Geology Reference
In-Depth Information
actions of rodents, such as the beaver (
Castor canadensis
) in building dams and raising
water levels, may also cause trees to die and permafrost conditions to change (Lewkowicz
and Coultish, 2004).
In those areas where the treeline and the southern boundary of continuous permafrost
are in close proximity, as is the case for much of North America, the presence or absence
of trees may become critical as regards permafrost occurrence. In Alaska, for example,
even isolated white spruce may infl uence the energy exchange at the ground surface suf-
fi ciently for a small permafrost body to exist (Viereck, 1965).
5.7.4. Snow Cover
Snow cover, like vegetation and organic material, infl uences local permafrost conditions
because it is also a good insulator (Table 5.1). The snowfall regime, type of snow, and
length of time snow lies on the ground are critical factors. In general terms, a heavy snow-
fall in autumn or early winter inhibits frost penetration, while a winter of low snowfall
does the reverse. Also, if snow persists late into spring, ground thawing will be delayed.
In detail, variations in snow cover are controlled by site characteristics such as micro-
relief, vegetation, and direction of the dominant snow-bearing winds.
In areas north of treeline, snowfall amounts are more limited and the effects of snow
cover are generally regarded as less important. However, signifi cant differences in active-
layer conditions still occur. For instance, it is not uncommon for upland surfaces and
interfl uves to be blown clear of snow for much of the winter while large snow banks accu-
mulate in gullies and on lee slopes. Deeper frost penetration may occur on the uplands
and interfl uves, therefore, than in depressions and on lee slopes. On the other hand,
ground thawing at snow-bank localities may be delayed until late summer when the
snow bank fi nally disappears. As a consequence, the active layer is often thinner at snow-
bank localities than on uplands. The role of snow banks as regards slope hydrology,
mass wasting, and other processes in areas north of treeline is discussed more fully in
Chapter 9.
South of treeline, especially in the discontinuous permafrost zones, snow cover fre-
quently assumes great importance in terms of permafrost distribution. For example, a
study at Schefferville, northern Québec, indicated that the pattern of accumulation of the
seasonal snow cover is the controlling factor in the distribution of permafrost in that area
(Granberg, 1973; Nicholson and Thom, 1973). Permafrost occurs only in uplands where
an absence of trees prevents snow from accumulating. By contrast, in adjacent lowlands,
where a thick snow cover accumulates each winter, permafrost is generally absent. It
appears that a winter snow depth of 65-70 cm is suffi cient to prevent the development of
permafrost in the Schefferville region (Nicholson and Granberg, 1973).
To illustrate the complexity of the climate-vegetation-snow interaction, Table 5.5 sum-
marizes ground temperature data from near Churchill, Manitoba, where permafrost
occurs at three adjacent sites but not at a fourth. All are located within a 2 km
2
area.
Churchill is at the border between continuous and discontinuous permafrost and is a few
km north of treeline. The mean annual air temperature is
7.3 °C and the average perma-
frost thickness, where present, is between 40 and 60 m. Table 5.5 convincingly demon-
strates that the progressive increase in average ground temperature at sites 2, 3, and 4
correlates with an increase in peat thickness and snow cover, and a decrease in snow
density. The absence of permafrost at site 4, a depression adjacent to site 3, is related to
waterlogging, enhanced by the accumulation of snow in the depression.
Similar subtle differences in terrain conditions occur widely throughout the dis-
continuous permafrost zones in both North America and Eurasia. As a consequence, it
−
