A Seismic Refraction-Shooting Survey off the

North Coast of Cornwall

A. S. Memiweather*
(Received 1957 October 30)

Summary

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

A seismic refraction-shooting survey has been carried out along
four bearings from a hydrophone laid on the sea bed near Perranporth,
Comwall. The structure and acoustic properties of the sea bed have
been deduced, and the attenuation of ground-waves with range has
been measured both at this site and with a hydrophone suspended in a
well nearby. The results obtained suggest that the sea bed in this area
consists of three layers; a sand or sediment layer 80-100 ft thick with
velocity 4950-5500 ft/s, a high velocity rock layer 1300 ft thick with
velocity 14 250 ft/s, and a very thick high velocity layer with velocity
18 500 ft/s. The 14 ooo ft/s layer is correlated in velocity with schists
obtained from quarries near Perranporth, which suggests that this layer
is a northerly extension of the Palaeozoic rocks of Devon and Cornwall.
The attenuation of the ground-waves was found to be approximately
12 dB per double distance. A comparison of the results with the existing
theories of the propagation of explosive sounds over elastic and fluid
sea beds is made, and some recommendations for further research
proposed.

I. Introduction
This paper describes the results obtained during a refraction-shooting survey
carried out near Perranporth, Comwall, in July, 1956. The survey was undertaken
in order to determine the geological structure of the sea bed in this area, and the
seismic method used was essentially the same as that used on land. The theory of
the seismic refraction method is described by Bullard, Gaskell, Harland & Kerr-
Grant (1940), Heiland (1946), Leet (1938) and Willmore (1949).
The possibility of using hydrophones rather than geophones as receivers for
seismic refraction-shooting at sea has been described by Bullard & Gaskell (1941)
and Hill (1952).
Refraction-shooting profiles were determined along bearings 226"T, 270"T,
316"T and oooOT,from a hydrophone laid on the sea bed in a depth of 17 fathoms
and 2 miles off shore. I lb T.N.T. charges were fired on the sea bed along each bear-
ing at intervals of one-eighth of a mile to a range of 2 miles, quarter-mile intervals to a
range of 5 miles, and half-mile intervals to a maximum range of 10 miles. During
the runs along bearings 316"T and oooOT,the ground-wave signals were also
* Admiralty Research Laboratory, Teddington, Middlesex.
73
74 A. S. Merriweather
received on a hydrophone suspended 400 ft down a well located approximately
quarter mile south-east of Cligga Head. The nominal positions occupied by the
firing ship during the experiments are shown in Figure I.

Navigation was carried out using the Decca Navigator and by utilising the
green and purple lanes of the south-west British Decca Chain (which give almost
right-angled intersections in this area) very satisfactory accuracy was achieved.

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

FIG.I.-Chart showing the four refraction-shooting runs.

A comparison of the Decca positions and the positions calculated from the water-
wave travel-times indicated an accuracy of roc-150 yd. The firing ship gave thirty
seconds warning by radio before firing each charge, and the signal from the
explosion, received on an accelerometer clamped to the ship’s hull, was used to
give the time-of-firing by breaking a C.W. radio transmission.
During the four runs, continuous echo-soundings were also made in order to
determine accurate bottom profiles. The profiles computed from the echo-sound-
ings and corrected for tidal height variations using Admiralty Tide Tables are
shown in Figure 2. Bottom samples were also taken at 2-mile intervals along each
run, using a conical dredge. These samples were invariably found to consist of a
mixture of fine sand, small stones and broken shells.
 75
Seismic refraction-shooting s w e y off coast of Cornwall
The ground-wave measuring equipment used during the survey is shown in
the block schematic diagram Figure 3.

