Deep Geoelectric Structure and Its Relation to Seismotectonics of the Saurashtra Region, Western India

The Saurashtra Peninsula and its adjoining regions covered by Deccan Traps (DT) are one of the important parts of the Indian continental lithosphere with interesting geophysical anomalies, tectono-thermal evolution since the Mesozoic times. Knowledge on the deep structure beneath these formations is important for understanding the seismo-tectonics of the region. This region has gained importance after the occurrence of a major earthquake (7.9 north of Saurashtra, namely Bhuj earthquake during 2001. It is also observed that Saurashtra region has experienced several earthquake swarms limited to small regions. Accordingly, it is important to investigate the deep structure of the Saurashtra region from seismotectonics point of view. In our study, magnetotelluric results of the deep crustal structure along five NS oriented traverses are presented. The five traverses are—Halvad-Rohisa (HR), Sa-par-Iswaria (SI), Mota Dahinsara-Bamagadh (MB), Jodiya-Jamkhandorna (JJ) and VavBeraja-Devda (VD). The total length of these 5 traverses is about 670 km. The derived deep geoelectric structure is also compared and c orrelated with gravity data to get more confidence on the derived results. The 2-D geoelectric section has delineated anomalous high conductivity structure at places extending from 20 km to about 40 km. From the spatial correlation, anomalous high conductive structure derived from MT data with intense localized seismic activity is an interesting observation. In the present study, the results of magnetotelluric studies along with other geophysical results are presented.


Introduction
At the western continental margin of India, a prominent feature-Saurashtra Peninsula (SP)-formed as a unique feature is known to have experienced different faces of stretching, rifting and uplifting during the geological past [1] [2]. The rifting pattern is believed to be developed initially from the north and sequentially extended to the south around the Saurashtra peninsula. The boundaries of the SP are the Kachchh rift towards the north and the Cambay rift towards the east with the other two sides occupied by the Arabian Sea. It is believed that the tectonics of the Saurashtra peninsula is controlled by three major Precambrian trends-Dharwar trend (NNE-SSW), Aravali-Delhi trend (NE-SW) and Satpura trends (ENE-SWS). The Aravalli trend cuts across the Cambay graben before entering SP. The major rifting events occurred in different geological periods from Gondwana in Late Jurassic or Early Jurassic after breaking up of the Indian landmass from Antarctica. The region is believed to be influenced by the thermo-tectonic events since the Mesozoic period due to the interactions of the Indian plate with Reunion plume. It is also believed that the massive volcanic eruption has occurred 65 Ma due to this tectonic event from the studies of Courtillot [3]. The duration of the eruption might be less than 1 Ma. Due to shallow mantle in this region high thermal gradients are attributed. A huge pile of basalts has erupted during Upper Cretaceous Tertiary boundary and spread over more than 100,000 sq·km [4]. Such a large spread of basalts also covered the SP. Additionally, there are a few volcanic plugs exposed in SP towards the southeast and southwestern part (see Figure 1). This can be clearly seen in the form of locally elevated hill like structures and also manifested in the form of large circular gravity anomalies. It is also believed that due to interaction of Reunion plume with the Indian Plate, a triple junction has been formed with Cambay rift as one of the arms, the Narmada rift as another arm towards the east and the Kachchh rift towards the west. High heat flow (67 mW/m 2 ) is reported along the triple junction and could probably be due to the outburst and trace of the Reunion plume [5]. Towards the northeastern part of SP large pile of sediments belonging to Dhrangadhra and Wadhwan formations are reported. From paleo-river channel studies, it is believed that the exposed sediments might also be present below the trap cover of SP as these sediments are older than the exposed Deccan traps. Additionally, if we detect any deep seated intrusive structures, those locations can be considered as possible occurrence of future earthquakes as they might be reactivated due to tectonic forces.
It is in this context that magnetotelluric studies have been taken up to understand the seismotectonics of the region. Anomalous conductive features that exist in the upper and lower crustal depths can easily be mapped using the magnetotellurics method. Magnetotellurics is well known for probing the earth from shallow level to large depths of even 50 -100 km from the observation of the natural variation of the electromagnetic signals. The details of data acquisition, processing and modeling procedures followed are described in Ph.D. thesis [6]. Here we discuss the importance of the results derived from the study from seis-motectonics point of view. Figure 1. Location of magnetotelluric stations along 5 traverses, namely HR, SI, MB, JJ and VD shown on the regional geological map of Saurashtra region. V-indicates the Deccan Traps, o, * and vertical lines indicate the exposed sediments, cluster of diamond symbols are the volcanic plugs.
In short, in the present study, the results of magnetotelluric studies are presented along five traverses-HR, SI, MB, JJ and VD. In Figure 2 these five traverses are shown on regional gravity anomaly map of the region. In the following the subsurface section derived along these profiles is presented. Figure 2. Five NS traverses shown on the regional gravity anomaly map. Five major closed group of contours-three towards the SE corner and two towards the western part-are the indication of the volcanic plugs in the region (modified Bouguer gravity anomaly map, 2006).

