Bangladesh and adjoined Indian Shield, and northern Bay of Bengal regions were not studied together as they belong to different physiographic domains and political territories (Fig. 1). The contiguous regions as one hold very critical and useful information for better understanding the tectonic evolution of various geological provinces and their assemblage. Otherwise, studying individual regions in isolation wouldn’t provide a holistic view of the evolution of this complex region of the northeast Indian subcontinent. In order to unravel the subsurface details from this contiguous area, we collated Bouguer gravity anomaly data of Bangladesh and surrounding Indian Shield and free-air gravity anomaly data of the northern Bay of Bengal from different data sources and prepared an assimilated gravity anomaly contour map with an interval of 4 mGal (Fig. 2). The gravity map is the first of its kind covering the land and sea domains in the region of northeast Indian subcontinent (Fig. 1). Bouguer gravity anomaly data of Bangladesh region is considered from the published Bouguer Gravity Anomaly Map of Bangladesh33. The Surrounding Indian Shield Bouguer gravity anomaly data are considered from published sources34–37 and unpublished reports of ONGC. As both Bouguer gravity anomaly data of onshore and Free-air gravity anomaly data of offshore regions present the corrected gravity data to the same reference of the datum plane, we considered both of these to prepare an assimilated gravity anomaly map for the contiguous onshore and offshore regions. Accordingly, we considered satellite (cryostat-2 and Jason-1) derived free-air gravity anomaly data of the northern Bay of Bengal38, and then combined with the Bouguer gravity anomaly data of Bangladesh and adjoining Indian Shield.
The Bouguer gravity anomaly maps of Bangladesh and adjoined Indian Shield regions were digitized using WINDIG freeware. The digital data are converted to gridded data and combined with the gridded data of free-air gravity data of the northern Bay of Bengal for preparing a combined database and to generate a combined contour map of both onshore and offshore regions (Fig. 2). Two transects running E-W (23.4°N latitude) and N-S (91°E longitude) directions covering the prominent anomaly signatures are chosen for modeling the subsurface structures. The GM-SYS software is used for determining the 2-D crustal density model for unraveling the nature of rocks within the study area (Fig. 3) and tectonic processes that the regions experienced during the past 120 Myr. The derived crustal models are more plausible as they were well constrained by the plate reconstructions models7, Seaward Dipping Reflectors39,40, magnetic doublet signatures41, Continent-Ocean Boundary line7,2, sediment thickness details42,43,10 and depths to the Moho discontinuity44 from onshore region, and basement and Moho topographies derived from seismic reflection and gravity results of the offshore region1,2.
Gravity signatures and crustal models - northeast Indian subcontinent and northern Bay of Bengal
Bouguer gravity anomaly data of Bangladesh and surrounding Indian Shield, and Free-air gravity anomaly data of the northern Bay of Bengal are combined (Fig. 2), thereby discussing detailed gravity responses of the area, where different geomorphic units are adjoined. The gravity data of the study area are quite variable ranging from -190 mGal to 55 mGal (Fig. 2), associates the negative value with the Himalayan Mountain Range because of deep burial of the Moho discontinuity at an average depth of 60 - 80 km45 and positive value with the Bangladesh shelf-slope region because of the rise of seafloor and basement towards the coastline2. A significant E-W elongated positive gravity anomaly reaching up to 28 mGal is observed accompanying more than a km elevated Shillong Plateau as the Moho discontinuity beneath it is shallower than that beneath the Sylhet Trough/ Surma Basin in the south and Brahmaputra Basin in the north46,28,47. The Sylhet Trough lies to the south of the Shillong Plateau, is associated with E-W oriented bulged negative gravity anomaly reaching up to -80 mGal, showing the response of 13 - 18 km thick clastic sediments deposited from the Tertiary to recent times within the trough4,48. In eastward between the coastline and Shillong Plateau, the gravity anomaly falls in W-E direction from -20 to -116 mGal with two different gradients having a lower one of about 30 mGal/ 50 km on deformational front and a relatively higher one of about 50 mGal/50 km in the vicinity of Burmese Arc (Figs. 1 and 2), representing the geometry of subducting slab of the Indian plate beneath the Burmese platelet. Towards the west, a prominent NNE-oriented narrow low gravity strip with three lower value closures within and broad positive gravity anomaly trends on either side of the strip is observed (Fig. 2). On northward around 24°N latitude, the low gravity strip takes a turn to NE direction and finally connects the E-W oriented Dauki thrust fault gravity anomaly. Following the geophysical observations of narrow low gravity