Three-dimensional lithospheric mapping of the eastern Indian shield: A multi-parametric inversion approach

Dear Author, Please, note that changes made to the HTML content will be added to the article before publication, but are not reflected in this PDF. Note also that this file should not be used for submitting corrections.

UNCO

RRECTED P

RO

OF

1Q1 Three-dimensional lithospheric mapping of the eastern Indian shield: A2 multi-parametric inversion approach

3Q2 A.P. Singh a,⁎, Niraj Kumar a, H. Zeyen b

4 a CSIR-National Geophysical Research Institute, Uppal Road, Hyderabad 500007, India5 b Départment des Sciences de la Terre, UMR 8148 GEOPS, Université Paris-Sud, CNRS, Orsay, France

6

a b s t r a c t7 a r t i c l e i n f o

8 Article history:9 Received 24 March 201510 Received in revised form 2 July 201511 Accepted 30 September 201512 Available online xxxx

13 Keywords:14 Gravity15 Geoid16 Topography17 3D inversion18 Lithosphere–asthenosphere boundary

19We analyzed satellite gravity and geoid anomaly and topography data to determine the 3D lithospheric density20structure of the Singhbhum Protocontinent. Our density model shows that distinct vertical density heterogene-21ities exist throughout the lithosphere beneath the Singhbhum Protocontinent. The crustal structure identified in-22cludes a lateral average crustal density variation from 2800 to 2890 kg/m3 as well as a relatively flat Moho at2335–40 km depth in Singhbhum Protocontinent and Bastar Craton. A similar Moho depth range is found for the24Mahanadi, Damodar, and Bengal basins. In the northern part of the area, Moho undulates between more than2540 km under the confluence of Mahanadi–Damodar Gondwana basins and the Ganga foreland basin, and 36–2632 km under the Eastern Ghats Mobile belt and finally reaches 24 km in the Bay of Bengal. The lithosphere–27asthenosphere boundary (LAB) across the Singhbhum Protocontinent is at a depth of about 130–140 km. In28the regions of Bastar Craton and Bengal Basin, the LAB dips to about 155± 5 kmdepth. The confluence of Maha-29nadi and Damodar Gondwana basins toward the north-west and the foreland Ganga Basin toward the north are30characterized by a deeper LAB lying at a depth of over 170 and 200 km, respectively. In the Bay of Bengal, the LAB31is at a shallower depth of about 100–130 km except over the 85 0E ridge (150 km), and off the Kolkata coast32(155 km). Significant density variation aswell as an almostflat crust–mantle boundary indicates the effect of sig-33nificant crustal reworking. The thin (135–140 km) lithosphere provides compelling evidence of lithospheric34modification in the Singhbhum Protocontinent. Similarities between the lithospheric structures of the35Singhbhum Craton, Chhotanagpur Gneiss Complex, and Northern Singhbhum Mobile Belt confirm that the re-36peated thermal perturbation controlled continental lithospheric modification in the Singhbhum Protocontinent.

37 © 2015 Published by Elsevier B.V.

3839

40

41

42 1. Introduction

43 Cratons are generally considered to be stable tectonic units and44 underlain by thick (~200 km) lithospheric roots that are chemically45 and physically distinct from the surrounding mantle. Buoyancy, as46 well as the refractory nature of Archaean sub-continental lithospheric47 mantle (SCLM), offers a simple explanation for thickness and longevity48 of the Archaean lithospheric keels (Jordan, 1988;O'Reilly et al., 2001). In49 contrast, recently discovered significantly shallower lithosphere–as-50 thenosphere boundary (LAB) in several cratons such as the North51 China Craton (Zhu et al., 2012; Wang et al., 2014), the western52 Wyoming Craton (Lee et al., 2000), the western Brazilian Craton (Beck53 and Zandt, 2002), the Saharan Metacraton (Abdelsalam et al., 2011),54 and the southern Granulite Terrane of Dharwar Craton (N. Kumar55 et al., 2013) open the questionwhether the SCLMhad a notably thinned56 root since the time of its stabilization or if subsequent tectonic events57 have modified it (Lee et al., 2011). Mapping, with reasonable accuracy,58 the present-day structure of the lithosphere and hence the depth to

59the LAB is for that reason a critical factor for understanding the tectonic60processes responsible for the evolution of cratons. Geological proxies61such as xenoliths and xenocrysts may offer complimentary constraints62on lithospheric modification mechanism.63The eastern Indian shield is one of the critical examples in the cur-64rent debate (Fig. 1). The Singhbhum Craton, one of the oldest nuclei65(~3.6 Ga old) of the eastern Indian shield that stabilized at about663.0 Ga (e.g., Roy and Bhattacharya, 2012), possibly amalgamated into67a reasonably compact, large continental mass including Chhotanagpur68Gneiss Complex by Proterozoic time and remained a coherent entity69throughout the Phanerozoic (e.g., Sharma, 2009). This triangular70amalgamated large continental region is referred to hereafter as the71Singhbhum Protocontinent. Major tectonic and magmatic events of72the region include Palaeoproterozoic thermal perturbations in73Singhbhum Craton (Mazumder, 2005), sandwiching of the Northern74Singhbhum Mobile Belt, and hence welding of the two terranes,75Singhbhum Craton and Chhotanagpur Gneiss Complex (Meert et al.,762010), Mesoproterozoic mafic as well as ultramafic intrusions in77Singhbhum Craton (Bose, 2009) and extrusion of large-scale Rajmahal78Traps in north-eastern part of the Chhotanagpur Gneiss Complex and79lamproites in Damodar Gondwana Basins at around 117±2 Ma

Tectonophysics xxx (2015) xxx–xxx

⁎ Corresponding author.E-mail address: [email protected] (A.P. Singh).

TECTO-126809; No of Pages 13

http://dx.doi.org/10.1016/j.tecto.2015.09.0380040-1951/© 2015 Published by Elsevier B.V.

Contents lists available at ScienceDirect

Tectonophysics

j ourna l homepage: www.e lsev ie r .com/ locate / tecto

Please cite this article as: Singh, A.P., et al., Three-dimensional lithospheric mapping of the eastern Indian shield: A multi-parametric inversionapproach, Tectonophysics (2015), http://dx.doi.org/10.1016/j.tecto.2015.09.038

http://dx.doi.org/10.1016/j.tecto.2015.09.038

mailto:[email protected]

Journal logo


http://www.sciencedirect.com/science/journal/00401951

www.elsevier.com/locate/tecto


UNCO

RRECTED P

RO

OF

80 (Kumar et al., 2003). Several competing geodynamic scenarios have81 been proposed to explain the tectono-magmatic evolution of the82 SinghbhumProtocontinent, starting from a sequence of collision tecton-83 ics (e.g., Sarkar, 1982; Saha, 1994; Rekha et al., 2011) inducing subcrust-84 al metasomatism (Banerjee, 1981; Chalapathi Rao et al., 2013) and85 resulting in EMI-typemantle (Roy et al., 2004).Most of these competing86 geodynamic scenarios rely on models of uncertain lithospheric struc-87 ture derived mostly from geological proxies.88 Geophysical observations provide high-resolution data that can be89 used to obtain refined images of the lithospheric structure. The few iso-90 lated geophysical studies undertaken in the SinghbhumProtocontinent,91 however, show ambiguous lithospheric structure due to their own set of92 assumptions and limitations. Published estimated thermal lithospheric93 thicknesses is ~65 km (Pandey and Agrawal, 1999), electrical litho-94 sphere varies from 58 to 95 km (Roy et al., 1989; Shalivahan et al.,95 2014), and lithospheric estimates using S-wave receiver functions

96range between 100 and 115 km (P. Kumar et al., 2007, 2013). In97contrast, the regional-scale estimates of the lithospheric thickness98derived from shear-wave velocity as a function of depth indicate a99180–220 km thick continental lithosphere under the Singhbhum100Protocontinent (Priestley and McKenzie, 2006). Global seismic data101demonstrate that the 95 ± 4 km interface likely represents a boundary102in composition, melting, or anisotropy, the LAB being otherwise103expected to be much deeper in cratonic regions (Rychert and Shearer,1042009). Owing to scanty data and large uncertainty in resolution of the105individual geophysical proxies, the delineated LAB structure of106Singhbhum Protocontinent seems elusive and the argument that the107lithospheric mantle is modified and recycled needs to be ascertained.108In view of this, we make an attempt to delineate a simplified 3D litho-109spheric density structure encompassing the Singhbhum Protocontinent110and adjoining part of the Bay of Bengal combining three distinct111geophysical proxies, namely, gravity, geoid, and topography data

Fig. 1. Locationmap of the principal structural units of the eastern Indian shield and adjoining Bay of Bengal. I: Eastern GhatsMobile Belt, II: Baster Craton, III: SinghbhumCraton, IV: NorthernSinghbhumMobile Belt, V: Chhotanagpur Gneiss Complex, VI: Damodar Basin, VII:Mahanadi Basin, VIII: Ganga Basin, IX: Bengal Basin. [1] Phanerozoic sediments; [2] Rajmahal/Deccan Traps;[3] Gondwana sediments (of E-W trending Damodar and NW-SE trending Mahanadi basins); [4] Proterozoic volcanics (Dalma, Dhanjori, and Simlipal) / Alkaline rocks; [5] Archaean / Prote-rozoic Granite / Granitoid; [6] Schist belts; [7] Charnockites and khondalites; [8] Proterozoic Basins; [9] Older Metamorphic Group; [10] Iron Ore Group; [11] Chhotanagpur Gneiss Complex;[12] Archaean Granite gneiss. Location of MT profiles, Heat flow (in mW/m2) observations, Kimberlites/Orangeites and Proterozoic dolerite dykes are also given for the ready reference. Thebroad-band seismic stations are *1 VISK: Visakhapatnam, *2 BWNR: Bhubaneswar, *3 CAL: Kolkata, *4 DHAN: Dhanbad, and *5 BOKR: Bokaro. MD stands for the Mahanadi Delta.

