Rit (Vísindafélag Íslendinga) - 01.06.1971, Side 24

24
GUÐMUNDUR PÁLMASQN
scatter in the reduced values. Some of this scatter is probably due to
random causes, such as shot point variability, ground-seismometer
conditions, etc. It has, however, been observed in some cases that a
relatively regular variation in reduced amplitude can be related to
a structure in the upper boundary of the refractor. The amplitude
curves may therefore be of value in combination with the travel time
curves, to aid in constructing a correct model for the interpretation
of the seismic data.
In studies of continental crustal structure (e.g. Steinhardt and
Mater, 1961) it has been common practice to use amplitude ratios
rather than the observed amplitudes themselves. This eliminates
some of the remaining variabihty mentioned above. In the present
work this procedure has not been found to be possible, because two
waves of the same kind (P or S) are not usually observed together
over the range of distances required. The ratio of P and S waves could
on the other hand be used for some studies, but no special emphasis
has been placed on the study of S-wave amplitudes in the present
work.
5. GENERAL FEATURES OF THE RESULTS
5.1. Layering of the crust.
A total of over 80 refraction profiles have been measured so far.
They are distributed over most parts of Iceland including the shallow
water off the west and south coasts. Relevant data on profiles 1-30
have been given by Pálmason (1963), while data on the remaining
profiles are given in Tables 5.1—5.2. The velocities are apparent velo-
cities and the thicknesses are computed on the basis of horizontal
boundaries. Although these calculations are not in accordance with
the real geological situation, they serve nevertheless to give a “nor-
mal” structure, on the basis of which geographical variations can be
discussed.
Tables 5.1—5.2 and Pálmason (1963) furthermore give S-wave
velocities, where they have been determined, and the calculated
values of Poisson’s ratio.
The histogram in Fig. 5 shows the distribution of apparent velo-
city values. The peaks correspond to the various layers. Variations
within each group are due partly to dips in the interfaces and partly
to real variations in the velocity values, especially in the upper
layers. The absence of a clear peak corresponding to layer 4 is due