Active Faults of the Northwest Himalaya: Pattern, Rate, and Timing of Surface Rupturing Earthquakes

The 2005 Kashmir earthquake (Mw 7.6) is the only Himalayan earthquake to rupture the surface since the 15th to 16th century A.D. when >Mw 8.5 earthquakes ruptured the Himalayan Frontal thrust (HFT) in the central Himalaya. Megathrust-type earthquakes like these seem to relieve a majority of the accumulated interseismic strain and concentrate permanent strain across a narrow width at the deformation front (faults within the orogen appear to accommodate little strain). The 2005 within-plate rupture in Kashmir may be a clue that a different seismotectonic model applies to the northwest Himalaya where active deformation occurs on faults distributed more than 120 km across the orogen. An asymmetric anticline marks the deformation front in Kashmir where the HFT is inferred to be blind, though ~20 m-high escarpments suggest that unrecognized thrust fault(s) may reach the surface locally. Folded river terraces and dip data also suggest that this frontal fold contains a SW-dipping back thrust. In Pakistan the Salt Range thrust system (SRT) defines the thrust front. New mapping and preliminary OSL dates from deformed Holocene sediments exposed along the westernmost SRT reveal that the fault slips at 1-7 mm/yr and last ruptured within the last several thousand years. Within the orogenic wedge to the north of the deformation front, active shortening occurs along a system of surface-rupturing reverse faults, extending from the Balakot-Bagh fault (source of the 2005 Kashmir earthquake) to the Reasi fault (RF) in Indian Kashmir to the southeast. One strand of the RF displaces a 350 m-high, 80 ± 6 ka (preliminary OSL age) fluvial terrace, yielding a minimum shortening rate of 3-5 mm/yr. Trenches excavated across the RF nearby reveal a distinct angular unconformity that likely formed during a surface rupture ~4500 yrs BP. Farther north, three northeast-dipping reverse faults cut Quaternary terraces on the southwest side of the Kashmir Valley. Trenches expose evidence for at least 2 surface rupturing events in the latest Quaternary and a shortening rate of 0.3 to 1.3 mm/yr. The active structures described above can account for 15 to 50% of India-Asia convergence, with up to ~20% of the shortening occurring on structures within the orogenic belt. Seismicity in the NW Himalaya is also broadly distributed but tends to concentrate in several places (e.g., the Indus-Kohistan and Hazara Lower seismic zones). Like in the central Himalaya, the zones of seismicity in the NW Himalaya may locate regions where interseismic strain accumulates, possibly in the middle crust along thrust ramps, and is released during large (>Mw 7.5) events. These relatively infrequent earthquakes likely activate portions (all?) of the plate boundary detachment fault and/or the within-plate fault systems. It may be possible for the region to generate earthquakes as large as >Mw 8.5, taking into account a reasonable average slip value and maximum possible rupture area. Recognition of internal surface-rupturing reverse faults indicates probabilistic models for seismic hazards in the NW Himalaya ought to account for great earthquakes on the Main Himalayan thrust (the basal detachment), moderate earthquakes on upper plate faults, and potentially events in the down-going Indian plate.