2 A.P. Singh et al. / Tectonophysics xxx (2015) xxx–xxx



UNCO

RRECTED P

RO

OF

112 (Motavalli-Anbaran et al., 2013; Kumar et al., 2014). Integrating the113 available xenolith information, we will then attempt to address the114 extent to which the evolution of the Singhbhum Protocontinent can115 be related to the inferred lithospheric structure.

116 2. Geological setting

117 The Singhbhum Protocontinent consists of two Archaean to Protero-118 zoic tectonic domains, named, the Singhbhum Craton and the119 Chhotanagpur Gneiss Complex, separated by the Northern Singhbhum120 Mobile Belt (Fig. 1). It is bordered by the NW-SE trending Mahanadi121 Graben with Gondwana sediments in the west and the charnockite122 and khondalite province of the Eastern GhatsMobile Belt and the recent123 coastal alluvium in the south. It is engulfed in the north by vast alluvium124 of the Ganga foredeep basin and in the east by Quaternary sediments of125 the Bengal Basin. The Rajmahal Traps of upper Jurassic to lower126 Cretaceous basalts occupy the contact zone of the Chhotanagpur Gneiss127 Complex and the Bengal Basin. To avoid the boundary effects, a rectan-128 gular area bigger than our area of interest partly integrates the Bastar129 Craton in the west and the Bay of Bengal in the south-east. The broad130 geology of our study area is characterized by a series of allochthonous131 terranes that were amalgamated to form the eastern Indian shield.

132 2.1. Singhbhum Craton

133 Singhbhum Craton at the center is composed mainly of Archaean134 granitoid batholiths, enclaves of the Older Metamorphic Group and135 Iron Ore Group, and amphibolites with TTGs along the contacts. The nu-136 cleus is punctuated by Paleoproterozoic Simlipal and Dhanjori volcanics137 and finally attested by the intrusion of ultramafic varieties of the mafic138 dyke swarms, generally known as Newer Dolerite Dyke swarms, rang-139 ing in age from 1600 to 950 Ma (Bose, 2009; Sharma, 2009).

140 2.2. Chhotanagpur Gneiss Complex

141 The Mesoproterozoic Chhotanagpur Gneiss Complex comprises142 mainly granitic gneisses and numerousmetasedimentary rocks. The lin-143 ear Gondwana Damodar Basin is dotted by swarms of mafic and ultra-144 mafic igneous intrusions, which occur mostly as dykes and sills145 (e.g., Chalapathi Rao et al., 2014 and references therein). The146 ultrapotassic dykes of the Ranigunj coal field are lamproites (sensu147 lato) having a commonparentalmagmawhichwas generated frompar-148 tial melting of the ancientmetasomatised lithospheric mantle (Mitchell149 and Fareeduddin, 2009; Chalapathi Rao et al., 2014). Srivastava et al.150 (2009) have, however, proposed that a geodynamic setting involving151 (i) metasomatically veined and thinned lithosphere located at themar-152 gin of the Singhbhum Craton, and (ii) the inheritance of Archaean153 subducted component have played a significant role in deciding the di-154 verging petrological and geochemical characters – those of kimberlites155 (orangeites) and lamproites (cratonic signature) and even those of156 aillikites (rift-related signature) – displayed by the Jharia potassic rocks.

157 2.3. Northern Singhbhum Mobile Belt

158 The late Palaeoproterozoic Northern SinghbhumMobil Belt basically159 consists of highly deformed shallow-marine shelf deposits metamor-160 phosed to kyanite-grade dotted by the ultramafic-mafic Dalma161 volcanics. The belt has been interpreted as representing island arc162 magmatism, an ophiolitic assemblage at the suture arising out of the163 Singhbhum “microplate” in the south subducting under the164 Chhotanagpur continental block (Sarkar, 1982), and a rift filling in a165 marginal basin or back-arc basin in a subduction-related environment166 (Bose, 1994).

1672.4. Bastar Craton

168The Bastar Craton dominantly comprises a granite gneiss complex of169Archaean age, unconformably overlain by supracrustal, intracratonic170Meso- to Neoproterozoic sedimentary sequences and coeval mafic171dyke swarms (French et al., 2008). Kimberlite fields with six diamond-172iferous pipes of late Cretaceous (Lehmann et al., 2010) are dated173synchronous with the two subsurface mafic dykes intrusive into the174sedimentary rocks of the Mesoproterozoic Chhattisgarh Basin175(Chalapathi Rao et al., 2011).

1762.5. Rajmahal Traps

177The Rajmahal Traps of late Jurassic to early Cretaceous basalts occu-178py the palaeocontinental margin of the eastern Indian shield blanketing179an area of about 4100 km2. When identified subsurface correlative180basalt in seismic and drilling results is considered, the Rajmahal Traps181extend southward below the surface for at least 100 km, beneath the182Tertiary sediments of the Bengal Basin covering an area of about183200,000 km2. Available age data (of ca. 117 ± 2 Ma) further suggests184that the magmatism is likely to be related to the Kerguelen mantle-185plume activity (Kumar et al., 2003).

1862.6. Bay of Bengal

187The Bay of Bengal evolved around 130 Ma with the breakup of the188Indian continent from eastern Gondwanaland. Since the rise of the Hi-189malayan Mountains, an enormous amount of sediments has been erod-190ed at the headwaters of the Ganges and the Brahmaputra rivers and191deposited in the Bay of Bengal to form the world's largest delta-fan192complex.

1933. Methodology and data

194Themethod used is a direct, linearized, iterative inversion procedure195in order to determine lateral variations of Moho and LAB depths and196average crustal and mantle density (Motavalli-Anbaran et al., 2013).197The area of interest is subdivided into rectangular columns of constant198size in E-W (X) and N-S (Y) direction. In depth (Z), each column is199subdivided into four layers: sea water (with known thickness, i.e.200bathymetry, and a density of 1030 kg/m3), crust, lithospheric mantle,201and asthenosphere. The model parameters we are imaging are Moho202and LAB depths with respect to mean sea level and average crustal203and mantle density in every column.204The topography is calculated in one dimension (1D) assuming local205isostatic equilibrium as a function of thickness and average density of206the lithosphere along vertical columns of the model (Turcotte and207Schubert, 1982):

ε ¼ fρa−ρl

ρaZl þ εcal

� �ð1Þ

209209where Zl is the thickness of lithosphere, ρl the average lithosphericdensity, ρa asthenospheric density (taken as 3200 kg/m3) and εcal is a

210calibration constant that corresponds to the depth of a free astheno-211sphere with respect to sea level (−2380 m; Lachenbruch and Morgan,2121990). The factor f is equal to one for positive topography and213corresponds to ρa/(ρa − ρi) in the case of negative topography, where214ρi is the density of basin infill (here taken as sea water).215Since the crust is treated as a single layer, no distinction between216sediments, upper and lower crust is made. The density distribution in217the crust is modeled using a linear vertical density increase with218depth. We fixed the density at the crust–mantle boundary (Moho) at2193000 kg/m3 and inverted for the average crustal density. We did220tests using different Moho densities between 2950 and 3050 kg/m3

221that hardly changed the results, whereas using a constant density

3A.P. Singh et al. / Tectonophysics xxx (2015) xxx–xxx



UNCO

RRECTED P

RO

OF

222 for the whole crust does. In the mantle, density varies linearly with223 temperature:

ρ Tð Þ ¼ ρ T0ð Þ 1−α T−T0ð Þð Þ ð2Þ

225225 whereα is the thermal expansion coefficient (3.5×10−5 K−1), T0 is a ref-erence temperature, taken as 1300 °C at the LAB, and ρ(T0) is the density

226 at this temperature (here: asthenospheric density ρa = 3200 kg/m3).227 In order to make the inversion procedure viable in terms of comput-228 ing time, we calculate the temperatures like topography in a 1D ap-229 proach. Geotherms are calculated using fixed temperatures at the230 surface and the LAB (10 °C and 1300 °C, respectively) and assuming con-231 stant heat production and thermal conductivity in the crust (1 μW/m3

232 and 2.5 W/(K*m), respectively, in the upper half of continental crust,233 0.2 μW/m3 and 2.2 W/(K*m), respectively, in the lower half of conti-234 nental as well as in oceanic crust; no heat production and a thermal235 conductivity of 3.3 W/(K*m) in the mantle). In continental crust, it is236 thus implicitly assumed that upper and lower crusts have the same237 thickness. Oceanic crust with properties of the lower continental crust238 is assumedwhen bathymetry is deeper than 2000 m. Gravity anomalies239 are generally used to deduce variations in massdensity and hence sub-240 surface geological structure for a wide variety of geophysical applica-241 tions. Whereas the gravitational potential decreases with distance242 from the surface of the Earth at a slower rate than gravity (i.e. potential243 α (1/R) and gravity α (1/R2), where R is radial distance from the density244 anomalies), the geoid tends to reflect deep mass anomalies better than245 the gravity field. To understand the upper-mantlemass distribution, the246 geoid anomaly is therefore used to derive the bands of mass excess and247 deficiency in the Earth's mass structure as complementary to gravity248 data (Bowin, 1983). Analytical formulas of Gallardo-Delgado et al.249 (2003) and Fullea et al. (2009) are used to calculate the gravity and250 geoid effects, respectively, for rectangular blocks with a linear vertical251 density variation.252 The inverse problem involves minimizing the following cost func-253 tion in an iterative way:

F ¼ d− f pð Þð ÞTC−1d d− f pð Þð Þ þ λ p−p0ð ÞTC−1

p p−p0ð Þ þ μ pTCsp ð3Þ

255255 where p is a vector with the model parameters (average crustal densi-ties, followed by theMoho and LAB depths), d is the data vector (topog-

