Comparison of seismic reflection data to a synthetic seismogram in a volcanic apron at Site 953

Abstract

The volcanic apron of Gran Canaria at Site 953 is characterized by numerous, closely spaced reflectors, allowing a high-resolution stratigraphic correlation. The calibration of the presite survey seismic data (during the Meteor Cruise 24) with regard to the lithology and stratigraphy found at the drill site was achieved by computing a synthetic seismogram serving as the link between seismic and borehole data. Because logging data were available for only 53% of the hole, velocity and density mea-surements taken from the recovered cores were used in the missing intervals to obtain a complete synthetic seismogram. Most reflectors in the upper ~900 m of the sequence (lithologic Units I–V) turned out to be thin volcaniclastic layers intercalated to the nonvolcanic background sediments. Their thicknesses are generally <2 m, and the reflections from their tops and bases overlap, forming a single reflection. The limit of the seismic detection of such interbeds is on the order of several decimeters and thus requires special care for the processing of the velocity and density data to avoid destruction of the signal from these thin layers.

Full text

Weaver, P.P.E., Schmincke, H.-U., Firth, J.V., and Duffield, W. (Eds.), 1998
Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 157
3
1. COMPARISON OF SEISMIC REFLECTION DATA TO A SYNTHETIC SEISMOGRAM
IN A VOLCANIC APRON AT SITE 9531
Thomas Funck2 and Holger Lykke-Andersen3
ABSTRACT
The volcanic apron of Gran Canaria at Site 953 is characterized by numerous, closely spaced reflectors, allowing a high-
resolution stratigraphic correlation. The calibration of the presite survey seismic data (during the Meteor Cruise 24) with regard
to the lithology and stratigraphy found at the drill site was achieved by computing a synthetic seismogram serving as the link
between seismic and borehole data. Because logging data were available for only 53% of the hole, velocity and density mea-
surements taken from the recovered cores were used in the missing intervals to obtain a complete synthetic seismogram. Most
reflectors in the upper ~900 m of the sequence (lithologic Units I–V) turned out to be thin volcaniclastic layers intercalated to
the nonvolcanic background sediments. Their thicknesses are generally <2 m, and the reflections from their tops and bases
overlap, forming a single reflection. The limit of the seismic detection of such interbeds is on the order of several decimeters
and thus requires special care for the processing of the velocity and density data to avoid destruction of the signal from these
thin layers.
INTRODUCTION
The volcanic apron of Gran Canaria consists of volcaniclastic de-
posits with generally high P-wave velocities and densities intercalat-
ed to the hemipelagic background sediments with lower velocities
and densities (Schmincke, Weaver, Firth, et al., 1995). Although the
volcaniclastic layers are generally <1 m thick, they are conspicuous
in the seismic record by their continuity outside the chaotic or discon-
tinuous slope facies proximal to the island. Because prominent re-
flectors can be used as marker horizons for the mapping of the apron,
it is worthwhile to calculate synthetic seismograms for correlating the
seismic data with the lithology and stratigraphy at the drill sites.
Site 953 (Fig. 1) is the deepest site around Gran Canaria (maxi-
mum penetration 1159 meters below seafloor [mbsf]), reflecting the
entire volcanic and erosional evolution of the island. It thus repre-
sents a key site for the calibration of seismic data. The computed syn-
thetic seismogram is compared with the high-resolution seismic re-
flection Profile 134 (Fig. 1) where Site 953 is located. This profile
was collected during the presite survey of the vessel Meteor Cruise
24 (Schmincke and Rihm, 1994).
The objective of this study was to calculate a complete synthetic
seismogram to benefit as much as possible from the drilling results
for the seismic interpretation. Special care had to be taken in process-
ing the density and velocity data to preserve the signal of the thin vol-
caniclastic interbeds.
DATA PREPARATION
The standard method of obtaining synthetic seismograms is to use
impedance logs from downhole measurements. Impedance is the
product of P-wave velocity and bulk density. However, because Site
953 was not logged intensely (only 53% of the hole was logged), the
logging data set alone was not sufficient to compute a synthetic seis-
mogram for the entire hole. Gaps in the logging data therefore were
filled with laboratory measurements carried out on the recovered sed-
iments aboard ship.