Geoelectric Section along the HR Traverse
The shallow section derived along the HR (Figure 3(a)) traverse has shown a thin (<250 m) trap cover from Chotila to Jasdan (site A16 to A23) and sharply increases to about 1.5 km from Amreli and further south. The basement is about 1.5 km north of Sara (site A8) and increases to about 2 km further south from Chotila. From the resistivity section it is observed that the top layer of the basement with a thickness of half a kilometer is less resistive (100 Ohm·m) as compared to the high resistive basement (>2000 Ohm·m) below.  . Deep geoelectric section along HR traverse along with regional gravity observed data and modelled response can be seen. C1 -C4 are conductors and R1 -R4 are resistors. Boundaries of the resistive formations, deep seated faults can be observed.

Geoelectric Section along the SI Traverse
The geoelectric section derived from 2-D modeling along SI traverse and the interpreted shallow geological section is presented in Figure 3(b). The exposed trap cover is thin (<500 m) from south (site 357) and maintains the same level and R3. The resistive feature can be interpreted as a block structure within the basement. Figure 5. Deep geoelectric section along SI traverse along with regional gravity observed data and modelled response can be seen. Here four layers with varying density values required to match the observed gravity data.

Geoelectric Section along the MB Traverse
As before the smoothed geological section derived from shallow geoelectric section is presented in Figure 3(c). Thin trap cover (<100 m) can be seen below the sites K1 -K4, which gradually increases to about 1 km below the site K7 until to mid crustal depths. Figure 6. Deep geoelectric section along MB traverse along with regional gravity observed data and modelled response can be seen. As before a four layered model with varying density is required to match the gravity model response with the observed data. A deep seated fault is inferred near the K12 MT station.

Geoelectric Section along the JJ Traverse
The smoothed section presented in Figure   . Deep geoelectric section along JJ traverse along with regional gravity observed data and modelled response can be seen. Good correlation of gravity high response with high resistive feature can be seen towards the southern part of the traverse. This could be due to the presence of volcanic plug buried below.

Geoelectric Section along the VD Traverse
The interpreted geoelectric section based on the derived shallow geoelectric section is presented in Figure 3(e). The exposed basalt layer is 1 km thick towards  Figure 3(e). Interestingly, a significant observation is that the Mesozoic sediments are concentrated more on the northern side of the five traverses (with respect to A23, close to Jasdan of HR traverse) and gradually become thinner at depths towards the south. The geoelectric section is drawn lengthwise, starting from the HR traverse towards the north to VD towards the west for more clarity. A progressive trend in sedimentary thickness is perceived from north to west as well as thinning of the sediments to the south.  . Deep geoelectric section along VD traverse along with regional gravity observed data and modelled response can be seen. Two resistive blocks towards the southern part of the traverse are co-relatable with gravity highs. Good correlation of intrusive conducting body near the station E7 with a gravity high is an indication of the presence of an intrusive body at mid-crustal depth.
In order to understand the trap thickness, sediment thickness and the basement depth features the results derived from all the above five traverses are combined to provide the regional distribution of the above formations; the details are described in the following: From the 2-D geoelectric section along the five profiles presented earlier in simplified shallow part of the geoelectric section described in the earlier section, the information available on all the individual profiles have been combined to provide a regional picture of the distribution of the geological formations namely the Deccan traps, the sediments and depth to the basement configuration ( Figure 9).

Sediment Thickness Map
Similar to the preparation of Deccan trap thickness map, the sediment thickness map has also been prepared and presented in Figure 9(b). As can be seen the sediment thickness is very thin towards the southern part of the study region and appears to be thicker towards the northeast and also the northwest of the study region. It is known from the geological studies that the Dhrangadhra and Wadhwan formations are exposed to the north and northeastern part of the study region. These sediments might have migrated towards the northwest through some transportation system, perhaps through Paleoriver channels.

