Jökull - 01.12.1966, Page 17

HOO m
Fig. 4. Funnel and sheet of volcanic material expected to be formed in an eruption that works
its way up through a 400 m thick ice cover. For comparison the table-mountain Skriða (broken
contour).
4. mynd. Trektarjylling og lag, sem vcenst er að myndist í gosi, sem brotizt hefur upp í gegnum
400 m þykkan jökul. Til samanburðar er stapafjallið Skriða (strikalína).
Finally, the difference between a sub-glacial
and a sub-marine eruption is now clear. In the
latter there is no solid ceiling that gives strong
counter-pressure and eventually stops the verti-
cal ílow of lava.
At the present time there are two active sub-
glacial volcanoes in Iceland, Katla and Gríms-
vötn. The conditions are insofar very special
as the volcanic fissures are situated at the foot
of a vertical wall of rock, and a more or less
open fracture in the ice, a kind of bergschrund,
must be contemplated. The ice is, furthermore,
not comparable in thickness with a Pleistocene
ice-cap. In the surroundings of Katla the ice
is much fractured because of rapid flow over
steep and uneven ground. Tliese favourable
conditions enable the eruption, in form of the
outrush of very fine fragments, to penetrate the
ice. But even these circumstances have not led
to anything remotefy related to a table-moun-
tain.
Even in these volcanoes there are possibilities
that the ice has in a nurnber o£ cases „suffocat-
ed“ an eruption in the initial stage. On the
basis of an empirical rule that lias been valid
for centuries an eruption of Katla has been
due since about 1953. In 1955 tliere was some
disturbance of the ice in this area and a big
water rush (jökulhlaup) came from Katla, but
there was no visible eruption. It seems possible
that there was really the very beginning of an
eruption that was then stopped by the ice. In
this case the next Katla eruption would be
due, according to the rule, after some 60 years
from 1955.
Quite similar things have happened in the
case of Grímsvötn. For a long time an erup-
tion, accompanied by a great rush of water,
happed every 10 years. But since 1934 only the
waterrushes have come regularly while tliere has
been no visible eruption. It is of course pos-
sible that these floods are only due to attain-
ment of a certain level of the lake Grímsvötn,
as this rises regularly between the floods. But
Thorarinsson (1966), in a recent review of the
data, considers this explanation unsatisfactory.
The possibility should then be kept in mind
that eruptions each time start the floods, but
are immediately suppressed.
APPENDIX.
a. Lifting of a long beam, cf. Fig. 3.
We consider the right half of the beam. The
left end is uplifted by u1(, is subjected to the
vertical force K and the moment M„. At a
distance x from the left end the uplift is u
and the moment M = — M0 + K • x — px2/2, p
being the uniform load. If E is Young’s mo-
dulus and I the moment of inertia of the sec-
tion of the beam, we have E ■ I • d2u/dx2 =
— M0 + Kx — px2/2.
Integration gives
E • I•du/dx =
— Mux + K • x2/2 — px3/6,
as du/dx = 0 for x = 0.
From du/dx = 0 for x = L we get
— M0 + KL/2 — pL2/6 = 0.
Further we take for x = L:
M = 0 = — M0 + K • L — p • L2/2,
and from these last equations we find
K = 2pL/3 and M0 = pL2/6.
JÖKULL 171