We identify possible sources of seismic anisotropy beneath India by synthesizing 2064 well-constrained shear-wave splitting parameters determined from a consistent analysis of waveforms recorded at 357 broadband seismic stations. Our effort includes compilation of previous results, reanalysis of old data, analysis of new data from previous networks and new stations. Our results reveal that the average delay time for entire India and its constituent tectonic provinces is 0.83 s suggesting moderate strength of anisotropy. Although the fast polarization azimuths (FPAs) are scattered, a NE trend appears dominant. Due to significant correlation of FPAs with the APM direction and lack of correlation between i) splitting parameters and backazimuths and ii) average delay times and lithospheric thickness, we conclude that the major contribution to anisotropy is from shearing in the upper part of the asthenosphere or a transitional layer from the base of the lithosphere to the upper part of the asthenosphere. Further, we postulate that a weakly anisotropic lithosphere in northern, central and south-eastern India is due to frozen anisotropy from past tectonic events. Northern and central India, Arunachal Himalaya and southern part of Burmese arc have simple anisotropy. Application of the spatial coherency technique reveals a source depth of 290 km for northern India. However, for south-eastern India and northern part of the Burmese arc, a two-layer model, with frozen-in and present-day anisotropy in the upper layer, and shearing and mantle flow in the lower layer, respectively, fits the anisotropy. In southern India, a large deviation of the FPAs from APM suggests imprints of deformation related to past tectonic events. A two-layer model, with frozen-in anisotropy in the upper and lower layers, is plausible. Variation in FPAs in the central part of the Indian shield is attributed to deflection in mantle flow at the northern edge of the lithospheric keel.

中文翻译：

---

深入了解印度下方上地幔形成地震各向异性的机制


我们通过综合 2064 个约束良好的剪切波分裂参数，确定了印度地下地震各向异性的可能来源，这些参数是通过对 357 个宽带地震台记录的波形进行一致分析而确定的。我们的工作包括汇编以前的结果、重新分析旧数据、分析来自以前网络和新站的新数据。我们的结果表明，整个印度及其组成构造省份的平均延迟时间为 0.83 秒，表明各向异性的强度中等。尽管快速偏振方位角 (FPA) 是分散的，但 NE 趋势似乎占主导地位。由于 FPA 与 APM 方向显着相关，并且 i）分裂参数和方位角与 ii）平均延迟时间和岩石圈厚度之间缺乏相关性，我们得出结论，对各向异性的主要贡献来自于软流圈上部的剪切或从岩石圈底部到软流圈上部的过渡层。此外，我们假设印度北部、中部和东南部岩石圈的弱各向异性是由于过去构造事件的冻结各向异性造成的。印度北部和中部、阿鲁纳恰尔喜马拉雅山和缅甸弧南部具有简单的各向异性。空间相干技术的应用揭示了印度北部 290 公里的源深度。然而，对于印度东南部和缅甸弧北部，两层模型，上层分别具有冻结和现今各向异性，下层分别具有剪切和地幔流，符合各向异性。在印度南部，FPA 与 APM 的巨大偏差表明与过去的构造事件有关的变形印记。 上层和下层具有固定各向异性的两层模型是合理的。印度地盾中部 FPA 的变化归因于岩石圈龙骨北缘地幔流的偏转。