The Role of Fluids in Triggering the 2005 Nias Earthquake Induced by the 2004 Sumatra-Andaman Earthquake

The 26 December 2004 M9.3 Sumatra-Andaman and the 28 March 2005 M8.7 Nias earthquakes ruptured adjacent portions of the plate boundary interface in the Sumatra-Andaman subduction zone (Bilek, 2007: Vigny et al., 2005). The spatial and temporal proximity of these events suggests that the earthquakes were mechanically coupled. The M9.3 earthquake ruptured an approximately 1200-km-long segment of the plate boundary interface, from near Simeulue Island northward through the Andaman Islands. About three months later, the M8.7 earthquake ruptured an adjacent segment, just to the south of the M9.3 rupture. The spatiotemporal separation of the second earthquake suggests a poroelastic delay and triggering mechanism.

We propose a two-step process to explain the M8.7 rupture as a suspended along-strike migration of the M9.3 rupture. First, the M9.3 earthquake on the plate interface was the initial trigger. Second, it produced a pulse of pore pressure that migrated laterally along the plate interface of the Sunda trench. This pore pressure pulse disturbed the stability of the segment south of the M9.3 rupture, leading to failure of the M8.7 segment. Preliminary numerical results suggest the presence of a fluid-rich permeable subducting oceanic crust sandwiched between the less permeable mantle regions of the down-going slab and overriding wedge. The isolated and permeable oceanic crust favors along-strike migration of excess coseismic pore pressure. This deep lateral migration of pore pressure along the subducting oceanic crust explains the delayed triggering at the hypocenter of the M8.7 earthquake.