Potential of the Sumatran megathrust for generating future giant earthquakes

A 1650-km long section of the Sunda megathrust, from the Equator to about 14°N, ruptured during the 2004 and 2005 giant earthquakes. Immediately south of the 2005 rupture, from about the Equator to 0.5°S is a short section of the megathrust that has failed only by aseismic slip and moderate earthquakes throughout the past 250 years. From 0.5° to about 5.5°S, however, the megathrust has failed repeatedly over the past 700 years during giant earthquakes. Large uplifts in about AD 1360 and 1600, and in 1797 and 1833 show that the interval between episodes of activity is just over two centuries. The past-50-year geodetic history from coral microatolls and 3-year Sumatran cGPS Array data show that this large patch of the fault is locked. Uplift on the outer-arc islands associated with the 1797 rupture reached values as great as 1 m and extended from about 0.5° to 3.2°S. The magnitude of the earthquake was in the mid-8s. Historical accounts indicate that tsunami runup was between 5 and 10 m at Padang, a city of nearly a million people on the mainland west coast. Uplift during the 1833 earthquake ranged as high as 3 m and extended from about 2° to 5.5°S. The magnitude of the event was about 8.9. It seems clear that the section of the Sunda megathrust between about 0.5° and 5.5°S is late in its typical cycle of strain accumulation and that a giant earthquake is likely within the next few decades.