Geography Reference
In-Depth Information
consuming, restricted in spatial extent, and do not depict the spatial variability of
permafrost occurrence. Modeling techniques using GIS and DEM variables have been
used to estimate the distribution of mountain permafrost (Etzelmüller et al. 2001a,
2001b; Heginbottom 2002). Automated mapping of alpine permafrost began with the
creation of the PERMAKART and PERMAMAP empirical models for parts of the Alps
(Keller 1992; Hoelzle et al. 1993). PERMAKART utilizes three variables: (1) mean an-
nual ground temperature (MAGT), derived from mean annual air temperature (MAAT);
(2) potential solar radiation; and (3) thickness of snow cover. PERMAMAP is based on
the relationship among BTS measurements, MAAT, and potential direct solar radiation
(Hoelzle et al. 1993).
Similar models have been formulated using other physiographic, climatological,
and biologic variables (Imhof 1996; Frauenfelder 1997; Gruber and Hoelzle 2001).
PERMAMOD joins topoclimatic information with biogeographical features, such as cold
water, absence of marmot burrows, perennial snow patches, and sites of rock glaciers
(Frauenfelder 1997). Stocker et al. (2002) developed PERMEBAL, which simulates snow
cover persistence and ground temperatures of snow-free points, based on meteorolo-
gical and site-specific data. Guglielmin et al. (2003) designed the PERMACLIM model,
which calculates ground temperatures for DEM points by including data from a climat-
ic database and snow thermal characteristics. Janke (2005c, 2005d) used rock glaci-
er topographic information (elevation, slope, and aspect) combined with a land cover
weighting procedure to model distribution in the Front Range of Colorado. Despite the
variety of models, uncertainty still exists as to which is most useful and effective for
accurately representing the distribution of permafrost (Frauenfelder et al. 1998; Lugon
and Delaloye 2001).
Modeling results need additional verification in the field. For example, several stud-
ies suggest that the D-l site (3,739 m with a MAAT of −3.5°C) along Niwot ridge, Co-
lorado, contains permafrost (Ives 1974; Ives and Fahey 1971; Janke 2005c). Modeling
results suggest a 63 percent probability of permafrost occurrence at D-l (Janke 2005c).
However, through an electric resistivity tomography survey, Leopold et al. (2010) found
no evidence of ice from the surface to a depth of 10 m near D-l. Permafrost occurrence,
however, is defined by freezing temperature for at least two consecutive years; thus, it
can be dry or ice-free. Boreholes lined with temperature data loggers would provide ad-
ditional direct measurement of possible permafrost occurrence.
PERMAFROST AND CLIMATE CHANGE
Since the definition of permafrost is based on thermal criteria, scientists are naturally
concerned with the implications of global warming. General circulation models (GCMs)
usually take into account solar radiation, greenhouse gases, or boundary conditions to
predict future climates. Predicting climate change in mountain regions is always diffi-
cult because of the need for high temporal and spatial resolution, which is often lost in
GCMs (Barry 1994; Beniston 1994). In fact, the Alps are not even perceived by some
models, nor is the physical influence of mountains (gravity-wave drag). Nested models,
in which the GCM focuses on a detailed region, provide a method for improving pre-
diction in the mountains. However, the feedback mechanisms of what appears to be a
chaotic system still create a great deal of uncertainty.
Search WWH ::




Custom Search