2. The analysis of the seismic results

The main purpose of the survey was to determine the depths to, and velocities
of, compressional waves in the different layers of strata of the sea bed, by measur-

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

s
Q
0 101 1
.I

I I /
0 I 2 3 4 f 6 7 B 9 1
0
Bearing ciwT
Abscissae : Range from hydrophone (nautical miles).
-- -
-
-- Echo sounder depths
Depths reduced to chart datum

FIG.2.-Bottom profiles along four bearings from hydrophone position. Hydrophone position
Lat. 50° 22' zoa N,Long. os0 I I' 49" W.

ing the variations of ground-wave travel-time with range. As is usual with the
analysis of seismic refraction-shooting results in an area which in depth is geologi-
cally unknown, it has been assumed that the compressional wave velocity increases
discontinuously with depth. There is no evidence in this area to indicate that
conditions exist which would produce a decrease of velocity with depth, By plot-
ting the ground-wave travel-time against water-wave travel-time, the velocities of
compressional waves in the various horizontally stratified layers of the sea bed
and the depths of these layers can be determined from the following well-known
equations :
C
l G W l = --tWW,
Cl
76 A. S. Merriweather

where :
tww = travel-time of water-wave.
t ~ w=l travel-time of ground-wave along c1 interface.
tGW2 = travel-time of ground-wave refracted at c2 interface.
tGWg = travel-time of ground-wave refracted at c3 interface.

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

c = velocity of sound in water.
c1 = velocity of compressional waves in the 1st sea bed layer.
c2 = velocity of compressional waves in the 2nd sea bed layer.
c3 = velocity of compressional waves in the 3rd sea bed layer.
hl = depth of layer with velocity c1.
hz = depth of layer with velocity c2.

The three equations above represent straight lines, the gradients of which
give the velocities in the layers, and the intercepts the depths to the layers. The
theory of the interpretation of seismic travel-time curves in horizontal structures
is described by Lehmann (1937), Muskat (1933) and Slichter (1932).
It frequently occurs that the line of lowest velocity on the time-distance diagram,
for which there are ground-wave arrivals, does not pass through the origin. For
the four refraction-shooting profiles described in this paper this occurred in every
case, and the line of lowest velocity gave a positive intercept on the ground-wave
travel-time axis. The positive intercept indicates the existence of a low-velocity
sand or sediment layer overlying the first layer from which ground-wave arrivals
were obtained. This layer probably represents Recent deposits but unfortunately
it is impossible to determine its depth without a knowledge of the velocity of com-
pressional waves in the layer. Therefore in this paper it has been assumed that
the velocity lies between 1.0and 1 . 1 times the velocity of sound in water. These
velocity values represent the limits found from many measurements made on ocean
sediments in America and the British Isles.
I n reading the beginning of the ground-wave, the “earliest” definite indication
of a disturbance of the ground-wave trace was taken. The time difference between
the time-of-firing signal and the beginning of the disturbance gave the ground-
wave travel-time. Due to the fact that the charges actually exploded slightly
before the time-of-firing signal was transmitted, it was necessary to apply a cor-
rection to the apparent travel-times of the water-wave and ground-wave signals.
A plot of ground-wave travel-time against water-wave travel-time then enabled
the structure of the sea-bed to be determined.
Unfortunately, it was not possible to fire any reverse shots during the survey
and therefore the component of dip along the four bearings has not been established.
However, the value for the velocity of compressional waves in the second rock layer
on bearing 270”Twas found to be 4 per cent greater than the corresponding values
obtained along the three other bearings, and in the theoretical analysis which
follows in a later section of this paper, an attempt is made to determine whether
this high value is consistent with an upward slope of this layer of strata.
 Bruel & Kjaer
Level Recorder

b P6i'
SoIartron Multiple Time of
B.40 a
Pre-Amplifier
- B.

c)
B
Pre-Amplifier -- a.
Power Pack -- 8
& Timing B
Unit 0
". L.F. L.F. H.F. T0.F
\Hydrophone)
0-60 C I SLow Pass P
a
aL
Filter c
6 . 3 ~ .D.C. Heater
4 Pen 1
Ediswan Ediswan Recorder Y
D.C. Driver -
- Twice Normal Paper
Panel Speed

Ediswan
LF. Oscillator Solartron
Pre-Amplifier Power Pack
LF. Calibration I
I
Solartron
Oscillator
for HF. Calibration
I pu
q Hydrophone

FIG.3.-Block schematic diagram of ground-wave mrmning equipment.

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

78 A. S. Merriweather
The travel-time curves derived from the experimental data are shown in Figures
4-9. The best straight lines were fitted to the points, using the method of least
squares, and the final results of the analysis are given in Table I.