strip, magnetic doublet33 and Seaward Dipping Reflectors39,40,7,2 delineated the presence of continent-ocean boundary that separates the Indian Shield from oceanic crust underlying Bangladesh. It is observed that broad positive gravity signatures lie on either side of the low gravity strip, in general, correspond to the responses of the shallow Precambrian rocks of the Indian Shield on the west and shallow Moho discontinuity on the east beneath Bangladesh. The continuity of broad positive gravity anomaly over the Indian Shield and Shillong Plateau has been notched by lower anomalies at two locations (Fig. 2), which can be interpreted as a response of paths of the Ganges and Brahmaputra river systems. The broad gravity signature over the Bangladesh region consists of near N-S oriented alternate bands of about 200 km wide positive and negative anomalies (Fig. 2).
Keeping the prominence of gravity signatures of the study area in view, we choose two transects in the E-W direction along 23.4°N latitude and N-S direction along 91°E longitude (Fig. 2) as representative profiles for determining the crustal structure models. The E-W transect is considered to take into account gravity signatures associated with the Indian Precambrian rocks, Bangladesh, and Burmese Arc. While the N-S transect runs through the gravity responses of the Brahmaputra Basin, Shillong Plateau, Dauki Fault, Sylhet Trough, Bangladesh, and north Bay of Bengal. For building initial crustal models, we considered details of available geological and geophysical information as detailed in the previous section. Average density values for various units of sediments, crust, and upper mantle rocks are considered from published results49–52.
The crustal models derived along both transects depict the geometry of sediment, crust, and upper mantle layers, thicknesses, and density of rocks (Fig. 3a and b). The E-W crustal model reveals the presence of ~36 km thick continental crust on eastern side of the Singhbhum craton, thereupon the crust tappers off westward and adjoins 8-9 km thick oceanic rocks. On eastern side, the location of trench is marked, where the oceanic lithosphere of the Indian plate subducts with a low gradient of about 3.0° to 3.7° beneath the Burmese continental lithosphere (Fig. 3a). While beneath the Indo-Burma Ranges the dip of the subducting Indian plate has relatively increased to about 7.8° - 9.5°44. Further east, from the Indo-Burma Ranges to the Central Myanmar Basin, Zhang et al.53 imaged the slab of the Indian lithosphere being subducting at deeper depths with an increased dip angle of ~25°. It is also observed that the oceanic crust beneath the Bangladesh region is folded in near N-S direction with a wavelength of about 200 km and an amplitude of about 2 km (Fig. 3a). The N-S crustal model shows more than 40 km thick continental crust beneath the Shillong Plateau with a small rise at Moho discontinuity compared to that lies beneath the oceanic crust in Sylhet Trough (Fig. 3b). Earlier studies using gravity in conjunction with seismological results54 and joint inversions of receiver functions and surface wave dispersion11,55 indicated the absence of crustal root beneath the Shillong Plateau to support the isostatic compensation. Further, the model in the vicinity of present-day shelf edge reveals more than a km rise at levels of the seafloor, basement, and Moho discontinuity towards the coastline. The sediments, in general, were found to be thicker beneath the Bangladesh region reaching around 20 km in the regions of Sylhet Trough and depressions of the folded lithosphere (Fig. 3a and b). The E-W crustal model further reveals the structures of volcanic passive margin and ocean-continent collision, which were existing on surficial levels in the geological past at least up to Miocene time. Subsequently, the structures were being covered and buried under the huge thick sediments carried by the Ganges and Brahmaputra rivers from the Himalayas. The pre-Miocene structures of continental margin in the east and collision-related accretions in the west of the Bangladesh region are well comparable with the structures of present-day volcanic passive margin on the eastern margin of South America (Fig. 3c)56 and Chilean subduction zone on the western margin of South America (Fig. 3d)57. Both volcanic margins are depicting the presence of underplated materials and SDRs with nearly similar kinds of gravity responses. Beyond the margin, the gravity anomalies are differing as those over Bangladesh are associated with the folded lithosphere. The gravity anomaly map of northeast Indian subcontinent and northern Bay of Bengal, and derived crustal structures are shown in three-dimensional view for better visualizing the correlations between the structures and gravity responses (Fig. 4). It is observed that the NNE segment of the continent-ocean boundary is solely correlated with the foot-of-the-slope of continental margin, while the NE and E-W segments of the boundary are further accompanied by the Debagram-Bogra Fault and Dauki Fault lines, respectively (Figs. 1, 2 and 4).
