Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
3 Status of Planning for Proposed Research-Oriented Ocean Observatories This chapter offers an assessment of the readiness of scientific plan- ning and technical development for construction and installation of each the three main infrastructure components of the OOI. A summary of the scientific and technical readiness for each component is presented after each section (Box 3-2, Box 3-4, and, Box 3-6~. GLOBAL OBSERVATORY PLANNING Global observatory science can be loosely divided into two catego- ries, according to whether the data are obtained from the seafloor or the water column and sea surface. Data from the seafloor include geophysical information about earthquakes and the structure of the Earth, investiga- tions of volcanic and tectonic activity, and studies of life on or below the seafloor. The data collected in the water column and at the sea surface are used for research on weather and climate, interactions between the atmo- sphere and ocean, and ocean physics, chemistry, and biology. Global geophysicists are particularly interested in sites far from is- land-based or other land-based observatories in order to provide more uniform sampling of the Earth's interior. Oceanographers are more inter- ested in measuring air-sea fluxes, investigating water mass formation, measuring transport and variability of major current systems, and study- ing the variability of the ocean's interior. While the sites of greatest inter- est to these various disciplines often do not coincide, the cost and limited availability of observatories in the more remote parts of the world's oceans 37
38 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY argue strongly in favor of locating global observatories where many re- search goals can be concurrently and productively pursued. Studies that cannot be supported by fixed observatories will likely find support in Lagrangian and satellite observing systems. The global observatory component of the OOI will provide important long-term, interdisciplinary measurements at 15-20 widely separated sites in all of the world's major oceans. A high priority is occupying sites in the remote Southern Oceans that are currently not sampled. This network will complement and enhance other international global observatory ef- forts (see Chapter 6~. The following discussion summarizes the status of scientific planning and technical development for the OOI global obser- vatory network. Status of Scientific Planning The scientific rationale for a global network of ocean observatories is well defined and planning for this network is well advanced, including the identification of specific sites for these observatories (Table 3-1~. The primary drivers behind this planning effort are international scientist com- munities involved in climate, global carbon cycle, biological and bio- geochemical, and solid earth geophysical research. Major research pro- grams such as WOCE, IGOFS, CLIVAR, RIDGE, ODP, GSN, the Global Ocean Ecosystem Dynamics Program (GLOBEC), the Carbon Cycle Sci- ence Program (CCSP), and the Surface Ocean Lower Atmosphere Study (SOLAS), have all pointed to the establishment of long time-series sites as critical elements of their research strategy. International groups provid- ing guidance on the development of a global observatory network in- clude the Ocean-Observing System Development Panel (OOSDP), the Ocean Observations Panel for Climate (OOPC), the CLIVAR Ocean Ob- servations Panel (COOP), the DEOS steering committee, and the Interna- tional Ocean Network (ION). Each has laid foundations for developing global arrays. These groups recognize the unique attributes of fixed observatories, which include high temporal and vertical resolution, the ability to ob- serve from the sea surface to the seafloor, and their presence on site through episodic events and in conditions difficult for ships. Such groups have indicated that fixed observatories would be an essential element of the observational approach needed to address their science foci (see Box 3-1~. In each case, strong arguments have been made to justify the cost of establishing long-term observatories, using a global reference frame that applies modeling, remote sensing, and other, already-available sampling methods to complement the observatory network (National Research Council, 2000~. It is particularly important to test the success of models at
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 39 replicating the variability observed in long time-series collected around the globe in different characteristic regimes. For example, satellite remote sensing helps numerical weather prediction models predict the global pattern of weather systems with greater accuracy, but such models have biases and error in sea-level meteorology and air-sea fluxes. With only one OWS (a ship) now operating and few open ocean time-series of sur- face meteorology available from buoys, there is no way to systematically address the size of the errors, why the models fail, and how to improve the model physics and performance in different regimes (such as in stratus cloud deck regions, in the trade winds, in the tropics, in the Southern Ocean, etc.~. Global observatories would make available time-series cru- cial to developing better air-sea flux fields and improving atmospheric general circulation models. An international Time-series Science Team (TSST) now coordinates the planning efforts of many of the groups interested in these scientific questions. Formed by CLIVAR and GOOS (through the COOP and the GOOS OOPC), and endorsed by the Partnership for Observation of the Global Ocean (POGO), the TSST represents the ocean community's di- verse scientific interests and is developing consensus on the desired loca- tions for time-series stations. In addition, it is charged with beginning the implementation of different components of an integrated global ocean- observing system. The TSST has identified a number of potential sites located at the intersection of regions of interest for multiple disciplines, giving priority to those sites where a shared interdisciplinary infrastruc- ture will provide cost-effective observing systems that should be afford- able and functional for decades to come. The DEOS Steering Committee has used the work of the TSST to identify preliminary locations for 20 moored-buoy observatories that could comprise the global network component of the OOI (Figure 3-1, Table 3-1~. The criteria for selecting these sites are outlined in the DEOS Global Network Implementation Plan (DEOS Moored Buoy Observatory Working Group, 2003~. These 20 sites are a subset of a longer list identi- fied to date by the TSST (Appendix E). Some of the sites on that longer list are already funded; the OOI contribution of an additional 20 sites would represent a dramatic increase in capability as well as a major step toward a global array of ocean observatories. Status of Technical/Engineering Development and Planning There are two distinct technological approaches for providing power and two-way communication to instruments at ocean observatories (1) moored observatories linked to shore via satellite and (2) observatories linked to shore by cables. Moored systems presently in use include both
40 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY TABLE 3-1 Potential Global Network Sites Proposed by the DEOS Global Working Group Latitude/ Longitude Remarks Atlantic Ocean Sites 36N 70W Gulf Stream extension; flux reference at critical site for air-sea coupling; exchanges of heat; CO2; freshwater; water column instrumented for water mass-variability and modification studies. 32N 65W BATS/Station S/BTM; historical time-series record and testbed site; physical, meteorological, biogeochemical. 30N 42W North Atlantic multi-disciplinary site; instrumented for DEOS; geophysics, meteorology, physical, biogeochemical. Multidisciplinary mooring; site now occupied by the Pilot Research Moored Array in the Tropical Atlantic (PIRATA) surface mooring, getting surface meteorology and upper-ocean data; upgrade to more capable full water column and seafloor. Flux reference site; for Atlantic climate modes; as above, upgrade existing PIRATA mooring. 35S 15W South Atlantic multidisciplinary site; instrumented for DEOS; geophysics, meteorology, physical, biogeochemical. Pacific Ocean Sites 50N 145W Former Ocean Weather Station P ("PAPA"), known today as Site P; meteorology; physical, biogeochemical. ON 20W 10S 10W 40N 150E Kuroshio Extension; meteorology and air-sea fluxes, water mass variability. 48 30N 176 30W 9 50N 04 20W 0 145W Endeavor segment of Juan de Fuca Ridge; RIDGE Integrated Study Site (ISS); seafloor biology, hydrothermal vents, geophysics. East Pacific Rise RIDGE ISS site; tropical air-sea coupling, surface meteorology, water column, seafloor biology, hydrothermal vents, geophysics. Eq. Pacific multidisciplinary site; upgrade Tropical Atmosphere Ocean Project (TAO) site for flux; biogeochemistry, water column. 40S 115W South Pacific multidisciplinary site; instrumented for DEOS; geophysics, meteorology, physical, biogeochemical. 35S 150W South Pacific multidisciplinary site; instrumented for DEOS; geophysics, meteorology, physical, biogeochemical. Indian Ocean Sites 15N 65E Arabian Sea; meteorology, physical, biogeochemical. 12N 88E Bay of Bengal; meteorology, physical, biogeochemical. 10S 90E 90E ridge multidisciplinary site; surface meteorology, water column, seafloor. continued
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 41 TABLE 3-1 Continued Latitude/ Longitude Remarks 25S 97E Indian Ocean DEOS; geophysics, physical, meteorology, biogeochemistry. 47.7S 60E Kerguelen Fixed Station (KERFIX), part of the Kerguelen Islands Time- series Measurement Programme follow-on; physical, meteorological, biogeochemical. Initial Southern Ocean Sites 42S 9E SW of Cape Town; meteorology, water column instrumentation geophysics. 55S 90W Antarctic Intermediate Water (AAIW) water-mass formation region; meteorological, physical, CO2, geophysical. 47 S 142E South of Tasmania; meteorology, physical, biogeochemical, geophysics. SOURCES: Modified from Figure 1, DEOS Moored Buoy Observatory Working Group, 2003 and data from R. Weller, Woods Hole Oceanographic Institution, personal communi- cation, 2003.