Table I

Layer Velociiies and Depths

CllC C2 c3 h, h2
Bearing Velocity Velocity Velocity Depth to Depth to
from Contrast in 2nd in 3rd 1st layer 2nd layer
Hydrophone in 1st layer layer layer
ft/s €t/s ft ft

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

z26@T I '0 10700 18 300 II0 650
I 'I 1 0 700 18 300 I20 640

270"T I '0 14 zoo 19zoo 90 1280

I 'I 14zoo 19 zoo I00 I 280

3 I 6"T I '0 15 700 I8 500 130 I500

I 'I 1.5 700 I8 500 I 40 I500

OOO@T I '0 14 3 0 0 18400 90 1250

I 'I '4 300 18 400 I00 I250

It must be emphasised that the velocities shown in Table I are only apparent
velocities, owing to the fact that reverse profiles were not determined. If the strata
have a large dip, the velocities shown above may be considerably in error. Through-
out the analysis, a velocity of 4950 ft/s for the velocity of sound in sea water has
been assumed.
Bearing [email protected] travel-time curve for this bearing is shown in Figure 4.
It can be seen that the curve shows rather complicated variations at the closer
ranges, which probably indicate that the strata are not horizontal. This assumption
is confirmed by the fact that Boden rocks (which rise 50-100 f t above sea level)
are approximately on this bearing at a range of zQ miles. The deviations from a
straight line of the points covering the range 0.8 to 4 miles may possibly indicate
a slight upward slope of the 18 300 ft/s layer over this range.
Bearing 270"T. The travel-time curve for this bearing is shown in Figure's.
On the whole, the curve is in very good agreement with the simple ray theory,
but the deviations from a straight line at the longer ranges are quite marked.
A more detailed study of the curve indicates that the points covering the range
1.2 to 3-7 miles can be fitted to a velocity of 18 400 ft/s, and the points from 3.7
to 10 miles to a velocity of 19 150 ft/s. A possible explanation is that the third
layer is horizontal out to a range of 3-7 miles, and thereafter has an upward slope
of about 2". A dip of this magnitude would indicate that outcropping of this
layei should occur at a range of about 8 miles, and this is in good agreement with
the indications of rock given on the Admiralty Chart of the area.
Bearing 316"T.-The travel-time curve for this bearing is shown in Figure 6.
Although the curve tends to show a continuous change of gradient, the best
straight lines were again fitted to the points. However, although the points covering
 Seismic refiaction-shooting survey off coast of Cornwall 79

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

Water-wave travel time (sec)

FIG. 4.-Refraction-~hooting profile along bearing 226"T from the hydrophone. Hydrophone
position Lat 50' 22' 20" N, Long. 05' I I' 49@W.

FIG. 5.-Refraction-shooting
Water-wave travel time (sec)
t
profile along beating 270"T from the hydrophone. Hydrophone
position Lat. 50' 22' zol N, Long. 05" I I' 49" W.
80 A. S. Memweather
the range 1-2to 2 miles give a velocity of 18500 ft/s the remaining points (although
they give the same slope) are displaced upwards for ranges greater than 2 miles.
This vertical displacement can be interpreted as an increase in depth of the third
layer at this range, and is probably caused by faulting in the vicinity. These dis-
placements of travel-time curves are not unusual and other cases have been re-
ported by Day, Hill, Laughton & Swallow (1956).
Bearing ooo"T.-The travel-time curve for this bearing is shown in Figure 7.
The curve is again in good agreement with theory, but deviations from a
straight line do occur at a range of 3 miles. It is possible that this slight discon-
tinuity is associated with the decrease in water depth which occurs at this range.

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

Well hydrophone travel-time curves.-The travel-time curves for the ground-
waves received on the well hydrophone are shown in Figures 8 and 9.
The gradients of these two curves give an average value of 18150 ft/s for the
velocity in the third layer, which is in good agreement with the other results.
The close range results for bearing 226"T shown in Table I are significantly
lower than the results obtained on the three other bearings. If this low value is
disregarded, the other results give average values of 14 700 ft/s and 18500 ft/s
for the velocities in the second and third layers.