Paleo-continental margin and relic fragment of the Bay of Bengal
Most geological and geophysical investigations so far carried out in the present study area were in isolation, which lead to restrain understanding of evolution of all geological features in a holistic view. In order to overcome this, we combined gravity anomaly data of both land and sea regions and then modeled with tight constraints for the purpose of delineating the nature of rocks and their evolutionary history. The crustal models determined along the E-W and N-S transects (Figs. 3 and 4) vividly show the presence of continental margin structure on east of the Singhbhum craton and south of the Shillong Plateau, which adjoins the ancient oceanic crust that evolved when India was drifted from East Antarctica (Fig. 5a). The gravity signatures, surface morphology, and crustal structures (Figs. 2, 3, and 4) suggest that the NNE segment of the continental margin was evolved in hypo-extended rift process in association with the magmatic activity, while the E-W continental segment was controlled by the Dauki Fault and may have evolved in a shear rift process, similar to the one identified on southern segment of the ECMI17,2.
Considering the signatures of NNE oriented low gravity anomaly strip, magnetic doublet connecting the similar age (about 117 - 118 Ma) volcanic rocks of the Rajmahal and Sylhet, Seaward Dipping Reflectors, Talwani et al.7 proposed the second stage of a continental breakup between the northeast Indian Shield in the vicinity of Rajmahal - Sylhet Line and parts of the Kerguelen Plateau at around 120 Ma. At this stage, the Kerguelen hotspot had emplaced the Rajmahal and Sylhet traps on Indian Shield and facilitated another continental breakup between India and East Antarctica (Fig. 5a). The breakup not only splits the continental fragments (Elan Bank and parts of the Kerguelen Plateau) from the Indian Shield, but also transfers the conjugate parts of the oceanic crust accreted during M12n anomaly time - just older to M0 anomaly time to the Antarctic plate16,17,7. It is appropriate to mention that the paleo coastline in the northeast of India runs very close to the locations of Rajmahal and Sylhet volcanic emplacements, but currently lies far inside from the present-day Bangladesh coastline (Fig. 5a). This implies that there was a presence of passive volcanic continental margin segments east of the Singhbhum craton and south of the Shillong Plateau and an older fragment of the Bay of Bengal having an age of 120 Ma and less right beneath most of the Bangladesh territory. Both the E-W and N-S crustal models demonstrate the presence of thick continental crust adjoining the old oceanic crust of about 8-9 km thick (Figs. 3a and b). The basement rocks recovered in Maddhapara area of northwest Bangladesh (Fig. 1) show the presence of Paleoproterozoic tonalitic and granodioritic rocks having a magmatic age of 1722±6 Ma58. It is further found that the buried rocks in this area are separate and discrete micro-continental fragments58, which were emplaced when the Gondwana supercontinent was located in high latitudes of the southern hemisphere.
