank. Turbulence in the flow maintained a suspension of sediments at concentrations of 10–20 mg L−1 at 0.5 m depth. Low air temperatures (∼−15°C) and 5 m s−1 winds resulted in total upward heat fluxes in the range 140–260 W m−2. The cooling produced frazil-ice crystals up to 2 cm in diameter with concentrations up to 4.5 g L−1 at 0.5 m depth. Considerable temporal variability with time-scales of <1 min was documented. A close to constant portion of the smaller frazil crystals remained in suspension. After some hours the larger crystals, which made up most of the ice volume, accumulated as slush at the surface. Current measurements were used to calculate the turbulent dissipation rate, and estimates of vertical diffusion were derived. After 5–8 hours, sediment concentrations in the surface slush were normally close to those of the water. After 24 hours, however, they were 2–4 times higher. Data indicate that sediment entrainment depends on high heat fluxes and correspondingly high frazil-ice production rates, as well as sufficiently strong turbulence. Waves do not seem to increase sediment entrainment significantly.

In 1997 a 121 m ice core was retrieved from Lomonosovfonna, the highest ice field in Spitsbergen, Svalbard (1250 m a.s.l.). Radar measurements indicate an ice depth of 126.5 m, and borehole temperature measurements show that the ice is below the melting point. High-resolution sampling of major ions, oxygen isotopes and deuterium has been performed on the core, and the results from the uppermost 36 m suggest that quasi-annual signals are preserved. The 1963 radioactive layer is situated at 18.5–18.95 m, giving a mean annual accumulation of 0.36 m w.e. for the period 1963–96. The upper 36 m of the ice core was dated back to 1920 by counting layers provided by the seasonal variations of the ions in addition to using a constant accumulation rate, with thinning by pure shear according to Nye (1963). The stratigraphy does not seem to have been obliterated by meltwater percolation, in contrast to most previous core sites on Svalbard. The anthropogenic influence on the Svalbard environment is illustrated by increased levels of sulphate, nitrate and acidity. Both nitrate and sulphate levels started to increase in the late 1940s, remained high until the late 1980s and have decreased during the last 15 years. The records of δ18O, MSA (methane-sulphonic acid), and melt features along the core agree with the temperature record from Longyearbyen and the sea-ice record from the Barents Sea at a multi-year resolution, suggesting that this ice core reflects local climatic conditions.

Observation of the terminus behavior of 38 North Cascade glaciers, Washington, U.S.A., since 1890 shows three different types of glacier response: (1) Continuous retreat from the Little Ice Age (LIA) advanced positions from 1890 to approximately 1950, followed by a period of advance from 1950 to 1976, and then retreat since 1976. (2) Rapid retreat from 1890 to approximately 1950, slow retreat or equilibrium from 1950 to 1976, and moderate to rapid retreat since 1976. (3) Continuous retreat from 1890 to the present.

Type 1 glaciers are notable for steeper slopes, extensive crevassing and higher terminusregion velocities. Type 2 glaciers have intermediate velocities, moderate crevassing and intermediate slopes. Type 3 glaciers have low slopes, modest crevassing and low terminusregion velocities. This indicates that the observed differences in the response time and terminus behavior of North Cascade glaciers in reaction to climate change are related to variations in specific characteristics of the glaciers. The response time is approximately 20–30 years on type 1 glaciers, 40–60 years on type 2 glaciers and a minimum of 60–100 years on type 3 glaciers. The high correlation in annual balance between North Cascade glaciers indicates that microclimates are not the key to differences in behavior. Instead it is the physical characteristics — slope, terminus velocity, thickness and accumulation rate — of the glacier that determine recent terminus behavior and response time. The delay between the onset of a mass-balance change and initiation of a noticeable change in terminus behavior has been observed on 21 glaciers to be 4–16 years. This initial response time applies to both positive and negative changes in mass balance.

One of the questions still unanswered concerning the surge behavior of glaciers concerns their quasi-periodic occurrence. Some results on the phenomenological connection between local cumulative balance and surge initiation of Variegated Glacier, Alaska, U.S.A., are discussed here. Based on climate data from neighboring weather stations, an empirical relation between precipitation, temperature and local mass balance is established and used to reconstruct the annual balance at a location in the accumulation area back to 1905. Between the last four surges in 1946/47, 1964/65, 1982/83 and 1994/95, the ice-equivalent cumulative balance was 43.5 m on average, with a 1σ error of 1.2 m. Although the existence of a surge level cannot be directly interpreted in physical terms, it explains the variable length of the quiescent periods of Variegated Glacier by variations in the accumulation rate prior to the surge. We use the surge level to hindcast former unobserved surges, to compare the results with other surge datings obtained from photographs and to establish a complete surge history for Variegated Glacier for the 20th century.

Annual proglacial solute fluxes and chemical weathering rates at a polythermal high-Arctic glacier are presented. Bulk meltwater chemistry and discharge were monitored continuously at gauging stations located at the eastern and western margins of the glacier terminus and at “the Outlet”, 2.5 km downstream where meltwaters discharge into the fjord. Fluxes of non-snowpack HCO3−, SO42−, Ca2+ and Mg2+ increase by 30–47% between the glacier terminus and the Outlet, indicating that meltwaters are able to access and chemically weather efflorescent sulphates, carbonates and sulphides in the proglacial zone. Smaller increases in the fluxes of non-snowpack-derived Na+, K+ and Si indicate that proglacial chemical weathering of silicates is less significant. En3hanced solute fluxes in the proglacial zone are mainly due to the chemical weathering of active-layer sediments. The PCO2 of active-layer ground-waters is above atmospheric pressure. This implies that solute acquisition in the active layer involves no drawdown of CO2. The annual proglacial chemical weathering rate in 1999 is calculated to be 2600 meqΣ+ m−2. This exceeds the chemical weathering rate for the glaciated part of the catchment (790 meqΣ+ m−2) by a factor of 3.3. Hence, the proglacial zone at Finster-walderbreen is identified as an area of high geochemical reactivity and a source of CO2.