Content uploaded by Sergey A. Merkuryev
Author content
All content in this area was uploaded by Sergey A. Merkuryev on Dec 18, 2014
Content may be subject to copyright.
SPECIAL SECTION: MID-OCEANIC RIDGES
CURRENT SCIENCE, VOL. 85, NO. 3, 10 AUGUST 2003 334
Structure and evolution of the Carlsberg Ridge:
Evidence for non-stationary spreading on old
and modern spreading centres
S. A. Merkouriev* and N. A. Sotchevanova
SPbFIZMIRAN, Muchnov per, 2, Box 188, St. Petersburg, 191023, Russia
We present the results of magnetic and bathymetric
data collected on board Russian vessels during the last
decade. Our analysis shows that since late Cretaceous
the proto-Carlsberg Ridge (CR) was spreading at a faster
rate, prior to India’s collision with Eurasia. Since
Eocene, the present CR is characterized by slow spread-
ing. Our results depict two discordant systems of linear
magnetic anomalies. One corresponds with fast spread-
ing with respect to latitude axis of the proto-ridge and
the other with slow spreading with respect to modern
axis of the CR, suggesting that the two spreading sys-
tems are asymmetric both relative to each other and
relative to axial anomaly. We infer that during both
these periods the structure and spreading on the CR
was non-stationary.
THE Carlsberg Ridge (CR) extends from 2°S to 10°N,
forming the NW–SE trending slow accreting plate boun-
dary between the African and Indian Plates and continues
as highly segmented Sheba Ridge in the Gulf of Aden to
Red Sea
1,2
(Figure 1). In the region north of 10°N on the
CR, the Arabia–India–Somalia triple junction has been
evolving since last 16 m.y. as Ridge–Ridge–Ridge triple
junction whose one arm trending N80°E is the ultra slow
divergent boundary between Arabian and Indian Plate
3
.
The evolution of the northern-western Indian Ocean (NWIO)
is the last event of the break-up of Gondwana, when sea-
floor spreading progressively stopped in the Mascarene
Basin
4
. India and Seychelles started separating along the
Proto-CR
5–7
. The spreading rate was not constant and
the phase of ultra-slow spreading (< 8 mm/a) between
about A18 (40 Ma) and A7 (24 Ma) was detected
8
. CR
on the modern phase is characterized by slow spreading
and segmented by a few transforms and non-transform
discontinuities
9–11
. The recent results
12
discussed the
presence of an axial discontinuity at 3°32′N in terms
of a propagating ridge head. The objective of the pre-
sent study was to know the evolutionary history, the
nature of spreading and structure of CR since the late
Cretaceous.
Data acquisition and processing
Detailed bathymetric and magnetic survey of CR between
9°N–58°E and 2°S–69°E (Figure 1) was carried out mainly
in the 1980s during successive Russian expeditions. The
digitization and compilation of these datasets were com-
pleted during 1990s. All magnetic anomaly and bathy-
metric profile data were digitized, combined, totally
reprocessed and loaded into the coherent magnetic and
bathymetric database. The close spacing (5–6 km) between
profiles allowed building accurate bathymetric and mag-
netic grids at a 2 min spacing for mapping in colour
shaded relief and contours maps. The corresponding gravity
dataset is obtained from the global FAA deduced from
satellite altimetry map
13
. This geophysical dataset allows
*For correspondence. (e-mail: SAM@ns1480.spb.edu)
Figure 1. Gene
ralized map of the northwest Indian Ocean. The
Cen
tral Indian and Madagascar basins are conjugate basins as flanks of
CIR (Central Indian Ridge) and Arabia and Somalia basins as flanks of
Carlsberg Ridge. The full dashed line encloses the area where marine
magnetic anomaly profiles were obtained. Bathymet
ric contours at
3000 m and were taken from ETOPO5 map
23
.
SPECIAL SECTION: MID-OCEANIC RIDGES
CURRENT SCIENCE, VOL. 85, NO. 3, 10 AUGUST 2003 335
studying of almost 1300 km along ridge section and more
than 10 Ma across the ridge. In addition to our dataset,
we have extracted the data from NGDC
14
files to study
the early evolution of the proto-CR during the Palaeo-
cene–Eocene period.
Results and discussion
The magnetic anomaly map of the CR and Eastern
Somali and Arabian basins (Figure 2) depicts two discor-
dant systems of linear magnetic anomalies. One corres-
ponds with fast spreading with respect to latitudinal axis
of the proto-ridge, and the other corresponds to slow
spreading with respect to modern axis of the CR. The main
feature of the two spreading systems is the asymmetry
relative to each other and relative to axial anomaly.
