Greinar (Vísindafélag Íslendinga) - 01.01.1977, Blaðsíða 126

124
mantle strength at depth, cases 1) and 3) most likely combine
in explaining the undulations of the geoid.
By the application of plasticity theory, it was concluded
(93) that (when the Icelandic positive undulation was, for
lack of the now available oceanic data, taken to be not much
larger the country itself) under Iceland, the strength must
be less than 10 kg/cm2. This would apply to the upper mantle
and/or the Low velocity layer. The very extensive undula-
tion of which Iceland is a part, could not exist by such a low
strength. This seems to mean that causes of the undulations
cannot be in the upper part of the mantle, but may in part
be due to generally existing strength in the lower part, which
would allow intermediate earthquakes.
Deep sea trenches. These are in zones, where there certainly
is strength down to 700 km. Referring to the picture given
in (95) for the Tonga treneh, and considered to be typical for
other such trenches, we have a high-Q bent tongue of thick-
ness 50—100 km, which goes through a low-Q zone, reaching
a depth of 200—300 km, to enter again high-Q material, ac-
cording to present general results. The Benioff seismic zone
is the upper surface of this high-Q tongue. The trench is at
the upper end of the tongue, where the bending from the
horizontal high-Q continuation of the tongue, i.e. from the
lithosphere, is sharpest.
As already pointed out, it is quite meaningless to try to tell
the origin and history of this tongue on the basis of tempo-
rary seismic data. On the other hand, the concurrence of the
trench and its negative isostatic anomaly, with the sharp
bend of a high-Q layer, is what first strikes the eye. It only
depends on the strength of this high-Q layer, and the age of
the trenches and the present stress in the bend, whether these
quantitatively explain the trench, the gravity anomaly, and the
deep earthquakes.
As to the age of the trenches, it is just an ad hoc hypothesis
in Plate Tectonics, that it is as high as some 150 My.
We may not be far off the truth, if we assume that the