3. The dip of the strata

The value of 19 200 ft/s for the velocity in the third layer for bearing 270"T is
4 per cent greater than the average value obtained from the other three bearings.
The accuracy of the velocity measurements is known to be better than 2 per cent,
and therefore it would appear to be worthwhile considering whether this high
value is consistent with a dip of this layer. The standard formula for a two-layered
bottom without dip is:
c 2hl ~ 2 2 t
--
tGW2 = -tww+-
c2 c2 (
c? I) -
The variables used in the above equation have already been defined in a previous
section.
If the sediment layer is disregarded, the Perranporth area can be treated as a
two-layered case for dip considerations, and it can be shown that the above
formula is modified to that given below when the second layer is assumed to slope
upwards away from the receiving position.
c cos(e23+or) 2h2 COSCY ~ 3 2
tGm= -
~3 ~0~023
tww+ CS ( S - I r
where :
a = angle of dip.
c2 = true velocity in 14 ooo ft/s layer.
c3 = true velocity in 18 ooo ft/s layer.
823 = critical angle for ~ 2 1 interface
~ 3 = sin-1 CZ/C~.

Therefore, with an upward-sloping layer, the apparent velocity ca' as given

by the gradient of the refraction-shooting time-distance curves is :
cos 023
c3' = c3
+
cos (e23
 Seismic refraction-shooting survey off coast of Cornwall 81

4 or
Baring/zbl'
nlb. Ch- Firrd
on the Sa &d
3 5 I

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

Water-wave travel time (sec)

FIG. 6.-Refraction-shooting profile along bearing 3 I 6 T from the hydrophone. Hydrophone

position Lat. 50'' 22' 20" N, Long. 05'. XI' 49@W.

Water-wave travel time (sec)

FIG.7.-Refraction-shooting profile along bearing OOO'T from the hydrophone. Hydrophone

position Lat. soo 22' 20"N, Long. 05' I I.' 49. W.
82 A. S. Merriweather

"+t++

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

Hange trom well hydrophone (nautical miles)

FIG.&-We@ refraction-shooting profie for run along bearing 3 I 6"T from sea hydrophone
Well hydrophone position Lat. 50' 20' 13" N, Long. 05' 10' 18" W. Well hydrophone 400 feet
below ground level.

0
0 2 3 4 s 6 1

Range from well hydrophone (nautical miles)

r 10.9.-weit rejracrron-snoonng proptejor run atong oeanng 000- I jrom sea nyaropnone. well
hydrophone position Lat. 50' 20' 13. N, Long. 05' 10' 18"W. Well hydrophone 400 feet below
ground level.
 Seismic refraction-shooting survey off coast of Cornwall 83
For the particular case of the run along bearing 270'T, we assume that the
apparent velocity c3' of the third layer is 19 150 ft/s, and that the true velocity cg
is 18 500 ftls. Then by taking an average value of 14 250 ft/s for the velocity c2
in the second layer, the angle of dip is given by:
COS (3g042'+a) = 0'744
a = 2OIZ'.

This value of 2'12' is not unreasonable but unfortunately it cannot be checked

owing to the fact that no reverse shots were fired during the experiments.

4. Compressional wave velocities in rocks from neighbouring land areas

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

T o aid the interpretation of the velocities found from the refraction-shooting
data, the velocities of compressional waves in three rock samples obtained from land
areas close to Perranporth were measured in the laboratory of the Department of
Geodesy and Geophysics at Cambridge. The apparatus used for these velocity
measurements has been described in t f e paper by Day and others (1956). Approxi-
mately ten rocksamples were sent to Czmbridge, but most of these were found to be
rather badly weathered, and only three were considered to be representative of the
bulk of the country rocks in this area. The velocity measurements were all made at
atmospheric pressure and in order to investigate the change of velocity with
water-content, the rocks were wetted by evacuating air for several hours and then
aliowing water to enter the pore spaces. Table z below shows the velocities
obtained for the three good specimens in the wet and dry condition.

Table 2

Velocities in schists from the north coast of Cornwall

Specimen Thickness Velocity dry Velocity wet Remarks
cm ft/s ft/s

I 2.68 17 IOO 15 400

4.64 12 800 1 2 800
6'04 '3 500 14 400
I .06 9 800 9 800 Perpendicular to
lamination

2 4-30 13 IOO '3100 .

