Science @ CIRES  >  Science Reviews  >  NOAA Science Review, 2002

Abstracts: 18

Earthquake Hazards and the Collision between India and Asia

Rebecca Bendick, Roger Bilham, Frederick Blume, Grant Kier, Peter Molnar, Anne Sheehan and Kali Wallace

India continues to drive into Asia at a rate of approximately 47 mm/year, sustaining the elevation of the Tibetan Plateau, deforming the Himalaya, Altyn Tagh and Tien Shan mountains, and causing a steady, but unpredictable, sequence of earthquakes in Asia and parts of India. The widespread application of GPS (Global Positioning System) geodesy in India, Pakistan, Nepal, Bhutan, Tibet and China permits us to quantify the strain rates that drive these earthquakes. The rate of contraction across the 150-km-wide Himalaya (18 mm/yr) is at least 4 times that determined across the 1500 km wide Indian subcontinent (<4 mm/yr), yet recent disastrous earthquakes in India have occurred largely within the "undeforming" continent where strain rates are apparently 4 orders of magnitude lower. One possible reason for seismicity in India is that deformation does not occur as net contraction, but as flexural strain associated with the Himalayan collision. In its northward progress, India streams through a flexural bulge that locally stresses the plate. The M=8.1 Assam 1897 and M=7.6 Bhuj 2001 earthquakes for example, ruptured from the base of the plate to depths of order 9 km.

We have conducted special studies of a dozen historic Indian earthquakes in the Himalayan collision zone and on the subcontinent. The Himalayan studies are historically complete only for the past two centuries, and are increasingly fragmented before that time. But a clear picture is emerging: historical events in the Himalaya in the past 300-300 years have occurred in only one third of the arc, suggesting that others will occur in the remaining regions. By combining the geodetic convergence rate with the time since the last known earthquake we may quantify the size of the earthquake that would occur, were one of these regions to slip today. In six locations we infer that a M>8 could occur. In the central part of the Himalaya where in 1505 a great rupture destroyed Agra, and Tibetan monasteries along a 600 km long zone in Tibet, the earthquake could be as large as M=8.2. Our findings caused considerable concern in India where we estimate 40 million people are at risk from Himalayan earthquakes.

The Bhuj earthquake was investigated by GPS geodesy teams from Boulder. Geodetic data confirmed that the event was deep (below 9 km) and that 3-5 m of slip was confined to a 40x40 km rupture zone penetrating to the base of the crust. We found moreover, that the Rann of Kachchh may have contracted 2 m in the past century. The success of this study led us to the 1905 Kangra rupture zone where also 20,000 people were killed by a M7.8 earthquake. There we confirmed that the rupture was smaller than some have supposed with slip less than 4 m. In contrast a geodetic study of the M=8.1 1897 rupture that is still ongoing revealed a mean slip of 16 m! Finally, in the past two years we installed and operated a broadband seismic network in Nepal, and continuous GPS tracking sites in Bhutan and Pakistan, with GPS remeasurements in The Altyn Tagh, eastern Tibet and the Ladakh Himalaya.