42 o be o o au V) o o 5- o
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 43 subsurface and surface moorings, but neither type has a hard-wired power or data connection to the ocean floor and thus cannot power or control instruments on the mooring or on the seafloor. Cabled observatories can be further subdivided into those utilizing new cables and those re-using cables retired by the telecommunications industry. Global observatories based on new-cable infrastructure are probably beyond the scope of the OOI given the high cost of the cables needed to reach remote observatory sites. Re-using retired telecommuni- cations cables for observatories is feasible in some circumstances, and is already occurring with the retired TPC-1 cable. Other cables are being used by the Earthquake Research Institute in lapan, the H20, and A Long- Term Oligotrophic Habitat Assessment (ALOHA) Observatory (Appen- dix D). Planning for the OOI global observatory network has thus far focused primarily on moored buoy systems. The opportunity to re-use retired telecommunication cables for some of these sites should be thoroughly investigated, however, as it might offer a cost-effective alternative to moored buoy systems in some locations. Moored Buoys At present, well-instrumented moored observatories using surface buoys are operational at tropical and mid-latitude sites (e.g., TAO, BATS, and the Pilot Research Moored Array in the Tropical Atlantic PIRATE. Surface meteorological sensors are mounted on the buoy and suites of oceanographic sensors are mounted on the mooring line, spaced as little as 5 m apart throughout the upper ocean, and extending into the interior of the ocean at wider spatial intervals (Plate 4~. Those surface moorings designed to survive strong currents and high seas, as in the Arabian Sea, have used a scope (the ratio of mooring line length to water depth) close to 1.4 and have included synthetic rope (often nylon), which stretches and prevents the anchor from dragging and the surface buoy from submerg- ing. The basic 3 m discus buoy hull developed at WHOI, used at many sites for research, has also been used by the National Data Buoy Center (NDBC) for collecting weather observations at many sites around the U.S. Exclusive Economic Zone (EEZ). Many aspects of surface and subsurface mooring technology are very well developed for surface and water column measurements. Servicing intervals are dictated primarily by the need to maintain sensor quality in the face of biofouling and corrosive marine air. While some subsurface moorings can be installed and left unattended for up to five years, servic- ing intervals for surface expression moorings are typically six to twelve months for current generation moorings. Increasingly sophisticated in-
44 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY strumentation is also being developed for moorings, including multidisci- plinary instruments, water samplers, in situ analyzers, and profiling in- struments that move up and down the mooring line. Current oceanographic moorings, however, have important limita- tions that the OOI should address. These moorings have a very limited capability for telemetering data to shore (samples per minute), since they rely on very low-bandwidth satellite systems. The buoy systems in use today also possess very limited power generation capabilities, as they rely generally on batteries or solar panels. Present-day moored observatories have no capacity to power or communicate at high data rates with instru- mentation on the mooring or the seafloor. Inductive coupling of instru- ments to steel mooring cable and acoustic telemetry have been used to communicate with underwater instruments, but data rates are low. Sur- face-to-bottom electro-mechanical (EM) or electro-optical-mechanical (EOM) cables are as yet unproven and their reliability in this application is unknown. Strong ocean currents, high sea states, freezing spray, and floating ice all challenge the survival of even the most robust surface moorings and cables available today. The large drag of strong currents can cause anchor dragging or buoy submersion. High waves lead to large cyclic tensions and mechanical fatigue, and may affect the performance of directional satellite antennas unless the surface buoy is large and designed to mini- mize pitch and roll. To sustain the quality of meteorological and air-sea flux observations in high sea states, buoy motion and mean tilts will need to be measured, or meteorological sensors will need to be mounted in gimbals. In addition, meteorological sensors more resistant to hostile en- vironments must be developed. Mitigating the effects of freezing spray requires heated meteorological sensors and, by extension, buoys with significant power generation capabilities. Subsurface moorings can survive longer in such hostile environments as they are not exposed to surface wave motion, can be deployed under- neath ice cover, and are less likely to be vandalized. However, subsurface moorings are sometimes damaged by fishing activity. Furthermore, the lack of surface expression limits data return to using pop-up capsules that communicate via satellite at relatively low data rates, absent a direct con- nection to a seafloor cable (e.g., ALOHA observatory). The DEOS Committee investigated new, more capable moored buoy systems, and found two particular concepts to be most promising for further development (Table 3-2) (DEOS Moored Buoy Observatory Work- ing Group, 2000~. The first option is a cable-linked, high-bandwidth spar or discus buoy that uses an EOM cable to connect seafloor and moored instruments to the surface (Figure 3-2~. This system is designed to deliver approximately 500 W of power to the seafloor and would telemeter data
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 45 TABLE 3-2 Technical Specifications of Moored Buoy Systems Under Development Low-Bandwidth, Acoustically-Linked Low-Bandwidth EOM Cable-Linked High-Bandwidth, EOM Cable-Linked Buoy type Discus buoy Mooring design Taut mooring EOM cable No Data throughput 2.4 kb/s Power to sensors None Junction box No (acoustically-linked) Discus buoy S-tether mooring Yes 9.6 kb/s 20W Yes Spar buoy Tri-moored Yes 64 kb/s 500W Yes NOTE: Data volume delivered to shore will depend on commercial telemetry charges rather than bandwidth of telemetry link (e.g., Iridium telemetry costs are currently $1/m~nute). SOURCE: Data from DEOS Moored Buoy Observatory Working Group, 2003. to shore using a 64 kb/s C-Band satellite telemetry system with a stabi- lized, directional antenna. The second concept is a low-bandwidth, discus buoy system that could be implemented using either of two different approaches (Figure 3-2~. One option uses acoustic modems to transfer data intermittently from instruments on either the seafloor or the mooring to the buoy at rates of up to 5 kb/s and a low power, omni-directional satellite system to telemeter data from the buoy to shore at rates of 2.4 kb/s. A second approach uses an EOM cable to deliver approximately 20 W of power to instruments on the seafloor and to provide two-way data communication between the buoy and a benthic node. A low power satellite telemetry system would then deliver data from the buoy to shore at rates of up to 9.6 kb/s. See Table 3-2 for a summary of the technical specifications of the systems discussed above. The following discussion offers an assessment of the technical readi- ness of each of these systems based on the recent DEOS Global Network Implementation Plan (DEOS Moored Buoy Observatory Working Group, 2003). Low-bandwidth, acoustically-linked system This design builds upon previous successful discus buoy deployments of up to one year in a variety of locations in the world's oceans. An acoustically-linked moored observatory has been used successfully by the NOAA Deep-Ocean Assessment and Reporting of Tsunamis program (DART) (www.pmel.noua.gov/tsunami/Dart/). While there are no high-risk
46 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY RING BUFFER ARCHIVE ....................... :;: ................................................... :;: .......................... ~9= RING BUFFER AR&HIVE PIS .......................... 4:~ ................................................... Pl'S FIGURE 3-2 Two moored-buoy design concepts being examined for the OOI. Top: A low-bandwidth discus buoy system uses acoustic modems to transfer data intermittently from instruments on the seafloor or mooring to a surface buoy and from surface to shore via a low-power, omni-directional satellite sys- tem. Bottom: a high-bandwidth moored observatory that uses an electro-optical cable to deliver power and two-way data communication to a seafloor junction box and is linked to shore via a 64 kb/s C-Band satellite telemetry system. Figure courtesy of John Orcutt, Scripps Institution of Oceanography.
