National Academies Press: OpenBook

Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories (2003)

Chapter: 4 Implementation of a Network of Ocean Observatories for Research

« Previous: 3 Status of Planning for Proposed Research-Oriented Ocean Observatories
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 72
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 73
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 74
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 75
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 76
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 77
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 78
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 79
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 80
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 81
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 82
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 83
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 84
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 85
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 86
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 87
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 88
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 89
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 90
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 91
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 92
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 93
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 94
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 95
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 96
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 97
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 98
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 99
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 100
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 101
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 102
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 103
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 104
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 105
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 106
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 107
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 108
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 109
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 110
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 111
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 112
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 113
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 114
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 115
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 116
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 117
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 118
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 119
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 120
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 121
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 122
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 123
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 124
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 125
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 126
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 127
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 128
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 129
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 130
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 131
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 132
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 133
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 134
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 135
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 136
Suggested Citation:"4 Implementation of a Network of Ocean Observatories for Research." National Research Council. 2003. Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories. Washington, DC: The National Academies Press. doi: 10.17226/10775.
×
Page 137

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.

4 Implementation of a Network of Ocean Observatories for Research Many issues, some of them complex, must be addressed for the suc- cessful implementation of a seafloor observatory network for research. The following sections address many of these issues, including program management, infrastructure and sensor needs, factors affecting construc- tion and installation, operations and maintenance, data management, and education and outreach. It is beyond the scope of this report to develop a comprehensive implementation plan for ocean observatories; instead, the goal of this chapter is to highlight some of the most important issues to be addressed in such a plan. The first task of the observatory management organization described below, should be the development of a detailed and comprehensive project implementation plan for each of the three major components of the OOI and the review of these plans by knowl- edgeable and independent experts. PROGRAM MANAGEMENT Even though the formal start of the OOI is not planned until FY 2006, a large amount of work must be done during the intervening years in- cluding detailed (node level) scientific planning, technical development, and exhaustive testing of critical observatory sub-systems. These activi- ties will ensure that (1) the risks associated with the construction and installation of the more advanced observatory systems are minimized, (2) the initial science experiments at individual nodes are identified so as to provide an opportunity for an early scientific payoff once the observato- 72

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 73 ries are in place, and (3) research scientists, educators, and the public have ready access to the data generated. For this reason, a management struc- ture should be established as soon as possible, and in any case well before the initiation of the OOI. Goals of the Management Structure The development of a network of ocean research observatories will require a large initial investment in excess of $200 million (National Sci- ence Foundation, 2002~. For this reason alone, the management system will be under intense scrutiny by Congress, NSF senior management, the U.S. Inspector General, and the marine science community (which has concerns about other programs being cut to cover observatory cost over- runs), as well as international partners, who must satisfy the concerns of their own funding agencies. Therefore, the first tasks of the management structure should be to: · develop a detailed implementation plan for the OOI; · generate defendable cost estimates; · put in place oversight mechanisms and fiscal controls to ensure that implementation tasks are completed on time and within budget; · establish a scientific and technical advisory structure to obtain com- munity input; and · work collaboratively with international partners to seamlessly in- tegrate complementary international activities. Design, Construction, and Installation Phase With respect to scientific planning and observatory installation, the management structure would oversee the following: · defining science-based performance goals (based on broad com- munity input); · producing an annual program plan and budget; · overseeing design, development and manufacture of observatory components and selecting contractors for those tasks; · selecting contractors for installation of observatory systems; · providing experienced oversight of contractors; · managing liability issues; · facilitating the development and implementation of standards (e.g., for user power; communications, and timing interfaces; metadata require- ments; system, sub-system and component reliability; and information management and archiving);

74 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY · facilitating seamless system integration; · ensuring comparability and inter-calibration of observations made by the OOI; and · ensuring that the scientific and technical components of interna- tional collaborations are coordinated effectively. Operations Phase As observatories become operational, the management structure must take on these additional tasks: · selecting observatory operators and putting in place appropriate review procedures; gets; tently; · ensuring that the observatory infrastructure supports the highest quality science and provides researchers with the best available technol- ogy; including calibrated sensors and instrumentation, at the lowest cost consistent with the safe, efficient operation of the facility; · establishing an appropriate budgetary balance between observa- tory operations and maintenance and enhancements to observatory infra- structure; and · ensuring the program has a strong and innovative education and outreach program. · managing the operations, maintenance and administration bud- · ensuring that access to observatories is dealt with fairly and consis- The Driving Philosophy The philosophy of the OOI management structure should be one in which the day-to-day operation of different components is the responsi- bility of entities (academic or commercial) with appropriate scientific and technical expertise. The role of the program management organization should be one of coordination, oversight, and fiscal and contract manage- ment. The management structure will need to work with the scientific community to select, support, and periodically evaluate "community" experiments; define access requirements; provide technical support for individual investigator-initiated experiments; facilitate education and out- reach access to selected data streams and products; develop protocols for scientists not involved in deploying experiments to access databases and archives; and negotiate access agreements with other users (such as for- profit entertainment industries and value-added enterprises). Operating rules for the observatories will have to take into account the needs of the

