Earl E Davis’s research while affiliated with Natural Resources Canada and other places

Seafloor and subseafloor temperature and pressure monitoring has been carried out with CORK (“Circulation Obviation Retrofit Kit”) sealed deep-ocean borehole observatories since their original design and deployments in 1991 in hydrologically and tectonically active settings. In this paper, we review results from 20 installations at six subduction zones that have provided novel insights into the thermal and hydrologic state, time-dependent deformation, and elastic properties of the subseafloor. Monitoring depths reached the levels of the megathrust faults beneath the outer subduction prisms at Nankai, Costa Rica, and Barbados, but only at the last were substantially elevated average pressures observed (excess fluid-pressure ratios of ∼0.3 and 0.5). Otherwise, all records are characterized by fluid pressures only slightly super-hydrostatic. Transient pressure anomalies are observed to be common within both the subduction prisms and the incoming subducting plates. These reflect deformation (volumetric strain changes) caused by distant seismogenic fault slip at the time of large earthquakes, and by more local slow slip that is seen to propagate to the outermost reaches of the prisms from greater depths at rates of several kilometers per day. No such activity is seen at the Cascadia prism site, however, suggesting that its subduction fault is more completely locked. Where CORK observatories are co-located with or near seismometers and other seafloor monitoring instruments and are connected to cable systems for power, real-time high-rate data acquisition, and precise timing, the observations allow relationships among seismic ground motion, seafloor pressure, and formation pressure to be defined. These, in turn, constrain formation elastic properties, and confirm the efficiency and fidelity with which strain is converted to formation pressure.

We review the contributions to our understanding of the hydrogeology of the oceanic crust as gained by monitoring subseafloor temperatures and pressures in CORK (“Circulation Obviation Retrofit Kit”) sealed hole observatories that penetrate through marine sediments into igneous basement. CORKs were installed during 1991–2012 in 17 holes drilled by the Ocean Drilling Program and Integrated Ocean Drilling Programs in igneous oceanic crust (aged 0.4–24 Ma) in the thickly sedimented eastern Pacific Ocean and in the thinly sedimented north Atlantic Ocean. Results indicate that the igneous crust in all these settings is highly transmissive and supports vigorous lateral fluid flow and (or) convection beneath the sediment cover driven by very small horizontal pressure gradients. Inter-CORK experiments suggest that the horizontal permeability of the young oceanic basement at these sites may be anisotropic, with higher values in the direction parallel to the axis of spreading and tectonic fabric and lower values in a ridge-normal cross-strike direction. Tracer experiments applied to ridge-parallel flow suggest that only a small fraction of the crust is well connected. CORKs also record the response of subseafloor formation fluid pressures to seafloor tidal loading, and, once the tidal responses are filtered out, response to tectonic strain events. Interpretations based on these responses and original two-dimensional modelling of ridge-normal fluid circulation in the crust suggest a scale dependence of permeability of igneous oceanic crust (higher values at larger scales up to tens of km), although recent three-dimensional models suggest a more limited scale effect. CORKs have also provided important geochemical and microbiological results from the igneous oceanic crust, although they are not reviewed in this paper.

The Cascadia subduction megathrust off the Pacific Northwest follows an “end member” seismogenic behavior, producing large (up to moment magnitude 9) but infrequent (every several hundred years) earthquakes and tsunamis. Crustal deformation associated with the ongoing plate convergence has been characterized by land‐based geodetic observations, but the state of locking across the full breadth of the seismogenic fault is poorly constrained. We report results of offshore monitoring of borehole fluid pressure, as a proxy for formation volumetric strain, at a site ∼20 km landward of the Cascadia subduction deformation front since 2010. The multi‐depth pressure records were plagued by hydrologic noise, but noise at the deepest monitoring level (303 m sub‐seafloor) abated in 2015. Subsequently, including at the times of regional large earthquakes that caused significant dynamic stressing, no persistent pressure transients are present above a threshold of 0.08 kPa imposed by unremovable oceanographic signals, corresponding to a strain detection limit of ∼16 nanostrain. Simple dislocation models using local megathrust geometry suggest a resolvable slip of <1 cm along a trench‐normal corridor beneath the borehole for a range of slip‐patch dimensions. A large slip patch can be well resolved even at considerable along‐strike distances from the borehole; for instance, ∼10 cm slip is detectable over a 200‐km strike range for a slip‐patch radius of ∼50 km. This high sensitivity for detecting slip, along with the lack of observed events, stands in stark contrast to observations at other subduction zones, and suggests that the Northern Cascadia megathrust is most likely fully locked.

Ground shaking caused by earthquakes is accompanied by seafloor and sub-seafloor formation fluid pressure variations in offshore areas, but there have been few collocated observations of these signals. In this work, we report seismic and high-sampling-rate fluid pressure records of the 2021 Mw 8.2 Alaska earthquake by the Ocean Networks Canada (ONC) NEPTUNE observatory at an epicentral distance of ∼2,200 km in the northeast Pacific Ocean. The system comprises observatory nodes in various tectonic environments, with each node including buried broadband seismometers, seafloor pressure sensors, and, at two nodes, borehole pressure sensors. Seismic and tsunami waveforms of the Mw 8.2 earthquake were documented in detail. Seismic seafloor pressure variations (Psf) were dominated by Rayleigh waves of periods between 5 and 50 s, with peak amplitudes of 3–4 kPa at most sites. Waveform similarity and the linear scaling between Psf and vertical ground acceleration indicate forced acceleration of the water column being dominant in governing Psf during long-period surface-wave arrivals, with an additional component of elastic oscillation occurring at higher frequencies (>0.1 Hz) causing extra pressure signals. Analysis of formation pressure variations due to various types of ocean loading of distinctly different frequencies (e.g., tides, tsunami, and infragravity waves) shows stable one-dimensional vertical loading efficiencies that depend on lithology at each borehole site, with loading response being strongly influenced by the presence of free gas at shallow depths within the Cascadia accretionary prism. Inter-site comparisons of seismic and seafloor pressure waveforms demonstrate a key role of sediment thickness in the amplification of surface wave amplitudes.