At Site 953, a huge amount of overlapping velocity and density
measurements was available. Velocity data were provided by the P-
wave logger of the multisensor track (MST; depth interval 0−192
mbsf), digital sound velocimeter (DSV; 0−76 mbsf), Hamilton frame
(187−1159 mbsf), and sonic log (372−963 mbsf). Density data were
available from the gamma-ray attenuation porosity evaluator
(GRAPE) sensor of the MST (0−1159 mbsf), the index properties (0−
1159 mbsf), and the density log (372−987 mbsf). These data had to
be processed and merged in a trial and error process to find the best
correlation between seismic Line 134 and the computed synthetic
seismogram.
1Weaver, P.P.E., Schmincke, H.-U., Firth, J.V., and Duffield, W. (Eds.), 1998. Proc.
ODP, Sci. Results, 157: College Station, TX (Ocean Drilling Program).
2Graduiertenkolleg, GEOMAR, Wischhofstr. 1-3, 24148 Kiel, Federal Republic of
Germany. (Present address: Department of Oceanography, Dalhousie University, Hali-
fax, Nova Scotia, B3H 4J1, Canada. tfunck@is.dal.ca).
3Department of Earth Sciences, University of Århus, Finlandsgade 8, 8200 Århus
N, Denmark.
15˚30
´
W 15˚00
´
W 14˚30
´
W
28˚00
´
N
28˚30
´
N
29˚00
´
N
P134
SITE 953
Gran Canaria Fuerteventura
50 km
1000
2000
3000
Figure 1. Generalized bathymetry chart of the volcanic apron north of Gran
Canaria (contours at 500-m intervals), showing the location of Site 953 and
the crossing reflection seismic Line P134 of the Meteor Cruise 24.

T. FUNCK, H. LYKKE-ANDERSEN
4
In the end, the MST data (0−192 mbsf) were used for the first part
of the final velocity function (Fig. 2), because the data density of the
overlapping DSV data was only ~1 per 5 m. The erroneous values at
the ends of each measured core section (1.5 m length) were deleted.
Furthermore, a median filter (filter length 0.25 m) had to be applied
because the broad scatter in the data introduced artificial reflections.
The second part of the velocity function consists of Hamilton frame
data (192−372 mbsf) as no other data were available. The third part
was provided by the sonic log (372−963 mbsf). Because the main part
of reflections is produced by only thin (meter-sized or even less) in-
terbeds of volcaniclastic material, the best results were achieved us-
ing raw data, although some spikes were probably erroneous; in fact,
filtering the logging data seemed to overly flatten the characteristics
(amplitudes became too low). The last part of the velocity function
was made again using the Hamilton frame (963−1157 mbsf); howev-
er, a velocity correction had to be applied because the synthetic seis-
mogram did not fit well with the measured data in the lowermost part.
A systematic offset was detected below ~720 mbsf, where the Hamil-
ton frame velocities were systematically higher than the logging ve-
locities. The reason for this deviation is unclear; possible explana-
tions include
1. Systematic error in the Hamilton frame measuring instrument
(improbable);
2. Selective coring, (i.e., softer sediments with a lower velocity
tend to be washed out of the core);
3. Selective measuring, that is, the measured points are not repre-
sentative for the cored sediments; and
4. Anisotropy, because the velocity measured using the Hamilton
frame was made perpendicular to the core axis (i.e., in the bed-
ding plane and not along the core axis).
Nevertheless, the average velocity for the interval 720−963 mbsf
was 306 m/s higher for the Hamilton frame than for the sonic log so
that 306 m/s was applied as the velocity correction below 963 mbsf.
