Jökull - 01.12.2006, Qupperneq 68

David W. McGarvie et al.
lar relevance to the model in Figure 4 are steep-sided
chimneys that have been observed in the deeper ice
beneath the shallow-angled basin of the ice cauldron
(for example at 1996 Gjálp eruption, Guðmundsson
et al., 2004; Guðmundsson, 2005). A scaled-up ver-
sion would provide the two forms of receptacle that
could explain the two tiers of lavas seen at the best-
developed tuyas, with the (confined) lower lava tier
being erupted within the steep-sided vertical chim-
ney (Figure 4a) and the (unconfined) upper tier be-
ing erupted above this (Figure 4b). Given that the bot-
toms of recently-observed ice cauldrons in Vatnajök-
ull have been up to 200 m below the original ice sheet
surface (Guðmundsson, 2005), and that lava B at rhy-
olite tuyas is only a few tens of metres thick at most
(Tuffen, 2001; Tuffen et al., 2002), it is reasonable to
suggest that the original initial ice sheet elevation at
rhyolite tuya eruptions was at least 100 m higher than
the summit of the tuyas, and possibly as much as 200
m higher (see Figure 4b). Consequently, it can be con-
cluded that the height of a rhyolite tuya summit will
represent only the minimum ice sheet thickness at the
time of eruption.
As well as enabling estimates of ice thicknesses
to be made for each of the three locations where these
tuyas were erupted, if corrections due to Holocene re-
bound are taken into account (so that the likely origi-
nal elevation of the tuya can be estimated) then an es-
timate of the minimum elevation of the ice sheet sur-
faces can also be made. Using reasonable densities of
920 kgm−3 for ice and 3100 kgm−3 for the Icelandic
mantle (Sigmundsson, 1991), and assuming a simple
linear relationship between ice thickness and depres-
sion of the lithospheric, a value of c. 0.3 the ice thick-
ness is derived. This simple relationship (see Table 2)
indicates that during the eruption of the ring fracture
rhyolites of Illihnúkur and SW Rauðfossafjöll the ice
sheet elevation at each tuya was 1093 m and 1087 m
respectively, which is consistent with the idea that the
ice sheet surface may be relatively insensitive to local
bedrock topography, which has long been recognised
at the Vatnajökull and Mýrdalsjökull ice caps (Björns-
son, 1988).
The older tuya of Gvendarhyrna provides less re-
liable information on ice thickness. An Ar-Ar age of
278 ka suggests that Gvendarhyrna has been eroded
during three glacial maxima (the Drenthe, Warthe,
and Weichselian), with an unknown thickness of ma-
terial removed from its summit area (e.g. pumiceous
carapaces are absent). Nevertheless the height differ-
ence between the base of the tuya (580 m) and the
summit (879 m) indicates a minimum ice thickness of
300 m. Given that the tuya was eroded by three ice
sheets, it is possible that the original summit altitude
may have been higher than the present one.
Figure 4. A schematic diagram showing one pos-
sible relationship between rhyolite tuya morphology
and ice cauldron morphology, after Tuffen (2001): (a)
shows a cone-shaped ice cauldron developed in an ice
sheet, with confined rhyolite lava A being erupted into
a steep-sided deep chimney; (b) shows further upward
development of the tuya lava cap, with the emplace-
ment of the relatively unconfined and thin lava B. Not
drawn to scale. – Skýringarmynd sem sýnir hvernig
höfundar sjá fyrir sér súrt gos undir jökli.
66 JÖKULL No. 56