RESEARCH COMMUNICATIONS
33. Kumar, K. and Loyal, R. S., J. Palaeontol. Soc. India, 1987, 32,
60–84.
34. Rana, R. S. and Kumar, K., in Cretaceous Event Stratigraphy and
the Correlation of Indian Non-marine Strata (eds Sahni, A. and
Jolly, A.), Contributions from Seminar cum Workshop IGCP 216
and 245, Chandigarh, 1990, pp. 55–57.
35. Sahni, A., Bhatia, S. B. and Kumar, K., Boll. Soc. Paleontol. Ital.,
1983, 22, 77–86.
36. Nanda, A. C. and Kumar, K., Wadia Inst. Him. Geol. Spl. Publ. 2,
1999, 1–85.
37. Srivastava, R. and Kumar, K., Palaeogeogr. Palaeoclimatol.
Palaeoecol., 1996, 122, 185–211.
38. Tong, Y., Vertebr. Palasiat., 1979, 17, 65–70.
39. Gingerich, P. D., Univ. Mich. Pap. Paleontol., 1976, 15, 1–140.
40. Kumar, K., International Conference on Distribution and Migration of Tertiary Mammals in Eurasia, Univ. of Utrecht, 2001, pp.
29–31.
41. Gingerich, P. D., Smith, B. H. and Rosenberg, K., Am. J. Phys.
Anthropol., 1982, 58, 81–100.
ACKNOWLEDGEMENTS. We are grateful to Prof. Ashok Sahni,
Panjab University, Chandigarh and Dr A. C. Nanda, Wadia Institute of
Himalayan Geology (WIHG), Dehradun for reading an earlier draft of
this paper and suggesting improvements. Thanks are also due to Dr
Asit Jolly for help and cooperation. Facilities for this research were
kindly provided by the Director, WIHG, Dehradun and are thankfully
acknowledged.
Received 8 August 2002; revised accepted 5 October 2002
Implications of novel results about
Moho from magnetotelluric studies
Shalivahan and Bimalendu B. Bhattacharya*
Department of Applied Geophysics, Indian School of Mines,
Dhanbad 826 004, India
Moho boundary is well established on the basis of
seismological studies. Magnetotelluric (MT) studies
have not been able to establish this boundary due to
the presence of highly conducting continental lower
crust (CLC). MT studies over the Eastern Indian
(3.3 Gyr) and Slave (4.03 Gyr) cratons resolve the
crust–mantle boundary due to the absence of highly
conducting CLC. The upper mantle beneath these two
cratons cannot be of pure olivine as resistivity of the
uppermost mantle is much less than the laboratory
studies on pure olivine. From these two studies it can
be speculated that a dynamics other than the plate
tectonics possibly existed up to the mid-Archaean.
THE electrical conductivity of rocks is more closely related to geochemical changes than other bulk physical
properties such as acoustics impedance, density and
*For correspondence. (e-mail: [email protected])
CURRENT SCIENCE, VOL. 83, NO. 10, 25 NOVEMBER 2002
seismic velocity. The magnetotelluric (MT) method used
for the investigations of deeper structures, therefore, provides complementary, sometimes supportive but other
times alternative interpretations of geochemical characters in geological units. It has been suggested on the basis
of laboratory studies that for dry rock assemblages there
may be an observable difference in conductivity between
deep crustal mafic and upper mantle ultramafic rocks1 .
Ionic conduction is a dominant type of conduction in
the crust and upper mantle. It is responsible for conduction in fluids and also for olivines at high temperatures.
Dry silicate rocks are highly resistive. The electronic
conduction in most solids is a thermally activated process
governed by the appropriate activation energy for the
material, Boltzmann constant and the absolute temperature. A region of enhanced electrical conductivity therefore represents interconnected network of fluid and/or
mineral conducting phase. The enhanced conductivity of
the continental lower crust (CLC) remains one of the
puzzles regarding the earth about which comparatively
little is known. This characteristic of the deep crust has
been observed globally, using MT method, but explanations for its existence remain controversial. The enhanced
conductivity of CLC can be explained either by interconnected brine below the brittle–ductile transition2–6 or by
interconnected, thin, grain-boundary carbon film7,8. The
MT method becomes unsuitable in determining the depth
of Moho due to the presence of highly conducting CLC.