The final density function contains only index properties mea-
surements. The GRAPE density was eliminated because the neces-
sary filtering of the values for the upper interval (0−192 mbsf), in
combination with the smooth-filtered MST velocities, did not pro-
duce the necessary impedance variations to fit the observed reflec-
tions. In the lower interval down to 1159 mbsf, the GRAPE densities
are erroneous because the applied rotary drilling technique resulted
in core disturbances and variable core diameters. The decision
whether to use index properties or the density log was made on the
basis of a comparison of the correlation with the seismic data. The in-
dex properties proved to be slightly better.
The average density and velocity increase with depth because of
the increasing consolidation and compaction of the hemipelagic
background sediments. Volcaniclastic interbeds show generally pro-
nounced deviations from this average, with a maximum of 3.1 g/cm3
in density and 5.3 km/s in velocity. Figure 2 shows that peaks in ve-
locity do not always correlate with peaks in density, because such in-
tervals are not sampled by both data sets, especially where core re-
covery is low.
SYNTHETIC SEISMOGRAM
The depths of velocity and density function derived from merging
the different data sources were converted to time using the time-depth
relation calculated from the velocity. The data were then sampled at
1 m/s intervals, and the reflectivity was derived from two successive
samples (Fig. 2). The reflectivity was then convolved with the source
0
100
200
300
400
500
600
700
800
900
1000
11001.0 1.5 2.0 2.5 3.0
Density (g/cm
3
)
TWT (ms)
1 2 3 4 5
Hamilton frame
Hamilton
frame
MST
Sonic log
Velocity (km/s)
-0.5 0.0 0.5
Reflectivity Synthetic
Figure 2. Computation of the synthetic seismogram
at Site 953. The bulk density is derived from the
index properties. The P-wave velocity is composed
of multisensor track, Hamilton frame, and sonic log
data. The vertical axis shows the TWT below
seafloor. The synthetic seismogram results from the
reflectivity convolved with the source wavelet used
during the Meteor Cruise 24 presite survey. The
AGC is applied to the synthetic traces.

COMPARISON OF SEISMIC REFLECTION DATA
5
wavelet used during the reflection seismic survey. The wavelet was
computed from stacking 100 direct arrivals using the same sleevegun
cluster (4 × 0.65l) as during the Meteor Cruise 24 (Larsen, Saunders,
Clift, et al., 1994). The computed synthetic seismogram does not
comprise internal multiples, even though some peg-leg multiples be-
tween the volcaniclastic horizons are likely. However, because gaps
in the velocity and density data sets occurred preferentially in the
coarse-grained volcaniclastic intervals, a number of erroneous multi-
ple reflections would be introduced to the synthetic seismogram.
To match the real and synthetic data as close as possible, the same
time-variant band-pass-filter and automatic gain control (AGC), as
used in the processing of the real data, was applied to the synthetic
data (Fig. 3). Furthermore, random noise was added to the synthetic
data. Differences between the synthetic seismogram and the seismic
profile occur where the data density was not adequate to detect thin
interbeds. Discrepancies shallower than 5.0 s two-way traveltime
(TWT) are, for example, caused by an insufficient sampling rate for
the index properties measurements, caused by the rapid flow of cores
in the laboratory. Further downcore (down to ~5.15 s and below 5.65
s TWT), many intervals had low core recovery, probably masking
some reflectors. Nevertheless, the synthetic seismogram fits the seis-
mic data surprisingly well considering that only half of the site was
logged.
CORRELATION OF SEISMIC DATA
WITH LITHOLOGY
By means of the synthetic seismogram computed for Site 953, a
number of reflectors could be assigned to specific lithologies, mainly
volcaniclastic interbeds in the hemipelagic background sediments.
Reflectors identified in the seismic data and in the synthetic seismo-
gram are listed in Table 1, and, when possible, the reflecting lithology
is given. Several of these lithologies are labeled with question marks,
indicating that the reflection could not be determined unequivocally.
For example, this might be the case where amplitudes did not fit very
well or where lithologies in the vicinity were not sampled but could
have caused the reflection. Where only one lithology is given, the re-
flection is caused by the interference of the top and bottom reflection
of a thin interbed surrounded by sediments with a different imped-
ance. These layers are generally thinner than 2 m, often <1 m.