256 raphy data followed by free-air and geoid data), f represents the257 calculated effect of the actual model (forward model) results, Cd a diag-258 onal matrix containing the variances (squared estimated uncertainty)259 of each data point and at the same time is used to normalize the variabil-260 ity of the different data types. Likewise, the estimated variability of261 model parameters is contained in Cp, a diagonal matrix in which con-262 stant variances are used for each parameter set. It has twomeanings:263 first, it is used to regularize the inverse problem. Moreover, it allows264 the introduction of a priori information on certain parameters. For265 instance, if the Moho depth is known for a given column then the266 variance of the corresponding parameter is reduced by a certain267 factor. Finally, the matrix Cs is responsible for the smoothing of the268 results.269 Minimizing this cost function ultimately takes the following form270 (Motavalli-Anbaran et al., 2013):

pkþ1 ¼ AT � C−1d � Aþ λC−1

P þ μCS

� �−1� AT � C−1

d � d− f pk� ��

þ pk ð4Þ

272272 where k is the number of iteration and A is the Frechet matrix in which

the element at aij contains the derivative ∂ f iðpÞ∂p j

. More details on both

273 forward and inverse problems can be found in Motavalli-Anbaran274 et al. (2013).275 For the inversion, we extracted the data from available global276 free-air gravity anomaly and topographic data grids (both available

277at ftp://topex.ucsd.edu/pub, Sandwell and Smith, 1997) and geoid278anomaly from EGM2008 global model (available at http://earth-info.279nga.mil/GandG/wgs84/gravitymod/egm2008/index.html) complete to280spherical harmonics of degree and order 2159 (Pavlis et al., 2012). To281avoid effects of sub-lithospheric density variations on the geoid, we282have removed the geoid signature corresponding to the spherical283harmonics developed until degree and order 10. The mass anomalies284contributing to the degree 2–3 field are inferred to be the result of to-285pography at the core-mantle boundary whereas the degree 4–10286geoid anomaly appears to lie mostly in the lower mantle depths287(Bowin, 1983, 1991), although some influence of the lithosphere has288also been observed (Root et al., 2015). Attribution of geoid above degree289and order 20 to near surface mass density anomalies (Christou et al.,2901989) is valid; however, it is not always valid to attribute geoid less291than degree and order 10 to deep-Earth mass density anomalies only.292The differing conclusions seem to depend upon the approach taken293and, at present, it cannot be proven which, if either, is more accurate294(Featherstone, 1997). Here, we assumed that the majority of long-295wavelength geoidal undulations are due to the deeper source and filter-296ing the degree and order 10 of spherical harmonics probably retains the297effects of density anomalies shallower than ~400 km depth.298The three data sets (topography, free-air gravity, and residual geoid)299were then projected onto a 10 × 10 km Cartesian coordinate grid300resulting in 108 × 112 grid points (Fig. 2). Data at 10 × 10 km interval301in each block has been coalesced into vertical columns of 30 × 30 km.302We built a starting model with a 1D algorithm using only topography303data and geoid anomaly (Fullea et al., 2007) (Fig. 3a–c). For this 1D304model, we initially used a constant crustal density of 2840 kg/m3

305which is close to the average crustal density predicted for a 40 km306thick crust (Zoback and Mooney, 2003). For the subsequent 3D inver-307sion, crustal densities are allowed to increase with depth, and we308chose density to vary linearly from 3000 kg/m3 at Moho decreasing309upwards. In order to stabilize the inversion procedure, we used two310complementary approaches: on the one hand, we used global data on311crustal thickness CRUST1.0 as a priori information for the starting312model, reducing the variability of Moho depth during the iteration313process. On the other hand, we applied slight lateral smoothing of all314model parameters (Table 1). Other physical properties used are given315in Table 2. The three computed parameters (crustal density and depths316of Moho and LAB) are finally obtained on a grid of 36 × 37 columns.

3174. Inversion results

318Westarted our 3D inversion froma1Dmodel and used a priori crust-319al thickness map at 10 × 10 grid interval extracted from the “CRUST1.0”320(Laske et al., 2012). Using the CRUST1.0 model helps to fix average321Moho depths in the solution. In order to investigate the uncertainties322of our lithospheric model, we performed many inversions using a323range of values for critical parameters such as data and model parame-324ter variances, smoothing, a priori Moho depths within their uncer-325tainties and, from a suite of more than 50 models created in this way,326we selected the model that best fits the observed gravity and geoid327anomaly and topography data. Figs. 3d–f show the final model parame-328ters after 8 iterations for which themodel results are stable (see Table 3329for details of the results). The differences between the a priori given330Moho depth and the final one in blocks with a priori information are331on average ±1.4 km, which is acceptable as the uncertainty on 1D332Moho depths is usually assumed to be 2–3 km. In some areas with333good coverage, the accuracy is probably better than 3 km like at Bhuba-334neswar, Dhanbad, and Kolkata (Kayal et al., 2011; P. Kumar et al., 2013).335A larger misfit is obtained in the continental slope area where strongly336varying Moho depths over short distances can be found. The modeled337response of the 3D density model demonstrates a good fit with the ob-338served gravity and geoid anomalies and topography data (Figs. 2d–f).339The misfit standard deviations between observed and calculated data340are 7 mGal for the gravity data, 0.19 m for the geoid data and 37 m for



ftp://topex.ucsd.edu/pub

http://earthnfo.nga.mil/GandG/wgs84/gravitymod/egm2008/index.html

http://earthnfo.nga.mil/GandG/wgs84/gravitymod/egm2008/index.html


UNCO

RRECTED P

RO

OF

341 the topography data (Table 3). Exceptions are some short-wavelength342 misfits, which can be attributed to local density contrast and shallow343 crustal structures at scales smaller than 30 km and therefore cannot

344be adequately reproduced by the 3D lithospheric density modeling.345Using an isostatic hypothesis for estimating the lithospheric thickness346suffers from the incomplete isostatic model and that the observed

Fig. 2. a) Free-air gravity data from ftp://topex.ucsd.edu/pub. b) Geoid data from EGM2008 without spherical harmonics until degree and order 10. c) Topography data from ftp://topex.ucsd.edu/pub. d) Differences between calculated and measured free-air data for final model. e) Differences between calculated and measured geoid data for final model. f) Differencesbetween calculated and measured topography data for final model. Roman numbers are as in Fig. 1 except the E-W trending Northern Singhbhum Mobile Belt which is shown by thedashed line.







UNCO

RRECTED P

RO

OF

Fig. 3. Initial and final crustal and lithospheric models from 3D inversion and location of two orthogonal lithospheric cross-sections along 220 N and 860 E. a) Initial model for Moho depths(in km) resulting from 1D geoid and topography inversion. b) Initial LAB (in km) model from 1D inversion. c) Initial crustal density was a constant value of 2840 kg/m3. d) Final Mohodepth in km. e) Final depth of the LAB in km. f) Inverted average crustal density distribution in kg/m3. The pixel size corresponds to the block size of 30 × 30 km. Red lines correspond to limitsof tectonic units, black lines are isolines. Romannumbers are as in Fig. 2. (For interpretation of the references to color in thisfigure legend, the reader is referred to theweb version of this article.)




UNCO

RRECTED P

RO

OF

347 gravity anomaly is not only generated by the topographic signal but also348 by non-isostatic effects which are disregarded in most of the applica-349 tions of isostatic models. For example, in Airy–Heiskanen isostatic hy-350 pothesis we assume that all the topographic masses are compensated351 by variable lithospheric thickness, which is not the case in reality. A por-352 tion of small-wavelength gravity anomaly is due to intra-crustal density353 distribution and small- and intermediate-wavelength free-air gravity354 anomalies may be due to the mechanical strength of the lithosphere.

355 4.1. Average crustal density distribution

356 Inversion of gravity signals into the average crustal density distribu-357 tion in the Singhbhum Protocontinent and adjoining regions shows a358 strong lateral density variation in the crust (Fig. 3f). These density distri-359 butions are coherent with the regional geology observed at the surface360 and boundaries of regional tectonic units (Fig. 1). Deep sedimentary361 fill in the Sonahat sub-basin (23.40 N, 82.50 E) at the junction of Gond-362 wana Damodar andMahanadi basins, the Ganga Basin and theMahana-363 di delta are characterized by lowdensity. Relatively low-density is found364 associated with the N-S trending fault-controlled basement depression365 that connects the easternGanga Basinwith the Bengal Basin. In contrast,366 relative high-density anomalies exist over Simlipal volcanics, Rajmahal367 Traps, Garo hills/Malda plateau (250 N, 88.50 E), and high-grade meta-368 morphic rocks exposed along the Eastern Ghats Mobile Belt. An excess369 of crustal mass beneath the Chhotanagpur Gneiss Complex is a signifi-370 cant feature of this map. Particularly, the one at 23.5 0N, 84 0Ewhich co-371 incides with a high topography and free-air anomaly (Fig. 2). This may372 be explained by a slower/hotter upper mantle observed by Kayal et al.373 (2011) using ISM-Dhanbad broad-band seismic data.374 The oceanic part of themap shows a locus of strong density lows cor-375 responding to the thick pile of sediments along the continental shelf.376 The high-density anomaly to the south of Kolkata is attributed mostly377 to the ridge-like structure in the basement, a possible extension of the378 Garo hills/Malda plateau (250 N, 88.50 E) in the Bay of Bengal (Tiwari379 and Jassal, 2003).