The presence of enhanced conductivity in CLC limits the
ability to resolve the uppermost mantle conductivity
structure and only a maximum limit can be placed on its
value9 . Therefore, in such cases, upper mantle resistivity
and the nature of olivine as upper mantle constituent cannot be determined. Thus, when the enhanced conductivity
of CLC is absent, only then can one resolve the conductivity structure of the uppermost mantle. Here we report
the definite identification of crust–mantle boundary due
to the lack of a conducting lower crust over Eastern Indian Craton (EIC)10–12 and Slave province, Canada13 using MT data and discuss its implication in delineating the
crust–mantle boundary.
The eastern part of the Indian Precambrian shield is
characterized by Archaean nucleus of Singhbhum Granite
(SG) batholithic complex and ancient supracrustals surrounded by several elongate and arcuate Proterozoic
belts. The major units of the EIC are shown in Figure 114 .
This Archaean nucleus is bounded by the arcuate Copper
Belt thrust zone (or Singhbhum shear zone) in the east,
north and northwest and Sukhinda thrust in the south (not
shown in Figure 1). Geochronologically, the Archaean
nucleus is at least ~ 3.3 Gyr old15 , represented by
tonalite–trondhjemite–gneiss (TTG) and amphibolites of
older metamorphic group (OMG). The OMG rocks are
intruded by SG of ~ 3 Gyr age which occupies most of
the craton. Both OMG and SG are unconformably overlain by supracrustals of Iron Ore Group (IOG) rocks con1259
RESEARCH COMMUNICATIONS
sisting of shales, schists, phyllite, banded haematite jasper and banded haematite quartzite. Along the southern
fringes of the Singhbhum craton, outcrops of charnockites and khondalites (graphite-bearing) occur to the
north and south of the Sukhinda thrust respectively.
Remote reference (RR) MT measurements were carried during 1996–97 field season over a part of the southern Archaean nucleus of EIC to map the electrical
conductivity of the crust and upper mantle with reference
to a fixed remote site16,17 (inset Figure 1). The measurements, both at the local and RR sites, were carried out for
18 h using GPS clock synchronization. Out of the 18 h of
measurements, 3 h were for high range (30 to 7.5 Hz) and
15 h for low range (6.0 to 0.00055 Hz). Time series for
all the five components for the entire frequency range
were converted to tensor impedances using least square
technique. Apparent resistivities and phases were then
obtained from impedance data.
All the data collected were processed by hybrid robust,
i.e. combination of robust separately with coherency
weighted estimates (CWE)18,19 and rho-variance19 . These
schemes improve the data quality in all the frequency
range, including the ‘dead band’ (5.0 to 0.1 Hz) which
lies between the frequency range of two types of sources,
thunderstorm and micropulsations. In this range the noise
Figure 1. Geological map of Singhbhum granite and adjoining region
(after Saha14 ). 1, Older Metamorphic Group; 2, Older Metamorphic
Tonalite gneiss; 3, Singhbhum Granite-Phase I; 4, Singhbhum GranitePhase II; 5, Iron Ore Group lavas, ultramafics; 6, Iron Ore Group
shales, tuffs, phyllites; 7, Singhbhum Granite-Phase III; 8, Quartziteconglomerate-pelite of Dhanjori Group; 9, Simlipal lavas; 10, Proterozoic Gabbro-anorthosite-ultramafics; 11, Mayurbhanj Granite. AB: The
line along which MT 2D model has been presented. Inset shows the
location map.
1260
is either equal or stronger than the signal. The static shift
was applied following Jones and Dumas20 by shifting the
apparent resistivity (ρa) curve so that the E-polarization
apparent resistivity data had the same long-period asymptote of 8000 ohm-m at 1 Hz. The corrected data are displayed in pseudosection form (Figure 2). The apparent
resistivity pseudosection (top) shows a resistive model,
specially in the central part of the profile. The corresponding phase values (bottom) are uniformly around 0°.
These two indicate a thick electrically homogeneous resistive model. In the southeastern part of the profile, the
lower resistivity is observed at relatively higher frequencies where corresponding phase values are high. Similar
features are also seen in the northwestern part of the profile. Occam inversion21,22 scheme which uses the finite
element forward modelling technique23,24 was applied to
the hybrid processed RR MT data for 2-D conductivity
modelling. The inversion scheme determines the smoothest model which matches with the observed data with
certain tolerance. One determines how well to fit the
data, and the inversion determines how much structure
the model requires. If the required tolerance is too small,
there will be no opportunity to smooth the data.