4-89 I5 I 0 0 15 800
I -08 g 800 9 800 Perpendicular to
lamination

3 3 80 15 700 15 I 0 0
4 65 1 3 IOO 13 100

The figures given in Table z indicate that there is a rather wide divergence
between the three directions of measurement. In particular specimens I and z
gave very low values when the measurements were made in the direction perpen-
dicular to the laminations. The discrepancy between the measurements probably
lies in the fact that the specimens were not pressurized. There is normally a rapid
change of velocity with initial pressure and therefore the in situ velocity in these
rocks is probably less sensitive to the direction of propagation than the laboratory
84 A. S. Memweather
measurements indicate. Therefore, from the point of view of propagation of
refracted waves, a value near the upper end of the range has been taken as the most
appropriate.
In order to obtain the best average value for the velocity of compressional waves
in these three rock samples, the results shown above, excluding the two low values
measured perpendicular to the laminations, were averaged and the standard
deviation calculated. The final results are as follows :
Mean velocity (dry) = 14 300 & 600 ft/s.
Mean velocity (wet) = 14 200 & 400 ft/s.
Although the accuracy of the above figures is not very great, the mean values
are in very good agreement with the value of 14 250 ft/s obtained from the refrac-

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

tion-shooting data.
Unfortunately, no specimens of rock corresponding to the 18000 ft/s layer
were found in the cliffs and quarries near Perranporth.

5. Ground-wave frequencies
A full analysis of the ground-wave frequencies and the dispersion of the received
signals has not been carried out, but from the results obtained it appeared that the
low frequency component preceding the water-wave arrival had a frequency of
10.2-11.5 CIS with invariably some third harmonic also present. Two typical
ground-wave signals for shots fired on the sea bed at ranges of 0.7 and 1.5 miles
from the hydrophone are shown in Figure LO.

c Water-Wive Arrival Time

Bearing 226"T.
Range 1.5 N nulea
Hydrophone position Lat. 50' 22' ;o" N, Long. of I I' 49" W.
Hydrochone on the aea bed
Charges fired on the sea bed
Ediswan paper speed I t cm/s.

W*tcr-Ware Arrival Time

B F U ' z2b'T
~~
Range 0.7 N milee.
FIG.~ o . - T w o typical ground-wave recordi.

The longer range shots showed considerable amplitude modulation of the

10 CISwave (presumably caused by a multiplicity of arrivals), and the ground-wave
lasted for a considerable number of seconds after the arrival of the water-wave.
In most records, a high frequency arrival of 40-45 CIS also appeared after the water-
wave arrival. This frequency was usually only present over part of the range, and
tended to disappear at the long range firing positions. At the longest ranges, the
ground-wave was a nearly sinusoidal modulated 10CIS wave; all the higher fre-
quency components having disappeared. No frequencies lower than 10c/s were
observed during the experiments.
During spring tide periods at Perranporth there is a 20 per cent change in
water depth at the position where the hydrophone was laid for these experiments.
 Seismic refraction-shooting survey off coast of Cornwall 85
This change should cause a detectable variation in the ground-wave frequency;
it would therefore be of great interest to fire charges at a fixed position throughout
a complete tidal cycle in order to investigate the variation of ground-wave fre-
quency with time.

6. Ground-wave amplitudes
Owing to the fact that the ground-wave consisted of a complicated modulated
10 c/s wave, some criterion had to be used in order to obtain consistent results for
the variation of amplitude with range. In this paper, the variation in amplitude
of the first modulation maximum after the first indication of the beginning of
the ground-wave has been taken. This corresponds to the "first arrival"

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

maximum amplitude, and not to the overall maximum amplitude of the wave.
The variations in amplitude with range and bearing are shown in Figures 11
and 12.
In all cases, the points show a rather large scatter, but a mean line through them
indicates that the attenuation is approximately 12 dB per double-distance along
bearings 270"T, 316"T and oooOT, and 9-10 dB along bearing 226"T. Along
bearings 226"T and 270"T the attenuation tends to increase with range, but for
bearings 316"T and O O O O T it remains reasonably constant at 12 dB per double-
distance. A loss of 12 dB per double distance corresponds to an intensity-range
law of Y-Z which is in good agreement with the ground-wave attenuations quoted
in Worzel & Ewing (1948).

