No Great Earthquake in the Central Himalaya Since 1505: a Possible Future M>=8.2 event?

The re-evaluation of the past several centuries of damaging Himalayan earthquakes has largely decreased their magnitudes and/or rupture areas, with one exception. An earthquake in 1505 that simultaneously destroyed Indian cities near Agra, and Tibetan monasteries between longitudes 78° and 84° appears to be larger than any known hitherto. It occurred exactly one month after a catastrophic earthquake in Kabul, and accounts from the two earthquakes have sometimes been confused. Although the data in Tibetan accounts are sparse the event appears to have had equal violence along the 600 km northern Himalaya and in the northern plains of India. From this we infer a rupture zone possibly twice as long as that associated with recent Himalayan earthquakes, corresponding to the segment that has hitherto been termed the Central Himalayan Gap. An enigmatic observation is that surface ruptures have been exhumed in trench investigations but have not been reported from the past two centuries of 7.8<M<8.1 earthquakes in the western and central Himalaya. These ruptures are unlikely to have developed aseismically since recent geodetic data indicate that creep processes beneath the Lesser Himalaya are negligible. A possible interpretation of the absence of recent surface ruptures is that they are associated only with the very largest Himalayan earthquakes, such as is inferred to have occurred in 1505. Geodetic data suggest that present convergence between India and southern Tibet of 16-18 mm/year is developing as elastic strain in the Greater Himalaya. Should this have prevailed since 1505 the so-called Central Himalayan Gap may have accumulated as much as 9 m of potential slip, sufficient to drive a Mw=8.2 earthquake. The infancy of systematic trench investigations, and the absence of a long continuous written history in the Himalaya, prevents conclusions about the timing of this inferred future M=8.2 event. No forecast is currently possible, but since convergence adds roughly 1.8 m/century to the future slip potential, a longer delay will result in a correspondingly larger event.