Jökull - 01.01.2005, Blaðsíða 84
Kate T. Smith and Hreinn Haraldsson
yr BP from Fljótshlíð down to the coast, which is un-
usual for presently vegetated areas in Iceland.
This may mean that before that time, the region
west of Mýrdalsjökull was relatively frequently over-
flooded by jökulhlaups depositing sandy and granular
sediments, as has been the case with the areas south
and east of Mýrdalsjökull in later times (Skógasandur
and Mýrdalssandur). Although sedimentary records
in the Markarfljót to date do not indicate such high
frequency events this may predominantly reflect the
low preservation potential of small-scale (Drumbabót-
sized) flood sediments and landforms. It is most likely
that jökulhlaups of this size flowed westward from
the Katla area more often than the terrestrial records
found so far can indicate and that these smaller events
had a significant impact on shaping the landscape dur-
ing this time.
CONCLUSIONS
The Drumbabót site in the Fljótshlíð area in the Mark-
arfljót sandur plain is a key site in the story of Markar-
fljót floods and for understanding issues of flood im-
pact and reconstruction. First studied by Haraldsson
during the 1970s–1980s, reconsideration by the au-
thors in 2000 has placed the site in a wider context
of Markarfljót floods and relates it to more general is-
sues in palaeoenvironmental reconstruction. This pa-
per presents a new radiocarbon date on a bark sam-
ple taken from a tree at Drumbabót south of Fljóts-
hlíð indicating that an extensive mature birch wood
was killed at 1230±35 14C yr BP, before the deposi-
tion of the Landnám tephra and the settlement of Ice-
land. Evidence from across the sandur area suggests
that these trees had been growing across the region for
some centuries.
Analysis presented here of the deposits surround-
ing these trees suggests that they were inundated by a
jökulhlaup fromKatla travelling down theMarkarfljót
valley, confirming earlier work by Haraldsson (1993).
Importantly the Drumbabót site points out the devas-
tating impact a medium-scale flood event can have on
the environment. Holocene records of flood activity
predominantly limited to sedimentary evidence at val-
ley margins may underestimate the frequency of such
events. Were they to happen today they could present
a significant hazard.
Acknowledgements
KTS was supported during doctoral study by Natu-
ral Environment Research Council (UK) studentship
GT/04/99/ES/77 from 1999 to 2003. Logistical and
financial support for fieldwork was provided by the
University of Edinburgh, Fornleifastofnun Íslands,
Icelandair (UK) Ltd and Vango Outdoors Ltd. Warm
thanks are extended to Guðrún Larsen, Dr. Hreggvið-
ur Norðdahl and Dr. Mike Church for comments and
discussion in preparation of this paper. KTS would
like to thank Prof. David Sugden and Dr. Andrew
Dugmore for helpful discussions. Dr. Gordon Cook
(Scottish Universities Environmental Research Cen-
tre, East Kilbride, U.K.) provided advice and help
with AMS radiocarbon dating. Electron Microprobe
Analysis was carried out at University of Edinburgh
with the support of Dr. Peter Hill and Dr. Anthony
Newton. We would particularly like to thank Markús
Runólfsson for help with access to the site and useful
discussions. Eleanor Green, Amelia Hunt and Fiona
Murray are acknowledged for their help in the field.
ÁGRIP
Ummerki forsögulegs stórhlaups á Markarfljóts-
aurum
Fyrir rúmum 1200 árum (1230+35 C14 ár) eydd-
ist fullvaxinn birkiskógur sunnan Fljótshlíðar á Suð-
urlandi af völdum vatnagangs og sandburðar í jökul-
hlaupi til vesturs frá Kötlu. Þetta hlaup var hið síð-
asta margra forsögulegra jökulhlaupa sem flætt hafa
um Markarfljótsaura og Landeyjar á nútíma. Grein-
ingar á setsýnum benda til að hlaupið hafi tengst gosi í
Kötlu, svipuðu jökulhlaupumaustar með suðurströnd-
inni sem þekkt eru frá sögulegum tíma. Ný kolefnis-
aldursgreining, gjóskulagarannsóknir og athuganir á
trjástubbum sem eftir standa sýna að birkitrén voru
60–100 ára gömul þegar þau voru færð í kaf. Gróð-
urfar var sennilega svipað á mestum hluta láglendis í
Landeyjum á þessum tíma. Setlög og skúraðar klapp-
ir ofar með Markarfljóti, á svipuðum stað í jarðlaga-
skipaninni, geta tengst þessu hlaupi frá vestanverð-
um Mýrdalsjökli niður Markarfljótsdal. Með tilliti til
áhættu á svæðinu er mikilvægt að hafa staðfestingu á
84 JÖKULL No. 55