7. Comparison of results with theory

The ground-wave results described in this paper indicate that the sea bed at
Perranporth consists of a low velocity layer of sand or sediment 80-100 ft thick in
which the velocity of compressional waves is probably 4950-5500 ft/s. Beneath
this top layer of sediment are two high-velocity layers of rock. The layer directly
beneath the unconsolidated sediment is approximately 1300 ft thick and the
velocity of compressional waves in it is 14 250 ft/s. The third layer is probably
very thick and the velocity of compressional waves in it is 18 500 ft/s.
Using the above data, and the existing theories of the propagation of explosive
sounds over elastic and fluid sea beds, an attempt will now be made to predict
the main characteristics of the ground-waves in the Perranporth area. The theory
for a sea bed which can be treated either as a semi-infinite fluid or a number of
fluid layers was fully developed by Pekeris (1948), and the case of a semi-infinite
elastic sea bed by Press & Ewing (1948, 1950) and Tolstoy (1954). Unfortunately,
none of these theories is directly applicable to the Perranporth case, where the sea
bed consists of a thin layer (which to a first approximation can probably be treated
as a fluid) overlying a semi-infinite elastic solid. However, even though the com-
parisons with theory are rather crude, they are certainly worthwhile making in
order to indicate the modifications required in order to obtain better agreement
between theory and experiment.
Pekeris (1948) shows that the predominant frequencies in the ground-wave
signal correspond to the cut-off frequencies of the normal modes. In most practi-
cal cases, the lower modes are more strongly excited than the higher modes, and
one would therefore expect the main ground-wave frequency to correspond to the
cut-off frequency of the first mode. The formulae shown below for the cut-off
86 A. S. Merriweather

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

+80

t7 0

"#
3
-e
U
+60.

9 +so
4
+40

d
2
d
+30

0.5 I .o 5.0 10.0

Distance from hydrophone (nautical miles).

Fro. I I .--Ground-wave transmission curvesfor bearings 226'T and 270"T. Hydrophone on the sea
bed at position Lat. 50' 22' 2omN, Long. 05" I I' 49" W.
 Seismic refraction-shooting survey off coast of Comwall 87

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

Distance from hydrophone (nautical miles)

FIG. ~z.-Ground-wave trannnission curves for bearings 316"T and ooo"T. Hydrophone on the sea
' N, and Long. 05' 1 1 ' 49. W.
bed at position Lat. 5 0 ~ ~ 2 220"
88 A. S. Memweather
frequency of the nth mode for a three-layered fluid model will be used to determine
the theoretical ground-wave frequencies to be expected in the Perranporth area.

fc=
(2n - 1)c
4Hu
.( I + ;TH)-l for-
h
< 0.5

and
(2n- I)CI h
fc = for - > 2.0
4hv H
where

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

f, = cut-off frequency of nth mode.
h = intermediate or sediment layer thickness.
H = depth of water.
p l = density of water.
pz = density of intermediate or sediment layer.
c = velocity of sound in water.
c1 = velocity of compressional waves in sediment layer.
cz = velocity of compressional waves in underlying layer.

The following experimental data will be used to derive the 1st mode cut-off
frequency from the above formulae :
H = 105 ft. cz = 14 250 ft/s.
h = 110 ft. p l = 1-03 g/cm3.
C = 4950 ft/S. p z = 1-80gIcm3.
C l = 54.45 ft/s.

Using these figures, the cut-off frequency of the 1st mode is:
fc = 8.6 c/s for h/H < 0.5
fc = 8.7 c/s for h/H > 2.0.