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 47 technologies associated with this system, there are questions associated with the performance of the acoustic modems under various environ- mental conditions. The primary issues that need to be resolved are (1) the reliability of the acoustic communication link; (2) the average data rates possible, both for instruments directly below the buoy on the mooring and for instruments placed some distance away on the seafloor or in the water column on a near-by subsurface mooring; and (3) the maximum horizontal distance from the surface buoy at which a useful acoustic com- munication link can be maintained. Low-bandwidth, EOM-cab/e-/inked system The primary difference between this design and the other low-band- width system is that an EOM cable, rather than an acoustic-link, is used both to provide two-way communication to a seafloor junction box, and to provide small amounts of power (20 W) to the seafloor. In this design, the EOM cable moors the buoy. The lack of experience with using an EOM cable in this way makes the mooring the highest-risk sub-system in this approach. Conventional mooring designs with a wave-following buoy use the large scope of the mooring line to accommodate both the hydro- dynamic forces associated with the mean currents that would otherwise either submerge the buoy or cause the anchor to drag, and the flexural fatigue problems associated with surface wave motion. Further, on con- ventional designs, plasticjacketed wire rope extends down to about 1500 m in depth and then transitions to synthetic line. As the buoy heaves with surface waves and swell, many cycles of flexing occur, and there is mo- tion of the stiff steel cable relative to the synthetic line below, which must have its uppermost section supplemented with a strain relief boot to re- duce fatigue from bending and wear from chafing. With an EOM moor- ing cable, however, scopes of close to one times the water depth are planned in order to avoid slack line that might loop, flex, and bend as currents and sea state vary, thus causing failure of the optic fibers or strain hardening of the copper conductors in the EOM cable. In that case, however, in-line tensions could be high and cyclic variation in load sig- nificant enough to risk fatigue-related failure of the cable. The expected operational lifetime of EOM cables in this application, and the way in which factors such as sea state and currents affect their reliability, is not well known. Outstanding issues regarding EOM cable include (1) cable design and selection, (2) EOM termination design (including bending relief), and (3) design of the junction between the discus buoy and EOM cable.
48 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY High-bandwidth, EOM-cab/e-/inked spar buoy system This design is the most challenging of the proposed new systems and includes several new or high-risk sub-systems. Unlike the discus buoy design, the spar buoy system design does not require that the EOM cable be load-bearing. Nonetheless, there is little or no experience with EOM cables used to connect seafloor and moored instruments to the surface. The expected operational lifetime of the EOM cables, or the way in which factors such as sea state and currents affect their reliability is unknown. In addition to the EOM cable, two other sub-systems requiring further de- velopment and testing include (1) the high-bandwidth C-band satellite telemetry system and (2) the diesel power generation system on the buoy. A key issue is determining if these systems are capable of reliable, unat- tended operation for 12-month periods. While prototypes of the two low-bandwidth systems will be built and tested at sea during the next two years (see discussion below), a detailed system engineering study of the high-bandwidth spar design must be completed and a prototype built and tested before this system will be ready for installation as part of the OOI. The design should consider issues associated with the use of a spar buoy as a measurement platform and any possible way in which it might interfere with meteorological measurements. Similarly, the approach to making those near-surface mea- surements in the ocean that are important to studies of upper-ocean phys- ics, biology, optics, and air-sea fluxes needs to be thought through care- fully as the hull of the spar will disturb the flow of the critical near-surface layer of the ocean and complicate instrumentation. The seafloor junction box Both the low- and high-bandwidth EOM-cable-linked systems require a seafloor junction box between the various instrument systems and the EOM cable to the surface. The moored-buoy junction box should be de- signed to appear to the user as identical to the cabled observatory junc- tion box. The only difference would be the amount of power and band- width that will be available. Engineers developing the cabled and moored-buoy observatories will need to coordinate closely to ensure a common seafloor junction-box interface protocol. Operation of these systems in severe environments One of the most challenging goals of a global ocean observatory pro- gram is to establish observatories at high-latitude sites where high winds, high seas, and strong surface currents often prevail. Challenges range
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 49 from the operational efficiency of the satellite telemetry system to the survivability of the mooring itself. The large tri-moored, spar buoy design seems like a promising approach for these harsh environmental condi- tions. The British DEOS (B-DEOS) has also developed an alternative de- sign for a large, high-latitude buoy on a single-point S-tether mooring that behaves like a spar buoy at lower sea states and a wave-following buoy at high sea states. Before the OOI will be ready for deployment at high-latitude sites, the system design requirements for deployments in severe environmental conditions will need to be defined, and engineering and testing of various sub-systems will need to be completed. Testbeds Several efforts are under way to test various systems and sub-systems associated with the next-generation moored ocean observatories discussed above. Acoustically-linked ocean observing system A group of investigators from WHOI and the University of Washing- ton have received funding from the NSF to develop and test a prototype deep-water, acoustically-linked moored-buoy observatory called the Acoustically-Linked Ocean-Observing System. The purpose of this project is to demonstrate the technical capabilities and scientific potential of an acoustically-linked moored-buoy system for ocean observatory studies. In late 2003, the prototype system will be deployed for three months off the U.S. East Coast, and in late 2004 for 15 months in the Northeast Pacific, in order to test the reliability of acoustic communication and the buoy-to- shore satellite link, in a variety of seasonal conditions. Monterey Bay Aquarium Research Institute Ocean-Observing System (MOOS) MBARI is testing several high-risk technologies associated with an EOM-cable-linked moored buoy observatory at the MOOS test mooring site in 1,860 m of water in Monterey Bay, California. These tests include (1) deployment of a mooring to verify dynamic modeling of the mooring and to verify the survivability of an EOM cable that delivers power and communications to benthic and water column networked instruments, (2) use of an ROV to lay EOM cable between benthic nodes, and (3) evalua- tion of smart network technologies to provide automated device discov- ery, configuration, and operation when an instrument is plugged into a node. Testing of these prototype efforts will continue through 2003, lead-