The analysis of magnetic anomaly and bathymetric
data (Figure 3) shows the clear limit between the rough-
ness of the bathymetry and the regularity of the abyssal
hills in the Eastern Somali basin (0.5°N, 60°E) to the
NW. The position of the similar limit in the Arabian
basin was obtained by reconstruction. These linear struc-
tures observed in the magnetic field and bathymetry sug-
gest that the spreading geometry of the Proto-CR has
changed in the form of propagating rift from Southeast to
Northwest during anomaly time 24–20.
The studied part of CR has a complex structure and is
characterized by an en-echelon system of spreading cen-
tres. The comprehensive analysis of the magnetic and
bathymetric data reveals transform/non-transform discon-
tinuities (Figure 4). The observed variation in spreading
direction of the ridge subdivides the region in two distinct
parts on either side of 2.5°N–66°E. The northwestern part
is characterized by a general trend almost orthogonal to
the present spreading direction, while the southeastern
one has about 45° oblique trend compared to the spreading
direction. We have defined segments as well-individualized
bathymetric high and MBA (Mantle Bouger Anomaly)
low. The axis of CR is segmented into 26 accretion
segments (10–85 km length) separated by two major
100–150 km long transforms and several 2nd–4th-order
discontinuities with offsets of 30 km or no offsets. Along
the northwestern ridge section, the axis is made up of 20
segments (mean length of 50 km). Most of the disconti-
nuities have small or zero offsets, only two of them are
transform faults. Along the southeastern section, the ridge
consists of six segments separated by larger offset (up to
100 km) discontinuities, among which some are not trans-
form faults but oblique relays.
The bathymetric and MBA variation along axis (Figures 4
and 5) also show two different sections, the limit of which
is at (3.5°N–64°E) more than 200 km west of the ‘geo-
metric’ limit, discussed above. The southeastern part is
characterized by along axis MBA variations of smaller
Figure 2. Colour-scale image of magnetic anomalies in the NW In
dian
Ocean. Dashed lines correspond to pseudofaults of Carls
berg Proto
Ridge during westward propagating rift (see also Figure 3).
Figure 3. Shaded bathymetric map of the NW Indian Ocean with mag-
netic lineation. Russian dataset grided at 2
min spacing is merged with
predicted bathymetry
14
.
SPECIAL SECTION: MID-OCEANIC RIDGES
CURRENT SCIENCE, VOL. 85, NO. 3, 10 AUGUST 2003 336
amplitude with a mean axial MBA value being more
negative than along the northwestern part.
The ridge axis geometry, which is inherited from
the opening of the oceanic basins, appears to control the
morphology of the ridge, but do not control the deep
structure, as the MBA values suggest a more continuous
magmatic source beneath these chaotic morphology. This
difference in deep structure could be due to a combina-
tion of hot-spot influence and increasing spreading rate
southward. The analysis of axial magnetic anomalies,
spreading rate and direction confirm that the ridge seg-
mentation is not the adaptation of inherited plate bound-
ary geometry to changes in spreading conditions but the
effect of the accreting processes, focused magma upwell-
ing along discrete spots of the slow spreading axis. The
segment scale morphological variations and associated
axial MBA suggest along-axis variations in the magmatic
and tectonic processes.
Full sequences of magnetic anomalies up to A5 (10 Ma)
have been identified on the CR by generating synthetics
for each profile (Figures 6 and 7). The mean half-spreading
rate increases from 1.2 cm/yr to 1.5 cm/yr from NW to
SE. The spreading rate for the west and east flanks of CR
and the spreading rate distribution for anomaly A5 along
the CR (Figure 6) have been computed using A5-CR
finite rotation pole (24.2°N, 28.3°E). The synthetic model
has been computed by Tisseau and Patriat method
16
, using
magnetic reversal scale of Candy and Kent
16,17
taking
magnetized layer thickness of 1 km at depth increasing
Figure 4. Shaded bathymetric map over Carlsberg Ridge with mag-
netic lineation, transform faults (grey plain line) and axis ridge seg-
ments (thick black line). Magnetic picks are plot
ted as points and
magnetic lineation are plotted as colour lines and num
bered; dashed
light lines are traces of ridge propagation.
Figure 5. Mantle Bouguer anomaly over the Carlsberg Ridge. Grav
ity
dataset is obtained from the global FAA deduced from satellite alti
-
metry map
13
. Thick black line represents the axis of ridge segments.
