next up previous print clean
Next: Conclusions Up: Vlad: Ocean-bottom seismometers Previous: OBS logistics aboard R/V

Other OBS models and comparison with OBC technology

The first OBS, built by Maurice Ewing in 1937, used rock salt for the release mechanism, containers with gasoline (incompressible and more lightweight than water!) for flotation and a pocket Hamilton watch for timing. It went as deep as 4500 m and is on display at the SEG Virtual Museum.[*] Today, NME operates a catamaran with a sonar array that can deploy 106 short-period OBSs for three months and retrieves them in 99$\%$ of cases. The technology has changed in the meanwhile, spinning off along the way broadband submersible instruments in the seventies and OBCs in the mid-nineties.

Currently the U.S. National OBS Instrument Pool (OBSIP) administers 135 short-period, 2-C (hydrophone and vertical geophone) OBSs and 109 4-C large broadband ones, designed for extended deployments longer than one year at a time. These instruments have been built and are maintained by Scripps Institute of Oceanography, Lamont-Doherty Earth Observatory and Woods Hole Oceanographic Institute, and are available for US academic and industrial use.[*] Broadband (0.00277 Hz- 50 Hz) OBSs are also produced by Guralp Systems. The U.S. instruments that the Japanese ones resemble most are the 19 OBSs maintained by the University of Texas Institute of Geophysics (UTIG), developed there in 1976 by scientists formerly involved in the lunar seismograph program. Another significant pool of instruments exists in Germany. It consists of approximately 40 long-period OBSs built by Geomar GmbH by adding an external moving arm that drops the 3-C geophone on the seafloor to their Ocean-Bottom Hydrophone (OBH). The OBH was originally designed as a ocean-bottom buoy, tall enough to be moved by water currents, which did not matter for the hydrophone, but does for the geophones. The University of Cambridge, UK, has developed a 4-C short-period, short-deployment (20 days) OBS of which 25 or more have been built. It resembles the Geomar design--taller than it is wide, and with a external arm that drops the geophones so that they are not shaken by water currents together with the rest of the tall instrument. Dalhousie University from Canada has a small number of short-period 4-C OBSs developed in-house. University of Durham, UK, has six in-house built short-period OBSs and the Monterey Bay Aquarium Research Institute has a few ROV-operated OBSs developed in-house too.

With the exception of the Japanese OBS, among the seismometers described above only the broadbands (Guralps and the long-deployment OBSIP ones) have both 4-C recording capabilities as well as reliability and recoverability demonstrated on a large pool of instruments. However, they are very expensive. The OBSIP short-period instruments are cheap (approx. $\$20$k) and reliable ($\gt 99\%$), but are only 2-C (vertical geophone and hydrophone). The UTIG ones seem less rugged and reliable than the Japanese OBSs, with instrument loss rates such as 3 instruments out of 33 (1997 Iberia experiment). The German short-period OBSs exhibit a lower recovery rate than the Japanese ones and, like the Cambridge ones, have external moving parts in order to correct for a design-induced problem. Other instruments (Dalhousie, Durham) look promising but they are larger and more complex than the Japanese OBSs, and very few of them have been built, so that extensive field testing (hundreds of individual deployments) has not occured. Should the seismic industry need an existing deep water 4-C receiver technology solution, the Japanese OBS seems to be the most appropriate among the instrument designs known to the author.

A comparison between OBCs and Japanese OBSs yields interesting results. They share many advantages. Both can be used for data acquisition in areas with obstacles such as platforms, reefs, and transition zones that do not allow for streamer operation. Both record 4-C data, which can be used to image through gas clouds and in anisotropic media, to map fractures, and to predict overpressure. Neither is sensitive to noise: the seafloor is quiet and as a bonus the upward going signal and downward going noise (including the receiver ghost) can be separated by processing in the manner developed by Barr and Sanders (1989).

It is in the disadvantages areas that the comparison shows dissimilarities. OBCs are currently limited to water depths of 500 m, with the most rugged going down to 1500 m. OBSs in contrast typically have a depth limitation of 6000 m. Whereas the cables sometimes have problems with withstanding deepwater pressures for extended periods of time, the OBSs have no problem. While OBCs suffer damage by strain from repeated retrievals due to repairs and moving surveys, the small and rigid OBSs are not harmed at retrieval. A localized event that destroys part of an OBC (submarine landslide, shark bite) can affect the entire cable, but OBSs are totally independent from each other. OBCs sometimes need burial by ROV for good coupling, which increases the costs and troubles at retrieval, while the weight of an OBS is sufficient to couple it well. But OBSs have disadvantages too, the first being a higher cost per deployed receiver. This may be alleviated by economy of scale if more units are produced, and by the longer life of a OBS as compared to a OBC. Another issue is that of drift during sinking, especially in deep water. The position can be determined satisfactorily with sonar arrays, but for industry purposes the deployment precision needs to be increased.


next up previous print clean
Next: Conclusions Up: Vlad: Ocean-bottom seismometers Previous: OBS logistics aboard R/V
Stanford Exploration Project
7/8/2003