50 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY ing to an integrated system deployment in 2004 or 2005 in Monterey Canyon. This system will include: · a surface buoy generating 50 W DC power; · bi-directional Globalstar communication to shore; · an EOM cable delivering power and communications to a benthic network; · two or three benthic nodes connected to the EOM cable; · Plug-n-Play instrumentation providing automatic dynamic moor- ing configuration; and · a shore-side data system that receives data automatically associ- ated with its metadata. Rea/-time Observatories, Applications, and Data Management Network (ROADNet) Project Scripps Institution of Oceanography is currently testing the perfor- mance of a commercial C-Band antenna system, one of the key design features of the high-bandwidth buoy system. For these tests, a 2.4 m C-Band satellite antenna from Seatel, Inc., of Concord, California has been installed on the R/V Roger Revelle, and service leased from a commercial teleport to provide a 64 kb/s full-period connectivity between the ship and the public Internet (a cost of about $100/day for 64 kb/s). A proto- type shore-side teleport is located at the San Diego Supercomputer Center (SDSC) on the University of California San Diego main campus. The SDSC is a major node on the Internet, providing a convenient wide-band gate- way. This prototype network provides for real-time delivery of large quantities of shipboard data to shore for quality control, archiving, and real-time data availability. During the first year of tests, throughput has been about 82 to 85 percent of capacity. The performance of the system to date suggests that the motion requirements specified in the DEOS Buoy Design Study (DEOS Moored Buoy Observatory Working Group, 2000) of less than 10 degrees per second may be unnecessarily stringent because of the rapid response of the antenna servo (90 degrees per second). How- ever, the reliability of the system will need to be improved in order for one year of unattended operation to be possible. Cabled Systems Planning for cabled observatories in the remote oceans began with meetings in Honolulu, Hawaii in 1990 to discuss the use of retired tele- communications cables (shave et al., 1990), and in La Tolla, California in 1995 to plan for permanent ocean observatories (Purdy and Orcutt, 1995~.
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 51 The IRIS took the lead by forming the Cable Steering Committee, and the Ocean Seismic Network (OSN) planning effort (Dickey and Glenn, 2003~. Until very recently it seemed that only older co-axial cables would be retired in the next decade, given the rapid growth of the telecommunica- tions industry. However, the lag in increase in demand for bandwidth and the huge increase in available bandwidth in new optical cable sys- tems has led to the imminent retirement of the older optical cable systems well before their design lifetimes. For example, the Hawaii-4 cable, which will be retired in 2003, was installed in 1989. Since this development is very recent, little planning has been accomplished for the possible use of these cables for ocean observatories. In fact, the high cost of new cable and cable ships was believed to limit the use of cables for observatories to near shore observatories; the development of the OOI reflects this history. Using submarine optical cable systems at global observatory sites is advantageous, given: · their high data bandwidth available for data transmission (250 Mb/s or more); · their ability to transmit large volumes of data to users in real-time; · their relative immunity to weather problems; · their ability to function without routine maintenance; · their ability to provide large amounts of continuous electrical power (kW) to the ocean floor; · the costly shore connections already in place for many applica- tions; and · the high reliability of the technology provided by commercial re- search and development. On the other hand, new submarine cables are relatively costly (about $10,000/km depending on cable characteristics and market conditions), although this is not an issue for cable re-use. In addition, installation of long cables is expensive, often requiring cable burial near shore and the use of cable laying ships. Furthermore, new shore stations, required to provide power and data links to users when a new shore connection is required, are also expensive. More than 35,000 km of electro-optical telecommunications cables on the ocean floor, representing more than $500 million in cable assets, will be retired in-place by the industry within the next few years, and more during ensuing years. These cables come to shore at existing cable sta- tions, and the shore hardware systems and the cable stations could be available for observatory use at little cost. These cables present a one-time opportunity to the oceanographic community for re-use in observatory support. The newest fiber cable systems, which use optical amplifiers
52 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY rather than optical signal regenerators, will likely not be retired for de- cades, since they have extremely high bandwidths and can be upgraded in place. A few of the oldest optical amplifier systems, installed in the mid 1990s, have relatively low bandwidth compared to the newest systems, and could be retired within a decade. Such systems use less current than the repeatered systems, and thus would supply less power to observato- ries. The near-shore end of these cables often represents a liability to the telecommunications industry due to fishing concerns and anchoring. As a result, the retired cables are often removed from the continental shelf after retirement, rendering them useless for in-place observatories. When these older cables are gone, the opportunity for cable re-use for observa- tories will likely end for decades. There are a number of technical, logistical, and financial issues that need to be evaluated in order to determine the suitability of using retired cables for any particular global observatory site. However, in those cases where cables can be used in-place or moved only a short distance while still utilizing the original shore station, submarine cables potentially offer many advantages over the use of buoys or laying new cable. In this case, installation cost is very low, as it requires only the recovery and termina- tion of the cable. If sampling of the upper ocean and measurements of air- sea interaction are required, the installation of a surface or subsurface mooring may be necessary. Installation of the H20 on the retired Hawaii- 2 coaxial cable between Hawaii and California was accomplished in 1998 with a UNOLS vessel (Petitt et al., 2002) (Figure 3-3~. It is technically feasible to relocate cables to a new, distant location, but doing so requires recovering and relaying the cable, establishing a new shore connection, obtaining permitting for the landing site, constructing a shore facility, and acquiring access to power and data distribution infrastructure on shore. In such a case, the cost and feasibility of moving and installing a re- used cabled system needs to be carefully assessed and compared to the cost, reliability, and assets of buoy observatories. Initial estimates suggest that from four to eight observatory nodes could be attached to a single optical cable, supplying up to a kilowatt of power and 100 Mb/s band- width to each node. The NSF should appoint a committee with the appro- priate expertise to thoroughly evaluate the issues regarding cable reuse and recommend how best to utilize this potentially valuable resource for observatory science. Some key technical, logistical, and financial issues this cable re-use committee should address include: Technica/ · the feasibility of recovering these existing fiber-optic cables from the deep sea without damaging the cable and repeaters; and
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 53 FIGURE 3-3 The H20, the first deep-sea cabled research observatory, was es- tablished in 1998 using the retired Hawaii-2 coaxial telecommunications cable. A seafloor junction box in 5,000 m of water provides power and data communica- tion for up to six instruments including a broad-band seismometer. More than 35,000 km of electro-optical telecommunications cable will be retired in place by industry within the next few years, providing a potentially significant resource for ocean observatory science. Figure courtesy of layne Doucette, (redwoods Hole Oceanographic Institution. · the ability to produce appropriate hardware and software for use of these systems for observatories; Logistica/ · the problems surrounding the cooperation of foreign countries with established power and communications infrastructure, required to establish cabled observatories in remote areas;
54 Financia/ ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY · moving long sections of cable will require a commercial cable ship and experts to handle and terminate the recovered cable. The adequacy of the benefits of making high-bandwidth and power available to observato- ries in remote areas in offsetting the costs of the commercial cable ship and experts to handle and terminate the recovered cable in order to move long sections; · the affordability of shoring these cables, which will require ar- mored cable and cable burial in many cases; and · the possibility of using spare cable equipment and cable that will also be discarded by the telecommunications industry for future observa- tories, and any attendant liabilities.