The reflectors are numbered from the seafloor (No. 0) downcore
(Fig. 4). The age of the reflectors is taken from biostratigraphic and
paleomagnetic data (Brunner et al., Chap. 9, this volume). Some re-
flectors, which are suitable for regional mapping because of their
continuity and amplitude, are labeled with letters, corresponding to
volcanic phases on Gran Canaria (Schmincke, 1976, 1982, 1994; Ho-
ernle and Schmincke, 1993a, 1993b) and the neighboring island Ten-
erife in the west (Ancochea et al., 1990). The sediments causing the
reflectors M and F were deposited during the Mogan and Fataga
phase of the Miocene volcanism on Gran Canaria. Reflector H repre-
sents the transition to the volcanic hiatus on Gran Canaria, reflector
T is generated by sediments deposited during the shield stage of Ten-
erife, reflector RN corresponds to the Pliocene Roque Nublo volcan-
ism on Gran Canaria, and the sediments causing reflector Q were de-
posited in the Quaternary with volcanic activity both on Gran Canaria
and Tenerife. This reflector nomenclature was introduced by Funck
(1996). Comparison between reflectors found at the Deep Sea Drill-
ing Project (DSDP) Site 397 (Wissmann, 1979; von Rad, Ryan, et al.,
1979) or Ocean Drilling Program Sites 955 and 956 (Schmincke,
Weaver, Firth, et al., 1995) south of Gran Canaria is hampered be-
cause of the influence of the continental slope, whereas the northern
basin is more or less shielded from continental input by the East Ca-
nary Ridge. Furthermore, no direct seismic correlation from south to
north is possible across the channel between Gran Canaria and Fu-
erteventura/Tenerife, because the reflectors onlap the volcanic base-
ment and/or reflection patterns become chaotic in the proximity of
the islands. This explains why reflector names introduced for the
southern basin at DSDP Site 397 are not suitable in the north. How-
ever, the prominent reflector bands R7 and R3 at DSDP Site 397 (von
Rad, Ryan, et al., 1979) correspond to the reflector bands around M
and RN, respectively, based on their depositional age and seismic fa-
cies (Funck et al., 1996).
DISCUSSION AND CONCLUSIONS
The high-resolution seismic data of the Meteor Cruise 24 revealed
a large number of reflectors from which 55 (Table 1) could be iden-
tified by means of the synthetic seismogram at Site 953. The deposi-
tional setting of the volcanic apron of Gran Canaria with its thin
volcaniclastic layers intercalated to the nonvolcanic background sed-
iments put great demands on the preparation of the velocity and den-
sity data. For example, filtering can remove unrealistic spikes but
also can eliminate the signal of thin layers. Another difficulty that oc-
curred at Site 953 was the incomplete logging data set resulting in the
use of discrete measurements carried out on the cores. Apart from the
fact that these data were only available for the cored intervals, two
problems occurred in the case of the investigated volcanic apron:
1. The use of the closely sampled (0.5−3.0 cm) MST data with its
broad scatter resulted in artificial reflections, whereas filtering
destroyed almost all real signals from thin layers.
2. The typical sampling rate of one per section (1.5 m) for the in-
dex properties, the Hamilton frame, and DSV measurements
often proved to be insufficient to record all reflectors.
Nevertheless, the careful composition of the velocity and density
function allowed identification of most of the reflectors. The reflec-
tors in lithologic Units I–V (Fig. 4) represent thin layers, generally <2
m thick. This raises the question of the vertical resolution of the seis-
mic data. Badley (1985) states that an interbedded layer has to be
thicker than half the seismic wavelength to enable resolution of its
top and base. Thinner layers appear as one single reflector with a
maximum amplitude at one-quarter wavelength—the tuning thick-
ness. For thicknesses below one-quarter wavelength, the reflection
remains the same shape, but decreases in amplitude. Once the thick-
ness is about one-thirtieth wavelength or less, there is no detectable
response. This explains why thin volcaniclastic beds can be detected
as single reflectors. Applying this knowledge to the data reported
here, the signal contains frequencies of up to 230 Hz, and the velocity
in the volcaniclastic layers is typically between 1700 and 2000 m/s in
the upper 100 m. One-thirtieth of a wavelength thus corresponds to
only 25−30 cm, which is the minimum thickness for seismic detec-
tion of such layers. Frequencies around the maximum energy of the
source wavelet (~80−100 Hz) correspond to a detection limit in the
order of 50−80 cm. Weak and discontinuous reflectors seem to rep-
resent thin layers with a thickness on the order of one-thirtieth of the
wavelength. This is the case for the thin sand layers that are recogniz-
able in the upper 50 m of the seismic record in Fig. 3. Their thickness-
es at Site 953 are ~80 cm (Schmincke, Weaver, Firth, et al., 1995).