Figure 6. Spreading rate distribution for anomaly A5 along Carls
berg
Ridge. 1, Spreading rate obtained for the west (a) and east (b
) flanks of
CR and half-full spreading rate (c
), 2, Running average; 3, Theoretical
half-full spreading rate calculated using A5-
CR pole finite rotation
(24.2 N; 28.3 E); 4, Carlsberg spreading axis; 5, Mag
netic pick of
anomaly 5.
a
b
c
d
SPECIAL SECTION: MID-OCEANIC RIDGES
CURRENT SCIENCE, VOL. 85, NO. 3, 10 AUGUST 2003 337
with age
18
and the magnetization intensity adjusted for
matching the synthetic and observed amplitudes (Figure 7).
The asymmetrical model (Figure 7 c) is the best fitting
model with ridge jump at a distance of 30.4 km on west
flank of the CR, 2.5 m.y. ago, which matches with the
observed profile VG860242 and gives high correlation
(R = 0.78).
The intra- and inter-segmental variation in spreading
rate has been studied. The distribution of average spread-
ing rate for the last 10 Ma is not always symmetric. The
observed asymmetry is not attached to one flank or one
period but seems rather randomly distributed resulting
into the continuously changing offsets. This asymmetry is
the result of either inter-segment axis jump or intra-segment
propagation
19,20
.
We have used the principle of ordering (ranking)
21
for
building of the lithological correlation for magnetic ano-
maly identification and jumping ridge-axis location. This
principle consists in affirmation that a curve joining the
correlation points is a monotone function. The correlation
points are those where the highest coefficient of correla-
tion between an interval on one curve and all intervals
of the same length on other curve is obtained. The basic
idea of the algorithm used is the combination of the
ordering principle and similarity measure. Using this
technique of maximum cross-correlation of similar aligned
features, we have located the time and distance of the
ridge jump. The symmetrical model (Figure 8, top) shows
the low correlation between anomalies 2A and 3A, whereas
the asymmetrical model with axis ridge jump provided
good correlation. The best correlation (Figure 8, bottom)
was obtained after comparison between observed profile
and more than 200 models calculated by varying various
arguments of a spreading ridge jump. Thus, we find that
the method of maximum cross-correlation of similar
aligned fragments and resulting correlation figures deter-
mine the best time and distance of the ridge jump. The
identification of magnetic anomalies allows determining
isochrones on both flanks. However, the superposition
of two conjugate isochrones is not accurate. A variable
slope of the block limits coming from the 3-D ridge
structure could explain this misfit.
Conclusion
We find that prior to the Himalayan collision, seafloor
spreading on the Carlsberg proto-ridge was extremely
active. A major reorganization of the spreading plate
Figure 7. Possible identifications of the magnetic anomalies o
b-
served on the CR (profile VG860242, see location on Figure 6) accor
d-
ing to either a symmetrical (a) or an asymmetrical (b and c
) spreading
model. d, Sketch showing jumping spreading ridge.
a
c
b
d
Figure 8.
Objective of the magnetic anomalies identification by the
correlation figures method applied to the observed and calculated pr
o-
files of Figure 5. Each solid circle with its associated number is related
to a short section of the observed profile and indica
tes the location
along the model where the best value of correlation is obtained.
SPECIAL SECTION: MID-OCEANIC RIDGES
CURRENT SCIENCE, VOL. 85, NO. 3, 10 AUGUST 2003 338
boundaries in the Indian Ocean during Eocene, as a con-
sequence of the hard collision of India with Eurasia, led
to the slow spreading on the present CR. Our results
show that during both these periods the structure and
spreading on the CR was non-stationary.
1. Wiseman, J. D. H., Scientific Reports of the John Murrar
Expedition, Britizh Museum, Natural History, 1937, vol. 3, pp.
1–28.
2. Drolia, R. K., Indian Ocean Ridge System (IORS): Results update,
challenges and opportunities in the new millennium, National
workshop on ‘Trends in ocean sciences – 21st century’ at Regio-
nal Center, National Institute of Oceanography, Visakhapatnam,
September 2000, pp. 28–30.
3. Fournier, M., Patriat Ph. and Leroy, S., Reappraisal of the Arabia–
India–Somalia triple junction kinematics. Earth Planet. Sci. Lett.,
2001, 189, 103–114.
4. Bernard, A., Le bassin des Mascareignes et l’arc des Amirants,
temoins de l’uverture au Cretace de l’ocean Indien occidental:
apport de la geophysique. Ph D thesis, Univ. de Strasbourg, 1998,
pp. 1–276.
