Episodic deformation and inferred slow slip at the Nankai subduction zone during the first decade of CORK borehole pressure and VLFE monitoring

- Ten-year formation fluid pressure records from the Nankai subduction zone are presented.
- Records include coseismic elastic strain caused by Tohoku earthquake 950 km away.
- Post-seismic pressure anomalies and VLFEs suggest triggered slow slip at Nankai.
- Incremental shear stress responsible for triggering is low, ca. 7 kPa.
- Numerous concurrent slip and VLFE swarms show Nankai to be close to a state of failure.

With new data recovered in December 2011 and November 2012 from two Ocean Drilling Program (ODP) CORK (circulation obviation retrofit kit) borehole observatories in the toe of the Nankai subduction zone accretionary prism (Hole 808I) and in the subducting Philippine Sea plate (Hole 1173B) off Southwestern Japan, records of formation fluid pressure now span over 10 yr. Over nearly this same period of time the Japanese terrestrial HiNet array has enabled detection of small earthquakes across the breadth of the accretionary prism. The records include several formation fluid pressure anomalies and concurrent local very-low-frequency earthquake (VLFE) swarms. In the subducting plate, pressure anomalies are most commonly slow negative steps that are inferred to reflect dilatational strain associated with slow slip on the subduction thrust fault in the areas of VLFE activity. In three instances, concurrent positive impulsive anomalies are observed in the prism; these are inferred to reflect contraction when slip reaches the location of Hole 808I. The spatial distribution of VLFEs suggests that slow slip may occur in patches that cumulatively span the seaward half of the subduction prism. Two anomalies occurred at times of the largest earthquakes in the region, the September 2004 Mw 7.7 off-Kii earthquake 220 km to the northeast, and the March 2011 Mw 9.0 Tohoku earthquake roughly 900 km to the northeast. In the subducting plate, the observed changes in pressure are abrupt and consistent in sign with the expected coseismic strain (contractional and dilatational, respectively). In the case of the Tohoku earthquake, the co-seismic pressure decrease was followed 12 days later by an additional but slow decrease, and by an impulsive but complex and long-lived increase in the accretionary prism Hole 808I. In this instance, co-seismic cross-strike stress may have triggered post-seismic slip and the deformation seen in the plate and prism. Given the great distance from the Tohoku epicenter and small change in stress estimated at the Nankai observatory sites (ca. 7 kPa shear stress), triggering of local slip would require the outer subduction thrust and overlying prism in this region to be very fragile. To what degree this state might vary with time is not known, but tracking episodic slip in the seaward part of this and other subduction zones using formation-fluid-pressure and VLFE monitoring may provide valuable clues about the evolution of subduction faults through their thrust earthquake cycles.