380 4.2. The Moho depth

381 Startingwith a 1D inversion of geoid and topography data, and using382 a priori crustal thickness map at 10 × 10 extracted from the “CRUST1.0,”383 the estimated Moho depth of our best fitting model ranges from 23 km384 in the Bay of Bengal to 44 km in the Ganga foreland basin (Fig. 3d). For a385 better perception of 3D crustalmodel, two 2D orthogonal sections along386 220 N and 860 E across the Singhbhum Protocontinent showing litho-387 spheric structure along with elevation and principal structural units388 are presented in Fig. 4a–b. The thinnest crust is obviously beneath the389 oceanic region (assumed when bathymetry is deeper than 2000 m)390 varying from ~23 to 26 km. Further north, approaching continental391 India, the Moho varies between 26 and 33 km. Our value of about392 33 km along the shore line is close to the averagemean sea level crustal393 thickness value of 33 km calculated for India (Kumar et al., 2011). The394 estimated Moho depths in the Bay of Bengal are consistent with the395 gravity and seismological experiments. Recent 3D gravity inversion in-396 ferred a depth of ~18 km at 17° N latitude, which deepens sharply to397 33 km near the Bangladesh shelf (Radhakrishna et al., 2010). Inversion398 of the fundamental mode Rayleigh wave group velocity dispersion399 across the Bay of Bengal delineated similar Moho depths lying at400 20 ± 3 km in the central Bay and 30 ± 3 km in the northern Bay401 (Mitra et al., 2011).402 The Eastern Ghats Mobile Belt is characterized by 33–36 km deep403 Mohowhich remains flat to 36–38 km in the eastern part of Bastar Cra-404 ton. Except for a gentle gradient from continent to ocean, a featureless405 Moho is found all along the Mahanadi Gondwana Basin. Through the406 cross-correlation of interstation phase velocities of surface wave data,407 the crustal thickness of Bastar Craton is found to be 40 km408 (Bhattacharya et al., 2009). A crustal thickness of 32–34 km with a

409high-velocity/density (Vp = 7.5 km/s, ρ = 3050 kg/m3) lower crust is410reported beneath the Mahanadi delta (Behera et al., 2011). About41141 km deep Moho under the junction of Mahanadi–Damodar Gondwa-412na basins indicates an inverse correlation with the regional topography413excess mass which is compensated at Moho depth. Though testing of414our ~36–40 km deep Moho in Rajamahal Trap region is difficult due415to insufficient seismic data, it conforms well to the gravity-derived416Moho depth of 38 km (Singh et al., 2004). A comparable Moho depth417in the adjoining Bengal Basin is in harmony with the two available418seismic sections (Mall et al., 1999). Our deepest (∼44 km) Moho419beneath the Ganga Basin reflects the significant crustal thickening due420to a downward flexure of the Indian plate. A little shallower Moho421(~40 km) is, however, modeled by Tiwari et al. (2006) and Jiménez-422Munt et al. (2008).423The Older Metamorphic Group of Singhbhum Craton has metamor-424phosed at 5.5 kbar and 660–630 °C, indicating substantial crustal thick-425ening during Archaean (Sharma, 2009). In contrast, the elevation426change of 700 m from the coast to themajor plateau regions and a den-427sity contrast of 400 kg/m3 across the Moho imply that the isostatically428compensated Moho in Singhbhum Craton should be ~5 km deeper429than the average crustal thickness along the shore line (33 km). Inte-430grating isostasy at whole-lithosphere level our calculated Moho in the431Singhbhum Craton lies at a depth of about 35–38 km which deepens432up to 40 km in the Chhotanagpur Gneiss Complex. OurMoho depth cal-433culation further indicates that in the SinghbhumCraton observed higher434heat flow have no link with its crustal thickness. This may be due to the435diversity in crustal composition as well as the evolution of the heat436production in the crust. A 2D geoelectric section shows a 38 km thick437electrically homogenous granitic crustal layer of very high resistivity438followed by a uniform layer of 8 km thickness below the granitic439crust with relatively lower resistivity in the Singhbhum Craton440(Bhattacharya and Shalivahan, 2002; Shalivahan and Bhattacharya,4412005 and references therein). Receiver function analysis of ISM-442Dhanbad broad-band data establishes a crustal thickness of 41 km443(Kayal et al., 2011). At Bokaro, which is nearby to ISM-Dhanbad444broad-band station, the S-wave receiver function stacked traces for445the individual station show a Moho depth of about 42 km (P. Kumar446et al., 2013). A little higher crustal thickness (45 km) is, however, ob-447served by Kosarev et al. (2013). According to Ravi Kumar et al. (2001),448the crust beneath the Bokaro appears to be highly complicated where449Moho depth is inferred at about 54 km, assuming an average S-wave450velocity of 3.7 km/s.451The one-to-one correlation between the electrical/seismicMohoand452the computed Moho depth beneath the Singhbhum Craton shows 7–4538 km deeper crustal boundary than our isostatically calculated454(Fig. 7a). This is partly due to the fact that our approach is not strictly455valid to image an uncompensated crustal underplated structure as pro-456posed by others. Partly the thicker crust is also denser, which would457tend to suppress an increase in elevation associated with increasing458crustal thickness. This implicitly implies that a thick depleted litho-459spheric root is required to “pull” the crust down (Zoback and Mooney,4602003). A premise that will go against the purported lithospheric thin-461ning to about 100 km in the Singhbhum Craton (Pandey and Agrawal,4621999; P. Kumar et al., 2007, 2013; Shalivahan et al., 2014).

t1:1Table 1t1:2Parameters used for inversion.

t1:3Data Gravity Geoid Topography

t1:4Uncertainty (matrix Cd in Eq. 4) 5 mGal 0.3 m 100 mt1:5

t1:6Parameters Moho depth LAB depth Density

t1:7Variability (matrix Cp in Eq. 4) 100 m 1 km 1 kg/m3

t1:8Smoothing (matrix CS in Eq. 4) 0.2 0.2 0




UNCO

RRECTED P

RO

OF

463 4.3. The Lithosphere thickness

464 The 3D geometry of our final LAB derived for the Singhbhum465 Protocontinent and the adjoining Bastar Craton and Bay of Bengal is466 shown in Fig. 3e. Hot colors indicate thin lithosphere and cool colors in-467 dicate thick lithosphere. For a better perception of 3D lithospheric468 model, two 2D orthogonal sections along 220 N and 860 E across the469 Singhbhum Protocontinent showing lithospheric structure along with470 elevation and principal structural units are presented in Fig. 4a–b. In471 our 3D model, LAB varies in the depth range of 110–120 km under the472 oceanic region reaching more than 150 km depth beneath the 850 E473 Ridge. Further north, approaching the continental slope, the belt veers474 eastward and realignswith the palaeocontinental margin of the eastern475 Indian shield. A lithospheric thickness of ~100 km beneath the central476 part of the Bengal Fan plausibly indicates the subsurface extension of477 theGaro hills/Malda plateau in the Bay of Bengal. Our lithospheric thick-478 nesses conformwell to recent surfacewave dispersion estimates of a lit-479 tle over 100 km (Mitra et al., 2011) and ~110 km by S-wave velocity480 experiments (Bhattacharya et al., 2013) in the northern part of Bay of481 Bengal.482 Under the Eastern Ghats Mobile Belt, we find a lithospheric483 thickness of ~120 km, which thickens further to 160 km beneath the484 Bastar Craton. The Singhbhum Protocontinent along with the Damodar485 Gondwana Basin is underlain by a comparatively thin continental litho-486 sphere lying at a depth of ~135 km. The LAB lies at a deeper depth under487 the junction of Mahanadi–Damodar Gondwana basins (~165 km) and488 in the Bengal Basin (~160 km) than in other regions controlled by489 extensional events. The calculated lithosphere thickens sharply under490 the foreland Ganga Basin and attains a depth of over 200 km; though491 a lithospheric thickness reaching up to 160–190 km is imaged below492 the eastern and central part of the Ganga Basin by Jiménez-Munt et al.493 (2008) and Robert et al. (in Press), respectively. In this region, the494 possible component of flexural support to the lithospheric bending,495 evidenced by large sedimentary thickness accumulation, could also496 affect the validity of our approach resulting in an overestimation of497 our modeled lithospheric thickness. Our result in Bastar Craton appears498 to be in agreement with the seismological estimations where a litho-499 sphere thickness of 140 km was proposed on the basis of observed500 low S-wave velocity (Bhattacharya et al., 2009). Similarly, the converted501 wave techniques (P- and S-wave receiver functions) applied to the data502 of Bilaspur station (22.10 N, 82.10 E) of the region reveals the LAB at503 142 km (P. Kumar et al., 2013).504 Most of the values of lithospheric thickness of adjoining Bay of505 Bengal, Bastar Craton, and Ganga foreland basin fall within the range506 of seismically determined thickness i.e., between 100 and 200 km. It507 significantly differs in the SinghbhumCratonwhere thinned lithosphere508 (~100 km) is inferred (Pandey and Agrawal, 1999; Kumar et al., 2007;

509Shalivahan et al., 2014). In particular, thin lithosphere is expected in510young and active areas such as Saharan Metacraton (Abdelsalam et al.,5112011). Anomalously thin lithosphere thickness values (~140 km) are512also predicted for some Proterozoic shields. These thin lithosphere513values correspond to the relatively high densities computed for the514Proterozoic lids (Zoback and Mooney, 2003 and references therein). In515contrast, the diversified crustal composition, since the crust in the516SinghbhumCraton contains importantminerals and pervasive granitoid517batholiths, indicates a relatively low mantle density. It may be attribut-518ed tobasalt depletion associatedwith a partialmelting of theuppermost519mantle due to higher mantle temperatures since the late Archaean.