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 75 scientific community, agencies interested in using or supporting the use of the facilities; international partners and collaborators, and other users, including the public. Proposed Management Model: Roles and Responsibilities An example of a management structure capable of addressing the goals and issues summarized above is presented in Figure 4-1. This struc- ture is modified from a draft management structure developed by the NSF and the DEOS steering committee and is modeled after the highly successful management structure of the international ODP. The ODP management model guided a complex program now in its fourth decade. It has shown itself to be flexible and capable of evolving in response to changing circumstances, yet stable enough to keep a multina- tional program operating year after year. This model, while an excellent starting point, does not fit the circumstances of the OOI exactly, and so has been modified to reflect the following important differences: · While the structure of the ODP often tends toward prescriptive technical requirements, the OOI will need performance-based require- ments, as the OOI will consist of many separate observatories using mul- tiple technologies in pursuit of different objectives. In addition, disparate parts of the system will be in different stages of development at any given time. · For the most part, the ODP utilizes standard technology developed for the oil exploration industry. The OOI will be utilizing more advanced technologies adapted specifically for ocean observatories and is likely to place a much greater emphasis on new technology development. As a result, the OOI technical advisory and management structure will need expertise to provide oversight of the operation of technologically sophis- ticated systems and engineering development projects. · The NSF is the only agency in the U.S. providing ongoing support for the ODP. It is likely that the OOI will have a number of agency sup- porters at the federal (and possibly even state) level, particularly during the operational phase. Many of these agencies will be interested in sup- porting only a few observatories or a single observatory type. · In the ODP, all international operating funds flow through the NSF and are managed as a single commingled pool. International funding for the OOI will mostly occur at the individual observatory level, or for a particular observatory type, and will likely be spent at the national level as part of a coordinated rather than commingled pool. · In the ODP, all of the international partners belong to the entire program. In the OOI, international partners may be interested in partici- pating only in a subset of the observatory system.

76 u) ~ a, ·,. 'O Z Q U) o o LL! U) ~ ·O Z Z ~ ~ O TIC An ~ .c ~ Z O ZO U) O O o F ~ Q ~ a) Z \ it_ / / / / / / / ~ / a: / a)0_ ~ U) ~ U) o 8 , robs' ~ At\ ~X ~ a: \ ~ \ / ~ U) ~ I . U) o CO Cal U) ~ o o / ~ ·~ . U) o O: g CO U) o o .O ~ U) o . CO C~ U) o _ . ............................ ~ · --- U) ....... ~ ~ .... ~ ... s o (1) u) / (d o , ~ C 1 ° \ ~ ~ \ g au v) ¢ \ au o au o 5- bC o 5- o o o au 5- v) au bC au v) o o 5-

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 77 The NSF will be the lead funding agency for the OOI. Other agencies that fund marine research may elect (and should certainly be encouraged) to support the OOI, but their contributions should be funneled through the NSF to ensure that the program is well coordinated and efficiently managed with clear fiscal accountability. Coordination of the OOI with the IOOS, the COOS, and other na- tional and international observatory programs will be critical in the areas of infrastructure development, instrumentation, ship and ROV utiliza- tion, data management, and technology transfer. The NOPP in its role as a coordinator for federal agencies, academia, and industry, should be utilized to facilitate this organization among the different U.S. agencies supporting observatory research. The NSF will be responsible for devel- oping appropriate coordination agreements with potential international partners, perhaps through bilateral Memoranda of Understanding. The management structure of the OOI must ensure that the project is in compliance with NSF policies and procedures and other federal regula- tions. It is critical that a single entity have overall financial and manage- ment accountability for the program. In the case of the OOI, this could be the Ocean Research Observatories Program Center (OROPC). The OROPC would enter into a cooperative agreement with the NSF to manage the OOI. Ideally, the OROPC should be a community-based organization ac- countable to the scientific community it serves. It could be either a new 501(c)3 not-for-profit corporation formed specifically for this purpose, or a division of an existing 501(c)3 corporation with demonstrated expertise in managing technically complex research facilities. The latter is the pre- ferred alternative. The OROPC would be advised by a Board of Gover- nors, whose members would include senior industry leaders with experi- ence in managing complex marine engineering projects as well as leaders of scientific institutions with expertise in managing research consortia responsible for major facilities. In addition to its advisory role, the Board would be able to provide Congress and senior agency management with an independent assessment of the OROPC's fiscal and technical manage- ment performance. The OROPC's primary responsibility would be coordination and pro- gram oversight. It would have a comparatively small staff including; a Director, a Program Engineer, a Data Management Coordinator, an Edu- cation and Outreach Coordinator, a Public Affairs Officer, a Contract Manager (or equivalent) to oversee contracting and support annual au- dits, and other staff as necessary. The Program Director should be com- petitively selected and should be a person of the highest scientific and technical caliber with a demonstrated ability to manage an organization of this scope.