Another point to discuss here briefly is the boundaries of the litho-
logic units and their seismic expression. Most of the lithologic
boundaries are not characterized by reflections, at least not by reflec-
tions with a high continuity and amplitude (Fig. 4). The reason for
this is simply the fact that the main criterion to subdivide the se-
quence into lithologic units was the variation of the flux of coarse ma-
terial to Site 953 (Schmincke, Weaver, Firth, et al., 1995). This flux
was averaged for larger intervals (core length), and therefore the
lithologic boundaries typically do not correspond to a volcaniclastic
interbed and its associated reflection. The boundary between litho-
logic Units VI and V, separating the massive basaltic pedestal of
Gran Canaria from the overlying sediments, is characterized by a
high amplitude reflection and represents an exception to the other-

T. FUNCK, H. LYKKE-ANDERSEN
6
Figure 3. Part of seismic Line P134 is plotted together with the synthetic seismogram at Site 953. The seismic data are stacked, scaled with an AGC (50−200
ms), and band-pass filtered (30−230 Hz at seafloor, 30−160 Hz at 6 s TWT). Random noise was added to the 10 synthetic seismic traces. One common depth
point (CDP) corresponds to a horizontal distance of 3.125 m, and the TWT is given in seconds below sea level.

COMPARISON OF SEISMIC REFLECTION DATA
7
wise low correlation between lithologic boundaries and their seismic
expression.
The narrow spacing of the volcaniclastic deposits in conjunction
with the drilling allows a high temporal resolution of the apron. The
nonvolcanic background sediments represent an ideal contrast medi-
um for seismic detection of thin volcaniclastic layers. Hence, a seis-
mic investigation in a volcanic apron benefits from high background
sedimentation rates like around the Canary Islands.
ACKNOWLEDGMENTS
We are grateful to the crew and scientific staff of the research ves-
sels Meteor and JOIDES Resolution for collecting the data in a pro-
fessional manner. Some figures were generated with software provid-
ed by P. Wessel and W.F.H. Smith (1991). The work on this paper
was supported by the Deutsche Forschungsgemeinschaft (DFG-
Schm250/49 III GK, DFG-Schm250/54), the Bundesministerium für
Forschung und Technologie, and the European Union (EPOCH,
EVSV-CT93-0283, MAS2-CT94-0083).
REFERENCES
Ancochea, E., Fúster, J.M., Ibarrola, E., Cendrero, A., Coello, J., Hernán, F.,
Cantagrel, J.M., and Jamond, C., 1990. Volcanic evolution of the island
of Tenerife (Canary Islands) in the light of new K-Ar data. J. Volcanol.
Geotherm. Res., 44:231−249.
Badley, M.E., 1985. Practical Seismic Interpretation: Englewood Cliffs, NJ
(Prentice Hall).
Funck, T., 1996. Structure of the volcanic apron north of Gran Canaria
deduced from reflection seismic, bathymetric and borehole data [Ph.D.
dissert.]. Univ. Kiel.
Funck, T., Dickmann, T., Rihm, R., Krastel, S., Lykke-Andersen, H., and
Schmincke, H.-U., 1996. Reflection seismic investigations in the volca-
niclastic apron of Gran Canaria and implications for its volcanic evolu-
tion. Geophys. J. Int., 125: 519−536.