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 55 REGIONAL OBSERVATORY PLANNING Plate tectonics is a well-established paradigm for the slow geological processes of formation, movement, and destruction of the Earth's ocean crust. Crustal generation takes place along the crests of the huge under- water ridge systems that girdle the globe, while destruction occurs along convergent boundaries where oceanic lithosphere sinks back into the earth. However, these processes remain poorly observed: partly because they primarily take place in regions hidden by the oceans, and partly because to a large extent they occur through events that are highly inter- mittent in time (sub-sea volcanism in the case of crustal construction, subduction earthquakes in the case of destruction). Further progress in understanding these geophysical processes requires extended time-series measurements within the ocean over the scale of a tectonic plate. A plate- scale, regional observatory would enable acquisition of such extended time-series. In addition, cabled networks with real-time instrument con- trol allow for "interactive sampling," in other words, (re-)deployment of sampling resources at the time and location of significant events. While initially focused primarily on either such geophysical processes as the study of biological communities at hydrothermal vents, additional studies have identified significant contributions to other branches of oceanography that would result from the combined temporal and spatial resolution characteristic of a regional-scale observatory. Suggested uses include studying continental slope stability and subsea methane deposits and determining long-term variability of ocean circulation. The high power and bandwidth of a cabled observatory is particularly important to studies involving interactions between the physical environment of the ocean and its embedded marine ecosystems, studies not only of funda- mental importance but also of special urgency given evidence of climate change. Assessment of implementation readiness for a regional-scale cabled observatory requires consideration of the maturity of planning for identi- fied scientific opportunities, as well as the extent to which technical chal- lenges have been met or are likely to be met in the near future. Status of Scientific Planning The often unique scientific opportunities presented by high temporal resolution, long-term ocean measurements that can be provided only by cabled observatories have been extensively documented as a result of various broad-based community workshops and a series of recent reports (See Appendix C). The report Ocean Sciences at the New Millennium (Na- tional Science Foundation, 2001) identified six cross-cutting ocean science themes that are likely to provide the most important, most promising,
56 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY and most exciting areas for ocean discovery and understanding over the next decade (Box 3-3~. The Illuminating the Hidden Planet report (National Research Council, 2000) identified major science problems under each of these themes for which geographically distributed, long-term, time-series observations would be either useful or very useful. The recent SCOTS workshop report provides detailed documentation of specific scientific problems within these themes that would be best addressed using the strengths of cabled ocean observatories (Dickey and Glenn, 2003~. Many of these problems had also been identified during earlier science planning for the NEP- TUNE plate-scale observatory in both the U.S. and Canada (NEPTUNE Phase 1 Partners, 2000; NEPTUNE Canada, 2000~. A regional-scale obser- vatory will complement the global observatory network, providing the higher temporal and spatial resolution data necessary to interpret the global-scale data for each of the interdisciplinary science questions listed above. Although the general scientific rationales for a regional-scale obser- vatory can be considered as firmly established by the reports referenced above, detailed planning is just beginning for those specific, rather than thematic, scientific objectives that are crucially dependent on cabled in- frastructure. Additional disciplinary and interdisciplinary community workshops should be held to define "lead-off" science questions and in- novative experiments likely to generate the most exciting scientific and educational results during the first few years of operation of a regional- scale observatory. These workshops should not only provide definition for these science questions but also begin to specify the instrumentation that scientists will expect as part of observatory infrastructure and the
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 57 combinations of measurements that will be required at each experimental location. Status of Technical and Engineering Development and Planning A cabled observatory may use a network of modern telecommunica- tions cables to support a wide variety of instruments on and beneath the seafloor and within the water column. A regional-scale cabled observa- tory facility could involve thousands of kilometers of cable, several shore stations, many science nodes, and possibly multiple-loop cable topologies as well as branches. Together, these elements will provide (1) a power distribution system; (2) a communication network for command and con- trol, data transfer, and accurate timing; and (3) connections for long-term core sensors and community experiments as well as shorter-term experi- ment-specific sensors. While technologies and methods developed by the sub-sea telecom- munications industry provide a strong foundation for regional cabled ocean observatories, such systems are simple in comparison to the pro- posed NEPTUNE system. The design and implementation of such a com- plex system of sub-sea cabled observatories pose many additional techni- cal challenges (NEPTUNE Phase 1 Partners, 2000~. The commercial submarine telecommunications industry focuses on moving data from one shore landing to another. In contrast, a sub-sea observatory must gather data from sensors distributed throughout an ocean volume, as well as on and under the seafloor, and deliver these data to shore while simultaneously enabling real-time control of power and control of a wide variety of instrumentation. These requirements demand challenging technological capabilities, including the following: · Branching: a requirement for constructing arbitrary topologies (spur, ring, mesh, etc.) for optimizing sampling array design, allowing future addition of cable runs that might be necessary to adapt initial observatory design to changing scientific questions, and providing re- dundancy in routing signals and power, thus increasing the reliability of the system; · Undersea nodes: highly reliable junction boxes for bringing signals and power to and from the primary cable; · Power supply: the ability to provide different and time-varying amounts of power to multiple nodes on the cable; · Communications protocols: the ability to add and drop two-way information at each node on the cable; · Plug-n-Play instrumentation: standardized scientific instrument interfaces, dynamic network detection of changed (added/deleted/
58 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY replaced) sensors, automatic addition of necessary meta-data to sensor files, and automatic addition of data from new sensors to the archival stream; · Accurate time base: highly accurate time registration, as will be required by certain classes of experiments; · Failure detectionlisolationlrecovery: timely detection (and if pos- sible, mitigation) of failure in individual system components, so that fail- ure does not propagate through the entire system; and · Commandlcontrolfunctionality: operational means of allowing the "owner" of an individual scientific instrument or experiment to control it in a way that is reasonably transparent, while protecting other deployed instruments and experiments and the network itself. Technical planning for regional-scale observatories has primarily taken place within the framework of the NEPTUNE project, a proposed joint U.S./Canada plate-scale observatory designed to encircle and cross the fuan de Fuca tectonic plate in the northeast Pacific with about 3,700 km of fiber-optic/power cables (Figure 3-4~. Experimental sites, estab- lished at approximately 30 nodes along the cable, would be instrumented to study geological, physical, chemical, and biological phenomena that vary on multiple scales of space and time over an observatory lifetime expected to be at least 30 years. Both spatial and temporal scales thus vastly exceed those of any scientific activity that has yet been attempted in the ocean. Since the scientific and technical challenges faced by NEP- TUNE will be present in any alternate proposals for regional observato- ries, the results of the NEPTUNE technical planning process (NEPTUNE Phase 1 Partners, 2000) will be used to provide necessary specifics in the following discussion of the major areas of engineering development re- quired for the successful implementation of a regional-scale observatory. These areas include power supply, data transmission, communications control and timing, data management and archiving, and sensors. Power Supply The NEPTUNE project is considering two options for power distribu- tion. The first option is based on the conclusion of the 2000 U.S. NEP- TUNE Feasibility Study that neither AC power nor the constant current serial DC power systems used in transoceanic telecommunications cables are appropriate for a submarine cabled observatory (NEPTUNE Phase 1 Partners, 2000~. The feasibility study concluded that the desired system of multiple nodes, each with time-varying power requirements, would be best served by a parallel DC system capable of providing constant voltage to each scientific user. At each node, high voltage (6-10 kV) DC power
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 59 FIGURE 3-4 Backbone cable structure and primary seafloor nodes proposed for NEPTUNE, an international USA-Canada sub-sea observatory spanning the Juan de Fuca plate off the western coasts of British Columbia, Washington, and Ore- gon. Figure courtesy of the NEPTUNE Project (www.neptune.washington.edu) and produced by the Center for Environmental Visualization at the University of Washington.
