To assess the response of lake Freeze-up and break-up dates to changes in atmospheric forcing, a physically based computational model of the coupled lake, lake-ice, snow and atmosphere system has been developed. Model performance is validated using meteorological and lake-ice observations from Great Slave Lake in northern Canada (1991/92) and St Mary Lake in Glacier National Park, Montana, (1992/93). Model integrations with modified atmospheric forcing indicate that air-temperature changes of ±4°C can delay or speed up the freeze-up and break-up dates by as much as 4 weeks for St Mary Lake, and 2 weeks for Great Slave Lake. For both lakes, break-up date is more sensitive to air-temperature changes than is freeze-up. Changes of ±3/10 cloud-cover fraction produce a shifting of break-tip dates by 1 week. Changes in wind speeds of ± 3 m s−1 modify the maximum ice depth of the lakes by 5–10 cm. For Great Slave Lake, lower wind speeds produced a surface temperature low enough to delay the onset of break-up by 2 weeks.

Bolshoi Lyakhov is one of the group of the New Siberian Islands in the Laptev Sea. The permafrost of the island is of an extremely low temperature regime, polygonal wedge ice being the most specific feature. The geomorphological level considered is a so-called edoma, presumably of late-Quaternary origin: polygonal ice wedges are more than 10 m wide and up to 25 m deep on this level, and about 1 m × 1.5 m on the peat bogs of Holocene age. Sixty-six samples of underground ice were taken on both surfaces. The macro-ion content was analyzed, i.e. Ca, Mg, Na, K, HCO3, Cl, SO4. A significant difference in ion content was found between the older and the younger ice. The late-Quaternary wedge ice is characterized by the predominance of Ca and HCO3, while the Holocene ice contains considerably higher proportions of Na and Cl. This may be attributed to different environmental conditions during wedge-ice growth: more continental in the late Quaternary and more maritime in the Holocene.

Models are developed to simulate changes in permafrost distribution and glacier size in mountain areas. The models exclusively consider equilibrium conditions. As a first application, the simplified assumption is used that one single parameter (mean annual air temperature) is changing.

Permafrost distribution patterns are estimated for a test area (Corvatsch-Furtschellas) and for the whole Upper Engadin region (eastern Swiss Alps) using a relation between permafrost occurrence as indicated by BTS (bottom temperature of the winter snow cover) measurements, potential direct solar radiation and mean annual air temperature. Glacier sizes were assessed in the same region with data from the World Glacier Inventory database. The simulations for the glaciers are based on the assumption that an increase or decrease in equilibrium-line altitude (ELA) would lead to a mass-balance change. Model calculations for potential future changes in ELA and mass balance include estimated developments of area, length and volume. Mass changes were also calculated for the time period 1850–1973 on the basis of measured cumulative length change, glacier length and estimated ablation at the glacier terminus.

For the time period since 1850, permafrost became inactive or disappeared in about 15% of the area originally underlain by permafrost in the whole Upper Engadin region, and mean annual glacier mass balance was calculated as −0.26 to −0.46 m w.e.a−1 for the larger glaciers in the same area. The estimated loss in glacier volume since 1850 lies between 55% and 66% of the original value. With an assumed increase in mean annual air temperature of +3°C, the area of supposed permafrost occurrence would possibly be reduced by about 65% with respect to present-day conditions and only three glaciers would continue to partially exist.