5. Matthews, D. H., Vine, F. J. and Cann, J. R., Geology of an area
of the Carlsberg Ridge. Geol. Soc. Am. Bull., 1965, 76, 675–682.
6. McKenzie, D. P. and Sclater, J. G., The evolution of the Indian
Ocean since the late Cretaceous. Geophys. J., 1971, 24, 437–
528.
7. Todal, A. and Edholm, O., Continental margin off Western India
and Deccan large igneous province. Mar. Geophys. Res., 1998, 20,
273–291.
8. Mercuriev, S., Patriat, Ph. and Sochevanova, N., Evolution de la
dorsale de Carlsberg: Evidence pour une phase d’expansion tres
lent entre 40 et 25 Ma (A18 a A7). Oceanol. Acta, 1996, 19, 1–
13.
9. Ramana, M. V., Ramprasad, T., Kamesh Raju, K. A. and Desa,
M., Geophysical studies over a segment of the Carlsberg ridge,
Indian Ocean. Mar. Geol., 1993, 115, 21–28.
10. Kameh Raju, K. A., Kodagali, V. N. and Fujimoto, H., Three
dimensional gravity and magnetic studies over a segment of the
Carlsberg Ridge, Indian Ocean. 35th Annual Convention of Indian
Geophysical Union held at National Institute of Oceanography,
Goa, December 1998, pp. 29–30.
11. Mudholkar, A., Kamesh Raju, K. A., Kodagali, V. N., Afzul-
purkar, S. and Ambre, N. V., Exploration of the Carlsberg Ridge.
InterRidge News, 2000, 9, 32–33.
12. Mudholkar, A. V., Kodagali, V. N., Kamesh Raju, K. A.,
Valsangkar, A. B., Ranade, G. H. and Ambre, N. V., Geomor-
phological and petrological observations along a segment of slow-
spreading Carlsberg Ridge. Curr. Sci., 2002, 82, 982–989.
13. Sandwell, D. T. and Smith, W. H. F., Marine gravity anomaly
from Geosat and ERS 1 satellite altimetry. J. Geophys. Res., 1997,
102, No. B5, 10039–10054.
14. Tisseau, I. and Patriat, Ph., Identification des anomalies mag-
netiques sur les dorsales a faible taux d’expansion: methode des
taux fictifs. Earth Planet. Sci. Lett., 1981, 52, N2, 381–396.
15. Cande, S. C. and Kent, D. V., Revised calibration of the geo-
magnetic polarity timescale for the late Cretaceous and cenozoic.
J. Geophys. Res., 1995, 100, 6093–6095.
16. Cande, S. C. and Kent, D. V., A new geomagnetic polarity time
scale for the late Cretaceous and Cenozoic. J. Geophys. Res.,
1992, 97, N.B10, 13917–13951.
17. Parsons, B. and Sclater, J. G., An analysis of the variation of
ocean floor bathymetry and heat flow with age. J. Geophys. Res.,
1977, 82, 803–827.
18. Merkouriev, S., Rommevaux-Jestin, C., Patriat, Ph. and Sotcheva-
nova, N., Large scale segmentation of a 1300 km long section of
the Carlsberg Ridge from bathymetry, gravimetry and magnetism.
Eos Trans. AGU Fall Meeting November 16, 1999, 80, N 46,
F1049.
19. Merkouriev, S. and Sochevanova, N., Asymmetry of magnetic
anomalies and model of jumping spreading axis on an example
of Carlsberg Ridge. in Materials of the 27th session of an
International seminar by Uspensky, Moscow, 2000, pp. 122–123.
20. Jekhowsky, B., La methode des distances minimales, nouveau
procede quantitatif de correlation stratigraphique; example d’applica-
tion en paleontologie. Rev. Inst. Franc. Petrole, Paris, 1963, 18,
N5, 629–653.
21. Wessel, P. and Smith, W. H. F., Free software helps map and
display data, EOS. Trans. Am. Geophys. Union, 1991, 72, 445–
446.
22. ETOPO5 Relief map of the Earth’s surface. EOS Trans. AGU,
1986, 67, 121.
23. NGDC/MGG-National Geophysical Data Center for Marine Geo-
logy and Geophysics. Bolder, 2002.
ACKNOWLEDGEMENTS. We thank captain and crew of Russian
Expeditions for collection of magnetic and bathymetric data. The criti-
cal reading of the manuscript and suggestions of R. K. Drolia improved
the manuscript. The permission accorded by the Director, SPbFIZ-
MIRAN, St. Petersburg, Russia is gratefully acknowledged. The figures
were drafted using GMT software
22
.