Hoernle, K., and Schmincke, H.-U., 1993a. The role of partial melting in the
15-Ma geochemical evolution of Gran Canaria: a blob model for the
Canary Hotspot. J. Petrol., 34:599−627.
————, 1993b. The petrology of the tholeiites through melilite nephelin-
ites on Gran Canaria, Canary Islands: crystal fractionation, accumulation,
and depths of melting. J. Petrol., 34:573−597.
Larsen, H.C., Saunders, A.D., Clift, P.D., et al., 1994. Proc. ODP, Init.
Repts., 152: College Station, TX (Ocean Drilling Program).
Table 1. Correlated reflectors at Site 953.
Notes: ? = reflection not fully determined. Reflector numbers are shown in Figure 4. Unless specified, the reflections are caused by thin layers of the given lithology intercalated to
background sediments with a different impedance.
Reflector Depth (mbsf) TWT (ms) Age (Ma) Reflection causing lithologies (and comments)
0 0.00 4745 0.00 Seafloor
1 11.45 4760 0.36 Clayey medium to coarse-grained silty pumice sand
2 20.60 4772 0.48 Silty fine medium-grained foraminifer pumice sand
3 23.90 4777 0.53 Calcareous sand
4 30.17 4785 0.61 Basaltic sand
5 45.85 4805 0.82 Silty crystal lithic sand (amplitude difference due to low recovery)
6 55.49 4818 0.95 Clayey nannofossil ooze (weak reflector)
7 60.77 4826 1.03 Crystal lithic sand
8 72.61 4840 1.19 Calcareous sand with volcanic lithics
9 96.40 4871 1.51 Foraminifer lithic sand
10 130.73 4916 1.97 Massive lithic to calcareous sand
11 149.74 4940 2.23 Massive foraminifer sand with lithics (weak reflector, low continuity)
12 170.53 4967 2.62 Foraminifer silt with lithics and crystals
13 176.00 4974 2.70 Clayey nannofossil ooze?
14 206.23 5010 3.12 Foraminifer sandstone
15 225.70 5030 3.39 Foraminifer lithic sandstone
16 237.59 5041 3.56 Lapillistone
17 257.54 5058 3.83 Lapillistone (weak reflector, low continuity)
18 267.67 5070 4.24 Silty, foraminifer nannofossil chalk (weak reflector)
19 278.07 5081 4.48 Nannofossil chalk (weak, discontinuous reflector)
20 296.19 5099 4.91 Density/velocity increase in a nannofossil chalk sequence
21 313.12 5118 5.32 Lithic crystal sand
22 325.53 5130 5.61 Indurated nannofossil ooze (weak reflector)
23 337.50 5144 5.90 Clay with nannofossils (grading to lithic crystal sand), weak reflector
24 350.81 5157 6.21 Slump unit consisting of ooze, clay, silt, mixed rock
25 361.35 5167 6.21 ?
26 385.85 5185 7.60 Foraminifer lithic crystal sand (weak, discontinuous reflector)
27 399.73 5200 8.28 Indurated clay with nannofossil (weak and discontinuous reflector)
28 409.80 5210 8.46 Calcareous sandstone?
29 417.79 5219 8.60 Indurated clay?
30 442.71 5242 9.03 Silty claystone
31 455.00 5254 9.25 Claystone?
32 470.39 5271 9.52 Nannofossil chalk?
33 495.30 5293 9.96 Lithic crystal sandstone?
34 516.44 5312 10.33 Nannofossil claystone
35 556.83 5350 11.04 Lithic crystal sandstone
36 594.71 5385 11.71 Lithic crystal sandstone
37 601.86 5390 11.82 Lapillistone
38 621.49 5408 11.82 Lithic crystal sandstone?