The tipper value for frequencies less than 1 Hz for the
data set ranged between 0.2 to 0.3. Therefore, a 2D modelling25 to fit the data was considered. The strike direction of the structure was obtained from Swift’s strike26 ,
polar diagrams and rotation angle plots. The average
strike direction was N25°W. Data at eight frequencies
were used in the range of 1 Hz to 1000 s. At first, 1-D
inversion of the processed MT data was carried out. A
two-dimensional cross-section consisting of a number of
side by side one-dimensional models, commonly called a
‘stitched’ section was prepared (Figure 3) along section
AB (Figure 1). The prepared stitched section served as
the initial model for 2-D inversion. The 2-D subsurface
along section AB was approximated by 67 grid points in
y direction and 46 grid points in z direction. The vertical
mesh was created with rows which increased in thickness
with depth.
Figure 4 gives a 2D geoelectric model along AB10–12
which shows 38 km thick electrically homogenous granitic crustal layer of very high resistivity (30,000 ohm-m)
below the EIC. The model fit the data to within 2.8° in
phase and 0.0217 in log ñ a. A uniform layer of 8 km
thickness below the granitic crust with relatively lower
resistivity is found at a depth of 38 km. The resistivity of
this layer is ≈ 8500 ohm-m. Keeping in view the outcrops
of charnockites along the southern fringes of the Singhbhum craton north of the Sukhinda thrust and of khondalites (graphite-bearing) to the south of the Sukhinda
thrust, these rocks may be candidates for the 8500 ohm-m
layer. This implies a greater crustal thickness, i.e.
46 ± 2.1 km below the craton12 . Seismic data in this area
are not available for comparison. However, an early work
based only on Bouguer anomaly and elevation data27
CURRENT SCIENCE, VOL. 83, NO. 10, 25 NOVEMBER 2002
RESEARCH COMMUNICATIONS
Figure 2.
Pseudosection of the corrected MT data: Apparent resistivity (top) and Phase (bottom).
shows the depth to the Moho near EIC around 40 km
against 46 ± 2.1 km obtained from this MT study over
EIC.
The resistivity of the upper mantle at a depth of
46 ± 2.1 km12 is about 750 ohm-m showing a decrease in
resistivity by an order from that of the lower crust of
about 8500 ohm-m. Constable et al.28 obtained the conductivity of pure olivine to a reasonable approximation in
the temperature range 720–1500°C as:
σ = 102.402e–1.60eV/kT + 109.17e–4.25eV/kT,
(1)
where T is the temperature, k is the Boltzmann constant
and eV is the electron volt.
CURRENT SCIENCE, VOL. 83, NO. 10, 25 NOVEMBER 2002
From eq. (1) the resistivity of upper mantle at a depth
of 46 km (temperature gradient 25°C/km = 1150°C)
should have been 4000 ohm-m. However, the resitivity
from MT studies at this depth is 750 ohm-m only.
In the southeastern end of the profile, the thickness of
the granitic crust remains 38 km. But at a depth of 8–
10 km to 20–25 km, a low resistivity zone is seen. This
low resistivity zone extends beyond the profile of measurement. Similarly, in the north western end of profile a
low resistive zone is seen at a depth of about 18 km and
extends up to 33 km. The source area for basic enclaves
and intrusive basic dykes in SG14 can explain conducting
zones of the northwestern and southeastern parts of the
profile. As compared to southeastern end, the granitic
1261
RESEARCH COMMUNICATIONS
Figure 3.
Figure 4.
1262
Stitched section along profile AB over Eastern Indian Craton.
2-D Occam Model over Eastern Indian Craton.
crust is around 23 km in north western part towards the
Dalma volcanic belt. 2-D modelling also was carried out
using Very Fast Simulated Annealing (VFSA) technique19 . Moho depth obtained by VFSA is 46 ± 2.6 km12 .
The models obtained by using Occam and VFSA inversion schemes are in agreement. The thinning can be attributed to a much younger (1.6 Gyr) plume magmatism in
Dalma volcanic belt29,30. The presence of a mantle plume
has important implications for the breakup of continents
and evolution of volcanic margin. Plume arrival beneath
a continent is not marked by sudden onset of mafic volcanism. The initial thermal input of the plume promotes
gentle doming and low-extension-factor rifting, possibly
accompanied by thinning of the subcontinent lithosphere29 .
Bhattacharya et al.12 obtained the conductivity of
0.00012 S/m for CLC under EIC which is in broad agreement with the experimental results31 . A possible explanation for the CLC below EIC is the absence of imbrication
of sedimentary material and underplating of mafic crust
related to subduction processes. About 8 km thick lower
crustal layer occurring at a depth of 38 km, and extending
CURRENT SCIENCE, VOL. 83, NO. 10, 25 NOVEMBER 2002
RESEARCH COMMUNICATIONS
Table 1.
Relevant parameters for EIC and Slave Craton
EIC10–12
Parameters
Age
Data
Frequency (Hz)
Frequencies used for modelling
Dimensionality
Table 2.