Six-year records of ocean bottom water temperatures at two locations in an isolated, sedimented deep-water (∼4500 m) basin on the western flank of the mid-Atlantic Ridge reveal long periods (months to >1 year) of slow temperature rises punctuated by more rapid (∼1 month) cooling events. The temperature rises are consistent with a combination of gradual heating by the geothermal flux through the basin and by diapycnal mixing, while the sharper cooling events indicate displacement of heated bottom waters by incursions of cold, dense bottom water over the deepest part of the sill bounding the basin. Profiles of bottom water temperature, salinity, and oxygen content collected just before and after a cooling event show a distinct change in the water mass suggestive of an incursion of diluted Antarctic Bottom Water from the west. Our results reveal details of a mechanism for the transfer of geothermal heat and bottom water renewal that may be common on mid-ocean ridge flanks.

Plain Language Summary Hydrothermal circulation in the oceanic crust plays fundamental roles in the exchange of water, heat, and chemical constituents between the oceanic hydrosphere and lithosphere, in the hydration of the crust and upper mantle, and in the support of the subseafloor microbial biosphere. Water is carried primarily in faults and fractures created at the time of seafloor spreading, and thus constraints on their distribution and orientation are important. Such information has been provided by active‐source seismic surveys, and results have shown that compressional wave speeds in the directions of spreading—thus across local faults and fractures—are typically <20% lower than those along the structural fabric. Here, we report more direct determinations of the anisotropy at the 3.6‐Ma Juan de Fuca plate, using formation fluid pressure variations in sealed subseafloor boreholes caused by seismic surface waves from distant earthquakes. These observations are sensitive to formation compressibility of the upper couple of hundreds of meters of the igneous crust. The observed azimuthal variation in fluid pressurization reveals a contrast in compressibility across versus along the structural trend of roughly a factor of five, and a much higher degree of seismic anisotropy (>50%) than that determined from standard seismic velocity measurements.

Ground‐motion amplification in the outer Cascadia subduction zone accretionary prism has been documented previously by comparing earthquake and microseismic vertical ground‐motion and pressure records from an ONC/NEPTUNE Canada cabled observatory site on the outer prism to those from a site on the flank of the Juan de Fuca Ridge. Since then, four additional instruments became operational, and data, now spanning 10 years, include more than 100 large (Mw > 7) distant earthquakes. Well‐tuned response at the outer prism sites is observed in both vertical and horizontal components, with peaks in the spectral ratio of vertical velocity relative to nearby ocean crustal and continental bedrock sites (V/Vref) at 9 s, and in the intra‐site horizontal to vertical spectral ratio horizontal‐to‐vertical spectral ratio at 14 s, both with band widths of ±10% to 15% at half amplitude. The vertical response is consistent with ¼ wavelength compressional wave reinforcement, while the horizontal motion tuning most likely reflects the effects of the low velocity prism sediments on surface‐wave propagation. At the periods of maximum relative motion, outer prism surface‐wave vertical and horizontal accelerations are enhanced by factors of up to 15 and 25, respectively. Similar behaviour is seen in microseismic records from these sites, and, to a lesser extent, in earthquake records from temporarily deployed Cascadia Initiative seismometers along the outer Cascadia prism to the south. Such tuning and amplification must be accounted for when assessing the dynamic behavior of the prism and its basal fault at the time of large local earthquakes.

We report 4 years of temperature profiles collected from May 2014 to May 2018 in Integrated Ocean Drilling Program Hole U1364A in the frontal accretionary prism of the Cascadia subduction zone. The temperature data extend to depths of nearly 300 m below seafloor (mbsf), spanning the gas hydrate stability zone at the location and a clear bottom-simulating reflector (BSR) at ∼230 mbsf. When the hole was drilled in 2010, a pressure-monitoring Advanced CORK (ACORK) observatory was installed, sealed at the bottom by a bridge plug and cement below 302 mbsf. In May 2014, a temperature profile was collected by lowering a probe down the hole from the ROV ROPOS. From July 2016 through May 2018, temperature data were collected during a nearly two-year deployment of a 24-thermistor cable installed to 268 m below seafloor (mbsf). The cable and a seismic-tilt instrument package also deployed in 2016 were connected to the Ocean Networks Canada (ONC) NEPTUNE cabled observatory in June of 2017, after which the thermistor temperatures were logged by Ocean Networks Canada at one-minute intervals until failure of the main ethernet switch in the integrated seafloor control unit in May 2018. The thermistor array had been designed with concentrated vertical spacing around the bottom-simulating reflector and two pressure-monitoring screens at 203 and 244 mbsf, with wider thermistor spacing elsewhere to document the geothermal state up to seafloor. The 4 years of data show a generally linear temperature gradient of 0.055°C/m consistent with a heat flux of 61–64 mW/m². The data show no indications of thermal transients. A slight departure from a linear gradient provides an approximate limit of ∼10⁻¹⁰ m/s for any possible slow upward advection of pore fluids. In-situ temperatures are ∼15.8°C at the BSR position, consistent with methane hydrate stability at that depth and pressure.