East Antarctica is considered as a stable, cratonic block with a Precambrian basement and has prominently figured in Proterozoic configurations of supercontinents including Columbia, Rodinia and Gondwana (e.g., Dalziel 1991; Borg & DePaolo 1994; Rogers et al. 1995; Rogers & Santosh 2004, 2009; Meert & Lieberman 2008; Satish-Kumar et al. 2008; Santosh, Maruyama & Sato 2009; Santosh, Maruyama & Yamamato 2009; Boger 2011). Central Dronning Maud Land in East Antarctica comprises mainly high-grade metamorphic rocks intruded by voluminous igneous bodies that form coastal and inland mountainous outcrops (Dallmann et al. 1990). The Neoproterozoic evolutionary history of this region is marked by two major tectonothermal events at about 1100 million years ago (Mya) and between 560 and 490 Mya (Jacobs et al. 1999; Jacobs et al. 2003). The younger event has been correlated to the East African Orogeny and involves an early collisional event at approximately 560 Mya, followed by large-scale extension associated with voluminous granitic magmatism. The ca. 630 My ages obtained from the coastal outcrop at Schirmacher Oasis have led to the suggestion of a different evolutionary history for this area in the late Neoproterozoic compared with that of the inland mountains (see Satish-Kumar et al. 2008 and references therein).

Schirmacher Oasis is located near the Princess Astrid Coast, central Dronning Maud Land, East Antarctica (Fig. 1). This region (70°44'–70°47'S and 11°25'–11°55'E) runs parallel to the east–west trending coastline between the inland ice and the shelf ice (Tingey 1991). The coastline lies at a distance of about 100 km north of Schirmacher Oasis. The Precambrian crystalline basement of Schirmacher Oasis forms part of the East Antarctic shield with granulite facies rocks occurring as remnants within a dominantly amphibolite-facies terrane (Ravikant et al. 2004). Metapelites, exposed in central and eastern Schirmacher Oasis and the nunatak Vettiyya, contain layers and boudins of metanorite dykes. Veins of orthopyroxene-bearing pegmatite intrude both metanorite layers and metapelite. In central Schirmacher Oasis, the major rock exposed is a foliated metanorite. This mafic unit, in turn, contains enclaves of melanocratic metagabbro, metapyroxenite, and spinel-bearing and rarely garnet-bearing metawebsterite. Structurally, the main deformation was associated with the amphibolite facies metamorphism accompanied by migmatization, isoclinal folding and formation of major east–west striking overthrusts (Kampf & Stackebranwd 1985; Sengupta & Bose 1997). From a comparison of ages and geological features, Ravikant et al. (2004) proposed that Schirmacher Oasis is a possible segment of a klippen of the Lurio Belt of south-east Mozambique and a possible extension of the East African Orogen into Antarctica. Satish-Kumar et al. (2008) considered Schirmacher Oasis as an extension of the Mozambique Belt in south-eastern Africa, with the inland mountains representing part of the crust generated during the final amalgamation of East Gondwana.

Voluminous mid-Jurassic magmatic provinces characterize the Western Antarctic part of Gondwana with a probable correlation to major plume activity (e.g., Storey 1995). The influence of a modern mantle plume in Antarctica has also been invoked in various studies (e.g., Wannamaker et al. 2004) confined to the polar and western part of Antarctica. Older plume-related mafic magmatism has been widely recorded in Antarctica and correlated with the spreading of the Karoo-Maud plume (e.g., Sushchevskaya et al. 2008) which triggered the final break-up of the Gondwana supercontinent and the formation of the Indian Ocean, to the east—from Queen Maud land towards Schirmacher Oasis. Although the geochemical and isotopic characteristics of the dolerite dyke in Schirmacher Oasis suggest crustal contamination during plume upwelling and emplacement of the mafic magmas into the continental crust, it is not reflected in crustal electrical structure delineated from this study. This might suggest that the plume activity might be very old or the thermal signatures might have calmed down quickly in the polar region. Numerous dykes of lamprophyre and basalt dissect the metamorphic complex of Schirmacher Oasis. Kaiser & Wand (1985) and Wand et al. (1988) distinguished two age groups of basalts; Palaeozoic and Mesozoic. Conventional K-Ar isotope determinations yielded ages of 223 and 354 My, respectively.

An anomaly measured by Magnetic Field Satellite over Schirmacher Oasis indicates a low as compared to its northern and southern sides, implying a region of weakly magnetized crust (Wagner & Lindner 1991). The nunataks close and around Schirmacher Oasis have a similar lithological set up and are probably extensions of Schirmacher Oasis (Mukerji et al. 1988). The general trend of geological features in the region is ENE–WSW with steep southern to sub-vertical dips. In the former, S-planes defined by foliation, strike towards nearly east–west direction, the attitude varying between WNW–ESE and WSW–ENE. Foliation planes dipping at 35–50° either northwards or southwards (Mukerji et al. 1988) correspond to the east–west strike. The influence of the plume in western Dronning Maud Land is stronger than in central Dronning Maud Land (Schirmacher Oasis). Schirmacher Oasis probably was in a peripheral position in the mantle-plume region, and the western Dronning Maud Land was located close to its axis. Apatite fission-track data and the high-elevation margin morphology of central Dronning Maud Land suggest a typical passive nature for the region (Meier 1999).