Features
Crust
Conductivity of CLC
Conductance of CLC
Moho depth
Upper Mantle Resistivity
Upper Mantle Rock Type
Slave Craton 13
3.3 Gyr
Remote reference MT
320–10 –4 Hz
Eight frequencies in the range
of 1 Hz – 1000 s
2D
4.0 Gyr
Remote reference MT
10 4 Hz – 10 –3 Hz
27 frequencies in the range of
1000–0.1 Hz
1D
Features of the obtained geoelectric models for EIC and Slave Craton
EIC10–12
Slave Craton 13
Electrically homogeneous
~ 0.00012 S/m
~ 1S
46 ± 2.1 km
~ 750 ohm-m
Not pure olivine
horizontally for more than 40 km as revealed by this MT
study will be difficult to explain by sub-horizontal disposition along a thrust. Near subduction boundary the
down-going slab usually becomes steeper. Plume magmatism can potentially explain the horizontal disposition
and layered pattern of the crustal structure of the region12 . Indeed Nd-isotope studies of the TTGamphibolites of OMG rocks strongly suggest their
derivation from a depleted plume source15 .
Jones and Ferguson13 presented the result of the MT
survey conducted as part of Lithoprobe’s Slave-northern
Cordillera lithospheric evolution (SNORCLE). The RR
MT data were acquired in the frequency range of 10000–
0.0001 Hz on the south-western corner of the Archaean
Slave Craton – the oldest craton (~ 4.03 Gyr).
The acquired time series data at each site were processed using a robust multi-remote-reference algorithm
(method 6, ref. 32). The effect for local distortions of the
electric field of the estimated responses were corrected33 .
The conductivity varies with depth only as the contoured
MT phases for all the sites in two orthogonal direction
shows uniform lateral behaviour. Therefore, 1-D inversion of the processed data using Occam inversion
scheme34 was carried out by them. In this work, 27 frequencies in the range 1000–0.1 Hz with minimum
assigned errors of 3.5% in apparent resistivity and 1° in
phase were used.
The model obtained over Slave Craton shows a low
conductivity uppermost layer of < 1 km underlain by
a more conductivity layer to a depth of 2–3 km. Below
this there is a region of very low conductivity
(< 0.000025 S/m) to some tens of kilometres depth,
beneath which is a moderately conductive (0.00025 S/m)
homogeneous basal layer. The obtained Moho depth was
35.8 ± 1.5 km. The obtained result is consistent with the
CURRENT SCIENCE, VOL. 83, NO. 10, 25 NOVEMBER 2002
Electrically homogeneous
~ 0.000025 S/m
~ 1S
35.8 ± 1.5 km
~ 4000 ohm-m
Not pure olivine
obtained depth from seismic reflection, refraction and
teleseismic studies. The resistivity of upper mantle at the
Moho depth as obtained in the Slave Craton using eq.
(1)28 should be 300,000 ohm-m which is much higher
than the value of 4000 ohm-m obtained from MT studies.
The comparison of relevant parameters and features of
the geoelectric models for EIC and Slave Craton are
given in Tables 1 and 2.
It is emphasised that these are the only two rare instances where the absence of conducting CLC, i.e. lower
crustal conductor (LCC) is observed. This enabled identification of definite crust–mantle boundary, upper mantle
resitivity and the nature of olivine as the upper mantle
material.
The conductance of CLC under EIC12 and Slave
Craton13 are less than 1S as against the minimum conductance of 20 S generally obtained for mid- to late-Archaean cratons13 . MT study over Slave province, indicates
that in the early Archaean rocks (~ 4 Gyr) absence of
conducting CLC suggest a quite different dynamics than
the plate tectonics process13 . One speculates that this difference addresses questions of early earth development
and tectonic processes, and the applicability of platetectonic theory to the early- to mid-Archaean.
1. Haak, V., in Numerical Data and Functional Relationship in Science and Technology (ed. Angenheister, G.), Springer, Berlin,
1982, vol. 1, pp. 291–307.
2. Hyndmann, R. D. and Hyndmann, R. W., Earth Planet. Sci. Lett.,
1968, 4, 427–432.
3. Brace, W. F., in Structure and Physical Properties of the Earth’s
Crust, AGU Geophys. Monogr. Ser., 14, 1971, pp. 206–210.
