Locked Frontal and Lateral Ramps on the Main Himalayan Thrust beneath NW Himalaya illuminated by precisely located seismicity

The Kashmir ‘seismic gap’ in NW Himalaya, between the 1905 Kangra and 2005 Muzaffarabad earthquake rupture zones, has been replete with moderate-to-small earthquakes. GPS geodetic measurements across the Himalayan-arc reveal arc-normal convergence of ~11 mm/yr, which reduces towards the foreland in the India-fixed reference frame. In 2013 the Jammu And Kashmir Seismological NETwork (JAKSNET), and later the Himachal Pradesh Seismological NETwork (HiPSNET) was established to study the seismological characteristics of this ‘seismic gap’. Using continuous waveform data from these networks an earthquake catalog has been created using the Regressive ESTimator (REST) algorithm. Following this, seismic phases were manually picked from ~1100 earthquake records to determine the accurate arrival-times. A subset of these events based on the quality of picked phases are relocated using a probabilistic Non-Linear Location (NLL) method. These earthquakes have magnitudes between 0.5 and 4.5, and are distributed throughout the crust, with the majority concentrating at shallow (<25 km) depth. These shallow earthquakes are concentrated beneath the Higher Himalaya with lateral variations south of the Kishtwar window and to a region to its east. In arc-normal cross-section, the hypocenters lie on and above the MHT and the depth increases hinterlandward. Two distinct clusters of seismicity with increasing depth coincides with the mid-crustal frontal ramp observed in Vs structure beneath the Kishtwar window. The arc-parallel cross-section shows two eastward dipping hypocenter-clusters on and above the MHT. The one west of the Kishtwar window coincides with the lateral ramp observed in the Vs model. We conjecture that the one to the east also illuminates a similar transverse structure within the Himalayan wedge. Comparison of our hypocentral distribution with GPS velocities across this region reveal a frictionally locked shallow segment of the MHT, with the down-dip unlocking-zone highlighted by the across-arc clustering of seismicity beneath the Higher Himalaya. The locked-to-creep transition occurs immediately north of the mid-crustal frontal-ramp. We compute strain-rate from the sparse GPS data which reveals a predominant NE-SW compression and high strain-rates in regions of clustered shallow-seismicity. We are in the process of further refining the hypocentral locations using a double-difference relocation method, results of which will be presented.