78 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY The OROPC would be advised by both an Executive Committee com- prised of scientific and technology leaders, which would focus on policy issues, and by a Science Committee, which would provide scientific and technical advice derived from appropriate standing and ad hoc subcom- mittees channeled through standing Scientific, Technical and Operations Advisory Committees. In the tradition of the ODE, the OROPC would disregard advice from these committees only under exceptional circum- stances and then only with the concurrence of the NSF. The Science Committee and its subcommittees and panels would consist of a broad, diverse, and interdisciplinary membership selected on the basis of ex- cellence and creativity within their respective fields. Members could be selected from academia, industry, government, and the international com- munity. International representation on the top four committees (Execu- tive, Science, and the two Advisory Committees) should be determined by NSF agreements with international partners. The actual design, development, manufacture, construction, installa- tion, and operation of the observatories that compromise the OOI will be subcontracted by the OROPC to consortia, individual institutions, or pri- vate companies as appropriate. This decentralized management structure will promote maximum creativity and the tailoring of the management of each observatory system to the specific scientific goals and operational requirements of that particular system. Each operating entity will have flexibility of implementation (to encourage innovation), but will have to meet certain performance criteria in such areas as user interfaces, data management and archiving, access to education and outreach users, and maintenance and upgrade strategies. This structure will ensure that the entire system of observatories is more than just the sum of its parts and that research and educational users of both the facilities and data streams can move easily from one observatory to another. The OROPC will be responsible for working with the advisory structure to develop system- wide performance goals that balance the need for flexibility to encourage innovation with the desire to maintain maximum system functionality. International participation could be at the level of the entire research observatories program, or with specific components of the program and might range from simple coordination of independently funded and man- aged efforts to an integrated, jointly funded observatory program. The management of the OOI will need to be flexible enough to accommodate these different modes of international participation, as long as the integ- rity and transparency of the entire system are not put at risk. In the case of coastal research observatories, it is clear that the uni- verse of potential state, local, industry, and other federal partners is much larger than for the open-ocean observatories. The partners will also differ

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 79 markedly from region to region and will likely include operational obser- vatories in some cases. Again, the NSF and the OROPC will have to be flexible in negotiating cooperative agreements that encourage broad par- ticipation without endangering the connectivity and synergism of the entire system. Over time, some observatories that are not part of the ini- tial OOI may wish to join this research observatory network. The pro- gram will need to develop a process by which this incorporation can be undertaken, as well as different options for program participation (e.g., some existing programs may only wish to utilize the data management system while others may wish to become a full partner in the scientific planning, operation, and maintenance of the observatory network). DEFINING "INFRASTRUCTURE" FOR OCEAN OBSERVATORIES As noted earlier in this report, funding for the OOI is being sought through the NSF's MREFC account. This account was established to pro- vide funding for major science and engineering infrastructure including the acquisition of: state-of-the-art tools that are centralized in nature, integrated systems of leading-edge instruments, and/or distributed nodes of information that serve as shared-use, networked infrastructure in advancing one or more fields of scientific study. (National Science Foundation, 2003, p. 1). [Note that in the MREFC context, "infrastructure" is used inter-changeably with "tools"]. While the three major components of the OOI described in Chapter 3 of this report clearly satisfy this definition of "infrastructure," there has been some confusion over what this infrastructure includes. From the OOI's inception, some of its proponents have argued that use of the MREFC should focus on acquiring the basic elements of an ocean obser- vatory system (e.g., cables, moorings, junction boxes, shore stations, and facilities for data distribution and archiving), not on acquiring the instru- mentation that would eventually utilize this infrastructure. Implicit in this approach is the assumption that funding sources other than the MREFC would provide support for instrument development and acquisi- tion as well as the deployment and maintenance of these instruments at various observatory nodes. The mechanism for obtaining this crucial sup- port would vary from project to project depending on the nature of the experiment, the sponsoring funding agency, and the role of the instru- ments in the overall observatory network. This mechanism would likely involve peer review, thus ensuring that only the most useful and best- justified instrumentation would be incorporated into the observatory sys- tem.

80 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY The rationale usually given for the above approach is that basic infra- structure funds are difficult to obtain. As a result, as much as possible of these funds should go toward acquiring basic hardware (e.g., cables, buoys, moorings and junction boxes). Risks of such an approach include a possible lack of available funding from other sources to acquire observa- tory sensors and/or the absence of a full suite of instruments that can utilize the observatory infrastructure once the construction and installa- tion phase of the OOI is completed. Others, therefore, have argued that some sensors and instrumentation should be included as part of the ob- servatory "infrastructure" even if doing so restricts the availability of resources for acquiring the cables, moorings, and junction boxes that com- prise the facility. The difficulty with this alternate approach is determin- ing which instruments and sensors should be included as part of the basic observatory infrastructure. The instrument lists developed in various ocean observatory workshop reports are long and costly. Including all, or even a significant fraction, of these instruments as part of observatory infrastructure could significantly reduce the number of observatory nodes that could be established with the limited funds available through the MREFC account. Another danger is that incorporating too many instru- ments into the basic infrastructure could discourage innovative produc- tion of new and better instruments. In considering this trade-off, certain analogies with oceanographic research vessels should be considered. In the case of a research vessel, the ship itself is the basic infrastructure. Scientists typically bring their own specialized instruments and install them on the ship for each expedition, using the ship as a platform for acquiring their data. Most ships, however, also include a basic suite of instrumentation that is required by most investigators (e.g., a GPS, an echo sounder, and a conductivity-tempera- ture-depth [CTD] profiler). By providing this basic instrument suite the ship operators ensure that every research vessel has a certain minimum scientific capability. This baseline capability is likely to be even more important at an ocean observatory, since the value of a node for special- ized and in some cases shorter-term scientific experiments may be very dependent on the availability of long time-series data of certain basic physical, chemical, and biological properties at the site. The critical question is thus not whether sensors and instruments should be considered as part of the basic observatory infrastructure (in- deed they should), but deciding which sensors or instruments should be part of the basic observatory infrastructure (thus funded through the MREFC account), and which sensors should be acquired by the scientific programs utilizing the observatories (thus funded through the Research & Related Activities account at the NSF or by other agencies supporting ocean research).