5204.4. Parameter uncertainty

521The parameter uncertainty corresponding to the propagation of data522errors into the parameter values is generally rather small, inducing523uncertainties in the order of hundreds of meters for the Moho depth, a524few km for LAB depth and a few kg/m3 for crustal density. Other param-525eters, however, linked to the starting model or inversion parameters526influence much more the uncertainty of modeling results. In order to527study the full uncertainty, we did more than 50 inversions with differ-528ent initial models (different parameters for the 1D inversion, different529average crustal densities (2830–2850 kg/m3) and a prioriMoho depths530varying within their uncertainty limits estimated at ±2 km as is typical531for refraction seismic models) and with different smoothing (from 0.1532to 0.3) parameter. We found that the parameter uncertainties,533calculated as standard deviation of the results of the different inversion534runs, are in the order of several km for the Moho and LAB depths, and53550–60 kg/m3 for average crustal densities (Fig. 5). The strongest536variability is thus concentrated in the crust, the LAB depth being rather537stable.

5385. Comparison of our lab depth with other geophysical proxies

539In contrast to the crust–mantle boundary, the LAB corresponds to a540rheological rather than to a compositional contrast, which explains541that its location and properties are more elusive (Eaton et al., 2009).542We therefore compiled the few previously published LAB data mainly543from the thermal, magnetotelluric, and seismological studies to test544the reliability of ourmodeling approach (Figs. 6a–b, 7b). Our LAB depths545in the Singhbhum Protocontinent are inconsistent with most of the546thermal, seismological, and magnetotelluric results. Intersection of547mantle solidus with the peridotite incipient melting curve from the548heat flow data reveals a highly deformed lithosphere having a thickness549of merely 66 km beneath the Singhbhum Craton (Fig. 6a) (Pandey and550Agrawal, 1999). The high heat flow (Fig. 1) can, however, be explained551on the basis of higher radioactive heat production of the diversified552crustal composition since the crust in this area contains important553minerals and pervasive granitoid batholiths (Rao et al., 1976; Rao and554Rao, 1983).555Electrical conductivity (or its inverse, resistivity) provides another556important constraint on mantle structure that is independent of other557geophysical techniques. The lithospheric mantle represents a relatively558resistive layer (1000–10,000 Ohm-m) beneath a (typically) conductive559lower crust. Typical values quoted for the resistivity of the electrical560asthenosphere are in the range of 5–25 Ohm-m (Eaton et al., 2009).561Early 1D inversion of magnetotelluric signal indicated that the electrical562LAB beneath the Singhbhum Craton varies from 58 to 76 km (Roy et al.,5631989). Recently, a deeper electrical LAB at about 95 km has been deter-564minedwhere resistivity falls from750to about 80Ohm-m. Interestingly,565the thin, 10–15 km thick relatively conductive layer at 95 km depth is566followed by a resistive (1000 Ohm-m) layer up to a depth of about567130 km. At this depth, the resistivity falls to about 200 Ohm-m and568remains the same up to 175 kmwhere it falls down further and touches569the typical asthenospheric resistivity of 25 Ohm-m (Fig. 6b). Their570modeling further shows that the anisotropic mantle is at a depth of

t2:1 Table 2t2:2 Fixed physical properties.

t2:3 Upper crust Lower crust Mantle

t2:4 Thermal conductivity [W/(K×m)] 2.5 2.2 3.3t2:5 Heat production [μW/m3] 1 0.2 0t2:6 Thermal expansion coef. [K−1] 0 0 3.5 × 10−5

t3:1 Table 3t3:2 Resulting standard deviation of data misfits for different models.

t3:3 Gravity (mGal) Geoid (m) Topography (m)

t3:4 Data 36.80 4.30 978.07t3:5 1D initial model calculated in 3D 20.63 0.86 2.37t3:6 1D model with modified Mohot3:7 (a priori information included)

17.37 0.99 232.28

t3:8 3D final model 7.61 0.19 37.50




UNCO

RRECTED P

RO

OF

571 about 175 kmwhere the phase-sensitive strike deviates from the direc-572 tion of the present-day Indian plate motion by a minimum of 430

573 (Shalivahan et al., 2014). The magnetotelluric experiments designed574 to delineate the electrical structure of the lithosphere beneath the575 Slave Craton had discovered a similar conductive region at depths of576 80–120 km (the Central Slave Mantle Conductor) that is spatially coin-577 cident with a geochemically defined ultra-depleted harzburgitic layer578 (Jones et al., 2001, 2003). Using the average magnetotelluric response579 for stations distributed throughout the Slave Craton showed the580 electrical LAB located on average at 260 km. It is quite likely that the581 electrical conductor delineated by Shalivahan et al. (2014) at the582 depth of about 95 km indicates such a first-order Singhbhum Craton583 Mantle Conductor and the average electrical LAB beneath the584 Singhbhum Craton lies apparently at greater depth.585 Results from surface wave tomography are mostly limited to penin-586 sular India partly because sufficient station data are not available for587 other areas. Studies using S–P wave receiver functions cover much of588 India and give estimates in the range of 70–140 km (P. Kumar et al.,589 2007, 2013). Incidentally, the base of the electrical conductor discussed590 above coincides with the seismic LAB depth (112 km) at Bokaro in591 Chhotanagpur Gneiss Complex (Fig. 7b) (Kumar et al., 2007). The592 converted wave techniques (P- and S-wave receiver functions)593 subjected to the data of Vishakhapatnam, Bhubaneswar, and Kolkata

594broad-band stations of the adjoining regions also reveal a lithospheric595thickness varying around 110 ± 2 km (P. Kumar et al., 2013). The59695 ± 4 km interface likely represents a boundary in composition,597melting, or anisotropy, the LAB is otherwise expected to be much598deeper in cratonic regions (Rychert and Shearer, 2009; Yuan and599Romanowicz, 2010). The refractory harzburgite and dunite layering is600a potential candidate that can explain the electrical and seismic mid-601lithospheric discontinuity at about 100 km depth observed within the602cratons (Rey et al., 2014). The thinned lithosphere observed beneath603the Dharwar Craton (Kumar et al., 2007) has already been interpreted604as a mid-lithospheric discontinuity by Bodin et al. (2013). The hypothe-605ses of lithospheric thinning to ~100 km in the Singhbhum Craton also606therefore seem untenable and together with the magnetotelluric607resistivity-depth curve our LAB estimates establish a minimum litho-608sphere thickness for the Singhbhum Protocontinent of at least 135 km.

6096. Geological signatures of lithospheric modification

610Results of isostatic analyses indicate that cold, thick lithosphere of611Archaean age be chemically depleted (Jordan, 1975; Boyd, 1989;612Foley, 2008) and preserved through geologic time (Doin et al., 1997).613However, earlier geophysical studies documented that the Singhbhum614Protocontinent experienced large-scale lithospheric modification

Fig. 4. a)West to east (220 N), and b) south to north (860 E) lithospheric cross-sections across the SinghbhumProtocontinent representing the topography and the depths of theMoho andthe lithosphere–asthenosphere boundary (LAB). Crossing point of the two profiles is indicatedwith vertical arrow. Principal structural units are as in Figs. 1 and 2, respectively. BB: BengalBasin, BC: Bastar Craton, BoB: Bay of Bengal, CGC: Chhotanagpur Gneiss Complex, GB: Ganga Basin, MB: Mahanadi Basin, MD: Mahanadi Delta, NSMB: Northern SinghbhumMobile Belt,and SC: Singhbhum Craton.




UNCO

RRECTED P

RO

OF

649649649649649649649649649649649649649649649649649649wherein more than 100 km of lithospheric mantle has been removed650(Pandey and Agrawal, 1999; P. Kumar et al., 2007, 2013; Bhattacharya651et al., 2009, 2013; Shalivahan and Bhattacharya, 2005; Shalivahan652et al., 2014). Relative to IASP91 a 5–10% reduction of S-wave velocity653in the upper most mantle further indicates a reworked (rejuvenated)654Archaean and early Proterozoic lithosphere in the Singhbhum Craton655(Kosarev et al., 2013). Our 3-D lithospheric density model of the656Singhbhum Protocontinent based on multiproxy geophysical data657shows a ~135 km thick lithosphere composed of ~38 km thick crust658and ~100 km thick subcrustal mantle (Fig. 3e). Though being thicker659than that inferred by earlier studies, our upper mantle model derived660here is still thinner than for many other Archaean cratons (Artemieva661and Mooney, 2001; Artemieva, 2006). A 135 ± 15 km thick thermal662lithosphere (defined as the intersection of the mantle geotherm with663the 1300 °C mantle adiabat) indicates reworked Archaean terranes664(Artemieva, 2006). It is therefore imperative to understand the possible665thermotectonic processes responsible for the lithospheric modification666of the Singhbhum Protocontinent.667As early as 1979, the relative positive gravity anomaly over the668eastern Indian shield has been attributed to thermal mantle processes669such as the Rajmahal Traps and a thinner lithosphere was proposed670(Kailasam, 1979). Other earlier geophysical studies, namely, thermal671(Pandey and Agrawal, 1999), magnetotelluric (Shalivahan and672Bhattacharya, 2005), and S-wave velocity (Kosarev et al., 2013) together673with the elastic thickness (Ratheesh-Kumar et al., 2014) results suggest674the likely occurrence of amid-Cretaceousmantle-plumehead (associat-675edwith the Rajmahal Traps) on the palaeocontinentalmargin of eastern676India (around the Singhbhum Protocontinent) that caused the thermal677modification of the lower lithosphere. It is also quite likely that circula-678tion changes in the asthenosphere due to the Paleogene–Miocene679Himalayan orogeny could have played a major role in delamination of680the lithospheric roots of the eastern Indian shield leading to a thinned681lithosphere (Shalivahan et al., 2014).682Highly buoyant aswell as the refractory Archaean lithosphere is very683stable and cannot be delaminated through gravitational forces alone684(Jordan, 1988) unless it is physically disrupted (e.g. rifting, thinning,685and displacement) with associated thermal and chemical erosion686(O’Reilly et al., 2001). At the same time, it is difficult to visualize a687plume-related mechanism that could cause selective depletion of688some incompatible elements like Ba and HFSE like Nb but preserving a689primitive mantle-like ΣHREE concentration (Roy et al., 2004). Similarly,690delamination of the lower continental crust cannot significantly reduce691the Nb/La ratio of the upper mantle compared to the bulk continental692crust (Taylor and McLennan, 1985). These, along with low Sr ratios693(~0.702) and ΣHREE ~2/3 times chondrite of eastern Indian craton694ultramafics (i.e., younger dolerite dyke swarms) rule out the possibility695of both delamination or plumemagmatism as a cause of modification of696lithospheric mantle below the eastern Indian craton (Roy et al., 2004).697Refertilization of ancient SCLM may be an alternative explanation for698the asthenospherization of lower parts of the SCLM (Tang et al.,6992013a; Foley, 2008).700Changes tracked in the SCLM in several regions, such as the701Wyoming Craton (Lee et al., 2000) and the North China Craton (Zhu702et al., 2012), show that Archaean mantle can be transformed by703thermo-mechanical destruction (lithospheric thinning and rifting) and704refertilized (chemical re-enrichment) by episodic infiltration of upwell-705ing fertilematerial (O'Reilly et al., 2001; Foley, 2008; Zhang et al., 2009).706The refertilization via melt/fluid infiltration will change the geophysical707property of ancient SCLMand lead to the destabilization anddestruction708of the SCLM (Griffin et al., 2003; Windley et al., 2010). The North China709Craton is oneof themost typical cases for nearly complete destruction of710its Archaean keel through refertilization (Tang et al., 2013a,b and

