Interseismic stress accumulation at the locked zone of Nankai Trough seismogenic fault off Kii Peninsula

• Interseismic stress is numerically calculated in the Nankai Trough.
• Shear stress is largest at downdip of locked zone and gradually decreases seaward.
• Maximum shear stress is proportional to Young's modulus in the hanging wall.
• Maximum shear stress (~ 3 MPa) is comparable to subduction earthquake stress drop.
• At the updip edge, Effect of interseismic plate coupling does not exceed 0.1 MPa.

A primary source of stress variation prior to great subduction zone earthquakes is the tectonic loading due to plate convergence. We have calculated the elastic stress around the locked zone accumulated during an interseismic period numerically, using a 2D plane-strain finite element model in the Nankai Trough off Kii Peninsula, and examined its dependence on elastic properties and some boundary conditions. For the modeling, we used the geometry of subducting plate interface determined previously through seismic surveys and the physical properties that were obtained through IODP scientific drilling offshore Kii Peninsula. To construct a physically realistic model, a dislocation along the plate interface is calculated from a 5 m displacement imposed on one side of the subducting plate, with the locked zone defined at a depth of 10 to 25 km along the plate interface.Calculated shear stress is maximized at the downdip edge of the locked zone, decreases exponentially seaward, and is almost zero near the updip edge. Maximum shear stress is primarily proportional to the Young's modulus of the overlying section. Comparing the maximum shear stress with the average stress drop for subduction earthquakes (3–4 MPa), the most appropriate Young's modulus and Poisson's ratio ranges are 50–70 GPa and ~ 0.3, respectively, which are in good agreement with previous estimations from observed Vp/Vs. The stress accumulated during one seismic cycle can be reasonably explained by a linear elastic model with an appropriately sized locked region. We found that the envelope of simulated horizontal displacement on the surface agrees well with GPS velocity data, which also supports this inference.Disturbance near the updip edge is no larger than 0.1 MPa of stress or ~ 0.1μ-radian of tilt, which are difficult to detect with current borehole measurement techniques and likely complex borehole conditions.