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 81 In considering this question it is useful to define three classes of in- strumentation that may be installed at ocean observatories. The first class of instruments, "core" instruments, include a basic suite of engineering and scientific instruments that are essential to the functioning of the ob- servatory and its usefulness as a platform for basic research. Core instru- ment needs will vary widely for different classes of observatories and will depend on the scientific objectives of each node. Such instruments could include (1) engineering or system management instruments used to de- termine the system's operational status, and (2) commercial off-the-shelf (COTS) instruments that make basic physical, chemical, or biological mea- surements and provide essential scientific context for the observatory's effective use as a platform for scientific research. Data from core instru- ments should be available to anyone in real-time, or as soon as is practical, through the observatory data management system. Furthermore, the core instruments should be maintained and routinely calibrated to interna- tionally-agreed upon standards so these data can be integrated with ele- ments of other observing networks. A second class of instruments, "community instruments," consists of specialized scientific instruments critical to the longer-term scientific ob- jectives of a particular node. These typically will be proven and reliable ('observatory-capable') instruments that provide data of interest to a wide range of investigators and that need to be in operation over an extended period of time. Examples might include ocean bottom seismometers, cam- eras and video systems, mooring line 'crawlers,' or borehole fluid sam- plers. Data from community instruments also should be freely available in real-time or as soon as is practical. The third class of instruments that will be used at most observatories will be those associated with individual, investigator-initiated experi- ments. These "investigator owned" instruments may be new or develop- mental, or may be specific to a particular scientific study or experiment. Data from such instruments may be proprietary to the investigator for some specified time period consistent with the data policy of the sponsor- ing funding agency (e.g., two years for the NSF). Data from these instru- ments must still be submitted to the ocean observatory data management system and should be made publicly available after the embargo period ends. The core instruments, as defined above, comprise an essential ele- ment of the basic observatory infrastructure that should be supported through the MREFC even if that means a reduction in the overall size of the observatory facility. Shore-based facilities for data distribution and archiving are also part of the basic observatory infrastructure that should be supported through the OOI. In some cases a community instrument might also be important enough to the scientific rationale of a particular

82 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY observatory that it should also be considered as part of the basic infra- structure. In most cases, however, funding for those science programs (e.g., CLIVAR, Ridge 2000, GLOBEC) or groups of investigators using the facility would seek funding for "community instruments" from sources other than the NSF's MREFC account via peer-reviewed proposals sub- mitted to the NSF or other agencies supporting ocean research. Since core instrument needs will vary widely from observatory to observatory, it is inappropriate for this report to define a list of core instruments or to specify a certain percentage of MREFC funds that should be utilized for core instrument acquisition. The proponents of each observatory system will be in the best position to judge the trade-off between basic observa- tory hardware (i.e., the number of nodes) and the basic sensor require- ments for that hardware given the finite resources available through the MREFC. The expectation, however, is that every observatory will require some core instrumentation. Even if core instruments are included as part of the basic observatory infrastructure funded through the MREFC, they will constitute only a small portion of the longer-term instrument needs for ocean observato- ries. The total investment in sensors and instrumentation for observatory systems acquired through the OOI could, over time, approach the cost of the observatory infrastructure itself. The research community has ex- pressed concern that the funding to acquire these instruments may not materialize and that, as a consequence, access to the observatory infra- structure will be delayed and the full scientific potential of ocean obser- vatories will not be realized. The long-term scientific success of the re- search-driven observatory network will depend at least in part on the development of a program within the NSF's Ocean Sciences Division which will select peer-reviewed proposals for funding of new observa- tory sensors and instrumentation. Given the significant lead-time involved in constructing and acquiring new instrumentation, the NSF is encour- aged to establish an "Ocean Observatory Instrumentation Program" well in advance of when these observatories become operational. As instru- mentation needs at observatories will evolve continuously (see the fol- lowing discussion), such a program will be needed as long as ocean ob- servatories remain in operation. Other agencies with an interest in ocean research may also support acquisition of instrumentation for ocean obser- vatories and the NSF is encouraged to explore these options, perhaps through an interagency mechanism such as the NOPP program. SENSORS AND INSTRUMENTATION NEEDS Making integrated physical, chemical, and biological observations in the oceans presents challenges quite different from those faced by atmo-

