Abstract
The role of fluid injection on the occurrence and migration path for the aftershocks of 2004 Sumatra earthquake (Mw 9.3) and January 2005 Andaman earthquake swarm within the aftershock sequence is investigated here from the viewpoint of pore fluid diffusion process. The Sumatra earthquake created a regionally extensive crustal rupture plane exceeding 1,200 km length below the Andaman Sea. The r-t plots (Shapiro et al. 1997) are constructed for these aftershocks in order to examine the role of poroelastic effects as rupturing progressed with time. Their main results are as follows: the r-t plot corresponding to first 3 h of aftershock activity (when only 44 events of mb >= 4.5 originated) reveals that 95% of the data points occurred below the modelled parabola with relatively high D value of 20 m(2)/s, whereas a significantly low D value of 3.5 m(2)/s characterises the aftershock activity for the first 24 h (when 420 events of mb >= 4.0 occurred). Here, the Coulomb stress was transferred from the main shock with a rapid imposition of normal stress, thus inducing the pore-pressure change that started diminishing almost immediately by fluid diffusion, at a rate, defined by the diminishing D value. The modelling results for fault seismicity at far off distances from the main epicentre are interpreted here as potential indicators for large-scale sub-seabed rupturing-consequent to stress changes induced by bending of the Indian Ocean plate. Bathymetric slopes under the Andaman subduction zone are particularly amenable to sub-marine slides where crustal E-W hinge faults inferred seismically cut across the N-S trending regional thrust and strike-slip faults. Seabed rupturing appears to allow deep-slab hydration in these areas, producing pressure gradients along the normal faults. These features are important since they can herald marine geohazards in the Andaman region.