Within discontinuous permafrost, significant differences in ground thermal regimes, permafrost conditions and processes may occur over short distances, due to variations in
moisture conditions, vegetation cover, surficial sediment types, and in depth to bedrock. In these and other permafrost environments, studies of present-day processes and landforms provide information for paleoenvironmental and geomorphic
reconstructions, and applied geotechnical purposes including hazard and environmental assessments. Herein, we examine contrasting contemporary settings of sand wedges and peatland terrain developed in sand within the extensive discontinuous
permafrost environment in the Great Slave Lake region of southern Northwest Territories, Canada, where the historical (1980-2010) mean annual air temperature is approximately -4.1°C. Observations: The study area resides on a raised beach setting,
with active and stabilized eolian surfaces, along the western shoreline of Great Slave Lake. Seasonal ground cracking, in linear and polygonal patterns, is observed on bare to sparsely vegetated active eolian surface (insets Fig. 1A) in the area.
Sand wedges on a bare active eolian surface (Fig. 1A and B) are partially obscured by a surface cover of active eolian sand and surface lags of pebbles. They are tapered, extending to ~1-m depth and up to 1-m wide at surface. These sand wedges have
formed within thin surface cover of sand-sheet sediments of variable thickness, (up to 40-cm), and within underlying beach sediments. The underlying beach sediments are optically dated to about 7.5 ka, and sand within a wedge dates to less than ca.
1.7 ka (Fig. 1D). Laminations in host eolian sand-sheet sediments are upturned adjacent to the wedges. At a lichen-covered stabilized eolian site, numerous sand-wedge polygons, 6 - 12 m in diameter, are noted (Fig. 1A and C). The troughs are covered
by a mat of live reindeer lichen. In contrast to the wedges at the bare active eolian site, the underlying features extend to depths in excess of 1.2 m with maximum widths of 20 cm (Fig. 1E and F). Their widths are relatively equal below about 50-cm
depth, but taper upwards from 50-cm depth towards the surface. These features are interpreted as sand veins, and are hosted by eolian sand-sheet sediments. Host sand-sheet sediments are optically-dated to about 3.0 ka at 80-cm depth, as are the sand
wedge sediments (Fig. 1E). Overlying detrital charcoal within host eolian sediments dates to ca. 2.0 ka at 60-cm depth and less than ca. 300 years at 30-cm depth (Fig. 1F). Organic material infilling the trough 20 cm below the surface may be reworked
organic debris, as it dates to ca. 1.1 ka (Fig. 1F). Host eolian sand-sheet sediments are horizontal to upturned below 50-cm depth, but are downturned at 20- 40° within shallower adjacent sediments. In addition, overlying organic material from within
the troughs extend downward into the underlying sand veins. Open voids are apparent within the upper 50 cm of the sand veins. Ground temperature sensors installed to 4.4-m depth and augering to similar depths indicate that permafrost is not
currently present at the sand wedge or sand vein sites. However, a third site, which is in an open spruce forest peatland (Fig. 1A) with peat extending to 90-cm depth, is underlain by permafrost. Ground temperatures recorded for two years within the
underlying sand at 1-m depth indicate that the active layer is contained entirely with the overlying peat. Temperature data indicate significant differences in the thermal regimes between the peatland and sand wedge sites. In 2014-15 annual mean,
maximum and minimum air temperatures of the study area were -4.0, 25.2, -41.2°C, respectively. Corresponding ground temperatures at 1-m depth for respective sites were: bare sand wedges -0.9 (mean), 11.2 (max), -12°C (min); lichen-covered sand wedges
-0.2, 14.1, -14.7°C; and peatland -0.9, -0.1, -2.8°C. In addition, whereas ground temperatures at 50-cm depth within the peatland were no colder than -4.1°C in either the winter of 2013 or 2014, ground temperatures at both sand wedge sites were
colder than -20°C. Whereas snow cover on vegetated sites in this region typically exceeds 30 cm, cameras set up to record snow depths over the winter indicated the snow cover was typically less than 10 cm at the sand wedge sites. Discussion: These
observations reveal distinctively contrasting present-day geomorphic processes and ground thermal regimes, and past processes operating over short distances within this discontinuous permafrost setting. Sand wedges on the active eolian surface are
interpreted as having developed epigenetically within underlying beach sediments. However, the associated sand-sheet deposits, surface lag of pebbles and cobbles, as well as the broad width of the upper surface of the sand wedges suggests that these
wedges may be continuing to develop anti-syngenetically within the sediments. The shallow depth of these sand wedges suggest that they form from seasonal cracking, within less than 1-m depth where winter ground temperatures may be colder than -20°C.
Although it is uncertain if permafrost was present at the site in the past, they are not presently forming within permafrost. Sand veins on the lichen-covered stabilized eolian site are interpreted to have developed syngenetically within eolian
sand-sheet deposits as they aggraded, as is apparent by their narrow width and age relations with surrounding sediments. Upturning of adjacent host sediments at depth is indicative of infilling of the wedge by sand. However, significant downturning
in upper host sediments, infilling of the vein by overlying organic material, and thinning of the sand wedge width, are indicative of a termination of sand infilling. This is most likely attributed to the stabilization of eolian activity and
development of vegetation cover, and may have occurred within the last few centuries. Similar to the active eolian site, whereas sand veins may have initially formed in permafrost terrain in the past, they continue to operate within seasonally frozen
ground, with cracking similarly restricted to less than the upper metre of sediment based on the modern thermal regimes and the depth of observed downturning in sand veins sections. Open voids within sand veins may be the result of seasonal infilling
of cracks by water and formation ice, which soon after melts as the ground thaws in summer. This process would account for downturning of adjacent sediments and infilling by overlying organic material to fill open voids. Within the upper few metres
of sediment, these sand wedge and sand veins sites, which do not contain permafrost, are significantly colder than adjacent peatland and other permafrost-bearing environments within this region. Sand wedge formation at these sites occurs within
seasonally-frozen ground, typically at depths of less than 1 m. Although these sand wedges are not directly associated with permafrost, they are presently found in association with discontinuous permafrost terrain.
|Summary||(Plain Language Summary, not published)|
Within discontinuous permafrost, significant differences in ground thermal regimes, permafrost conditions and processes may occur over short distances,
due to variations in moisture conditions, vegetation cover, surficial sediment types, and in depth to bedrock. In these and other permafrost environments, studies of present-day processes and landforms provide information for paleoenvironmental and
geomorphic reconstructions, and applied geotechnical purposes including hazard and environmental assessments. Herein, we examine contrasting contemporary settings of sand wedges and peatland terrain developed in sand within the extensive
discontinuous permafrost environment in the Great Slave Lake region of southern Northwest Territories, Canada, where the historical (1980-2010) mean annual air temperature is approximately -4.1°C.