3:30 PM - 3:45 PM
[SCG58-11] Numerical modeling of three-dimensional fluid migration in subduction zones
★Invited Papers
Keywords:subduction zones, fluid migration, slow earthquakes
In subduction zones fluid plays critical roles in triggering various types of earthquakes through increasing pore fluid pressure and in generating melt by lowering solidus of rocks. Many previous studies have revealed where and how much fluid is released from the slab mainly based on the thermal structure predicted by numerical modeling and the maximum water content of each rock type constituting oceanic plate. Little has been investigated, however, on the fluid migration after its release. Recent numerical studies have demonstrated that the fluid does not simply go up vertically due to its buoyancy, but it can move horizontally a long distance. In this presentation we show two examples of three-dimensional (3-D) fluid migration in subduction zones.
The first type of model focuses on the effects of anisotropic permeability. In relatively warm subduction zones the serpentinite in the mantle wedge may be sheared significantly due to the slab subduction just beneath it. It leads to a high permeability anisotropy of serpentinite, which allows the fluid within the serpentinite to migrate in the direction subparallel to the slab surface until it reaches the forearc Moho. We construct a numerical model which incorporates the effects of 3-D slab geometry and permeability anisotropy of serpentinite. In the serpentinite the permeability is assumed to be the same in any direction parallel to the slab surface, and it is higher than that in the direction perpendicular to the slab surface. We find that the fluid released from the slab goes upward in the maximum-dip direction of the slab. 3-D fluid focusing occurs where the slab bends away from the trench and hence the fluid volume fraction increases there. The observed 3-D fluid migration and the resultant along-arc variation in the fluid volume fraction may explain the spatial variation in the average slip rate by short-term slow slip events in Cascadia and southwest Japan.
The second type of model focuses on the effects of a resistance to volume change of rocks. When the fluid mobility, which is defined as the ratio of reference permeability to fluid viscosity, is high, a large amount of fluid is trapped within the fluid source and it migrates in the direction subparallel to the source. The fluid volume fraction in this case is low. When a low fluid mobility is assumed the fluid moves in the direction of slab subduction as a whole, and the fluid volume fraction is high. In either case, 3-D fluid focusing and the resultant increase in the fluid volume fraction are observed at the bend of the slab. These findings may help us explain a deepening of interplate earthquakes and the upper plane of double seismic zone at the junction of northeast Japan and Kurile arcs. 3-D fluid focusing may also make the subduction angle shallower where the slab bends away from the trench through decreasing bulk density.
The first type of model focuses on the effects of anisotropic permeability. In relatively warm subduction zones the serpentinite in the mantle wedge may be sheared significantly due to the slab subduction just beneath it. It leads to a high permeability anisotropy of serpentinite, which allows the fluid within the serpentinite to migrate in the direction subparallel to the slab surface until it reaches the forearc Moho. We construct a numerical model which incorporates the effects of 3-D slab geometry and permeability anisotropy of serpentinite. In the serpentinite the permeability is assumed to be the same in any direction parallel to the slab surface, and it is higher than that in the direction perpendicular to the slab surface. We find that the fluid released from the slab goes upward in the maximum-dip direction of the slab. 3-D fluid focusing occurs where the slab bends away from the trench and hence the fluid volume fraction increases there. The observed 3-D fluid migration and the resultant along-arc variation in the fluid volume fraction may explain the spatial variation in the average slip rate by short-term slow slip events in Cascadia and southwest Japan.
The second type of model focuses on the effects of a resistance to volume change of rocks. When the fluid mobility, which is defined as the ratio of reference permeability to fluid viscosity, is high, a large amount of fluid is trapped within the fluid source and it migrates in the direction subparallel to the source. The fluid volume fraction in this case is low. When a low fluid mobility is assumed the fluid moves in the direction of slab subduction as a whole, and the fluid volume fraction is high. In either case, 3-D fluid focusing and the resultant increase in the fluid volume fraction are observed at the bend of the slab. These findings may help us explain a deepening of interplate earthquakes and the upper plane of double seismic zone at the junction of northeast Japan and Kurile arcs. 3-D fluid focusing may also make the subduction angle shallower where the slab bends away from the trench through decreasing bulk density.