Remote sensing, regional ground temperature and ground ice observations, and numerical simulation were used to investigate the size, distribution, and activity of ice wedges in fine-grained mineral and organic soils across the forest-tundra transition in uplands east of the Mackenzie Delta. In the northernmost dwarf-shrub tundra, ice wedge polygons cover up to 40% of the ground surface, with the wedges commonly exceeding 3 m in width. The largest ice wedges are in peatlands where thermal contraction cracking occurs more frequently than in nearby hummocky terrain with fine-grained soils. There are fewer ice wedges, rarely exceeding 2 m in width, in uplands to the south and none have been found in mineral soils of the tall-shrub tundra, although active ice wedges are found there throughout peatlands. In the spruce forest zone, small, relict ice wedges are restricted to peatlands. At tundra sites, winter temperatures at the top of permafrost are lower in organic than mineral soils because of the shallow permafrost table, occurrence of phase change at 0°C, and the relatively high thermal conductivity of icy peat. Due to these factors and the high coefficient of thermal contraction of frozen saturated peat, ice wedge cracking and growth is more common in peatlands than in mineral soil. However, the high latent heat content of saturated organic active layer soils may inhibit freezeback, particularly where thick snow accumulates, making the permafrost and the ice wedges in spruce forest polygonal peatlands susceptible to degradation following alteration of drainage or climate warming.

[1] Low-centered ice wedge polygons in the Big Lake Delta Plain of the outer Mackenzie Delta are unusual because their bounding ramparts appear to have a single ridge. Twenty-two ice wedges in the area were examined between 2006 and 2009 to describe their morphology and diagnose their growth processes. The ground above ice wedges had a subtle microtopography, with ridges of 0.12 m relief and 4.0 m total width, bisected by troughs only 0.05 m wide and 0.09 m deep. The troughs, initially obscured by vegetation growth and organic matter, were underlain by ice wedges with average widths that increased downward in the uppermost 1 m of permafrost from 0.03 to 0.95 m. "Shoulders" on the ice wedges indicated vertical growth stages. Temperatures near the top of permafrost were favorable to thermal-contraction cracking, and ice veins connected to the top of wedge ice were observed in the active layer at five sites. These observations indicate the ice wedges are syngenetic and active, although without dating control, we cannot unequivocally dismiss the possibility that the wedges are epigenetic features that were truncated by a recent thaw unconformity. Muted relief above the ice wedges, which is uncommon above epigenetic ice wedges, was largely due to aggradation of the surface. Secondary ice wedges have not developed within the polygons, suggesting that climate variability has not led to polygon network development in this area. Wedge ice occupied only about 1.5% of the uppermost 1 m of permafrost, a much smaller volumetric proportion than in epigenetic settings.

Permafrost underlies peatlands of the Great Slave region, Northwest Territories, Canada, but permafrost relations beneath other ecotopes of black spruce (Picea mariana), white birch (Betula papyrifera) and mixed forests remain unknown. Permafrost-ecotope relations examined over a 3 year period (2010-13) establish the occurrence and thermal state of permafrost under these different types of forest. Air temperatures and snow depths are regionally consistent. Ground temperature variation primarily reflects latent heat effects during the freezing season, with the duration of season-normalised active-layer freezeback explaining 76% of 1 m ground temperature variation among all sites except xeric peatland. Low apparent thermal diffusivities from substantial latent heat effects strongly attenuate ground temperature variation with depth, and yield zero annual amplitude depths of 7 m or less where annual mean ground temperatures range among sites from -1.4°C to 0.0°C. Extensive discontinuous permafrost conditions, related to the extent of forested ecotopes, are commonly in thermal disequilibrium. Whereas permafrost in peatlands may be ecosystem-protected, this represents only about 2% of the area of the region. Permafrost in other forested ecotopes, occurring in ice-rich unconsolidated sediments, is climate-driven and ecosystem-protected because of latent heat effects. Though the rate of permafrost degradation may be reduced, an eventual transition to isolated permafrost retained primarily within ecosystem-driven peatlands implies substantial reductions of permafrost extent in this region.