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 83 spheric, terrestrial, or space scientists. Remote sensing via satellite is im- possible in most of the ocean, excepting the very thin upper layer. To penetrate the ocean's depths, power for instrumentation and communica- tions must be delivered in an 'ocean-proof' package, for instance via a shore-side cable or limited-lifetime in situ battery packages. Due to the considerable technical challenges and the time required for converting new technologies into robust, seaworthy packages, as well as the different nature of economic forces that could motivate industry investment in such development, ocean-going instrumentation lags far behind state-of- the-art technologies in other fields. Even with all of the recent advancements in computer and sensor technology, the physical challenges of making measurements at the sea surface, on the seafloor and in the water-column remain vexing. Strong winds, high waves, platform motion, high salinity, pressure, and biologi- cal activity (i.e., biofouling) can all conspire to complicate sustained con- tinuous deployment of new technologies at ocean observatories. Iterative cycles of designing, building, field-testing, troubleshooting, redesigning, and redeployment are required before instrumentation becomes seawor- thy enough for routine use. The challenges of continuous and auton- omous operation are even greater in the context of open ocean observa- tories. A major, sustained, and well-funded effort will be required to develop a new generation of instruments and sensors for ocean observa- tory science. The on-going work of sensor calibration, as well as the con- siderable maintenance instruments often need after a long deployment at sea, will also require a considerable investment in staff and facilities. The development, calibration, and maintenance of new and robust instrumentation for quantifying the physical, chemical, and biological ocean will be a key element in achieving the true interdisciplinary prom- ise of ocean observatories and implementing integrated studies and mod- eling of such ocean systems in real-time. Current State-of-the-Art in Ocean Instrumentation To offer a rough analogy, ocean observing systems are similar to complex living systems. Cables, fiber, and wire provide the backbone for physical support and the nervous system to deliver information and en- ergy throughout the network. On-board computer systems perform brain functions, coordinate activity in the network, process information, and manage communications, both within the internal system as well as with external systems via cable, acoustics, or satellites. The sensors themselves see, hear, taste, and feel the ocean environment either directly, or via proxies, and report those data back through the observatory backbone and nervous system infrastructure. Thus sensors form the crucial part of

84 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY the observatory network that allows for the acquisition of new informa- tion about the ocean. Sensor development and implementation are difficult to discuss in a generic way, since the variety of sensors and instruments is large and will vary with specific problems, projects, and deployment scenarios. Never- theless, careful planning of the development and deployment of new sensors and instruments will determine the ultimate impact of ocean ob- servatories on knowledge of ocean processes and dynamics. A "spectrum of maturity" currently exists with respect to readily available and routinely deployable ocean observatory sensors. Physical sensors, the mature end of the spectrum, represent the most reliable, ro- bust, and routinely deployable instruments available. Instruments for measuring meteorological parameters, salinity, temperature, pressure, current speed, light quantity and quality, and seismic waves can be fairly routinely deployed in the ocean (Figure 4-2~. Further along the spectrum, instruments for specific detection and quantification of ocean chemical parameters represent a maturing, but not yet fully developed area. For example, it has recently become possible to acquire continuous real-time, in situ measurements of carbon dioxide (Friederich et al., 2002) or non- conservative bioactive compounds like nitrate Johnson and Coletti, 2002), but such data are not yet universally collected. More work is required in the area of chemical sensors that can measure a wider range of analyses with greater sensitivity, resolution, and reliability. The least well- developed instruments, though perhaps the most important for the com- ing century, are biological sensors. At present, only a few big-monitoring devices are available, and these are relatively crude with respect to their sensitivity and specificity. Existing in situ biological instruments are largely big-optical devices, measuring either light scattering as a proxy for biological particles or chlorophyll fluorescence as a proxy for phyto- plankton for instance. Although development efforts for autonomous bio- logical sampling and sensing instruments using molecular probes are underway (Scholin et al., 1998), the field is nascent and much more effort will be required to produce readily available and robust biological sensors. Ocean Observatory Sensor and Instrument Development The need for sensor development has been clearly articulated in a number of workshops and reports on the future of ocean observing sys- tems and in situ instrumentation. Examples of such discussions include the NSF report Ocean Sciences at the New Millennium (2001~; a previous NRC report on ocean observatories, Illuminating the Hidden Planet (2000~; a workshop on in situ sensors (RIDGE, 2000~; the SCOTS workshop (Dickey and Glenn, 2003~; and the CoOP workshop (Jahnke et al., 2002~. The rec-

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 85 FIGURE 4-2 A remote instrument node (RIN) under development at MBARI; the ROV Ventana is shown on the right. This RIN contains a variety of instru- ments for measuring water temperature, conductivity, and density as well as current speed, sediment load, and chlorophyll concentrations. This RIN will even- tually connect to the MARS cabled observatory testbed. Figure courtesy of Todd Walsh (a) 2002 MBARI.

86 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY ommendations and conclusions of these reports all emphasize the critical need for the development of new sensors and instrumentation for ocean observatory science. A sampling of these recommendations is given in Box 4-1.

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 87 Instrument Re/iabi/ity and Calibration Observatory-capable instruments must meet international standards for accuracy and must maintain their calibration and sensitivity for long periods of time, at least for a period between six months to a year (the expected service interval for many ocean observatories). Ocean instru- mentation often borrows and adapts technology from other developing fields, many with stronger economic drivers for development. As such, these instruments need to be adapted for the unique challenges of long- term deployment in the ocean environment. Problems common to nearly all ocean instrumentation include biofouling, corrosion, and physical damage due to the harsh ocean conditions. Generic solutions to these long-term deployment problems need to be addressed and, where pos- sible, shared across the community. The international ocean observing community should agree upon routine calibration standards in order to document instrument performance and support the use of OOI observa- tions for scientific research and integration with observations from other elements of global Earth- and ocean-observin~ networks. - -O Iterative Development and Deployment Cycles Iterative design, development, and deployment cycles are necessary to achieve robust instrument designs and reliable, maintenance-free in- strument operation. As a result, the development of new ocean instru- mentation can be a lengthy process, often taking five years or longer. Easy access to a mooring (e.g., the BTM) or a cabled seafloor junction box (e.g., LEO-15, FRF, MARS), located close to shore for testing new technology and instrumentation can greatly accelerate the pace of developing and proving new instrumentation. These testbeds also provide a platform for establishing the comparability of a new technique with older methods that are being phased-out. The creation of centers for instrument develop- ment may also facilitate this process, but it is essential that such centers work in close collaboration with the scientists who will be using these instruments. The NSF, working in cooperation with other agencies, also needs to ensure that ongoing deployment and calibration activities can be sustained and that the staff and facilities for maintenance and calibration are up to the task of meeting the additional ongoing demand created by instrument development and deployment cycles associated with the OOI. Multi-lastrument Interface and System Integration issues Some of the major goals of ocean observatories include integrating the data streams of individual sensors and instrumentation in real-time