60 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY would be converted to 400V and 48V supplies for scientific users. Dupli- cate, cross-linked power supplies in each node are planned to enhance reliability. Even so, the performance level required of individual compo- nents for the life of the system is still very high. A second option currently being considered aims to reduce this prob- lem by including circuit breakers in the branching unit where the side cable to a node joins the main (backbone) cable. Any fault that develops in a node or its branching cable would result in the isolation of that node without affecting the rest of the system. With such a system, the depen- dence on the reliability of individual nodes could be reduced (and could even vary throughout the network, depending on local scientific require- ments) without any effect on the availability or reliability of other nodes. The first option has recently undergone a concept design review, the results of which are available at http://neptunepower.apl.washington.edu. A comparison of the two options (which share most hardware and control features) will be completed in the near future and will be further tested through the development of prototype hardware for laboratory and field- testing. Communications, Control, and Timing The 2000 U.S. NEPTUNE Feasibility Study, which was based on the assumption that a new, custom-designed cable would be laid, concluded that the two-way communication and interactivity requirements of a cabled scientific research observatory would be best met by extending the Internet into the deep sea (NEPTUNE Phase 1 Partners, 2000~. Under NSF support, WHOI, in collaboration with Cisco Systems, has been develop- ing a Gigabit Ethernet system that will provide the communications for a multi-node, cabled observatory. Each node will have packaging and add/ drop transmission capabilities that will enable users to control their in- struments, retrieve data, and access a time signal that will time-stamp all observations with an accuracy of one millisecond or better. The node routers will also serve as amplifiers for the optical signals moving along the cable, obviating the need for expensive optical amplifiers. The current development status of this communications system is described in a draft conceptual design report (NEPTUNE Data Communications Team, 2002~. The virtual collapse of the transoceanic cable industry subsequent to completion of the NEPTUNE feasibility study initially raised the hope that surplus conventional submarine cable might become available at rela- tively low cost, though that does not appear to be the case now. Use of this type of cable, which includes embedded optical amplifiers, would elimi- nate the need to space nodes at approximately 100 km intervals to amplify the optical signal and opens the possibility of starting with a sparse set of
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 61 initial nodes that could be augmented as scientific demand and resources grow. However, the combination of Ethernet and conventional subma- rine telecommunications protocols introduce significant additional engi- neering complexity, and the cost of this alternative system may be signifi- cantly greater than the original NEPTUNE telecommunications design. A full comparison of the two options is currently under way, and will evalu- ate these options based on their life cycle costs, reliability, and flexibility and expandability by an independent panel of experts. A final decision on which design to pursue should be made by late 2003. Data Management and Archiving System Leadership of the Data Management and Archiving System (DMAS) planning for the proposed NEPTUNE regional-scale observatory has been provided by the Canadian National Research Council's Herzberg Insti- tute of Astrophysics in Victoria, British Columbia, which is one of three Hubble Space telescope data repositories. A conceptual design study (www.neptunecanada.ca/about/reports.htm) incorporates key lessons learned from the astronomical community's experience with large data streams. These lessons include: · the DMAS data management and archives must be science-driven; · instrument and observatory control system design must support data management; · collection, packaging, and validation of data and meta-data should be automatic; · data must flow simultaneously to both the Principal Investigator (PI) and the archive; · automating the process of quality control should be based on prac- tical experience; and · data archives should allow both public and privileged access to various data streams. Major additional challenges faced by regional cabled observatories lie in the heterogeneous data types that will be generated from a wide vari- ety of instrument suites and the need to support simultaneous execution of many experiments. Valuable experience will be provided by a proto- type based on this design study and scheduled for implementation dur- ing 2003-2004 as the DMAS for VENUS, a funded Canadian testbed obser- vatory. As part of this implementation, Canadian DMAS planners are working closely with engineers at MBARI, who are developing the Scien- tific Instrument Interface Modules (SIIMs). These interface modules will be inserted between scientific instruments and deep-sea cabled nodes and
62 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY both will add essential metadata to the data streams and allow the net- work to recognize and accept data from newly installed instruments. Testbeds In addition to the technical issues described above, there are other major technical challenges involved in making the transition from the current oceanographic research techniques to effective use of a regional cabled observatory. Scientists must learn how to design, install, operate, and recover complex experiments, while engineers must design, build, test, and install new hardware in a challenging operating environment. Information managers must learn how to use huge streams of data (on the order of 1 Petabyte 10~5 bytes per year) of many different types. Op- erators must learn how to install instruments, service an array with mini- mal down time, protect a network from failures in any one of many indi- vidual instruments, keep a network running with high reliability, restore service rapidly in the event of a failure, and provide users with the ability to interact with their experiments as needed. It is widely recognized that technical challenges are best met through testbed sites, a sequence of cabled systems that will build the experience necessary to attain the full potential of a regional cabled observatory. Two testbed systems are presently under development in support of the eventual establishment of regional-scale undersea observatories. Victoria Experimenta/ Network under the Sea Funded by the Canada Foundation for Innovation and the British Columbia Knowledge Development Fund for a start in 2003, VENUS will lay a short single-node cable in Saanich Inlet (an anoxic fjord), a cable with multiple nodes across the Canadian portion of the Strait of fuan de Fuca, and another into the middle of the Strait of Georgia south of Vancouver (Plate 5~. In addition to providing researchers with access to interesting and very different marine environments, VENUS will enable early testing of various technologies proposed for a regional observatory (more information available at: www.venus.uvic.ca/). Cables will be marine industry standard, with a central core steel tube containing eight optical fibers surrounded by steel strength wires, a single copper conducting sheath, and insulation. Installation will use commercial entities with ex- tensive local experience in submarine telecommunications. As planned, VENUS will use the major components of the power system proposed for NEPTUNE. The VENUS communication system will connect each science node directly to shore, using dedicated fibers and commercial off-the- shelf hardware. Where possible, science nodes and SIIMs will use designs
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 63 and components being developed for NEPTUNE and MARS. The VENUS Data Management System will be implemented by data managers at the Herzberg Institute of Astrophysics, based on initial planning carried out during the NEPTUNE feasibility study. The VENUS installations placed, in relatively shallow water and us- ing relatively short cables, will test the following components necessary for eventual establishment of a regional-scale observatory: · single-cable multiple nodes; · power system; · communications system (simplified version); · data management system; · node design; · SIIMs design; and · ROV installation and maintenance. Monterey Accelerated Research System Funded through MBARI by the NSF and the David and Lucite Packard Foundation, MARS will involve the installation of a cabled test bed adjacent to Monterey Canyon in Monterey Bay, California (more in- formation available at: www.mbari.org/mars/~. Planned for installation in early 2005, MARS will extend the VENUS experience to longer cable runs (62 km) and deeper water (1220 m), and test a spur topology involving a branching node supporting an "extension cord" to instrumentation at a distance of several kilometers from the backbone cable (Plate 6~. In this environment, the MARS testbed will further test both the DC power and communications systems being proposed by the MARS/NEPTUNE engi- neering group. Finally, successful deployment of MARS will (1) allow researchers to become accustomed to working with the type of nodes, SIIMs, and data management systems that will accompany a regional- scale installation; (2) provide the crucial experience required to develop operational skills (ROV-based installation and maintenance, system fault response and repair, and system control) that will be needed by operators and users of larger regional-scale observatories; and (3) supply education and outreach organizations with real-time and archived data products that can be used both immediately and in preparation for the more exten- sive data streams derived from regional-scale installations. Neither VE- NUS nor MARS provides in-water testing of the mesh (with branches and loops) cable topology or the high voltages and currents proposed by NEP- TUNE for a regional-scale observatory.