Fig. 5. a–c) Uncertainty estimate of model parameters for Moho depth (a), LAB depth (b),and the average crustal density (c). The pixel size corresponds to the size of the invertedblocks of 30 × 30 km. Roman numbers are as in Fig. 2.




UNCO

RRECTED P

RO

OF

711 references therein). The Archaean–Proterozoic component of the litho-712 spheric mantle beneath the Singhbhum Protocontinent shows first713 signs of metasomatic alteration by hot, fertile juvenile melts during714 late Archaeantimes (2.6 Ga) (Roy et al., 2004). Primitive-mantle-715 normalized trace element plots of ultramafic (harzburgite and716 lherzolite) dykes belonging to 1600–950 Ma newer dolerite dyke717 swarms display enrichment in large ion lithophile elements (LILE). Iso-718 topic, trace, and REE data together indicate that during 2.6 Ga, the near-719 ly primitivemantle below the eastern Indian shield wasmetasomatized720 by a fluid (± silicate melt) coming out from subducting early crust721 resulting in LILE and LREE enriched, Nb depleted, variable εNd, low Sri722 (0.702) and low δ18O bearing EMI-type mantle (Roy et al., 2004).

723Recurrent manifestations of temporally distinct depleted magma724types from the Archaean Older Metamorphic Group to Proterozoic725Newer dolerite intrusions in the Singhbhum Craton were noted to be726“remarkable” (Bose, 2009). The observed secular evolution in SCLM, at727least through Proterozoic time,may reflect the progressivemodification728of relict, buoyant Archaean lithosphere. This is additionally supported729by the cratonic geochemical signatures, depicted by ultrapotassic730rocks from Gondwana coal fields, which were emplaced during early731Cretaceous (Srivastava et al., 2009). Likewise, the trace element geo-732chemical and isotopic signatures of the Rajmahal Traps and Bokaro733dykes from the Chhotanagpur Gneiss Complex are shown to be the734products of a “lost” Indian lithosphere (Ghatak and Basu, 2013).

Fig. 6. (a) Thermal (Pandey andAgrawal, 1999) and (b) electrical (Shalivahan et al., 2014) lithosphericmodels available in the SinghbhumCraton are given for comparisonwith ourmodel.

Fig. 7. Comparison of our (a) crustal and (b) lithospheric models with the available seismic model (P. Kumar et al., 2013). Abbreviations are as in Fig. 1.




UNCO

RRECTED P

RO

OF

735 Therefore, it is likely that the late Archaean to early Cretaceous episodic736 thermo-mechanical destruction and chemical re-enrichment could737 have played a major role in the rejuvenation and modification of the738 lithospheric roots of the Singhbhum Protocontinent leading to a rela-739 tively thin LAB. The discovery of 65 Ma diamondiferous kimberlites in740 the Bastar Craton, synchronous with the Deccan flood basalts, however,741 demonstrates that lithospheric roots of the Bastar Cratonwere intact, at742 least partially, till 65 Ma (Chalapathi Rao et al., 2011). Distinct litho-743 spheric low-density anomaly imaged between 35 and 120 km depth744 in the Eastern Ghats Mobile Belt together with the shallow crust–745 mantle transition and significant Moho offset along the junction with746 Bastar Craton (Kumar et al., 2004) indicates continental collision and747 delamination resulting from gravity instability as a possible factor for748 the lithospheric destruction in this area.

749 7. Conclusions

750 Our integrated approach constrains the upper mantle structures be-751 neath the Singhbhum Protocontinent by using the free-air anomaly, to-752 pography, and geoid data. Results from gravity inversions, under the753 assumption of local isostatic equilibrium indicate widespread lateral754 density heterogeneities in the crust. More than 40 km deep Moho755 under the Ganga foreland basin flattens at 35–40 km in Singhbhum756 Protocontinent and Bastar Craton. It further undulates between 36 and757 32 km along the coast and finally reaches 24 km in the Bay of Bengal.758 The study provides a better LAB structure of the study region which759 deepens from ~100 km beneath the northern part of Bay of Bengal to760 a little over 130 km south of continental India. This is broadly consistent761 with the available seismic disposition and lends credence to the hypoth-762 esis of a continental like lithospheric structure of the Bay of Bengal. Our763 model further shows that significantly thinned lithosphere (~135 km)764 is widespread in the Singhbhum Protocontinent, while in Bastar Craton765 a relatively thicker lithosphere (mostly N150 km) still exists. LAB depth766 of more than 200 km in Ganga foreland basin indicates that the litho-767 spheric thickening is not restricted but extends hundreds of kilometers768 away from the Indo-Eurasian collision front. Relative lithospheric thin-769 ning in the Singhbhum Protocontinent, predicted by our model, is likely770 to be associated with, or reflection of, the thermo-chemical erosion re-771 lated to the widespread refertilization soon after the late Archaean772 time till at least the early Cretaceous.

773 Acknowledgements

774 The authors thank the Director, CSIR-National Geophysical Research775 Institute, Hyderabad for his encouragement and support for this study.776 Thanks are also due to Dr.M. Radhakrishna and an anonymous reviewer777 for their critical comments and suggestions that have helped greatly to778 improve the quality of the manuscript. The study was performed as a779 part of INDEX project of CSIR-National Geophysical Research Institute.780 Some of the figures are generated with the Generic Mapping Tools781 software (Wessel and Smith, 1995).

782 References

783 Abdelsalam, M.G., Gao, S.G., Liégeois, J.-P., 2011. Upper mantle structure of the Saharan784 Metacraton. J. Afr. Earth Sci. 60, 328–336.785 Artemieva, I.M., 2006. Global 1°×1° thermal model TC1 for the continental lithosphere:786 implications for lithosphere secular evolution. Tectonophysics 416, 245–277.787 Artemieva, I.M., Mooney, W.D., 2001. Thermal thickness and evolution of Precambrian788 lithosphere: a global study. J. Geophys. Res. 106 (B8), 16,387–16,414.789 Banerjee, A.K., 1981. Ore genesis and its relationship to volcanism, tectonism, granitic790 activity, and metasomatism along Singhbhum shear zone, eastern India. Econ. Geol.791 76, 905–912.792 Beck, S.L., Zandt, G., 2002. The nature of orogenic crust in central Andes. J. Geophys. Res.793 107 (B10), 2230. http://dx.doi.org/10.1029/2000JB000124.794 Behera, L., Sain, K., Reddy, P.R., 2011. Evidence of underplating from seismic and gravity795 studies in the Mahanadi delta of eastern India and its tectonic significance.796 J. Geophys. Res. 109, B12311. http://dx.doi.org/10.1029/2003JB002764.