Geophysical data are modelled and interpreted in terms of subsurface geology in two ways: a direct approach such as forward modelling and an indirect method of modelling through inversion. Additional constraints (Parker 1994) are needed to resolve the sharp structures and often the assumed discrete layered models are fitted in the least square way (e.g., Petrick et al. 1977). In the inversion here, a model of the subsurface is assumed as a combination of a fine grid mesh of suitable and appropriate size and the theoretical geophysical response is calculated. This is compared with the observed data and the inversion process is repeated until a minimum difference between the computed and observed response is achieved. In this study, the apparent resistivity and phase data for both TE and TM modes have been inverted. The vertical (Hz) component is, however, not considered here as reliable estimates because the component could not be obtained for all the sites due to noise caused by a strong breeze across the hard terrain.

A data set of six soundings (MT sites) has been considered here for two-dimensional modelling of the east–west oriented profile running for about 16 km. MT site 6 falls towards the eastern end of Schirmacher Oasis, close to the Russian station, Novolazareyskaya. MT site 1 is located on a nunatak on the western side. The electrical strike for the sites along the profile has been computed using the McNiece & Jones (2001) algorithm. A strike of ca. 15°E has been obtained for sites 2–6, as shown in Fig. 5 in the form of a Rose diagrams (individual and combined sites) for Schirmacher Oasis region with an ambiguity of 90^o. Interestingly, site 1, which is located on a nunatak and lying in the western part of Schirmacher Oasis, shows a different strike (ca. 70°E). Distortion in the observed MT data due to conductive inhomogeneities needs to be estimated and corrected. Also, the dimensionality needs to be understood before the modelling procedure. Following the analysis of Bibby et al. (2005), the dimensionality in this particular case is more of a two-dimensional nature as observed in Fig. 6. The corrections for the distortions have been applied following Becken & Burkhardt (2004) and the corrected data set subjected to inversion for arriving at the geoelectrical structure of the region.

The yx-component has been considered as the TE mode in the inversion, as the profile direction and geological strike are same. The MT data were inverted using the nonlinear conjugate gradients scheme implemented using the version 2.20 of the WinGlink software package (Rodi & Mackie 2001) to get a two-dimensional resistivity section. An initial model with half-space resistivity of 100 ohm m is considered and the model subjected to 200 iterations. Both TM and TE data with five decades (0.01–1000 Hz) are considered here with an error floor of 20% for resistivity and 5% for phase. A tau (τ) of 1, which indicates smoothness for the model, has been used here for the initial inversion. The obtained model is then studied for different parameters of τ, starting from 1 to 1000. Variation of the root mean square error (RMSE) with roughness for the model is plotted and an optimum value of 8 for τ is obtained from this L curve (Fig. 7a) following Hansen's (1998) procedure. The variation in the root mean square with iterations is given in Fig. 7b, which shows that the RMSE attains a steady value after the 40th iteration. A final inversion using a τ of 8 was carried for 200 iterations and an rms misfit of 2.528 was obtained. The rms values at different locations are shown in Fig. 8a. The geo-electric section for the profile is shown in Fig. 8(b).

This MT study has revealed a well-defined layer with an average thickness of 15 km and a highly resistive (8000–10 000 ohm m) upper crust below sites 2–5 along the profile (Fig. 8b). The thickness of this layer varies over the length of the profile, with 20 km on the eastern side of Maitri and a thinning (ca. 10 km) towards the western side. A less resistive (500–600 ohm m) lower crust has been delineated all along the profile. The same layer approaches shallower depths (ca. 6–8 km) between sites 1 and 2.

The fit between the observed and computed data for this profile in the form of apparent resistivity and phase pseudo-sections is presented in Fig. 9a and b for TE and TM modes, respectively. A reasonably good fit is observed for all sites. The conductive features in the geoelectrical section are well resolved as shown in the sensitivity map (Fig. 10) for the profile. As the strike at site 1 is different from the other five sites, the inversion has been attempted with and without site 1 to estimate the robustness of the subsurface structure. The contribution of site 1 is insignificant on the geoelectrical structure as seen from Fig. 11a (with site 1) and Fig. 11b (without site 1).

There are some limitations to our findings. (1) Even though Schirmacher Oasis represents a three-dimensional situation, the modelling has been confined to two-dimensions due to a limited data set (i.e., six stations in a line with 2- or 3-km spacing between stations). (2) A vertical magnetic field that yields tipper values could not be considered due to large noise from winds. (3) Due to accessibility problems, data could not be collected in a north–south oriented profile (i.e., strike perpendicular). Hence, the resistivity section shown here will indicate along-the-strike a lateral heterogeneity in resistivity in east–west direction over Schirmacher Oasis.