4. Gough, D. I., Nature, 1986, 323, 143–144.
5. Jones, A. G., Geophys. J. R. Astron. Soc., 1986, 87, 1215–1224.
6. Jones, A. G., in Continental Lower Crust, Elsevier, The Netherlands, 1992, pp. 81–143.
1263
RESEARCH COMMUNICATIONS
7. Duba, A., Huenges, E., Nover, G., Will, G. and Jodicke, H., Geophys. J., 1988, 94, 413–419.
8. Frost, B. R., Fyfe, W. S., Tazaki, K. and Chan, T., Nature, 1989,
340, 134–136.
9. Jones, A. G., Lithos, 1999, 48, 57–80.
10. Bhattacharya, Bimalendu B., Shalivahan and Sinharay, Rajib K.,
61st EAGE Conference, Helsinki, 1999, Extended Abstract Volume 1.
11. Bhattacharya, Bimalendu B., Shalivahan, Shaw, R. K., Mohanty,
P. R., Sinharay, Rajib K., Acharya, S. K. and Pathak, S., Research
Highlights in Earth System Science, DST Special Volume 1, 2000,
pp. 151–162.
12. Bhattacharya, Bimalendu B. and Shalivahan, Geophys. Res. Lett.,
in press.
13. Jones, A. G. and Ferguson, I. J., Nature, 2001, 409, 331–333.
14. Saha, A. K., Geol. Soc. India, 1994, Memoir 27.
15. Sharma, M., Basu, A. R. and Ray, S. L., Contrib. Mineral Petrol.,
1994, 117, 45–55.
16. Bhattacharya, Bimalendu B. and Shalivahan, Curr. Sci., 1999, 76,
1108–1113.
17. Shalivahan and Bhattacharya, Bimalendu B., J. Geophys. Res., in
press.
18. Egbert, G. and Livelybrooks, D. W., Geophysics, 1996, 61, 964–
970.
19. Shalivahan, Ph D thesis, Department of Applied Geophysics,
Indian School of Mines, Dhanbad, 2000.
20. Jones, A. G. and Dumas, I., Phys. Earth Planet. Inter., 1993, 81,
289–314.
21. deGroot-Hedlin, C. and Constable, S., Geophysics, 1990, 55,
1613–1624.
22. Constable, S. C., Occam v2.0 Documentation – A Computer Program for Nonlinear Smooth Inversion, Scripps Institute of Oceanography, La Jolla, USA, 1992.
23. Wannamaker, P. E., Stodt, J. A. and Rijo, L., Geophys. J. R.
Astron. Soc., 1987, 88, 277–296.
1264
24. Wannamaker, P. E., PW2DIS User Documentation – Finite Element Solution for Two Dimensional Magnetotelluric Modelling
Responses and Sensitivities for Two-dimensional Earth Resistivity
Structure, University of Utah Research Institute, Utah, USA,
1989, ESL-89043-TR.
25. Vozoff, K., Soc. Explor. Geophys., 1991, 641–711.
26. Swift, G. M., Ph D thesis, Department of Geology and Geophysics, Massachusetts Institute of Technology, Cambridge, MA,
1967.
27. Narain, H., Tectonophysics, 1973, 20, 249–260.
28. Constable, S., Shankland, T. J. and Duba, A., J. Geophys. Res.,
1992, 97, 3397–3404.
29. Kent, R. W., Geology, 1991, 19, 19–23.
30. Roy, A., Sarkar, A., Jeyakumar, S., Aggrawal, S. K. and Ebihara,
M., Gondwana Res., 2001, 5, 133–146.
31. Olhoeft, G., J. Geophys. Res., 1981, 86, 931–936.
32. Jones, A. G., Chave, A. D., Egbert, G., Auld, D. and Bahr, K.,
ibid, 1989, 94, 14201–14213.
33. McNeice, G. and Jones, A. G., Geophysics, 2001, 66, 158–173.
34. Constable, S. C., Parker, R. L. and Constable, C. G., ibid, 1987,
52, 289–300.
ACKNOWLEDGEMENTS. We thank Sobhan Pathak, Dr Ranjit K.
Shaw. Dr P. R. Mohanty, S. K. Acharya, Raja Mukhopadhyay and
Prakash Kumar of the Department of Applied Geophysics, Indian
School of Mines for assisting in collecting the MT data. Thanks are
also due to Drs A. Sarkar, D. Asthana and S. Mohanty, A. S.
Venkatesh, Department of Applied Geology, and ISM for discussion.
The processing has been carried out in the computer centre of UGC
SAP DSA of AGP. The work was funded by DST, Govt. of India, UGC
SAP DSA and All India Council of Technical Education (AICTE),
New Delhi.
Received 23 April 2002; accepted
CURRENT SCIENCE, VOL. 83, NO. 10, 25 NOVEMBER 2002