64 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY COASTAL OBSERVATORY PLANNING The coastal ocean is chronically under-sampled, which is especially problematic given mounting evidence that human activity is altering coastal waters by changing sediment deposition, erosion patterns, nutrient distributions, microbial food web structure, and fisheries (Hallengraeff, 1993; National Research Council, 1995; Jorgensen and Richardson, 1996; Rabalais and Turner, 2001~. These human-induced changes will increase in the coming decades with projected coastal development and will be compounded by climate change and global sea-level rise. Unfortunately, research efforts in coastal waters have been hampered by inadequate infrastructure and logistical limitations for individual or small groups of scientists who need to measure the full breadth of time (seconds to decades) and space (cm to km) scales that affect coastal processes. There is a community consensus that traditional monitoring strategies are inad- equate for studying many coastal processes of increasing societal rel- evance (Thornton et al., 2000; Jahnke et al., 2002~.
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 65 To overcome these sampling shortcomings, the coastal ocean research community has been conducting interdisciplinary studies focused on de- veloping integrated observation techniques. With the enabling technolo- gies that have evolved from these efforts, the ocean sciences are currently poised to address pressing questions regarding spatial and temporal het- erogeneity and change in coastal ecosystems. These technologies expand the range over which ocean phenomena can be observed and will make it possible to identify and study processes at previously impossible time and space scales. In addition, these envisioned systems offer the opportu- nity of in situ, real-time, interactive observations that cross conventional disciplinary boundaries, allow for adaptive sampling, and provide useful information to many marine-related user groups. The coastal community appreciates that to fully realize the potential of these new observation systems and to move understanding of the coastal environment to the next level, the community needs to leverage multi-disciplinary studies, sharing many of these resources along with the logistic costs. Status of Scientific Planning Scientific planning for the coastal component of the OOI has grown from nearly a decade of integrated research efforts conducted through the NOPP, the Department of Defense ONR, the NOAA EcoHAB Program and the NSF CoOP. Many of the central scientific problems for coastal waters were synthesized in two separate NSF sponsored workshop re- ports: Coastal Ocean Processes and Observatories: Advancing Coastal Research (lahnke et al., 2002) and the Scientific Cabled Observatories for Time-series (Dickey and Glenn, 2003~. In addition, the state of knowledge in near- shore processes and a community consensus on future near-shore re- search strategies were detailed in State of Nearshore Processes Research: II, from a 1998 workshop (Thornton et al., 2000~. Finally, Ocean.US, which is designing a national operational IOOS, has highlighted many of the tech- nical issues for coastal observatories in Building Consensus: Toward An Integrated and Sustained Ocean Observing System (2002) (Box 3-5~. While all these reports agree the OOI will provide major advances in our understanding of coastal processes and new tools for observing coastal processes, no clear consensus has yet emerged on the mix of obser- vational technologies and approaches that will best address the major scientific problems identified in these reports. These approaches include relocatable arrays of moored buoys and radars (Pioneer Arrays), cabled observatories, and fixed, long-term moorings (lahnke et al., 2002~. These different technological approaches are discussed below in the context of a multi-faceted, multi-dimensional approach to interdisciplinary coastal research.
