Jökull - 01.12.1989, Blaðsíða 7

12 3 4
RECEIVING SPREADS SHOTPOINTS
Table I. Seismic velocities in the surface layers of
the ice shelf as calculated by the Wiechert-Herglotz
method.
Taflal. Bylgjuhraði í yfirborðslögum íshellunnar
semfall afdýpi.
Fig. 3. Shotpoint and receiving spread arrangement
on survey lines. A spread between every two shot-
points is used for recording signals from shots at
both shotpoints.
Mynd 3. Fyrirkomulag skotpunkta og skjálftanema
á mœlilínum.
1. Velocity analysis, which is the basis for normal
moveout correction of the data as well as the
conversion of reflection times to depth.
2. Normal moveout correction.
3. Enhancement of the data which involved filtering
and editing.
VELOCITY ANALYSIS
Two separate methods were used for velocity
analysis. For the uppermost 40 m of the ice shelf a
210 m long reversed surface to surface refraction
profile was shot near the southem end of seismic
line 2. For the deeper layers of the ice shelf the X2t2
method was used (Dix, 1955; Sheriff and Geldart,
1983).
As illustrated by Joset and Holtzscherer (1953)
the Wiechert-Herglotz method for calculating the
velocity distribution with depth (Grant and West,
1965) is ideal for the uppermost layers of glaciers,
where the transformation of snow to ice takes place.
The results of applying the W-H method to the
refraction prohle are given in Table I and Fig. 4.
The velocity increases rapidly from 700 m/s at
7=0 m to 2000 m/s at 5 m. A velocity of 3000 m/s is
observed at about 22 m depth and 3500 m/s at 30 m.
Below 30 m the gradient is much smaller.
The main source of information on the seismic
velocity at depths below 40 m (Table II and Fig. 4)
consists of reflections from the ice-water interface.
To invert these times into velocity the X2t2 method
is used, which is based on the equation
Z (m) V (m/s) Z (m) V (m/s)
1.3 1131 15.7 2629
2.6 1548 25.3 3118
4.1 1833 29.8 3478
4.8 2000 36.1 3557
9.5 2250 43.3 3723
11.7 2500
t2 2 X2 - to + V2 ’ v rms (1)
where t is the reflection time measured at the surface at a distance X from the shotpoint and to is the
reflection time as measured at the shotpoint. The
root-mean-square velocity, Vrms, is defined by the
equation
V = v rms (XV?ti)/(Xti) i=l i=l (2)
The model of the subsurface is of n horizontal layers
where V; and t; are respectively the velocity in and
the traveltime through layer i. In order to calculate
the velocities of individual layers, the interval velo-
cities, the Dix formula is used
Vn =
n-1
(VL,n^ti-V^nSín_1 Xti)/tn
i=l i=l
(3)
The Dix formula follows directly from (2). It is
important to note that the formula is not valid for
large offsets, as it is assumed that the only difference
in raypaths for waves reflected from interfaces n-1
and n is the additional travel between the two inter-
faces. If the two interfaces are not parallel large
errors can occur and the Dix formula gives results
that are meaningless. This however, is not a prob-
lem in the present survey, as areas where the ice-
water interface is dipping can easily be identified
JÖKULL, No. 39, 1989 5