39 633.90 5419 11.98 ? (logged but not recovered)
40 689.14 5467 12.71 Siltstone
41 710.40 5485 12.99 Siltstone? (weak and discontinuous reflector)
42 749.72 5519 13.37 Nannofossil claystone, graded to crystal siltstone
43 755.34 5524 13.41 Claystone? (close to not sampled sandstone)
44 768.32 5535 13.50 Vitric tuff
45 787.20 5552 13.63 Calcareous vitric siltstone? (weak reflector)
46 794.45 5559 13.68 Lithic crystal vitric tuff
47 822.44 5584 13.90 Vitric rich claystone with sandy vitric tuff
48 842.17 5602 14.25 Nannofossil claystone grading to lithic crystal sandstone with pumice
49 863.22 5618 14.61 Claystone grading down to lithic crystal sandstone (weak reflector)
50 920.42 5660 14.80-15.80 Lapillistone with basaltic breccia
51 988.44 5709 14.80-15.80 Pebble- and granule-sized, fine-grained hyaloclastite tuff
52 1029.97 5737 >15.80 High velocity unit in hyaloclastite lapillistone
53 1066.83 5756 >15.80 Transition from hyaloclastite lapillistone breccia to hyaloclastite tuff
54 1110.61 5781 >15.80 Transition from basaltic hyaloclastite breccia to hyaloclastite tuff
55 1139.86 5797 >15.80 Transition from hyaloclastite breccia to hyaloclastite lapillistone

T. FUNCK, H. LYKKE-ANDERSEN
8
Schmincke, H.-U., 1976. The geology of the Canary Islands. In Kunkel, G.
(Ed.), Biogeography and Ecology in the Canary Islands: The Hague (W.
Junk), 67−184.
————, 1982. Volcanic and chemical evolution of the Canary Islands. In
von Rad, U., Hinz, K., Sarnthein, M., and Seibold, E. (Eds.), Geology of
the Northwest African Continental Margin: Berlin (Springer), 273−306.
————, 1994. Geological Field Guide: Gran Canaria (7th ed.): Kiel, Ger-
many (Pluto Press).
Schmincke, H.-U., and Rihm, R., 1994. Ozeanvulkan 1993, Cruise No. 24,
15 April−9 May 1993. METEOR-Berichte, Univ. Hamburg, 94-2.
Schmincke, H.-U., Weaver, P.P.E., Firth, J.V., et al., 1995. Proc. ODP, Init.
Repts., 157: College Station, TX (Ocean Drilling Program).
von Rad, U., Ryan, W.B.F., et al., 1979. Init. Repts. DSDP, 47 (Pt. 1): Wash-
ington (U.S. Govt. Printing Office).
Wessel, P., and Smith, W.H.F., 1991. Free software helps map and display
data. Eos, 72:441, 445−446.
Wissmann, G., 1979. Cape Bojador slope, an example for potential pitfalls in
seismic interpretation without the information of outer margin drilling. In
von Rad, U., Ryan, W.B.F., et al., Init. Repts. DSDP, 47 (Pt. 1): Washing-
ton (U.S. Govt. Printing Office), 491−499.
Date of acceptance: 6 January 1997
Date of initial receipt: 24 June 1996
Ms 157SR-100

COMPARISON OF SEISMIC REFLECTION DATA
9
4.8
5.0
5.2
5.4
5.6
5.8
I
II
III
IV
V
VI
VII
A
B
C
Lithologic UnitM24-LINE 134 M24-LINE 134 Reflector
Synthetic
seismogr.
TWT (s)
Water
S e d i m e n t s
(volcanic / non-volcanic)
Basaltic
pedestal
Gran Canaria
Hyaloclastite
debris flows
1159 mbsf
maximum
penetration
0
1
2
4
5 (Q)
6
8
9
10
11
12
14
15
16 (RN)
17
18
19
20
21 (T)
22
23
24
25
26
27
28
29
30
31
32 (H)
33
34
35
36
38 (F)
39
40
41
42
44
45
47 (M)
48
49
50
51
52
53
54
55
Age (Ma)
01020
age
Figure 4. Part of seismic Line P134 with synthetic seismogram at Site 953 and the lithologic units converted to TWT (in seconds below sea level). The two age
lines in lithologic Unit VII give minimum and maximum ages. The small numbers to the right of the seismic data refer to individual reflectors as listed in Table
1.