Greinar (Vísindafélag Íslendinga) - 01.01.1976, Síða 85

83
and emphasize the liighest values of the times for the stations situ-
ated in the rift zone of Iceland.
The absolute times at a distance of 200 km in the Jeffrey — Bullen
— continental travel time curve is equal to 32.5 sec and 25 sec for
the oceanic one (see Fig. 3). Thus the conclusion can be drawn that
from the type of the graph the areas under investigations can be
related to the continental crust.
Seismic Cross-Section
The seismic cross-section has been constructed in two asstmiptions
as to the nature of the first arrivals and velocity spread in the crust.
1. The first arrivals are weakly refracted (penetrated) waves.
The velocity in the crust continuously increases with depth and the
velocity gradient is likely to vary i.e. the second order boundaries
are possible. Discontinuity M is characterized by 8 km/sec velocity.
In the mantle the gradient is very little.
2. The first arrivals are head waves. Each travel time curve
corresponds to a seismic boundary oharacterized by a jump of the
velocity.
In the first assumption the technique developed by N. I. Pav-
lenkova was employed (24). It involves two stages: first the experi-
mental travel time curves are transformed into the system of the
reduced variable velocities. From the reduced travel time curves a
time cross-section is constructed, the values of the reduction velo-
cities being its parameters. Then the time cross-section is recalcu-
lated into a depth one allowing (with account) for the experi-
mental dependence of the average velocity variations on depth in
the region under investigations. The cross-section is represented
in the form of the velocity isolines. Then the cross-section is ascer-
tained by solving a direct kinematic problem. If the observed and
calculated travel time cmrves differ greatly the cross-section is
corrected and a direct problem is recalculated. This operation is
repeated until the difference between the calculated and recorded
travel time curves becomes less than a precalculated error. Nor-
mally either the second or the third approximation yields a good
agreement. A section on fig. 7.2 matches the observed travel time
curves with accuracy ± 0.2 sec in average. Only a few rather