88 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY and allowing for synoptic measurement and correlation of data streams. Common instrument interfaces for Plug-n-Play operation, as well as inte- grative data handling, are part of the solution to these issues. Neverthe- less the diversity of sensors that are likely to be developed, as well as the range of technologies utilized, will present challenges to full sensor suite integration. Obviously instrument and sensor developers will benefit greatly by communicating and integrating their efforts into the overall technological context of observatory infrastructure. Practically speaking, a range of 'interactivity' will probably evolve for different instruments, ranging from autonomous operation to highly integrative operation with other sensor suites. Both technology and the specific needs, goals, and science drivers for any particular sensor application will likely drive this range of interactivity. Data Stream Management As is the case with other system integration issues, data streams from particular sensors will be quite varied. Metadata, calibration and valida- tion data, and raw and processed data streams are all instrument-specific. Data management architecture and instrument interfaces are unlikely to ever be totally standardized. A flexible system is likely to evolve, span- ning a wide range of instrument interoperability and compatibility within the overall observatory infrastructure. Sensor Development Funding and Anticipated Future Needs National funding agencies have recognized the need for environmen- tal sensor and instrument development, and a number of programs cur- rently exist that could support sensor and instrument development for ocean observatories. These programs include the NSF's Directorate for Engineering and the Directorate for Computer and Information Science and Engineering; the NSF's Instrumentation Development for Environ- mental Activities (IDEA) program; the Oceanographic Instrumentation Development program within the NSF's Division of Ocean Sciences; the multi-agency NOPP; NOAA's Cooperative Institute for Coastal and Es- tuarine Environment Technology (CICEET); and the NOAA-funded Alli- ance for Coastal Technologies (ACT). It is clear that the agencies supporting ocean research in the U.S. have already taken the need for sensor and instrumentation development seri- ously. However, the establishment of ocean observatories will increase the requirements and potential user base for ocean sensor suites far be- yond those of today. Such increased demand will require new and signifi- cant resources for sensor and instrument development. A critical issue

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 89 revolves around the time required for iterative development and deploy- ment cycles before robust and mature instruments are available. If timed improperly, the observatory infrastructure could be obsolete or in need of replacement by the time many critical instrument and sensor suites are available. A major challenge for the national programs and funding agen- cies will be fitting the required iterative cycles inherent to instrument development into the fiscal cycles and current funding schemes for stan- dard awards (typically two- to three-year funding cycles). Development and engineering milestones and timelines are often different from parallel scientific milestones and timelines, and the review and award process should be responsive to these differences. If truly integrated interdisciplinary in situ observations of ocean sys- tems are to emerge from the OOI, then a much wider variety of sensor suites, in particular those for sensing chemical and biological phenomena, will need to be aggressively developed in the very near future. If the same standard oceanographic parameters continue to be the only items mea- sured, then only an incremental gain in scientific knowledge can be ex- pected, even with the enhanced spatial and temporal resolution that ob- servatories will provide. There will be little benefit from seafloor observatories unless a broad base of multidisciplinary ocean scientists use them to advance knowledge and understanding of the oceans. The re- search community as a whole has not yet attained this goal, and one consequence of the OOI should be to foster this evolution. Sensors and instrumentation are as important as cables and platforms, and their ag- gressive and immediate development is an absolute requirement if ocean observatories are to fulfill their potential promise. In addition sufficient support must be provided to maintain sensors and instrumentation, espe- cially to ensure via calibration their accuracy, reliability, and comparabil- ity across all platforms of the OOI and other ocean-observing networks. CONSTRUCTION AND INSTALLATION The construction and installation of ocean observatories give rise to many issues. Although there are significant differences in the require- ments for moored and cabled observatories, in both cases careful pre- installation planning and extensive shore-side and wet testing will be required to ensure the successful installation of these systems. Moored Buoys The oceanographic community has considerable experience with the fabrication and installation of both surface-expression and sub-surface moorings under a variety of conditions. Both types of moorings are easily

go ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY fabricated, and prior experience has indicated the materials required. Torque-balanced wire rope, which is plasticjacketed for corrosion resis- tance, is used for the upper part (1500 m) of a deep-ocean mooring to prevent fish bite and at all depths in shallow moorings. Nylon and dacron rope are used to allow for compliance and polypropylene rope is used where added buoyancy and stretch-resistance are needed. Select marine- grade, corrosion-protected fittings (shackles and swages) and chain; foam, aluminum, and steel buoy hulls; and synactic, steel, or glass subsurface flotation are also widely used. It should be noted that a range of manufac- turers can build metal buoy hulls; by comparison, only a few manufactur- ers can fabricate hulls from closed-cell foam. Mooring cable and rope and related hardware can be purchased from commercial sources. Cutting, terminating, and strength-testing lengths of cable has traditionally been done in-house. The bandwidth requirements for many oceanographic moorings are modest and can be accommodated by commercial satellite systems such as Inmarsat-B, Iridium, or Globalstar. Low-power transceivers and an- tenna systems are commercially available and can be purchased at very modest costs. Acoustic modems can be used to telemeter data from sen- sors on the mooring or on the seafloor to the buoy, although data rates are relatively low. Second-generation acoustic modems are available com- mercially and third-generation systems are under development. These modems can provide sustained data rates of 5 Kb/s (with error correc- tion) at ranges of up to several kilometers. With exception of the large, severe-environment surface buoys (see below), the current trend is to make all of the mooring components discussed above easy to ship by designing them to fit inside standard 20-40 ft sea containers. The presence of a dedicated winch (drum or traction) to deploy and recover the wire rope and synthetic line greatly facilitates mooring de- ployment. Deployment practices are well established and documented by groups such as the Mooring and Rigging Shop at WHOI. Installation can be handled by medium and large UNOLS vessels. Heavily-instrumented deep-water surface moorings may require an- chors approaching 10,000 lbs. That weight, plus the weight of the buoy hull (4,000 lbs with instruments), place constraints on crane and A-frame capacity and on the number of moorings that can be carried at one time. Deployments and recoveries are accomplished either over the side on the fantail or through the A-frame on the stern. Mooring deployments re- quire moderate to calm weather (15 knot winds or less and waves smaller than 6 ft). The greatest dangers faced in high seas are the possibility that the buoy hull will swing, hitting the ship during either lift-off or the transition back onto the deck and the possibility that surge loads could approach the breaking strength of the mooring lines. In addition, com-

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 91 plete cable-linked, low-bandwidth moorings will require an ROV for in- stallation of a seafloor junction and connection to the EOM riser. The high-latitude and/or high-bandwidth discus or spar buoy sys- tems that meet the specifications outlined in the DEOS Moored Buoy Ob- servatory Design Study (DEOS Moored Buoy Observatory Working Group, 2000) raise special construction and installation issues due to the size of the buoys and the weight of the mooring lines and anchors (Figure 4-3~. Fabrication of large spar and discus buoys would be contracted to a com- mercial company (moored systems typically very large spar buoys- have been in use for some time in offshore oil production). Several com- mercial VSAT C-Band antenna systems, designed primarily for shipboard use, are available and could be adapted for use on a large spar or discus FIGURE 4-3 Artist's conception of a 40-m spar buoy design for a high-band- width moored-buoy observatory. The main module (gray box) will contain com- partments for the electronics and communications electronics, generators, and batteries. A C-Band antenna radome is mounted atop this module. Fuel bladders are located in the spar. Deployment of this system will require two legs, one to install the mooring and spar buoy; a second to install the topside module. Figure courtesy of John Halkyard, Technip Offshore, Inc.

92 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY buoy. The generators required to power these systems are also commer- cially available, and there is considerable experience with their operation on unattended buoys. The high-bandwidth discus or spar buoys described in the DEOS Moored Buoy Observatory Design Study (DEOS Moored Buoy Observatory Working Group, 2000) also entail special deployment requirements. In the case of the large discus buoy design, the weight of the anchor and the diameters of the steel and synthetic mooring lines are greater than can be handled by standard UNOLS winches and wires. In order to deploy these moorings from a global-class UNOLS vessel, an independent winch sys- tem would have to be installed to deploy and recover mooring compo- nents. Even the largest UNOLS vessels cannot deploy the large 40-m spar buoy described in the DEOS Moored Buoy Observatory Design Study (DEOS Moored Buoy Observatory Working Group, 2000), primarily due to deck- size limitations and reel and winch capabilities (which fall short of the 20,000+ lbs lift capability required). An offshore supply or anchor han- dling boat would be required for launching the mooring and spar buoy, and would probably be the best option for deploying the large 5-m discus buoys as well. Installation of the spar buoy and mooring would require two separate stages: one for pre-installation of anchors and spring buoys, spar installa- tion, and hookup to the mooring and a second for topsides module instal- lation (DEOS Moored Buoy Observatory Working Group, 2000~. These steps have to be performed in sequence, but not necessarily at the same time. While the first step will require a heavy-lift vessel, the topside in- stallation of power, communications, and instrumentation modules could be performed by a global-class UNOLS vessel after the spar and mooring have been installed. These deployment operations will be weather sensitive, but can be performed in seas up to about 2 m of significant wave height. Given the unpredictability of the weather at high latitudes, even during favorable times of the year, significant contingency time will need to be budgeted for installation operations. Cabled Systems: New Installation In contrast to moored buoys, the oceanographic community has very limited experience with the design, fabrication, and installation of subma- rine cable systems. However, the commercial telecommunications indus- try possesses a wealth of experience and the knowledge of the exploration industry is also increasing. Both can be drawn upon in designing, con- structing, and installing research cable observatories. This section is in- tended to broadly outline the range of requirements and tasks required