797Bhattacharya, B.B., Shalivahan, 2002. The electric Moho underneath Eastern Indian798Craton. Geophys. Res. Lett. 29 (10), 1376. http://dx.doi.org/10.1029/2001GL014062.799Bhattacharya, S.N., Suresh, G., Mitra, S., 2009. Lithospheric S-wave velocity structure of the800Bastar craton, Indian Peninsula from surface wave phase-velocity measurements.801Bull. Seismol. Soc. Am. 99, 2502–2508.802Bhattacharya, S.N., Mitra, S., Suresh, G., 2013. The shear wave velocity of the upper mantle803beneath the Bay of Bengal, Northeast Indian Ocean from inter-station phase velocities804of surface waves. Geophys. J. Int. 193, 1506–1514. http://dx.doi.org/10.1093/gji/ggt007.805Bodin, T., Yuan, H., Romanowicz, B., 2013. Inversion of receiver functions without806deconvolution—application to the Indian craton. Geophys. J. Int. 196, 1025–1033.807Bose, M.K., 1994. Sedimentation pattern and tectonic evolution of the Proterozoic808Singhbhum basin in the eastern Indian shield. Tectonophysics 231, 325–346.809Bose, M.K., 2009. Precambrian mafic magmatism in the Singhbhum craton, Eastern India.810J. Geol. Soc. India 73, 13–135.811Bowin, C., 1983. Depth of principal mass anomalies contributing to the Earth's geoidal812undulations and gravity anomalies. Mar. Geod. 7 (1-4), 61–100.813Bowin, C., 1991. The Earth’s gravity field and plate tectonics. Tectonophysics 187, 69–89.814Boyd, F.R., 1989. Compositional distinction between oceanic and cratonic lithosphere.815Earth Planet. Sci. Lett. 96, 15–26.816Chalapathi Rao, N.V., Lehmann, B., Mainkar, D., Belyatsky, B., 2011. Petrogenesis of the817end-Cretaceous diamondiferous Behradih orangeite pipe: implication for mantle818plume-lithosphere interaction in the Bastar craton, Central India. Contrib. Mineral.819Petrol. 161, 721–742.820Chalapathi Rao, N.V., Sinha, A.K., Kumar, S., Srivastava, R.K., 2013. K-rich Titanate from the821Jharia Ultrapotassic rocks, Gondwana coal fields, eastern India, and its petrological822significance. J. Geol. Soc. India 81, 733–736.823Chalapathi Rao, N.V., Srivastava, R.K., Sinha, A.K., Ravikant, V., 2014. Petrogenesis of824Kergulean mantle plume-linked early Cretaceous ultrapotassic intrusive rocks from825the Gondwana sedimentary basins, Damodar valley, Eastern India. Earth-Sci. Rev.826136, 96–120.827Christou, N.T., Vanicek, P., Ware, C., 1989. Geoid and density anomalies. EOS Trans. Am.828Geophys. Union 70, 625–631.829Doin, M.-P., Fleitout, L., Christensen, U., 1997. Mantle convection and stability of depleted830and undepleted continental lithosphere. J. Geophys. Res. 102, 2771–2787.831Eaton, D.W., Darbyshire, F., Evans, R.L., Grütter, H., Jones, A.G., Yuan, X., 2009. The elusive832lithosphere–asthenosphere boundary (LAB) beneath cratons. Lithos 109, 1–22.833Featherstone, W.E., 1997. On the use of the geoid in geophysics: a case study over the834north-west shelf of Australia. Explor. Geophys. 28 (1), 52–57.835Foley, S.F., 2008. Rejuvenation and erosion of the cratonic lithosphere. Nat. Geosci. 1,836503–510.837French, J.E., Heaman, L.M., Chacko, T., Srivastava, R.K., 2008. 1891-1883 Ma southern838Bastar-Cuddapah mafic igneous events, India: a newly recognized large igneous839province. Precambrian Res. 160, 308–322.840Fullea, J., Fernàndez, M., Zeyen, H., Vergés, J., 2007. A rapid method to map the crustal and841lithospheric thickness using elevation, geoid anomaly and thermal analysis. Applica-842tion to the Gibraltar Arc System, Atlas Mountains and adjacent zones. Tectonophysics843430, 97–117.844Fullea, J., Afonso, J.C., Connolly, J.A.D., Fernàndez, M., Garcı´a-Castellanos, D., Zeyen, H.,8452009. LitMod3D: an interactive 3-D software to model the thermal, compositional,846density, seismological, and rheological structure of the lithosphere and847sublithospheric upper mantle. Geochem. Geophys. Geosyst. 10, Q08019. http://dx.848doi.org/10.1029/2009GC002391.849Gallardo-Delgado, L.A., Pérez-Flores, M.A., Gómez-Treviño, E., 2003. A versatile algorithm850for joint 3D inversion of gravity and magnetic data. Geophysics 68, 949–959.851Ghatak, A., Basu, A.R., 2013. Isotopic and trace element geochemistry of alkali-mafic-852ultramafic-carbonatitic complexes and flood basalts in NE India: origin in a heteroge-853neous Kerguelen plume. Geochim. Cosmochim. Acta 115, 46–72.854Griffin, W.L., O'Reilly, S.Y., Abe, N., Aulbach, S., Davies, R.M., Pearson, N.J., Doyle, B.J., Kivi,855K., 2003. The origin and evolution of Archaean lithospheric mantle. Precambrian856Res. 127, 19–41.857Jiménez-Munt, I., Fernàndez, M., Vergés, J., Platt, J.P., 2008. Lithosphere structure under-858neath the Tibetan Plateau inferred from elevation, gravity and geoid anomalies.859Earth Planet. Sci. Lett. 267, 276–289. http://dx.doi.org/10.1016/j.epsl.2007.11.045.860Jones, A.G., Ferguson, I.A., Chave, A.D., Evans, R.L., McNeice, G.W., 2001. Electric861lithosphere of the Slave craton. Geology 29, 423–426.862Jones, A.G., Lezeata, P., Ferguson, I.A., Chave, A.D., Evans, R.L., Garcia, X., Spratt, J., 2003. The863electrical structure of the Slave craton. Lithos 71, 505–527.864Jordan, T.H., 1975. The continental tectosphere. Rev. Geophys. Space Phys. 13, 1–12.865Jordan, T.H., 1988. Structure and formation of the continental lithosphere. In: Menzies,866M.A., Cox, K. (Eds.), Oceanic and Continental Lithosphere: Similarities and867Differences. J. Petrology Special lithosphere issue, pp. 11–37.868Kailasam, L.N., 1979. Plateau uplift in peninsular India. Tectonophysics 61, 243–269.869Kayal, J.R., Srivastava, V.K., Kumar, P., Chatterjee, R., Khan, P.K., 2011. Evolution of crustal870and upper mantle structures using receiver function analysis: ISM broad band871observatory data. J. Geol. Soc. India 78, 76–80.872Kosarev, G.L., Oreshin, S.I., Vinnik, L.P., Kiselev, S.G., Dattatrayam, R.S., Suresh, G., Baidya,873P.R., 2013. Heterogeneous lithosphere and the underlying mantle of the Indian874subcontinent. Tectonophysics 592, 175–186.875Kumar, Anil, Dayal, A.M., Padmakumari, V.M., 2003. Kimberlite from Rajmahal magmatic876province: Sr-Nd-Pb isotopic evidence for Kerguelen plume derived magmas.877Geophys. Res. Lett. 30 (20), 2053. http://dx.doi.org/10.1029/2003GL018462.878Kumar, Niraj, Singh, A.P., Gupta, S.B., Mishra, D.C., 2004. Gravity signature, crustal879architecture and collision tectonics of Eastern Ghats Mobile Belt. J. Indian Geophys.880Union 8, 97–106.881Kumar, Prakash, Yuan, X., Ravi Kumar, M., Kind, R., Li, X., Chadha, R.K., 2007. The rapid882drift of the Indian tectonic plate. Nature 449. http://dx.doi.org/10.1038/nature06214.



