PORE PRESSURE RECOVERY AND COULOMB STRESS EVOLUTION FOLLOWING THE 2004 M9.2 SUMATRA-ANDAMAN EARTHQUAKE

The 26 December 2004 M9.2 Sumatra-Andaman earthquake (SAE) ruptured a 1200 km segment of the plate boundary separating the Indo-Australian plate and Burma microplate. Three months later on 28 March 2005, the M8.7 Nias earthquake (NE) ruptured a 400 km segment adjacent to, and south of, the SAE rupture. The spatial and temporal proximity of these two earthquakes suggests that the earthquakes were coupled; that is, the evolution of stress and pore pressure induced by the SAE advanced the timing of the NE. We construct 3D finite element models (FEMs) to simulate the coseismic and postseismic deformation of the SAE for a problem domain having a distribution of material properties expected for the Sumatra-Andaman subduction zone (SASZ), based on seismic tomography, gravity models, and observed geologic structures. The coseismic slip distribution, having both thrust and strike-slip components, is estimated from near-field GPS data, FEM-generated Green's Functions, and damped-least-squares inverse methods. The coseismic rupture of the SAE perturbs the stress and pore pressure fields in the region, as described by poroelastic mechanics. Following the SAE, the excess pore pressure recovers to equilibrium via fluid-flow driven by the excess pressure gradients. Characterizing this interplay of stress and pore pressure along nearby faults is key to predicting earthquake coupling. Coulomb stress quantifies this interplay as the change in tendency for slip to occur along a fault and is defined as Δσc = Δσs + f(Δσn + ΔP), where σc is Coulomb stress, σs is shear stress, f is friction, σn is normal stress, and P is pore pressure. Preliminary results suggest that the SAE initially increased Coulomb stress near the location of the NE hypocenter by about 1.0 MPa, even though pore pressure decreased by a similar amount. Pore pressure recovered during the three months separating the SAE and NE. Consequently, Coulomb stress increased during this three-month interval by about 0.1 MPa near the NE hypocenter. Using FEM-based calculations of interseismic strain accumulation for the fault of the NE, we can estimate how the SAE advanced the timing of the NE.