Segmentation of the Himalayan megathrust around the Gorkha earthquake (25 April 2015) in Nepal

[Display omitted] •The rupture of Gorkha 2015 earthquake is bounded by structural barriers.•Structural lineaments and ramp/flat geometry of the Himalayan thrust system control its segmentation.•Stable barriers define five barrier-type earthquake families in Central Himalaya.•Various rupture histories are expected for each individual earthquake. We put the 25 April 2015 earthquake of Nepal (Mw 7.9) into its structural geological context in order to specify the role of the segmentation of the Himalayan megathrust. The rupture is mainly located NW of Kathmandu, at a depth of 13–15km on a flat portion of the Main Himalayan Thrust (MHT) that dips towards the N-NE by 7–10°. The northern bound of the main rupture corresponds to the transition towards a steeper crustal ramp. This ramp, which is partly coupled during the interseismic period, is only locally affected by the earthquake. The southern bound of the rupture was near the leading edge of the Lesser Himalaya antiformal duplex and near the frontal footwall ramp of the upper Nawakot duplex. The rupture has been affected by transversal structures: on the western side, the Judi lineament separates the main rupture zone from the nucleation area; on the eastern side, the Gaurishankar lineament separates the 25 April 2015 rupture from the 12 May 2015 (Mw 7.2) rupture. The origin of these lineaments is very complex: they are probably linked to pre-Himalayan faults that extend into the Indian shield beneath the MHT. These inherited faults induce transverse warping of the upper lithosphere beneath the MHT, control the location of lateral ramps of the thrust system and concentrate the hanging wall deformation at the lateral edge of the ruptures. The MHT is therefore segmented by stable barriers that define at least five patches in Central Nepal. These barriers influence the extent of the earthquake ruptures. For the last two centuries: the 1833 (Mw 7.6) earthquake was rather similar in extent to the 2015 event but its rupture propagated south-westwards from an epicentre located NE of Kathmandu; the patch south of Kathmandu was probably affected by at least three earthquakes of Mw⩾7 that followed the 1833 event a few days later or 33years (1866 event, Mw 7.2) later; the 1934 earthquake (Mw 8.4) had an epicentre ∼170km east of Kathmandu, may have propagated as far as Kathmandu and jumped the Gaurishankar lineament. This combined structural approach and earthquake study allows us to propose that the MHT in the central/ea