PART II: NERO PROJECT
INTRODUCTION
Seismic data from a worldwide network (WWSSN) established in the early 1960s accelerated advances in seismology and were a great resource of new discoveries up to the 1970s. During the past ten years, our knowledge of the processes of the deep Earth has been greatly improved by the development of new generations of global monitoring networks in seismology and geodesy and the continuation of long-term observations in geomagnetism (GEOSCOPE, IRIS, GeoFon on a global scale; and MedNet, Poseidon, CDSN, GRSN on a regional scale). While the quantity and quality of data have increased, this new information has revealed that there are large departures from lateral homogeneity at every level from the Earth's surface to its center. The intensive use of broadband data has provided remarkable seismic tomographic images of Earth's interior. These models are now routinely used in geodynamics for earthquake studies and to obtain the complex time histories of the inhomogeneous earthquake faulting related to tectonics. Improvements in the observatory locations for seismology, geodesy, and geomagnetics, particularly in the oceans, can greatly enhance our understanding of the Earth's interior.
Installing a reentry cone and casing down to basement is the first step toward the installation of a Geophysical Ocean Bottom Observatory (GOBO). The seismometer instrumentation will be installed at a later date. This observatory will be part of the future network of seafloor observatories proposed in the International Ocean Network (ION) program. The selected site on the Ninetyeast Ridge (Fig. 1, Part I) should not produce any technical problems, as previous holes in this area were drilled with a single bit (ODP Sites 756 and 757 during Leg 121 in June 1988). Establishing this cased reentry hole will require up to a week of ship time.
The installation of ocean bottom seismic stations, their maintenance, and the recovery of data on a timely and long-term basis represent a formidable technical challenge. However, different pilot experiments carried out by Japanese (Kanazawa et al., 1992; Suyehiro et al., 1992), French (Montagner et al., 1994a, b), and American groups (OSN1, Dziewonski et al., 1992; Orcutt, pers. com., 1997) demonstrate that there are technical solutions to all the associated problems.
The technical goal of the French Pilot Experiment OFM/SISMOBS (Observatoire Fond de Mer) conducted in April and May 1992 was to show the feasibility of installing and recovering two sets of three-component broadband seismometers (one inside an ODP borehole and another inside an OBS sphere in the vicinity of the hole). Secondary goals were (1) to obtain the seismic noise level in the broadband range 0.5-3600 s, (2) to conduct a comparative study of broadband noise on the seafloor, downhole, and on a continent, and (3) to determine the detection threshold of seismic events. A complete description of the experiment can be found in Montagner et al. (1994a) and a drawing is presented in Figure 3.
After the installation of both sets of seismometers, seismic signals were recorded continuously during 10 days. The analysis of these signals shows that the seismic noise is smaller in the period range 4-30 s for both ocean floor seismometers (OFS) and downhole seismometers (DHS) than in a typical broadband continental station such as spinning sidebands (SSB). The noise is still smaller than the noise at SSB up to 600 s for OFS. The noise on vertical components is much smaller than on the horizontal ones. The difference might be explained by instrument settling. It was also observed that the noise level tends to decrease as time goes by for both OFS and DHS, which means that the equilibrium stage was not yet attained by the end of the experiment (Beauduin et al., 1996a,b). The patterns of microseismic noise in oceanic and continental areas are completely different. The background microseismic noise is shifted toward shorter periods for OFS and DHS compared to a continental station. This might be related to the difference in the crustal structure between oceans and continents. The low level of seismic noise implies that the detection threshold of earthquakes is very low and it has been possible to correctly record teleseismic earthquakes of magnitude as small as 5.3 (Montagner et al., 1994b). It was also possible to extract the earth tide oceanic signal. Therefore, the experiment was a technical and scientific success and demonstrated that the installation of a permanent broadband seismic and geophysical observatory at the bottom of the seafloor is now possible and can provide the scientific community with high quality seismic data.
The Indian Ocean is considered to be the most complex of the Earth's oceans. Since the 1970s, magnetic anomalies, fracture zone information, and other geophysical information (McKenzie and Sclater, 1971; Norton and Sclater, 1979; Schlich, 1982; Royer and Sandwell, 1989) have been used to understand the tectonic history of the Indian Ocean, which is characterized by irregularities in kinematic behavior (e.g., ridge jumps, reorganization of the ridge system, asymmetric spreading, spreading velocity changes, and finally collision between India and Asia). Few tomographic investigations have been performed so far in the Indian Ocean (Montagner, 1986; Montagner and Jobert, 1988; Debayle and Lévêque, in press). These studies display a good correlation between surface tectonics and seismic velocities down to 100 km (Fig. 5), but there seems to be some offset at larger depth for the Central Indian Ridge, as a consequence of the decoupling between the lithosphere and the underlying mantle. This complexity at large depth is also present in global tomographic models. However, the lateral resolution is still quite poor and it makes it necessary to increase the station coverage of oceanic areas. The next step in tomographic techniques regards the simultaneous use of surface waves and body waves. By installing only one station in the Central Indian Ocean, it will be possible to obtain direct measurements of delay times and, therefore, unique and fundamental information on the local anisotropy (from SKS splitting), particularly for the 410 km and 660 km discontinuities (from converted seismic waves) and for pure oceanic paths. As shown in Figure 6, the future observatory is well surrounded by seismically active areas. This ensures there will be a reasonable amount of data within one or two years.
The approximate location of the observatory should be around 28°S, 90°E. It will complete the coverage of the Indian Ocean provided by stations RER, CRZ, PAF, AIS of the GEOSCOPE network. The precise location is not crucial. However, in order to secure the drilling of the borehole, we propose a site close to previous drilling sites. A second constraint on the site location is the need of a sufficient thickness of sediments. This last condition can be easily satisfied by a site on the Ninetyeast Ridge. Close to this point, two sites occupied on ODP Leg 121 fulfill these constraints:
ODP Site 756
Site 756 is located at 27°21.30'S, 87°35.85'E. This site was surveyed in September 1986 as part of the Robert Conrad Cruise 2708 (RC 2708). Site survey information is in the Leg 121 Initial Reports volume (Shipboard Scientific Party, 1989a,b). Conrad and JOIDES Resolution tracks are shown in Figure 7, and an example of a seismic profile is shown in Figure 8. At this site, sediment thickness is 139 m. The issue of basement penetration is largely dependent on the nature of the rocks and the need to avoid hydrothermal circulation. To facilitate the future installation of a GOBO, it is necessary to penetrate 200 m into basement.
ODP Site 757
Site 757 is located at 17°01'S, 88°11'E (DSDP Site 253 is located at 24°52.65'S, 87°21.97'E). The area near Site 757 was surveyed in August 1986 as part of Robert Conrad Cruise 2707 (RC 2707). Tracks and an example of seismic profile are presented in Figures 9 and 10. The thickness of sediments is about 370 m. Since the drilling conditions in this area were excellent, it is likely that basement penetration of 100 m should be sufficient. The hole must be cased down to basement with a reentry cone attached at the top. Whichever of these two sites is finally selected, the basement part of the hole will be cored.