Glacier Lake Outburst Floods (GLOFs)—Mapping the Hazard of a Threat to High Asia and Beyond - Impact of Global Changes on Mountains: Responses and Adaptation - page 333

Lake characteristics

One of the most important variables for analyzing the potential danger of
a GLOF is the change of the glacial lake. Today, normally these changes
are identifi ed using multi temporal space imagery (cf. Table 14.2). The
growth of a supraglacial or proglacial lake depends primarily on the glacier
characteristics and retreat. However, it has to be noted that also strong and
fast glacier advances (surges) from a side valley into the main valley can
block the river and also lead to PDGL. This situation is especially known for
the Karakoram (Hewit and Liu 2010) The volume of typical proglacial lakes
may be addressed based on the empirical formula (Equation 1) suggested
by Huggel et al. (2002) which is primarily based on 15 lakes with existing
depth measurements.

V = 0.104 A 1.42 Equation 1

It is, however, important to emphasize that this scaling formula
may only serve as a rough estimate, since the lake volume depends on
several variables. For instance, by applying this formula to the lakes in the
Kishi Almaty Valley of the northern Tien Shan with existing bathymetric
measurements (Kasatkin and Kapista 2009, Tokmagambetov 2009) reveals
an overestimation of the volume of up to 20% (Bolch et al. 2011a). Similar
uncertainties were also mentioned by Huggel et al. (2002).

Characteristics of adjacent glaciers

In the fi rst instance the recent outlines of glaciers have to be mapped based
on multi-temporal imagery. These outlines allow to assess the changes
of the glacier front in detail. Glacier velocity is usually estimated from
multitemporal optical or SAR imagery based on feature tracking using
cross-correlation techniques (cf. Kääb 2005, Berthier et al. 2005, Bolch et al.
2008). Here, the open source software 'Cosi-Corr' (Leprince et al. 2007) may
serve the automated estimation of the velocity of mountain glaciers as well
(Scherler et al. 2008). Due to the better contrast than in the higher resolution
panchromatic band, the use of the near infrared band is recommendable.
A glacier may be called stagnant in its distal part if the calculated velocity
is below the uncertainty of one pixel.
The average slope of the glacier surface gives a hint where glacial lakes
can develop or an existing lake can extend in the near future. A threshold of
2° for supraglacial lake formation on debris-covered glaciers in the Himalaya
(Reynolds 2000, Quincey et al. 2007, cf. Bolch et al. 2008) or of 5° for the
formation of proglacial lakes in overdeepening of debris-free glaciers in the
Alps (Frey et al. 2010) have been reported.