Basement Depth Map
Similar to the trap and sedimentary thickness maps, the basement depth contour map of the study area has been prepared and presented in Figure 9(c).
As can be seen from the figure, basement depth is shallow towards the north and gradually decreases to very shallow (1000 m) and increases towards the south to a depth of about 3500 m. Interestingly, towards the north western part of the study region, the basement depth is steeply dipping and increases to about 3.5 km and nearly oval-shaped depression in the basement can be clearly observed. The basement undulations might have played a key role in controlling the deposition of the sedimentary formations as well as the Deccan flood basalts. This can be understood more clearly from the 3-D representation of the basement depth map as described in the following.

Gravity Modeling along HR Traverse
Gravity variations along the traverse are good indicators for understanding the surface features. Short wavelength and long wavelength anomalies reflect the shallow and deeper features respectively. Majority of the sites along HR traverse fall in the zone of negative gravity as can be observed in Figure 2. Site A1 towards the north and A37 towards the south fall in positive gravity regions as can be observed from the figure. Bouguer gravity variations along the HR traverse are shown along with the derived geoelectric section in Figure 4. A gravity low is located at the center of the traverse as shown. WinGLink software has a forward modeling utility. Modeling of gravity for the 2-D geo-electric section has been carried out to bring out the subsurface in terms of reasonable layered density structure as shown in Figure 7. The inferred gravity values for traps, Meszoic sediments and basement are based on detailed gravity studies carried out by the gravity group of NGRI [10] [11]. Traps (2.790 g/cc) are underlain by sediments (2.350 g/cc) and high resistive basement (2.900 g/cc). The underlying structure is an interplay of conductive (C1, C2, C3 and C4) and resistive sections (R1, R2, R3 and R4) with Moho (30 -32 km). A good fit is observed between the computed and observed gravity for the assumed layered model for the deep structure from magnetotellurics.

Gravity Modeling along the SI Traverse
Bouguer gravity anomaly for the sites towards the north has higher values as compared to the sites towards the south as can be seen in Figure 5. The geoelectric section is characterized in terms of traps, sediments and high resistive basement. Density values for these formations have been evaluated from the gravity studies after the sample analysis from Lodhika bore well and characterization in the laboratory [10]. An upper density layer (2.79 gm/cc) corresponding to traps is followed by a sedimentary layer (2.35 gm/cc). This is underlain by consecutively increasing density layers, from 2.85 gm/cc (basement) to 3.3 gm/cc at lower mantle level.
Forward modeling of the Bouguer gravity anomaly, using the algorithm of Rodi and Mackey [12] and implemented on WinGLink platform is attempted here to understand the geoelectric section through layered structure. The eastern part is an uplifted region [8] as compared to the western region and Moho values of 32 km for the east have been used for the forward model. Figure 5 shows one such model where a good match between the observed and computed Bouguer values can be seen. A high gravity region (+12 mGal) is observed at sites K3, K4 on the northern end and is likely due to the conductive fault structure (F1) between the sites K3A and K5 as well as the underlying conducive zone C1.

Gravity Modeling along the Traverse JJ Traverse
Bouguer gravity variations for the sites along the JJ traverse have been plotted as shown in Figure 7. The geoelectric section for the traverse can be interpreted in terms of a layered structure using 2.5-D gravity program on WinGLink platform. Different density values are ascribed to traps, sediments and basement, following calibration standards prescribed by the gravity group [10]. A forward gravity model for the geoelectric section can be seen in the Figure 7. A steady rise in gravity (−5 mGal to 25 mGal) from the sites on the north to those on south is observed from the plot. A rising trend is observed at site R16, 222 till 267, probably due to the fault structure F4 and the conductive zone C2. One probable model is presented in Figure 7.

Gravity Modeling along the Traverse VD Traverse
Bouguer gravity variations for the sites along the traverse VD are plotted in Figure 8 along with the geoelectric section. A gradual rise in gravity value from −3 for the sites towards the north to positive values (+30 mGal) for the sites towards the south is observed. MT site D12 is occupied over the plutonic mass or volcanic plug at Barda and hence has high positive Bouguer value. The conductive region at site at E7 is seen to contribute to the rising trend in Bouguer gravity value towards the southern end and this could be extended signatures for the observed volcanic plug at Barda or another possible buried volcanic plug at E7. Another interesting comparison can be made between the seismic study and MT study from Lodhika borehole. Incidentally, a deep borehole exists near Lodhika and this location can be used as a calibration standard for testing various geophysical techniques. Two such attempts were made earlier from reinterpretation of the seismic data [13] [14]. From the reinterpretation of the seismic data, the existence of thin basaltic layer and another thin sedimentary layer just above the basement have been shown. However, from another study by Sain et al. the second thin sedimentary layer above the basement has been ruled out. Using joint inversion of the MT and deep resistivity sounding data, it is shown that joint inversion results showed a close match with the thickness of the Deccan traps and sedimentary formations. Geoelectric section along the MB traverse derived from MT studies near Lodhika well (K12) is presented in Figure 11 along with the results derived from seismic studies.
For comparison, a portion of the 2-D electric section near Lodhika borehole along with the results derived from seismics and borehole lithology can be seen from Figure 11. As can be seen from the figure, the results derived from 2-D geoelectric section shows a close match with borehole lithology.

