Jökull - 01.12.2006, Blaðsíða 8
Haraldsdóttir et al.
winds, causing new snow to be wind packed imme-
diately. It has been shown that the snowpack in Ice-
land has generally higher densities than the snowpack
in the Alps (Jóhannesson et al., 1998), but close to
the values found in the western part of Norway. The
densities are high due to wind packing, interchanging
periods of melting and freezing throughout the winter.
Salinity could also be of importance in some places,
close to the coast.
Most avalanche accidents occur during or shortly
after strong winds and heavy snowfall. In many coun-
tries the residential areas have already been perma-
nently protected, e.g. in the Alps. Increasing moun-
taineering during the winter has led to increased num-
bers of avalanches triggered by skiers or other traffic.
Two fatal avalanches in 1998 and 1999, when slabs
broke on weak layers during fine weather, were trig-
gered by traffic, one by a snow-scooter and the other
by a tractor. One person was killed in each accident.
REGISTRATION AND MAPPING OF
AVALANCHES
Jónsson (1957) conducted pioneering work on record-
ing the avalanche history of Iceland, including the
available information about avalanches causing dam-
age or fatalities from 1118 to 1957. A report on
avalanches from 1958 to 1971 with maps locating
the avalanches was published by Jónsson and Rist
(1972). The previously mentioned analysis of Björns-
son (1980) is partly based on these reports. Sigvalda-
son co-ordinated a new edition of Jónsson’s work
(Jónsson et al., 1992), based on previous work and
more recent registrations (Rist, 1975; Jónsson, 1981,
1983a, 1983b, 1984; Eyþórsdóttir, 1985; Ágústsson,
1987; Magnússon, 1988, 1989, 1991, 1992). Since
1990, information on avalanches has been collected
at Veðurstofa Íslands and published in reports. Jó-
hannesson and Arnalds (2001) presented an up to date
overview of avalanches, as well as information about
the costs of avalanche damage and protection mea-
sures.
In recent years, intensive work has been in-
vested in registering avalanches in the vicinity of
the towns presented in Figure 2. Reports for each
town have been published (Veðurstofa Íslands, 1997–
2004). The Public Roads Administration has also
registered avalanches on roads for many years. The
avalanche history provides necessary background in-
formation for hazard assessment (Arnalds et al.,
2004), permanent protection measures (Jóhannesson
et al., 1996), as well as evacuation plans (Veðurstofa
Íslands, 1997), avalanche forecasting, and develop-
ing and testing avalanche models (Jóhannesson et al.,
2001–2002).
Table 2 gives an overview of known avalanches
in 13 coastal towns and their immediate surroundings
up to 31 July 2003. The data is from the avalanche
database of Veðurstofa Íslands. The surroundings of
these towns and outlines of registered avalanches are
shown in Figures A1-A13. The maps are made us-
ing a digitally based Geographic Information System
(GIS). It is necessary to keep in mind that knowledge
regarding avalanches other than those which occurred
during the most recent decades, is limited in some of
the towns.
Evacuation in case of avalanche hazard
Evacuation plans for 15 towns endangered by
avalanches have been made at Veðurstofa Íslands
(Figure 2). They are: Ólafsvík, Patreksfjörður,
Bíldudalur, Tálknafjörður, Suðureyri, Flateyri, Súða-
vík, Ísafjörður, Hnífsdalur, Bolungarvík, Siglufjörður,
Ólafsfjörður, Seyðisfjörður, Neskaupstaður and Eski-
fjörður (Veðurstofa Íslands, 1997).
Dry slab avalanches are the primary hazard in
most of the towns, but some areas are threatened
by wet avalanches as well. There is a minor threat
of dry avalanches affecting Suðureyri and Tálkna-
fjörður, where there is more prominent hazard due
to wet avalanches and landslides. Evacuation plans
due to slush flow and landslide hazard are mostly
organised locally, in each case in cooperation with
Veðurstofa Íslands. Hazard zoning and hazard maps
of the towns have been made or are in preparation
at Veðurstofa Íslands. In addition, Þingeyri, NW-
Iceland, and Fáskrúðsfjörður, E-Iceland, are (in 2004)
considered for hazard mapping due to landslide risk.
Evacuation levels are defined according to
avalanche paths, run-out of known avalanches and
the results of avalanche modeling (Jóhannesson et al.,
6 JÖKULL No. 56