Seismic Anisotropy in Northwest Himalaya from core refracted shear (SKS) and direct S waves 

Abstract

Recent enhancement in instrumentation in the northwest (NW) Himalaya provides an unprecedented dataset to study the deformation due to the Indo-Eurasia convergence. The NW Himalaya is unique in terms of hosting a seismic gap and a flat-ramp geometry at its decollement. We determine azimuthal  seismic anisotropy using core refracted shear waves (SKS) and interpret the results to develop insight about the prevailing geodynamics. Here, 459 raypaths with moment magnitude (Mw) >= 5.5, recorded at 15 seismographs of the J&K Seismological NETwork (JAKSNET), operational between 2013 and 2022, are used. To avoid contamination from direct S and SKiKS phases, we analyze the data within the epicentral distance range of 90°-125°, filtered at 0.04-0.2 Hz. We perform a 2-D grid search over the splitting parameters (delay time and fast axis azimuth) and compute their optimum values, for which the energy of the transverse component is minimum (MTE) after correcting for the inferred splitting. Simultaneously, the Rotation Correlation (RC) method is employed to calculate the delay time and fast axis azimuth corresponding to the maximum correlation coefficient between the splitting-corrected horizontal components. We use selection criteria based on the quality factor and the signal-to-noise ratio (SNR) to determine the measurements to be used for station averaging. The quality factor depends on the similarity of results obtained from the RC and MTE methods, hence helps in avoiding subjective interpretation about the quality of the measurement. The non-null splitting measurements passing these selection criteria are then used for station averaging applying the circular mean method and the energy map stacking method. We observe mostly N-S to NE-SW trending fast axes azimuths (13 of 15 stations); this direction corresponds to the absolute plate motion of India in a no-net rotation frame. The two remaining stations show average NW-SE fast directions, which are parallel to the mountain front, but also these stations show somewhat contradictory single splitting measurements, and one of those two anomalous stations is located very close to stations with NE-SW fast measurements, so we will not interpret these. The mean delay times range from 1.5-3.3 s, with the majority of the stations exhibiting > 2s split time being situated on the foreland basin deposits of the Sub-Himalaya. The high absolute Indian plate motion of 51 mm/yr appears to align the upper mantle olivine beneath the orogeny and NE oriented fast axes track the mantle flow manifested by the basal shear of the plate motion. To complement the SKS data, which are dominated by results with eastern backazimuths and corresponding initial polarisation, we further measured splitting for direct S-wave with the reference station method. Here, the correlation between the horizontal traces of the target and reference stations is maximised after correcting the target trace with trial splitting parameters and differences in gain; the reference station trace has previously been corrected for SKS splitting. We will present direct S splitting measurements from 370 events with Mw>=5.5 and distance within 40°-80° with an interstation spacing of <120 km.