Previous studies have established that the crustal thickness in West Antarctica varies between 20 and 30 km in contrast to the crust that is ca. 40 km thick in East Antarctica, derived on the basis of seismic refraction profiles near the north–east coast, coupled with regional gravity interpretations (Bentley 1991). However, according to Wagner & Lindner (1991) the Moho depth in Schirmacher Oasis region has been interpreted to be ca. 35 km on the basis of geomagnetic data. Surface wave studies confirm a distinct contrast between the East and West Antarctic blocks, with the former showing more craton-like velocities on an average and with the highest lateral gradient in model velocities (Bentley 1991; Danesi & Morelli 2000; Ritzwoller et al. 2001; Harley 2003). Seismicity throughout Antarctica is generally low, which is consistent with a slowly moving plate almost entirely surrounded by mid-ocean ridges. However, the western part is ca. 10 times more seismically active than the east (Bentley 1991; Hole & Lemasurier 1994). Electromagnetic measurements conducted for bed rock investigations (Bhattacharya et al. 1987; Bhattacharya & Majumder 1987) have identified basement rocks below an ice cap 200–400 m in thickness south of Schirmacher Oasis.

A substantial and widespread sedimentary section with significant porosity or matured organic content has also been inferred below the ice sheet in previous regional MT studies in East Antarctica and has been correlated with similar results from seismic refraction data pertaining to the South Pole region (Wannamaker et al. 2004). Sedimentary basins ca. 6 km thick were also identified below the ice cover by Fournier (1994) from MT studies on Seymour Island and James Rose Island in the north-eastern part of Antarctica. Our traverse in Schirmacher Oasis reveals a thinning of the upper crust from ca. 20 km in the east to less than 10 km to the west within the short distance of 16 km.

Wannamaker et al. (2004) observed that the present-day deep resistivity structure of the South Pole region is markedly different from the typical cratonic lithosphere worldwide in exhibiting a low resistivity in the lower crust and upper mantle. They correlated this feature with a thermal anomaly generating high-temperature fluids or melts. Siegert (2000) proposed that heat flow through much of the East Antarctic lithosphere is in the order of ca. 50 mW m^-2 or more. As the degree of extensional activity in Antarctica today is limited, perhaps related to its being surrounded by oceanic ridges, the thermal anomalies are hypothesized to stem from mantle plume processes (Hole & Lemasurier 1994). Regional plume-related dynamics in the deep lithosphere have been postulated as the dominant trigger for this thermal activity, with a possible contributing influence from the adjacent West Antarctic Block, a process that has also been invoked in earlier studies to account for the regional uplift in East Antarctica such as in the Gamburtsev–Vostok areas (Studinger et al. 2003; Bell et al. 2006). However, a recent study by Ferraccioli et al. (2011) negated this hypothesis and interpreted the feature to be a compositional Airy compensation.

Continental flood basalts are known from western Dronning Maud Land and the Transantarctic Mountains which are regarded as manifestations of the Karoo-Maud mantle super-plume activities at about 180 Mya (e.g., Bormann & Fritzsche 1995), which enabled subsequent Gondwanaland break-up (e.g., Storey 1995; Raval & Veeraswamy 2003). Sushchevskaya et al. (2008) reported olivine-bearing dolerites and gabbro dolerites from the Dronning Maud Land originating in Jurassic trap activity. Otherwise, there is no evidence so far for Cenozoic igneous activity in this region. In relation, the Farrar era volcanics are quite old. Although thermal causes have been invoked locally elsewhere in Antarctica, solid phases can also generate such anomalies, as displayed in the data set from the Canadian lithoprobe programme, where values of a few 100 ohm m seem easily achievable without the presence of thermal anomalies. Wagner & Lindner (1991) suggested magnetization all the way to the Moho, implying a cool regime. Thus, the lower resistivity of the lower crustal layer observed in our MT traverse is consistent with a cratonic nature and solid-phase conductors (Wannamaker et al. 1996).

The MT study in Antarctica, as reported here, is the first of its kind under the Indian Expedition programme. Two-dimensional modelling of the data has delineated a laterally varying upper crust (10–20 km) with a resistivity of more than 8000 ohm m. This is underlain by less resistive (500–600 ohm m) lower crust. The resistive upper crust, which thins towards the west, appears terminated between sites 1 and 2. The geoelectrical structure of the Dharwar craton shows a highly resistive (3000–10 000 ohm m) upper crustal layer of 10 km followed by a relatively low (200–1000 ohm m) resistive layer down to depths of 30 km (figure 7 of Gokarn et al. 2004). The geo-electric structure obtained from Schirmacher Oasis resembles the same pattern with a highly resistive (>8000 ohm m) layer down to depths of 10–20 km followed by a low resistive (500–600 ohm m) layer down to depths of 40 km (Fig. 8b). In view of the similarities between Schirmacher Oasis and the Dharwar craton, Schirmacher Oasis may also be considered as a craton (stable).

It is also noted that the Mozambique Belt which runs along the east coast of South Africa, Madagascar, Southern India and Sri Lanka might extend into East Antarctica along a corridor which lies to the west of Schirmacher Oasis and whose eastern boundary appears to be controlled by the fault delineated in this study.