Jökull - 01.01.2020, Qupperneq 13

Hannesdóttir et al.
These studies are referred to in Table 2.
The age of glacial geomorphological features, in-
cluding moraines, may often be estimated from his-
torical accounts, surface texture, vegetation cover and
lichenometry, but remains uncertain in many cases, as
further discussed below. A (subjective) estimate of
the horizontal accuracy of the LIA glacier outline, as
derived mainly on the basis of glacial geomorpholog-
ical evidence, is on the order of 50–100 m, based on
the width of the geomorphological features. Problems
in identifying the crest or the outermost part of the
moraines from remote sensing imagery affect the ac-
curacy of the ∼1890 glacier outline. The accuracy of
the age of the terminal moraines, that is the time when
the glacier margin was last adjacent to the moraine,
where available from the above mentioned sources, is
presumably in the range of 10–50 years.
The smaller and thinner glaciers, often leave only
vague traces of their former extent, which results in an
uncertain LIA maximum glacier outline. The differ-
ent types of bedrock of the various glacierized areas
also affect the preservation of the moraines, trimlines
and glacier erratics. The glacial erosion is, for ex-
ample, different in the neovolcanic zone than within
the Tertiary basaltic bedrock in East and West Iceland.
This applies to the glacier outlets on the eastern side
of Vatnajökull, where glacial geomorphological evi-
dence is sparse on the eroded bedrock characteristic
of this area. This applies also to parts of the Dranga-
jökull ice cap and the northeastern side of Langjök-
ull. The total length of the uncertain LIA outline is
estimated to be a few km for eastern Vatnajökull and
negligible for Langjökull. The reconstruction of the
maximum LIA glacier outline of Drangajökull is cov-
ered in a separate section.
LIA glacial geomorphological evidence is not
available in some areas where repeated jökulhlaups
and braided glacial rivers have sculptured the land-
scape, washed away moraines or buried them with
sediments. In these areas and others were detailed
geomorphological evidence is ambiguous, we have
drawn a qualitatively estimated LIA maximum out-
line based on the geometry of younger outlines at
these locations.
For simplification, the timing of the LIA maxi-
mum is here referred to as “∼1890”. However, there
is firm evidence from historical documents (e.g. Þór-
arinsson, 1943) and lacustrine records (Harning et
al., 2016), that Drangajökull started receding already
around 1850. Also, it is worth noting that Skeiðarár-
jökull, the large southern outlet glacier of Vatnajökull,
reached its maximum extent during a surge in 1929
(Sigurðsson, 2005). An estimated maximum LIA or
∼1890 extent of virtually all glaciers in Iceland is
shown on the map of Icelandic glaciers published by
the IMO (Sigurðsson et al., 2017). The delineation
of this outline is mostly based on satellite and aerial
imagery, oblique aerial photographs and lidar DEMs,
as well as on historical information, used to identify a
“recent” maximum glacier extent visually, but without
field observations or dating of the identified features.
In this paper (see Figures 4–10), the LIA glacier out-
line has been updated in certain areas, taking into ac-
count more detailed mapping of the glacial geomor-
phological landforms where available (see references
cited in Table 2). Thus, the maximum LIA glacier ar-
eas have changed slightly from the values presented
on the map of Icelandic glaciers (Sigurðsson et al.,
2017).
Glacier extent in 1945/1946
The US Army Map Service (AMS) created topo-
graphic maps of Iceland at a scale of 1:50,000. These
were based on aerial photographs taken in 1945 and
1946 and had contour lines drawn at 20 m intervals.
During recent years, the original AMS aerial pho-
tographs, that are now stored at the National Land
Survey of Iceland (NLS), have been scanned, geo-
rectified and processed to create DEMs and orthoim-
ages utilizing the 60% acquisition overlap (Magnús-
son et al., 2016; Belart et al., 2019). These DEMs
have been used to digitize glacier outlines more ac-
curately than those displayed on the original topo-
graphic maps (see for example Belart et al., 2019,
and Andreassen et al., 2020). In Table 2, a distinc-
tion is made between 1945/1946 glacier outlines dig-
itized from the original maps and outlines derived
more recently from the scanned aerial photographs.
Two-thirds of the 1945/1946 outline for Vatnajökull
(from Skeiðarárjökull to the west and north to Lamba-
10 JÖKULL No. 70, 2020