Estimating the return times of great Himalayanearthquakes in eastern Nepal: Evidencefrom the Patu and Bardibas strandsof the Main Frontal Thrust

International audienceThe return times of large Himalayan earthquakes are poorly constrained. Despite historicaldevastation of cities along the mountain range, definitive links between events and specific segments ofthe Main Frontal Thrust (MFT) are not established, and paleoseismological records have not documentedthe occurrence of several similar events at the same location. In east central Nepal, however, recentlydiscovered primary surface ruptures of that megathrust in the A.D. 1255 and 1934 earthquakes areassociated with flights of tectonically uplifted terraces. We present here a refined, longer slip history of the MFT’stwo overlapping strands (Patu and Bardibas Thrusts) in that region, based on updated geomorphic/neotectonicmapping of active faulting, two 1.3 km long shallow seismic profiles, and logging of two river-cut cliffs,three paleoseismological trenches, and several pits, with constraints from 74 detrital charcoals and 14cosmogenic nuclide ages. The amount of hanging wall uplift on the Patu thrust since 3650 ± 450 yearsrequires three more events than the two aforementioned. The uplift rate (8.5 ± 1.5mm/yr), thrust dip(25° ± 5°N), and apparent characteristic behavior imply 12–17.5m of slip per event. On the Bardibas thrust,discrete pulses of colluvial deposition resulting from the coseismic growth of a flexural fold scarp suggest theoccurrence of six or seven paleo-earthquakes in the last 4500 ± 50 years. The coeval rupture of both strandsduring great Himalayan earthquakes implies that in eastern Nepal, the late Holocene return times of suchearthquakes probably ranged between 750 ± 140 and 870 ± 350 years