10:00 AM - 10:15 AM
[SSS07-17] What causes a local decrease in effective pressure in the source region of shallow slow earthquakes?
Keywords:effective pressure, hydraulic modeling, subduction zone, slow earthquake
Rock strength increases linearly with increasing effective pressure Pe (total pressure minus pore-fluid pressure) in the brittle regime. Particularly in plate subduction zones, the relationship between local decreases in Pe and slow earthquakes has been reported based on the seismological and geodetic observations. Although Rice (1992) realized constant Pe at depth by modeling an upward fluid flow along a vertical strike-slip fault zone whose permeability is a rapidly decreasing function of Pe, appropriate models for other tectonic settings have not yet been developed. The development of such models for subduction zones enables us to quantitatively evaluate the local decreases in Pe in subduction zones, and its consequence would deepen the discussion regarding the generation of slow earthquakes.
In order to quantitatively examine the causes of the local decrease in Pe on the shallow section of the subduction zone, we performed hydraulic modeling that incorporated mechanisms characteristic to subduction zones. We focused on the Kumano Transect in the Nankai Trough as a representative subduction zone, and referred to the experimental results using the core samples to determine the physical properties for the model calculations. Our basic model considers smectite dehydration and the mechanical effect (e.g., compaction) of subduction on the sediments. As in Rice (1992), physically sound solutions of the model show that the gradient of Pe remarkably decreases with increasing depth, whereas the realistic fluid properties rule out the nearly constant Pe at depth. We obtained a monotonic increase in Pe with increasing depth, and failed to generate a locally low Pe. It was also absent, despite considering fluid leakage through a splay fault. In the case with a local decrease in permeability, possibly owing to silica cementation, locally low Pe is realized around the precipitation area. Therefore, the local decrease in permeability is a possible candidate for the occurrence of shallow slow earthquakes. The water release caused by the dehydration reaction of smectite is not likely to be the dominant factor on Pe, although smectite dehydration releases silica and promotes its precipitation.
In order to quantitatively examine the causes of the local decrease in Pe on the shallow section of the subduction zone, we performed hydraulic modeling that incorporated mechanisms characteristic to subduction zones. We focused on the Kumano Transect in the Nankai Trough as a representative subduction zone, and referred to the experimental results using the core samples to determine the physical properties for the model calculations. Our basic model considers smectite dehydration and the mechanical effect (e.g., compaction) of subduction on the sediments. As in Rice (1992), physically sound solutions of the model show that the gradient of Pe remarkably decreases with increasing depth, whereas the realistic fluid properties rule out the nearly constant Pe at depth. We obtained a monotonic increase in Pe with increasing depth, and failed to generate a locally low Pe. It was also absent, despite considering fluid leakage through a splay fault. In the case with a local decrease in permeability, possibly owing to silica cementation, locally low Pe is realized around the precipitation area. Therefore, the local decrease in permeability is a possible candidate for the occurrence of shallow slow earthquakes. The water release caused by the dehydration reaction of smectite is not likely to be the dominant factor on Pe, although smectite dehydration releases silica and promotes its precipitation.