3. Borehole Instrument Package¹

Shipboard Scientific Party²

SYSTEM OVERVIEW

The lack of in situ long-term observations in the oceans, which cover areas of major tectonic activity, limits our understanding of presently active geological processes. Since the beginning of the Deep Sea Drilling Project, there have been many attempts to utilize the drilled holes for observatory purposes. Recent circulation obviation retrofit kit deployments (CORKS) to measure pressure and temperature changes in sealed boreholes are beginning to produce interesting results. The Ocean Drilling Program (ODP) continues to recognize the importance of observatory objectives (ODP, 1996).

Many ODP sites have been drilled at plate subduction zones to understand the accretion processes, to tackle the mass balance problems, or to construct the history of the making of island arcs, among other objectives. A major unknown in the plate subduction process is its episodic nature. Our main goal for Leg 186 was to install two permanent observatories on the landward side of the Japan Trench directly above the seismogenic zone of the subduction plate boundary. The two sites, Sites 1150 and 1151, are located in presently seismically active and inactive zones, respectively.

At a plate subduction zone, there is coupling between the two plates, but the definition of coupling can be confusing. One measure is the seismic efficiency, which is the percentage of plate motion released by interplate thrust earthquakes. Plate motion is determined from global plate velocity models. For the Japan Trench zone, the seismic efficiency is probably <~30%. Another measure of coupling is how locked a portion of a plate boundary is relative to adjacent portions. A 100%-locked condition means that there is normally no differential motion across a part of the plate boundary. Eventually slip must occur, the presumption being that it will be manifest as a sizable earthquake. Since 1994, Global Positioning System (GPS) observations have not revealed any slow earthquakes comparable to or larger than the one that accompanied the 1994 Far-off Sanriku earthquake. The stress transfer across the plate boundary (shear coupling) is a third measure of coupling. In the Japan Trench, the plate motion direction largely governs the tectonic stress field in a wide area across the island arc to the Sea of Japan.

The Japan Trench Borehole Geophysical Observatories provide seismic and deformation data to help quantify these characteristics. The observatory system is named 'Neath Seafloor Equipment for Recording Earth's Internal Deformation (NEREID).

This section will outline the borehole geophysical observatory (Fig. F1). Each component is detailed in a separate section. The observatory is designed to last for many years, first as a stand-alone system, then as a system cabled to the shore. Comprised of a strainmeter, a tiltmeter, and two broadband seismometers, the sensors reside near the bottom of the hole with each housed in a separate pressure vessel. Four cables connect the sensors uphole. Except for one seismometer, Guralp Systems Ltd.'s CMG-1T, the signals are sent in analog form. The relative orientation of the sensors is fixed and known. The absolute orientation of the installed systems was not known at installation time but will be determined to sufficient accuracy from the recordings of earthquake waves by the seismometers.

The multiple-access expandable gateway (MEG) seafloor package digitizes and combines the analog-to-digital (A/D) data from downhole to one serial data stream. It also distributes the power to individual modules. The data are stored in digital format in the storage acquisition module (SAM) via an RS232 link using Guralp compressed format (GCF) protocol. The SAM has four 18-GB small computer system interface (SCSI) hard disks equivalent to more than seven channels of 1-yr-long continuous data of 24-bit dynamic range at 100-Hz sampling rate. In this case, there are three channels each for the seismometers (100 Hz for the CMG seismometer and 20 Hz for the PMD seismometer), two channels for the tiltmeter (4 Hz), and nine channels for the three-component and three channels for the single-component strainmeter (50-Hz sampling rate). The borehole sensors are permanently cemented in as required for the strainmeter operation and to assure good coupling for the other sensors. The MEG, on the other hand, may be serviced by a submersible remotely operated vehicle (ROV). The MEG can be physically replaced and accepts commands and software upgrades through the SAM. The SAM needs to be replaced by a submersible ROV before the disks become full, which takes ~1 yr in the present design.

The back from ocean bottom module (BOB) provides a communication link to the borehole system while the station is being serviced by a submersible ROV. The BOB can extract data through the SAM, calibrate the clock, and check the health of the system.

Except for the BOB, all power is supplied by the seawater battery (SWB) system, which can provide ~18 W with more than 300-kWhr capacity. Its source of energy is the magnesium anode, which needs to be replaced once it is consumed by electrolytic dissolution. Replacement of the magnesium anode is also handled by a submersible ROV.

The Japan Trench is probably one of the most suitable sites for testing various ideas associated with the episodic nature of plate subduction and its relation to earthquakes and tectonics. Although the seismogenic zone itself cannot be reached by present drilling capability, the present target depth of ~1 km below seafloor provides stations within 10 km of the seismogenic zone of the plate boundary. Since the plate dip angle is very shallow (<10°), a wide area can be viewed by these stations. The nearest geodetic stations are coastal stations ~50 km above the plate boundary; that part of the plate interface may be below the lower-dip end of the seismogenic zone.

Site 1150 and Site 1151 can be considered as two end-members in terms of seismicity. The background seismicity is high around Site 1150 but the area around Site 1151 is aseismic. The former has experienced M7-class events, but there is no historical record of large earthquakes rupturing the area of Site 1151. Although both the microseismicity and large events have contrasting characteristics, there is no compelling evidence of structural differences between the two sites.

For high-sensitivity measurements, pressure fluctuations caused by ocean long waves or temperature changes can be sources of noise, as can water motion near the sensor. Thus the sites are located at key geophysical coordinates and in boreholes for noise reduction. Because of its principle of operation, the strainmeter needs to be grouted inside the borehole to respond accurately to strain changes in the surrounding rock.