11:30 〜 11:45
[SCG45-19] Constraining the State of Locking and Current Absence of Slip of N Cascadia Megathrust Based on Offshore Formation Fluid Pressure Monitoring
キーワード：subduction megathrust locking, offshore geodesy, fluid pressure monitoring, borehole observatory, lack of dynamic triggering, event detection threshold
The Cascadia subduction megathrust off the Pacific Northwest is known to host large (M~9) tsunamigenic earthquakes. Evaluating potential seismic and tsunami hazards requires a complete understanding of the present-day locking state of the fault. Existing land-based geodetic observations cannot resolve the shallow fault behavior far offshore, and direct sub-marine measurements are needed. Here, we report the usage of long-term (~12 year) monitoring of formation fluid pressure – as a proxy of volumetric strain – in IODP Hole U1364A, to constrain the near-trench locking state (slip deficit) of the northern Cascadia megathrust.
The borehole observatory is located at ~20 km landward of the subduction deformation front, with seafloor and formation-fluid pressures monitored at multiple levels (down to 304 m beneath seafloor), with sensors at the wellhead connected to permeable screens at depths. Following the installation of the observatory in 2010, pressure recordings were made at a sampling rate of once per minute, until its connection to Ocean Networks Canada’s NEPTUNE cable in 2017 when the sampling rate increased to once per second. The raw pressure record is dominated by tidal signals, with additional variations originating from smaller-scale oceanographic phenomena and hydrological perturbations. Large excursive pressure transients, sometimes with opposite signs between monitoring levels, were seen in the early monitoring period (several years). At shallower monitoring levels, pressure excursions have persisted, likely due to the dynamics of gas migration in the vicinity of a local gas-hydrate stability boundary. At the deepest monitoring level S1, however, pressure noise largely abated after mid-2015, and the subsequent record is of sufficiently high quality to detect potential tectonically driven pressure transients.
To assess the presence or absence of triggered slip, we examine the S1 record at the times of regional and distant large earthquakes, during which intensive ground shaking can cause high dynamic stressing. Our examination of multiple events, which caused estimated shear stress perturbations along the base of the accretionary prism of up to >20 kPa, reveals a lack of pressure transients or offsets. This stands in stark contrast with pressure monitoring by similar borehole observatories at other subduction zones such as Nankai, Costa Rica, and Hikurangi, where monitoring spanning comparable or shorter durations all have captured multiple tectonically driven pressure transients.
To better quantify the megathrust locking state, we study the noise floor of the processed level S1 pressure records. Statistical analysis of the data noise suggests a “threshold” of 0.08 kPa – equivalent to 16 nanostrain for shallow accretionary-prism sediments – for detecting tectonically driven pressure transients. We use this “detection threshold” with simple dislocation modeling using the local fault geometry, to determine the magnitude of “resolvable slip” along the megathrust. Model tests of hypothetical slip, done by systematically varying slip-patch size and location, suggest that our data can resolve minimal slip of <1 cm directly beneath the borehole. For large slip patches such as with a radius of ~50 km, the borehole data can resolve ~10 cm slip from 100 km away in the strike direction, covering an along-strike extent comparable with the size of Vancouver Island. This high detection sensitivity, and the lack of events detected over the eight-year noise-free record, indicate that the northern Cascadia subduction fault remains fully locked during the monitoring period. Our observation provides an important end-member case among the global observations of subduction fault slip behavior. While slow slip has been detected at many places, the Cascadia subduction zone appears to be rather unique in producing no slip at its “late inter-seismic” stage.
The borehole observatory is located at ~20 km landward of the subduction deformation front, with seafloor and formation-fluid pressures monitored at multiple levels (down to 304 m beneath seafloor), with sensors at the wellhead connected to permeable screens at depths. Following the installation of the observatory in 2010, pressure recordings were made at a sampling rate of once per minute, until its connection to Ocean Networks Canada’s NEPTUNE cable in 2017 when the sampling rate increased to once per second. The raw pressure record is dominated by tidal signals, with additional variations originating from smaller-scale oceanographic phenomena and hydrological perturbations. Large excursive pressure transients, sometimes with opposite signs between monitoring levels, were seen in the early monitoring period (several years). At shallower monitoring levels, pressure excursions have persisted, likely due to the dynamics of gas migration in the vicinity of a local gas-hydrate stability boundary. At the deepest monitoring level S1, however, pressure noise largely abated after mid-2015, and the subsequent record is of sufficiently high quality to detect potential tectonically driven pressure transients.
To assess the presence or absence of triggered slip, we examine the S1 record at the times of regional and distant large earthquakes, during which intensive ground shaking can cause high dynamic stressing. Our examination of multiple events, which caused estimated shear stress perturbations along the base of the accretionary prism of up to >20 kPa, reveals a lack of pressure transients or offsets. This stands in stark contrast with pressure monitoring by similar borehole observatories at other subduction zones such as Nankai, Costa Rica, and Hikurangi, where monitoring spanning comparable or shorter durations all have captured multiple tectonically driven pressure transients.
To better quantify the megathrust locking state, we study the noise floor of the processed level S1 pressure records. Statistical analysis of the data noise suggests a “threshold” of 0.08 kPa – equivalent to 16 nanostrain for shallow accretionary-prism sediments – for detecting tectonically driven pressure transients. We use this “detection threshold” with simple dislocation modeling using the local fault geometry, to determine the magnitude of “resolvable slip” along the megathrust. Model tests of hypothetical slip, done by systematically varying slip-patch size and location, suggest that our data can resolve minimal slip of <1 cm directly beneath the borehole. For large slip patches such as with a radius of ~50 km, the borehole data can resolve ~10 cm slip from 100 km away in the strike direction, covering an along-strike extent comparable with the size of Vancouver Island. This high detection sensitivity, and the lack of events detected over the eight-year noise-free record, indicate that the northern Cascadia subduction fault remains fully locked during the monitoring period. Our observation provides an important end-member case among the global observations of subduction fault slip behavior. While slow slip has been detected at many places, the Cascadia subduction zone appears to be rather unique in producing no slip at its “late inter-seismic” stage.