The observations came from nine temporary broadband seismic stations in and around the Satpura Gondwana Basin. The stations operated in three deployment phases between July 2023 and July 2025, temporarily increasing coverage across a region that had previously been sparsely instrumented for shear-wave splitting studies. The Central Indian Tectonic Zone is a Mesoproterozoic suture marking the collision of the northern Bundelkhand Craton with the Bastar, Dharwar and Singhbhum cratons of southern India.

The precise geometry of that convergence remains debated, including proposals for northward or southward underthrusting, double-sided subduction and subsequent changes in subduction polarity. Researchers screened 424 teleseismic earthquakes with magnitudes greater than 5.5 for SKS phases arriving at epicentral distances of 84–145 degrees and SKKS phases arriving at 84–180 degrees. After processing and quality control, the dataset contained 82 SKS and 47 SKKS splitting measurements from 87 earthquakes, plus six null measurements.

An SKS or SKKS wave crossing anisotropic material separates into fast and slow components with mutually perpendicular polarization directions. The fast polarization direction records the preferred propagation direction, while the time delay between the components reflects the accumulated strength of anisotropy along the receiver-side path. The team used transverse energy minimization to identify the splitting parameters, searching for the orientation and time lag that minimize energy on the transverse component after correction.

Rotation correlation was applied as an independent check, and polarization patterns were used to verify station orientation. Station-averaged fast directions were calculated with a von Mises circular mean, while delay times were averaged arithmetically. Across the network, fast directions ranged from 24.84 to 70.88 degrees, and most were oriented northeast–southwest.

The generally narrow range of back azimuths, with most events arriving from 90 to 120 degrees, limited the study’s ability to resolve the number and geometry of anisotropic layers. That coverage limitation prevented the researchers from constructing a two-layer anisotropy model. Such models require a clear 90-degree periodicity in splitting parameters across a broad range of azimuths, whereas the available measurements were concentrated in a narrow sector and were unevenly distributed among stations.

The study therefore did not attempt to assign separate layers to fast directions and delay times, instead presenting the results as evidence for spatially variable, possibly multilayer anisotropy. This is an important constraint: the data support a broad tectonic interpretation but do not uniquely resolve the number or detailed structure of the anisotropic strata. The limited coverage also means that the six null measurements cannot be used to establish a unique subsurface model.

A ray-piercing-point spatial-coherency analysis was used to estimate where the dominant anisotropy might reside. Because SK(K)S phases approach stations at steep angles, their splitting measurements have poor inherent depth resolution, so the researchers calculated the positions where the rays would cross assumed depth horizons from the surface to 400 kilometres. The resulting variation factor fell sharply toward a broad minimum at approximately 200–250 kilometres before flattening at greater depths.

This pattern was obtained with spatial averaging blocks of 0.30 and 0.35 degrees. The authors describe 200–250 kilometres as a promising average depth, not a precise boundary, because the method loses sensitivity as ray-piercing points become increasingly separated at greater depths. The inferred depth favors asthenospheric mantle flow as the principal source of the measured splitting rather than shallow crustal anisotropy or frozen fabric confined entirely within the lithosphere.

The crust can contribute to shear-wave splitting through aligned minerals, cracks and structural layering, but the observed delays greatly exceed typical crustal contributions. The crust is about 40 kilometres thick in the region, with local thickening toward the Narmada paleo-rift zone, and the authors note that only part of the mid- to lower crust is likely to be strongly anisotropic. Even the lowest station-average value, 0.76 second at SG02, is larger than the expected crustal contribution, supporting a dominant upper-mantle source.

Present-day flow alone, however, cannot explain every observation. The absolute plate-motion direction for the Indian plate is approximately 48 degrees, and five of the nine stations differ from it by more than 10 degrees. Deviations reach 26 degrees at SG06, 19 degrees at SG03, 16 degrees at SG04 and 12 degrees at SG08, while SG03’s station-average fast direction reaches 70.88 degrees.

These departures are consistent with some anisotropy being preserved from older deformation rather than generated solely by current plate-driven flow. The likely inherited signal comes from Mesoproterozoic convergence and collision within the tectonic zone, potentially modified by the uneven lithosphere-asthenosphere boundary beneath central India. The seismic data do not uniquely identify the relative contribution of each process.

Low delay times and null measurements offer further clues. Stations SG02, SG06, SG07 and SG09 have average delays below 1 second. SG02, which recorded four nulls, lies in the Narmada–Tapti region, where Cretaceous dykes provide geological evidence of magmatic activity.

SG09, with two nulls, is near the Deccan Volcanic Province, which also experienced substantial Cretaceous–Paleogene magmatism. The investigators interpret the reduced splitting as possibly reflecting modification or overprinting of pre-existing mantle fabric by plume-related thermal and deformation processes, but they caution that the spatial association does not by itself establish causation or a direct contribution from shallow dykes. Nulls may result from complex anisotropy, absent effective splitting, or an initial polarization aligned with a fast or slow axis; limited back-azimuth coverage prevents a more detailed interpretation.

At SG08, near the east–west-striking Tapti Fault, the fast polarization direction is parallel to the fault. The authors interpret this as evidence consistent with the fault being a deep-seated lithospheric weakness that may extend into the mantle and influence deformation there, although the seismic observations do not directly image its geometry. Overall, the study concludes that Central India’s upper mantle records a combination of northeast–southwest asthenospheric flow, inherited Mesoproterozoic orogenic fabric, fault-related structure and possible plume overprinting.

The results do not confirm the proposed slab breakoff or subduction-polarity reversal, and the authors say that tomography, splitting-intensity imaging and receiver-function analysis would be needed to test those deeper tectonic models.

Subject of Research: Upper-mantle seismic anisotropy beneath the Central Indian Tectonic Zone

Article Title: Mantle deformation patterns beneath the Central India Tectonic Zone: evidence from SK(K)S splitting measurements in the Satpura Gondwana Basin and adjacent areas

Article References: Bishoyi, N., Dubey, A. K., & Tiwari, A. K. (2026). Mantle deformation patterns beneath the Central India Tectonic Zone: evidence from SK(K)S splitting measurements in the Satpura Gondwana Basin and adjacent areas. Solid Earth, 17(10), 1201-1217. https://doi.org/10.5194/se-17-1201-2026

Image Credits: AI Generated

DOI: 10.5194/se-17-1201-2026

Keywords: seismic anisotropy, shear-wave splitting, Central India, upper mantle, Indian plate motion, Mesoproterozoic orogeny, Cretaceous magmatism, lithosphere, tectonic deformation, mantle flow, Mantle, deformation