The experimental cut-off frequency is 10.85 CIS and therefore the three-
layered fluid model gives values which are too low. However, by reversing the
process and putting the experimental cut-off value of 10.85 CIS into the above
formulae, the depth of the sediment layer can be found and compared with the
result obtained from the refraction-shooting data. If this is done, the following
values are obtained:
h = 38.0 f t for h/H < 0.5
h = 86-5 f t for h/H > 2-0.
The use of the above formulae to determine the depth of the sediment layer is
not strictly valid owing to the fact that at Perranporth h/H is 1.0. However, the
average value of 57 ft for the sediment depth agrees reasonably well with the value
given by the refraction-shooting data.
In order to see whether better agreement with the experimental results can be
obtained by assuming a semi-infinite elastic sea bed, the 1st mode cut-off frequency
 Seismic refraction-shooting survey off coast of Cornwall 89
for this case will be calculated. The formula for the cut-off frequency for the nth
mode over an elastic sea bed is as follows :

fc =
(212- r)C ( I--
;;)-I

4H
where B is the velocity of shear waves in the sea bed.
For the purposes of this calculation it will be assumed that the velocity in the
sediment layer is so close to that in water that the depth of water H in the above
formula can be taken as the true depth plus the depth of the sediment layer. If
this assumption is made and the following experimental data used:
H = IO5+ I 1 0 = 215 ft,

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

C = 4950 ft/S,
c2 = 2//3Bs
the first mode cut-off frequency is given by:
fc = 7.2 c/s.
The ratio 2/3 of compressional to shear-wave velocities in the sea bed used in
the above formula is the value assuming a Poisson's ratio of 0.25. The cut-off
frequency of 7.2 c/s using this value is low compared with the experimental value
of 10.85c/s. However, by reversing the procedure and using the experimental
value of 10.85c/s in the formula it is possible to determine the ratio of compressional
to shear-wave velocities in the sea bed. If this is done a ratio of 2-4is obtained,
which is a reasonable value for the ratio of compressional to shear-wave velocities
in the hard rocks near the earth's surface.
As a result of the calculations carried out above, a ground-wave of 7-9 CIS
with second and third modes with frequencies of 21-27c/s and 35-45 c/s would be
expected to occur in the Perranporth area. These frequencies are in fair agreement
with the experimental values of 10 c/s, 30 c/s and 40-45 c/s, but the discrepancies
are certainly significant and are due to the shortcomings of the existing theories and
the lack of accurate knowledge of the in situ acoustic properties of the sediment
layer. Good agreement between theory and experiment will probably only be
possible when the theoretical case of a thin fluid layer or a thin layer with very low
rigidity (in which the velocity increases with depth) overlying a semi-infinite
elastic solid is available. Many interesting cases of propagation in layered media
are discussed in a recently published book by Ewing, Jardetzky & Press (1957).
With regard to the ground-wave variation of amplitude with range, it has been
reported that an Y-1 law or greater is to be expected theoretically for the fluid bottom
case. The results reported in this paper show that an 1-2 law applies in the Per-
ranporth area, and this ground-wave intensity fall off of 12dB per double-distance
is in good agreement with ground-wave intensity losses quoted in Worzel & Ewing
(1948).
From the foregoing qualitative analysis, it appears that neither the simple
fluid nor the elastic sea bed theories can fully explain the ground-wave propagation
at Perranporth. The sediment depth of IOO f t is sufficiently small to allow the first
rock layer greatly to affect the cut-off frequencies, and therefore the predominant
periods of the refracted waves. The high rate of attenuation of the ground-waves
is probably due to the fact that these very low frequencies penetrate deep into the
sea bed, and therefore suffer heavy shear-wave damping at the sediment-rock
90 A. S. Merriweather
interface. The higher frequencies would probably not suffer such heavy damping
owing to their smaller depths of penetration, but would be attenuated due to
scattering and absorption. If theoretical group-velocity curves could be computed
using the Perranporth parameters reported in this paper, a much better insight
into the mechanism of ground-wave propagation in coastal waters would probably
be possible.

8. Conclusions
The structure of the sea bed near Perranporth, Cornwall, has been determined
with reasonable accuracy as a result of the refraction-shooting experiments des-
cribed in this paper. The top layer consists of a low-velocity sand or sediment
layer approximately 80-100 ft thick, in which the velocity of compressional waves