Comparison of the MT Results with Seismicity of the Region
It is interesting to study the seismicity of the Saurashtra region as intense seismic activity has been reported in Saurashtra region from the data observed for more than a decade. After the devastating Bhuj earthquake on 26 th January 2001, a major geophysical laboratory has been established in Gandhinagar to understand the earthquake phenomena of Kachchh and also other parts of Gujarat.
The institute since its inception has established a number of seismic stations to monitor the seismicity of the Kachchh region and also the Saurashtra region.
The network of observations made by this institute has helped to locate the earthquakes with greater precision of the epicentral locations of even micro-earthquake activity (0.5 -2.0 M). From the compilation of the seismic data observed over a period of nine years for Kachchh and Saurashtra region and also the adjoining regions, a map with epicenters for Saurashtra region has been prepared and is presented in Figure 12 [15]. structure derived from MT data, the intense localized seismic activity is an interesting observation. Another interesting observation is a localized gravity high that also spatially has correlation with the localized seismic activity. From such a spatial correlation of the gravity high, anomalous high conductivity located at depth of 20 km and micro-seismic activity it can be conjectured that there is a movement between the anomalous conductive structure with respect to the surrounding rock matrix. It is not very clear as to whether another volcanic plug located at a depth of 20 km might be trying to reach the surface. From the correlation of high gravity and high conductivity one can visualize for a possible presence of the rock matrix in a fluid state that might have close association with basic magmas.

Conclusions
Based mainly on the five geoelectric sections derived from five magnetotelluric traverses in Saurashtra basin and joint interpretation along with gravity and seismic data, the following major conclusions can be drawn.
1) The results derived from the present study gave a new concept on the deposition of sedimentation and also the direction of the lava flow, apart from mapping the complex basement undulations. It has also provided evidence for crustal scale fault features, besides delineating anomalous conductive features. The results derived from the present study also gave a new concept on deep tectonic features as well.
2) The geoelectric section along the HR traverse has shown thick sediments towards north and thinning of the sediments towards south with sharp change of its thickness from the middle of the traverse, Jasdan (Figure 3(a)). Similar signatures of variation in the thickness of sediments are observed on all the five traverses ( Figure 3).
3) The thickness of the Deccan traps along the HR traverse (Figure 3(a)) is thick towards south (2000 m) and thin towards north until the middle of the traverse and it becomes very thin or absent as we move further towards north.
Similar signatures can be seen along MB, JJ and VD traverses. The trap thickness along the SI traverse is thin (500 m) as compared to the other four traverses.
4) The basement variation along HR traverse is about 1500 m towards northern part, sharply increasing to 2000 m in the middle of the traverse and gradually increases to about 2500 m towards the south. While such is the case along HR traverse, the basement depth gradually increases for other MT traverses as we proceed from HR traverse to VD traverse indicating that the basement depth increases from east to west. 5) A 3-D representation of basement depth and also from the thickness contour maps of sediment, Deccan traps have shown interesting features. From the basement depth map prepared for the northern part of the study region ( Figure  3(c)), two prominent structural features can be derived. Shallow basement towards the east with a depth range varying from 1500 -2000 m with a nearly cir-cular shape feature, indicates a hill like structure buried below the study region. Another interesting feature is a sharp increase in basement depth towards the north-west of the study region with increasing basement depth reaching to as much as 3500 to 4000 m. The thickness of the sediments (Figure 3(b)) is large (1500 to 3000 m) towards the north and, north-east and north-west whereas it is thin (<500 m) in other parts of the region. The Deccan trap thickness is large (1000 -2000 m) towards south and NW part of the study region, whereas it is thin towards other parts. From such a signature, it is very clear that the basement playing a key role in the deposition of the sediments is from north and flow of Deccan traps is from south. Projection of such intrusive structures on the surface can be considered as a possible location for future earthquakes.