Upper mantle anisotropy and it's geodynamical implications in Sri Lanka region

Abstract

<p>Anisotropy in the Earth’s upper mantle is a signature of past and present deformation. Sri Lanka comprises four main lithological units, viz. the Highland Complex (HC), the Wanni Complex (WC), the Vijayan complex (VC) and the Kadugannawa complex (KC). To calculate the upper mantle anisotropy, we have collected the earthquake data from IRIS (Incorporated Research Institutions for Seismology) network. The upper mantle anisotropy beneath Sri Lanka is measured in the frequency band 0.01–0.15 Hz, with magnitude (Mw) of six or more and within the epicentral distance of 90°-140°. We have analyzed (the fast direction and delay time) shear wave splitting of SKS/SKKS phases at 3 stations, namely, MALK (WC), HALK (HC) and PALK (KC) in Sri Lanka. In this study, shear wave splitting measurements were done using high-quality seismograms (~30) of many earthquakes occurring in the region. We have used rotational correlation (RC) , minimum energy (SC) and eigenvalue techniques. The result of the shear-wave splitting measurement shows the presence of two anisotropic layers in the upper mantle. The upper and lower layer’s fast-polarization direction is found to be NE-SW and NW-SE, has the delay time varies from 0.4-0.5s in the upper layer, and 0.6-0.8s in the lower layer. We found two major fast directions in the upper and lower layers, viz. NE-SW in the upper layer of MALK and PALK and NW-SE for the HALK stations, and NNE-SSW in the lower layer beneath MALK and HALK stations and NW-SE in the PALK station. Overall, Fast direction for Sri Lanka region is found to be NE-SW in the lower layer and NW-SE in the upper layer. Our study suggests that fast axis direction of lower layer with an average delay time of 0.6 s depicts a ~67 km thick anisotropic layer with 4% anisotropy (from previous studies) beneath Sri Lanka region. However, if we assume an anisotropy range of 3–5%, then the calculated delay time of 0.6 s would correspond to thickness variation of 89.3 to 53.59 km, respectively, for the inferred anisotropic layer. Comparing from APM (Absolute Plate Motion) direction with our fast directions, we infer that the SAF(Simple Asthenospheric Flow) model prevails in this region and secondly, when S<sub>hmax </sub>(Maximum Horizontal stress) and the GPS (Global Positioning System) data compared with the fast direction we infer that there is partial contribution from lithospheric mantle. So, we confirmed that anisotropy in the region is mainly governed by asthenospheric flow and partially due to lithospheric mantle.</p>