http://dx.doi.org/10.1029/2000JB000124

http://dx.doi.org/10.1029/2003JB002764

http://dx.doi.org/10.1029/2001GL014062

http://dx.doi.org/10.1093/gji/ggt007

http://dx.doi.org/10.1029/2009GC002391

http://dx.doi.org/10.1016/j.epsl.2007.11.045

http://dx.doi.org/10.1029/2003GL018462

http://dx.doi.org/10.1038/nature06214


UNCO

RRECTED P

RO

OF

883 Kumar, Niraj, Singh, A.P., Singh, B., 2011. Insights into the crustal structure and884 geodynamic evolution of the Southern Granulite Terrain of India from isostatic885 considerations. Pure Appl. Geophys. 168 (10), 1781–1798.886 Kumar, Niraj, Zeyen, H., Singh, A.P., Singh, B., 2013a. Lithospheric structure of Southern887 Indian shield and adjoining oceans: integrated modelling of topography, gravity,888 geoid and heat flow data. Geophys. J. Int. 194 (1), 30–44.889 Kumar, Prakash, Ravi Kumar, M., Srijayanthi, G., Arora, K., Srinagesh, D., Chadha, R.K., Sen,890 M.K., 2013b. Imaging the lithosphere-asthenosphere boundary of the Indian plate891 using converted wave technique. J. Geophys. Res. 118, 5307–5319. http://dx.doi.892 org/10.1002/jgrb.50366.893 Kumar, Niraj, Zeyen, H., Singh, A.P., 2014. 3D lithosphere density structure of Southern894 Indian Shield from joint inversion of gravity, geoid and topography data. J. Asian895 Earth Sci. 89, 98–107.896 Lachenbruch, A.H., Morgan, P., 1990. Continental extension, magnetism and elevation;897 formal relations and rules of thumb. Tectonophysics 174, 39–64.898 Laske, G., Masters, G., Ma, Z., Pasyanos, M.E., 2012. CRUST1.0: An updated Global Model of899 Earth's Crust. EGU General Assembly, held 22-27 April, 2012 in Vienna, Austria,900 p. 3743.901 Lee, C.-T., Yin, Q., Rudnick, R.L., Chesley, J.T., Jacobsen, S.B., 2000. Osmium isotopic evi-902 dence for Mesozoic removal of lithospheric mantle beneath the Sierra Nevada, Cali-903 fornia. Science 289, 1912–1916.904Q3 Lee, C.-T., A., Luffi, P., Chin, E.J., 2011. Building and destroying continental mantle. Annu.905 Rev. Earth. Planet. Sci. 39, 59–90.906 Lehmann, B., Burgess, R., Frei, D., Belyatsky, B., Mainkar, D., Chalapathi Rao, N.V., Heaman,907 L.M., 2010. Diamondiferous kimberlites synchronouswith the Deccan Flood basalts in908 central India. Earth Planet. Sci. Lett. 290, 142–149.909 Mall, D.M., Rao, V.K., Reddy, P.R., 1999. Deep sub-crustal features in the Bengal basin: seis-910 mic signatures for plume activity. Geophys. Res. Lett. 26, 2545–2548.911 Mazumder, R., 2005. Proterozoic sedimentation and volcanism in the Singhbhum crustal912 province, India and their implications. Sediment. Geol. 176, 167–193.913 Meert, J.G., Pandit, M.K., Pradhan, V.R., Banks, J., Sirianni, R., Stroud, M., Newstead, B.,914 Gifford, J., 2010. Precambrian crustal evolution of Peninsular India: a 3.0 billion year915 odyssey. J. Asian Earth Sci. 39, 483–515.916 Mitchell, R.H., Fareeduddin, 2009. Mineralogy of the peralkaline lamproites from the917 Raniganj coalfield, India. Mineral. Mag. 73, 457–477.918 Mitra, S., Priestley, K., Acton, C., Gaur, V.K., 2011. Anomalous surface wave dispersion and919 the enigma of “continental-like” structure for the Bay of Bengal. J. Asian Earth Sci. 42,920 1243–1255.921 Motavalli-Anbaran, S.-H., Zeyen, H., Ebrahimzadeh Ardestani, V., 2013. 3D joint inversion922 modeling of the lithospheric density structure based on gravity, geoid and topogra-923 phy data—application to the Alborz Mountains (Iran) and South Caspian Basin region.924 Tectonophysics 586, 192–205.925 O'Reilly, S.Y., Griffin, W.L., Poudjom Djomani, Y., Morgan, P., 2001. Are lithospheres forev-926 er? Tracking changes in subcontinental lithspheric mantle through time. GSA Today927 11, 4–9.928 Pandey, O.P., Agrawal, P.K., 1999. Lithospheric mantle deformation beneath the Indian929 cratons. J. Geol. 107, 683–692.930 Pavlis, N.K., Holmes, S.A., Kenyon, S.C., Factor, J.K., 2012. The development and evaluation931 of the Earth Gravitational Model 2008 (EGM2008). J. Geophys. Res. 117, B04406.932 http://dx.doi.org/10.1029/2011JB008916.933 Priestley, K., McKenzie, D., 2006. The thermal structure of the lithosphere from shear ve-934 locities. Earth Planet. Sci. Lett. 244, 285–301.935 Radhakrishna, M., Subrahmanyam, C., Damodharan, T., 2010. Thin oceanic crust below936 Bay of Bengal inferred from 3-D gravity interpretation. Tectonophysics 493, 93–105.937 Rao, G.V., Rao, R.U.M., 1983. Heat flow in Indian Gondwana basins and heat production of938 their basement rocks. Tectonophysics 91, 105–117.939 Rao, R.U.M., Rao, G.V., Narain, Hari, 1976. Radioactive heat generation and heat flow in the940 Indian shield. Earth Planet. Sci. Lett. 30, 57–64.941 Ratheesh-Kumar, R.T., Windley, B.F., Sajeev, K., 2014. Tectonic inheritance of the Indian942 shield: new insights from its elastic thickness structure. Tectonophysics 615–616,943 40–52.944 Ravi Kumar, M., Saul, J., Sarkar, D., Kind, R., Shukla, A.K., 2001. Crustal structure of the945 Indian Shield: new constraints from teleseismic receiver functions. Geophys. Res.946 Lett. 28 (7), 1339–1342.947 Rekha, S., Upadhyay, D., Bhattacharya, A., Kooijman, E., Goon, S., Mahato, S., Pant, N.C.,948 2011. Lithostructural and chronological constraints for tectonic restoration of Prote-949 rozoic accretion in the Eastern Indian Precambrian shield. Precambrian Res. 187,950 313–333.951 Rey, P.F., Coltice, N., Flament, N., 2014. Spreading continents kick-started plate tectonics.952 Nature 513, 405–408.

953Q4Robert, A.M.M., Fernàndez, M., Jiménez-Munt, I., Vergés, J., 2015s. Lithospheric structure954in Central Eurasia derived from elevation, geoid anomaly and thermal analysis.955Geol. Soc. London Spec. Publ. (in Press). Q5956Root, B.C., van der Wal, W., Novák, P., Ebbing, J., Vermeersen, L.L.A., 2015. Glacial isostatic957adjustment in the static gravity field of Fennoscandia. J. Geophys. Res. Solid Earth 120,958503–518. http://dx.doi.org/10.1002/2014JB011508.959Roy, A.B., Bhattacharya, H.N., 2012. Tectonostratigraphic and geochronologic reappraisal960constraining the growth and evolution of Singhbhum Archaean craton, Eastern961India. J. Geol. Soc. India 80, 455–469.962Roy, K.K., Rao, C.K., Chattopadhyay, A., 1989. Magnetotelluric survey across Singhbhum963granite batholith. Proc. Indian Acad. Sci. (Earth Planet. Sci.) 98, 147–162.964Roy, A., Sarkar, A., Jeyakumar, S., Aggrawal, S.K., Ebihara, M., Satoh, H., 2004. Late Archaean965mantle metasomatism below eastern Indian craton: evidence from trace elements,966REE geochemistry and Sr-Nd-O isotope systematics of ultramafic dykes. Proc.967Indian Acad. Sci. (Earth Planet. Sci.) 113 (4), 649–665.968Rychert, C.A., Shearer, P.M., 2009. A global view of the lithosphere–asthenosphere bound-969ary. Science 324, 495–498.970Saha, A.K., 1994. Crustal evolution of Singhbhum, north Orissa, Eastern India. Geol. Soc.971India Memoir 27. Geological Soc. of India publication, p. 341.972Sandwell, D.T., Smith, W.H.F., 1997. Marine gravity from Geosat and ERS 1 satellite altim-973etry. J. Geophys. Res. 102, 10,039–10,054.974Sarkar, A.N., 1982. Precambrian tectonic evolution of Eastern India: a model of converging975microplates. Tectonophysics 86, 363–397.976Shalivahan, Bhattacharya, B.B., 2005. Electrical anisotropy of asthenosphere in a region of977window to mantle underneath Eastern Indian Craton. Phys. Earth Planet. Inter. 152,97843–61.979Shalivahan, Bhattacharya, B.B., Chalapathi Rao, N.V., Maurya, V.P., 2014. Thin lithosphere-980asthenosphere boundary beneath Eastern Indian craton. Tectonophysics 612-613,981128–133. http://dx.doi.org/10.1016/j.tecto.2013.11.036.982Sharma, R.S., 2009. Cratons and Fold Belts of India. Lecture Notes in Earth Sciences 127.983Springer-Verlag, p. 304. http://dx.doi.org/10.1007/978-3-642-01459-8_2.984Singh, A.P., Kumar, Niraj, Singh, B., 2004. Magmatic underplating beneath the Rajmahal985traps: gravity signatures and derived 3-D configuration. Proc. Indian Acad. Sci.986(Earth Planet. Sci.) 113 (4), 759–769 (Presently J. Earth System Sci.).987Srivastava, R.K., Chalapathi Rao, N.V., Sinha, A.K., 2009. Cretaceous potassic intrusives with988affinities from Jharia area: magmatic expression of metasomatically veined and989thinned lithospheric mantle beneath Singhbhum craton, Eastern India. Lithos 112,990407–418.991Tang, Y.-J., Zhang, H.-F., Ying, J.-F., Su, B.-Z., 2013a. Widespread refertilization of cratonic992and circum-cratonic lithospheric mantle. Earth Sci. Rev. 118, 45–68.993Tang, Y.-J., Zhang, H.-F., Santosh, M., Ying, J.-F., 2013b. Differential destruction of the North994China craton: a tectonic perspective. J. Asian Earth Sci. 78, 71–82.995Taylor, S.R., McLennan, S.M., 1985. The continental crust: its composition and evolution.996Blackwell, Oxford (312 pp.).997Tiwari, S., Jassal, G.S., 2003. Origin and evolution of the Garo-Rajmahal Gap. J. Geol. Soc.998India 57, 389–403.999Tiwari, V.M., Rao, M.B.S.V., Mishra, D.C., Singh, B., 2006. Crustal structure across Sikkim, NE1000Himalaya from new gravity and magnetic data. Earth Planet. Sci. Lett. 247, 61–69.1001Turcotte, D., Schubert, G., 1982. Geodynamics: Applications of Continuum Physics to1002Geological Problems. John Wiley & Sons, New York 0-471-06018-6 (450 pp.).1003Wang, X., Fang, J., Hsu, H., 2014. Three-dimensional density structure of the lithosphere1004beneath the North China craton and the mechanisms of its destruction.1005Tectonophysics 610, 150–158.1006Wessel, P., Smith, W.H.F., 1995. New version of the Generic Mapping Tools released. EOS1007Trans. Am. Geophys. Union 76, 329.1008Windley, B.F., Maruyama, S., Xiao, W.J., 2010. Delamination/thinning of sub-continental1009lithospheric mantle under Eastern China: the role of water and multiple subduction.1010Am. J. Sci. 310, 1250–1293.1011Yuan, H., Romanowicz, B., 2010. Lithospheric layering in the North American craton.1012Nature 466, 1063–1068. http://dx.doi.org/10.1038/nature09332.1013Zhang, H.F., Goldstein, S.L., Zhou, X.H., Sun, M., Cai, Y., 2009. Comprehensive refertilization1014of lithospheric mantle beneath the North China Craton: further Os–Sr–Nd isotopic1015constraints. J. Geol. Soc. Lond. 166, 249–259.1016Zhu, R.-X., Yang, J.-H., Wu, F.-Y., 2012. Timing of destruction of the North China Craton.1017Lithos 149, 51–60.1018Zoback, M.L., Mooney, W.D., 2003. Lithospheric buoyancy and continental intraplate1019stresses. Int. Geol. Rev. 45, 95–118.

1020



http://dx.doi.org/10.1002/jgrb.50366

http://dx.doi.org/10.1029/2011JB008916

http://dx.doi.org/10.1002/2014JB011508


http://dx.doi.org/10.1007/978-3-642-01459-8_2

http://dx.doi.org/10.1038/nature09332


Documents

Three-dimensional lithospheric mapping of the eastern Indian shield: A multi-parametric inversion approach