Jökull - 01.12.2006, Qupperneq 69

Pleistocene rhyolitic volcanism at Torfajökull, Iceland
By combining Ar-Ar ages with tuya heights (and
applying a minor correction for post-deglaciation re-
bound, see Table 2), it can be concluded that in the
Torfajökull area at 67–72 ka the surface elevation of
the Weichselian ice sheet was at least 1100 m, and the
surface elevation of the Drenthe ice sheet at 278 ka
was at least 790–840 m. This represents an impor-
tant first step towards developing a new proxy of using
rhyolite tuyas to provide estimates of ice thicknesses
for past glacial periods.
Ar-Ar ages, glacial/interglacial periods, and oxy-
gen isotope stages
Figure 5 shows oxygen isotope variations (a proxy for
ice volume) for the past c. 420 ka, onto which the Ar-
Ar ages and their uncertainties have been added. Ta-
ble 1 provides details of the glacial/interglacial period
indicated by the eruption age, as well as the oxygen
isotope stage(s) bracketed by eruption ages and their
1σ uncertainties.
The two youngest ages (67±9 ka and 72±7 ka),
from the ring fracture rhyolites, indicate eruption of
these tuyas during the last (Weichselian) glacial pe-
riod, probably during cold OI stage 4 (Table 1) which
started at 75–80 ka (Lowe and Walker, 1997). The
age for the suspected older tuya of Gvendarhyrna
(278±18 ka) suggests eruption during the Drenthe
glacial period, and its 278±18 ka age falls entirely
within cold OI stage 8 (Lowe and Walker, 1997).
Therefore for all three of these tuyas there is good
correlation between the field evidence that suggests
the presence of an ice sheet (i.e. sustained eruption
and ice confinement leading to tuya formation) and
the climatic conditions that oxygen isotopes indicate
would have prevailed at the time of eruption (i.e. the
considerably greater ice volume associated with cold
periods).
The two units on the caldera rim (TJ97-14 and
TJ97-9) provide less-useful palaeoclimatic informa-
tion, largely because they are not from tuya-forming
eruptions and so ice thicknesses are unknown. Al-
though the 83±6 ka age spans a wide temperature
range, even the minimum temperature in this range
indicates conditions at least 2◦C colder than those
of the present day (Figure 5). An 83±6 ka age in-
dicates eruption during OI stage 5, but the ±6 ka
uncertainty in the age precludes assigning this erup-
tion exclusively into either substage 5a (temperate) or
5b (cold). OI stage 5 is characterised by fairly rapid
fluctuations between cold and temperate conditions,
with two distinct cold periods (OI substages 5d and
5b) when there was rapid build-up of land ice (Mc-
Manus et al., 1994) followed by rapid deglaciation
(Lundqvist, 1986) during OI substages (interstadials)
5c and 5a. The 384±20 ka age spans the widest tem-
perature range of all five samples (Figure 5), from
temperatures close to those of the present day to those
that are -6◦C colder.
This large temperature range is partly a conse-
quence of this eruption occurring at a time when tem-
peratures were falling towards a glacial maxima at
c. 350 ka (Figure 5), and partly a consequence of
the higher uncertainties associated with Ar-Ar dating
older samples. Although a 384±20 ka age indicates
Table 2. Estimated minimum ice sheet thicknesses (i.e. difference in height between base and summit) based
on the assumption that the summit of the tuya was beneath the original pre-eruptive ice sheet surface. Orig-
inal ice sheet elevation values are less than present tuya summit elevations due to a minor correction for
post-deglaciation rebound (see text for details). – Mat á lágmarksþykkt ísaldarjökulsins þegar Illihnúkur, Rauð-
fossafjöll og Gvendarhyrna mynduðust, sbr. 4. mynd.
Tuya Age Base Summit Minimum ice Ice sheet
elevation elevation thickness elevation
Illihnúkur 72 ka 670 m 1131 m 460 m 1093 m
SW Rauðfossafjöll 67 ka 800 m 1174 m 374 m 1087 m
Gvendarhyrna 278 ka 580 m 879 m 300 m 789 m
JÖKULL No. 56 67