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

is probably 4950-5500 ft/s. Beneath this top layer of sediment there are at least
two high-velocity layers of rock. The layer directly beneath the sediment is approx-
imately 1300f t thick and the velocity in this layer is 14250 ft/s. The third layer is
probably very thick and the velocity of compressional waves in it is 18500 ft/s.
Refraction-shooting experiments carried out in the English Channel and
described by Hill & King (1953)and Day & others (1956)indicated the existence
of strata with velocity lying between 12ooo and 1 5 900 ft/s. These strata were
found to be correlated in velocity and on stratigraphical grounds with the Palaeozoic
terrain of Devon and Cornwall. The velocity of the first rock layer found by re-
fraction-shooting at Perranporth also lies between the limits found for the Channel
layer, and is also correlated in velocity with the rock samples from the north coast
of Cornwall. This evidence would therefore suggest that the Palaeozoic rocks of
Devon and Cornwall also form a continuous layer beneath the sea, north of the
Cornish peninsula.
The predominant frequency of the ground-wave at Perranporth was found
to be approximately 10CIS and the intensity loss as a function of range was found
to follow an r-2 law.
The experiments described in this paper have also shown that although the
refraction-shooting method is very useful for determining the deep structure of
the sea bed it gives very little information on the sediment layer, unless this
happens to be very thick. The minimum range of 250 yd used in the experiments
was not sufficiently small to enable the wave which had travelled through the
sediment to be resolved, and therefore an accurate value for the velocity in this
layer is still not known.
Fair agreement between the experimental results and the existing theories of
the propagation of explosive sounds over fluid and elastic sea beds has been ob-
tained, but accurate agreement will probably only be possible when new techniques
have been developed, which will enable the vital parameters of the upper sediment
layers of the sea bed to be determined.

Acknowledgments
It is with pleasure that the writer acknowledges the assistance given by the
following :
The Captain and Ship’s Company of H.M.S. Tiree for their wholehearted
co-operation throughout the experiments; Mr P. Preddy, Mr W. W. Reay, Mr D.
Smith and Mr G. Weardon, who were responsible for the refraction-shooting
measurements; Mr F. Pierce of the National Institute of Oceanography for his
design work on the modified Ediswan Encephalograph recorder; and Dr M. N.
 Seismic refraction-shooting survey off coast of Cornwall 9'
Hill of the Department of Geodesy and Geophysics at Cambridge for his encourage-
ment and interest in the results of the experiments, helpful suggestions regarding
the intepretation of the results, and his kind offer to carry out the velocity measure-
ments on the rock samples.
The permission of the Admiralty to publish this account is also gratefully
acknowledged.

Admiralty Research Laboratory,

Teddington, Middlesex.
1957September.

Downloaded from https://academic.oup.com/gji/article/1/1/73/572110 by guest on 11 March 2025

References
Bullard, E. C., Gaskell, T. F., Harland, W. B. and Kerr-Grant, C., 1940. Phil.
Trans. Roy. SOC.A., 239, 29-94.
Bullard, E. C. & Gaskell, T. F., 1941.Proc. Roy. SOC. A., 177,476-499.
Day, A. A., Hill, M. N., Laughton, A. S. and Swallow, J. C., 1956. Quart. J.
Geol. SOC.Lond., 112,15-44.
Ewing, M., Jardetzky, W. S. & Press, F., 1957. Elastic Waves in Layered Media,
McGraw-Hill Book Company, Inc.
Heiland, C. A., 1946. Geophysical Exploration, New York.
Hill, M. N., 1952. Phil. Trans. Roy. SOC.A., 244, 561-596.
Hill, M. N. & King, W. B. R., 1953. Quart. J. Geol. SOC. Lond., 19, 1-20.
Leet, L. D.,1938.Practical Seismology and Seismic Prospecting,Appleton-Century-
Crofts, Inc. New York.
Lehmann, I., 1937. Mon. Not. R. Astr. SOC.Geophys. Suppl. 4, 250-271.
Muskat, M., 1933. Physics, 4, 14-28.
Pekeris, C.,1948. Geol. SOC. of Amer. Mem., 27.
Press, F. & Ewing, M., 1948. Trans. Amer. Geophys. Un., 29, 163-174.
Press, F. & Ewing, M., 1950. Geophysics, 15, 426-446.
Slichter, L. B., 1932. Physics, 3, 273-295.
Tolstoy, I., 1943. Bull. Seismol. SOC.Amer., 4, 493-512.
Willmore, P. L., 1949. Phil. Trans. Roy. SOC. A., 242, 123-151.
Worzel, J. L.& Ewing, M., 1948. Geol. SOC.Amer. Mem., 27, 30.