As the expansion of the northeast Indian Ocean floor continued, the continental collision has occurred between India and Eurasia in the Early Tertiary period22 in the north, while in the Indian Ocean, the divergent plate boundary (Wharton spreading ridge) becomes extinct leading to the unification of Indian and Australian plates as a single Indo-Australian lithospheric plate59,21. At around Oligocene - Miocene time (~23 Ma), the continued process led to the rapid development of the Himalayan mountain range, resulting in origination of Ganges and Brahmaputra river systems for carrying huge deposits of sediments to the Bengal Basin and further to the Bay of Bengal for forming the delta and fan systems23,60. Even during the Oligocene - Miocene time, the present-day onshore Bengal Basin was under the Bay of Bengal waters, consequently, the Continent-Ocean Boundary paralleling the East Coast of India up to Mahanadi Basin follows the edges of Singhbhum craton, and Rajmahal - Sylhet Line (Fig. 5b). This implies that the present onshore Bengal Basin was once bordered by the segments of volcanic passive margin on west and sheared margin on north. As sediment deposition continued from Miocene through Holocene, the oldest oceanic lithosphere adjoining the continental craton, and volcanic emplacements (Fig. 5b) was buried under the clastic sediments derived from the Himalayas. The deposition process led to prograde the coastline oceanward to the present location, which eventually bury the west and north continental margin segments (Fig. 5b) and old fragment of the Bay of Bengal lithosphere, thereby forming a new margin segment completely built with the sedimentary rocks. The studies of eastern Himalaya drainage patterns and Ganges-Brahmaputra Delta evolution since the Oligocene allowed Betka et al.6 to map the location of drifted shoreline through time towards the ocean. Thus, the present-day Bangladesh margin is a unique feature exclusively formed by the progradation of huge sediment deposits over the oceanic crust of the proto-Bay of Bengal and remains an exceptional margin from the ones formed by the processes of continental rifting and breakup.
The near N-S trends of gravity anomaly signatures over the onshore Bengal Basin (Fig. 2) and gravity-anomaly derived crustal structures (Figs. 3a, 3b, and 4) clearly show the presence of approximately 200 km long folded oceanic crust and overlying sedimentary layers. The shortening of this oceanic crust is possibly driven by the resultant processes of ocean-continent collision at the Indo-Burman subduction zone, tectonics offered by the Burmese platelet, convergence across the Dauki Fault, etc. A few questions like plate convergence and shear motions along some of the faults existing in the regions of Bangladesh, northeast India, and Myanmar are still unclear. Analysis of the Harvard CMT data and GPS measurements from the Burmese Arc region led Rao and Kumar61 and Gahalaut et al.62 to point a possible cessation of eastward subduction, while, Steckler et al.29, Mallick et al.63 and Fan et al.64 opine that subduction process is still active at Indo-Burman plate boundary. However, there is a commonly accepted view among the researchers that the Indian plate subducts with a significant component of strike-slip motion along the edge of Burmese platelet, in addition, some of the fault/ thrust zones within Bangladesh and Burmese platelet are accommodating the east-west convergence. The process of excessive sedimentation in the Ganges-Brahmaputra Delta region prompted Mallick et al.47 to believe that the process led to collapsing the Sylhet Trough/ Surma Basin into the mantle, thereby affecting the westward propagation of the Indo-Burman Wedge since the Miocene. Previous studies on subduction process of the Indian lithosphere beneath Burma also suggested a westward trench migration65,45,66. Using the Block Model studies Mallick et al.63 posited that there is an accumulation of E-W compressional strain within the outer wedge of the deformational front (Fig. 1). The tectonic processes of an eastward moment of the Indian plate, westward propagation of the Indo-Burman Wedge, and convergence across the Dauki Fault together may have applied east-west compression to the oceanic lithosphere lying beneath Bangladesh (Figs. 2 and 4), resulting in formation of long-wavelength lithospheric folding.