66 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY Pioneer Arrays A relocatable observatory system would provide the coastal research community with a flexible observational capacity for collecting high- resolution, synoptic-scale measurements in a focused region spanning 100-300 km. The total observatory system, termed a Pioneer Array, would include an array of autonomous surface and subsurface moorings provid- ing real-time data, coupled with land-based, multi-static, high-resolution, surface current radars and integrated with other available remotely- sensed (land-based, airborne, and satellite) data streams (lahnke et al., 2002) (Plate 7~. The motivation for such a relocatable observatory is to provide an infrastructure that is not permanently located at one geo- graphic site, since coastal processes vary widely with location. For ex-
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 67 ample, to address the question of elemental cycling on the continental shelves, measurements should be acquired on both broad shallow (Mid- Atlantic-type) and short canyon (Monterey-type) coastal shelves. To in- crease knowledge of near-shore processes, beaches on both broad and narrow shelves need to be studied not only because of the differences in the incident wave and current conditions but also to increase the under- standing of the dominant processes on different beach genotypes (sandy, cobble, and muddy bottom material; open coast; and protected pocket). The participants at the CoOP meeting proposed the Pioneer Array as a means to entrain the wider coastal scientific community by providing an asset available to scientists in different geographic locations, thus al- lowing the research community to identify the best study site for a spe- cific process through peer-reviewed competitive grants (lahnke et al., 2002~. The geographic flexibility of the Pioneer Array is analogous to a ship on station for several years, providing data from an array of 20-40 real-time coastal buoys and multi-static arrays of high-resolution coastal radars. Coupled with the output of other available remote sensors, the Pioneer Array would provide continuous coherent data streams that could be used by data-assimilation models. Finally, the process studies con- ducted with the Pioneer Array could provide insight on the optimal place- ment of long-term time-series moorings. The technology required for the mooring aspect of Pioneer Arrays is mature, although further work will be required to establish techniques to counter biofouling and fishing losses. Given that the overall goal is to provide a well-sampled ocean for interdisciplinary studies, the sensor arrays on the buoys will include operational and preoperational IOOS instruments but would also provide the flexibility to integrate more ex- perimental instruments. This flexibility allows the OOI system to inno- vate components for the IOOS backbone. The CoOP workshop report recommended that three Pioneer Arrays be acquired as part of the OOI for process-oriented studies in the coastal regions of the U.S. (lahnke et al., 2002~. Surface-based radar technology for remote sensing of surface current fields has experienced rapid growth and acceptance within the scientific community over the last few years. Like satellites, radars provide two- dimensional maps of surface currents, allowing for spatially focused sam- pling efforts (Plate 8~. Experimental long-range systems demonstrated at several locations around the country can measure surface currents over a 200 km area with 6 km resolution. Higher resolution systems provide estimates for radial current vectors over 40 km areas with a resolution of 1.5 km. Some proposals for IOOS call for a national network of these long- range radar systems to support science, commercial shipping, and the U.S. Coast Guard. However, the 6 km resolution of this proposed national
68 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY network lacks the resolution necessary to resolve many of the scales that are relevant for scientific efforts in coastal waters (on the order of 0.5-2 km). Coastal applications will require continued engineering and devel- opment to optimize the number of radar measurements given cost, foot- print size, and resolution. Coastal Cabled Observatories Cabled observatories have been successfully deployed at several coastal sites for many years. Some examples are LEO-15, the FRF in Duck, North Carolina, and the MVCO (Appendix D). These systems offer ultra- high bandwidth and significant power. Their data acquisition systems have multiple user ports that allow a variety of instruments to be plugged into the system. Data are telemetered back to shore via twisted pair or fiber-optic cable. The MVCO is the most recently deployed coastal cable system, designed and built by the same group who built the earlier LEO- 15 system. Communication at MVCO occurs through commercial-off-the- shelf, Gigabit Ethernet network switches that communicate over a pair of fibers back to shore. The MVCO nodes include support for both serial and Ethernet connectivity for instrumentation, a guest-port management sys- tem, and a Gb/s link back to shore. The communications electronics have worked very reliably for the past two years. Scientists are able to easily connect to their instruments and collect data from them over the Internet, and can monitor and control their instruments via privileged services provided on an observatory website. The major advantage of cabled coastal systems is the availability of power for sensors that would have to limit their data acquisition sched- ules if deployed on autonomous moored buoys. This feature greatly en- hances the coastal community's ability to acquire synoptic data at the wide range of temporal scales (from seconds to storm events to decades) that drive coastal processes. Like a Pioneer Array, cabled observatories can provide a test-bed for new instrumentation because they are easily serviced by small vessels and divers. The SCOTS report strongly advo- cates the use of cabled observatories for coastal research (Dickey and Glenn, 2003~. Fixed, Long-Term Moorings A third essential component of a coastal observatory system is an array of fixed, long-term moorings designed to provide measurements of key physical, chemical, and biological parameters and their long-term variability in order to detect subtle but important trends of environmental change in the coastal ocean in the coming decades (Figure 3-5~. Indeed,
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 69 one of the major rationales for establishing a seafloor observatory net- work for basic research is to advance oceanographic science beyond the ship-based expeditionary approach, in order to obtain "long time-series measurements of critical ocean parameters" (National Research Council, 2000~. Such long-term observatories will likely be required at approxi- mately 20-30 locations in order to cover significant regions of the coastal ocean and Great Lakes of the U.S. The CoOP Workshop addressed the issue of long time-series data acquisition and concluded that the coastal IOOS backbone could provide the necessary long-term observations in the coastal zone. However, to serve as useful platforms for research, the IOOS backbone or "Sentinel" moorings may require a broader suite of instrumentation (e.g., pCO2 sensors, time-series sediment traps) than is presently envisioned. If the coastal component of the OOI was restricted to Pioneer Arrays and cabled observatories as recommended by the CoOP FIGURE 3-5 Launch of fixed, long-term coastal mooring in Monterey Bay. These buoys have provided detailed information on sea surface temperature, salinity, chlorophyll concentrations, and other data for more than a decade. The data are collected from sensors mounted on and suspended from the buoy, and are radi- oed to shore. Figure courtesy of Victor Kuwahara, C)2001 MBARI.
70 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY and SCOTS workshop reports, an important opportunity to establish long, continuous observations on the Atlantic, Pacific and Gulf Coasts and in the Great Lakes could be missed. There is a clear need for the coastal research community to determine a minimal list of key parameters and key sites needed to observe long-term changes in the coastal ocean. In addition the community should play an active role in the design and implementation of IOOS in order to ensure that the placement and instru- mentation of long-term coastal Sentinel moorings meet both the opera- tional and research needs of the ocean community. Status of Technical and Engineering Development and Planning The technologies required for coastal surface and subsurface moor- ings and cabled coastal observatories are relatively mature and offer the OOI immediate scientific results. At present, well-instrumented moored observatories using surface buoys are operational in multiple locations around the continental U.S. Long-term deployments are possible, but are not necessarily a cost-effective design criterion as coastal buoys can be easily serviced. The ease of accessibility is fortunate since biofouling rates are high in coastal waters. Corrosion rates are also significant and high wave stresses can dramatically reduce system lifetime. Finally, the prox- imity to coastal populations leads inevitably to encounters and losses, most commonly from fishing trawlers. The ability to cost-effectively service coastal buoys frequently has the additional benefit of allowing moorings to also serve as platforms for testing newly developed instrumentation. While data telemetry from open-ocean sites is often a problem, coastal sites can combine line of sight radio-frequency modems, cell phone, and Iridium satellite links for real- time data transmission. Pilot deployments of integrated sensor packages that have successfully collected physical, biological, and chemical data have occurred off the coasts of California and Oregon and in the Gulf of Maine. Traditional coastal radar sites are monostatic and consist of transmit and receive antennas. These antennas work together to measure the scat- ter of the transmitted signal off the ocean surface. Since these systems use the phase of the transmitted signal to interpret the signal from the re- ceiver, they require that the transmitter and receiver be physically con- nected. However, using the GPS satellite-timing signal should make it possible to synthesize the transmitted signal at the receiver, allowing the transmitter and receiver to be physically separated and converting the monostatic backscatter system into a bistatic forward-scatter system. This synthesis increases the footprint of the system and allows transmitters to
STATUS OF PLANNING FOR PROPOSED RESEARCH-OR/ENTED OBSERVATORIES 71 be placed on moorings. Additionally, GPS timing allows the transmitters and receivers to be simultaneously operated at the same frequency in monostatic and bistatic mode, forming a multistatic radar array. Multi- static operation thus increases resolution, decreases the geometric dilu- tion of precision errors nearshore, and potentially decreases the number of expensive monostatic systems that need to be purchased. For example, offshore points within view of multiple sites will experience a significant decrease in the expected error of the total current vectors because they contain N2 rather than N component estimates. The greater number of points available means smaller radii averaging circles in the total vector calculation used, enabling the network to better resolve fronts. These de- velopments will be necessary for coastal research efforts where effectively resolving currents is a key to success. The potential of radar arrays has been demonstrated over the past five years, and the OOI would allow for the development and deployment of high-resolution multi-static arrays, ideal platforms for collecting coherent surface current data in coastal waters.