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 93 for the planning, installation, and maintenance of a submarine cabled observatory system. The telecommunications industry's experience is particularly valu- able, as it has installed and operated submarine cable systems for over 100 years with an exceptional record of success and reliability. However, there are important differences between commercial systems and ocean obser- vatories. Commercial systems operate with land stations on both ends of the cable with no connectors or variable loads on the ocean floor. Nearly constant power can be supplied from either end of commercial cables with no loops and few branch points. All branches on commercial cables terminate on shore in order to make it possible to control the power system. All commercial systems operate on constant-current power sup- plies. By contrast, the power supplied to observatories will vary over a large range of current and voltages. Furthermore, large observatory sys- tems will have many connections and dead-end branches with variable power requirements, resulting in complex power and data systems and, as a result, likely lower reliability. Commercial systems lose millions of dollars per day if a failure oc- curs, resulting in the necessity for rapid and immediate repair of the system. A data loss caused by the failure of a research observatory system will not cause excessive economic impact and it will not be necessary or generally economically feasible to make repairs within days of a failure, as with a commercial system. As a result of the remote location of many observatory sites, schedul- ing a repair may take six months or more, depending on ship availability. The inaccessibility of many ocean observatory sites makes the reliability of the basic observatory infrastructure of paramount importance in the design, implementation, and operation of the proposed undersea net- work. While commercial and research cable systems share a common need for high reliability of the basic infrastructure (e.g., cables, nodes, junction boxes), at a research observatory scientists must be given the opportunity to experiment with sensors, instruments, and experiments that may fail. Thus the reliability of individual sensors and instruments on a research observatory may be less than would be acceptable for an operational observatory or commercial cable. While new sensors or instruments should be tested first on land and in prototype ocean testbeds before installation, a cabled research observatory must be designed to protect the basic infrastructure from failures of individual sensors or instruments. Commercial systems depend on highly explicit and rigid contracting to ensure uniformity, guaranteed compliance, and reliability. The failure of a contractor to perform a required task can result in costly legal action.

94 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY At research observatories this approach will be appropriate in those cases where the required tasks are routine and similar to industry procedures. Some procedures and systems at research observatories may be more experimental or developmental, however, in such cases a more flexible approach to contracting and performance will be required. Planning the Insta//ation and Maintenance of Submarine Cable Systems Many logistical details require consideration prior to laying any cable, plugging in instruments, or collecting data at nodes. This section summa- rizes the scope of work for planning the installation and maintenance of a submarine cable system. This discussion is geared toward commercial systems given a wealth of experience from the commercial telecommuni- cations industry and the exploration industry from which to draw. Detailed system design The characteristics of observatory systems will be driven by infra- structural capabilities. For example, the proposed communications band- widths for buoyed observatories are orders of magnitude less than those of cabled observatories, not because the science is less demanding but because the capabilities of satellite links are far less than those of optical cables. Science will adapt to the available support assets. For ocean obser- vatories, it is reasonable to take advantage of infrastructure that is avail- able without considerable development, and given to apply those capa- bilities to observatory science. Development of the system specifications possible given existing technology, or with moderate modification of ex- isting technology, will lead to reasonable performance expectations. Care should be taken to stay within reasonable development bounds to ensure high reliability and to prevent unexpected development costs. Cable characteristics and topology will constrain available power, thus early determination of cable lengths, necessary slack, and other pa- rameters are required before the power system can be modeled or power supplies and maximum loads determined. The power delivery of a cabled observatory system depends on the cable resistance, cable length, current, and voltage. Power loss from cable resistance increases as the cable length and as the square of the current, increasing the incentive to use low cur- rents and high voltages. High voltages, however, are difficult to use in the ocean and most ocean cables and connectors are rated at 10 kV or less. Loops and branches complicate the assessment of the maximum power available to an observatory network. Furthermore, in such a system an increase in the power supplied to one node could affect power availability for other nodes. Heavy loads at nodes distant from power sources could

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 95 also severely limit power available closer to shore. It might be more effi- cient to power nodes close to shore on separate cables from those power- ing distant nodes. Once the decision is made to go forward, detailed designs of elec- tronic systems and components of the cable must begin. One of the first priorities is to ensure that the network architecture contains compatible connecting equipment (the "Plug-e-Play" concept). To ensure this is the case, the cable provider must coordinate efforts with the operator to (1) determine complete fiber and copper requirements in the trunk line and all branches; (2) identify submerged electronics, repeaters, amplifiers, con- nectors and branching units; and (3) confirm transmission protocols and equipment specifications of sensors and data recording equipment. Building issues on land are also important and must be addressed to install all the physical property and equipment necessary to operate and maintain the land end of any cable system. Detailed planning should be undertaken to ensure that protocols exist to address concerns regarding security, heat, humidity, and flooding. Permitting One of the most time-consuming and difficult processes in planning construction of any cable system is securing the rights to place the cable, both on land and under the water. Obtaining landing rights to cross coast- lines and property easement rights to cross federal, state, county, city, and private lands can take months. To lay a cable one must first secure all federal, state, and local permits to land the cable, bury it along the land route, and construct terminal buildings. This may include performing Environmental Impact Assessments (EIA) and Environmental Impact Statements (EIS) for landing and sub-sea routes. The cable contractor or owner must also locate, and usually pay, all affected land owners. In some cases, finding all of the legal owners of a piece of land in coastal areas can be extremely difficult. In addition, determining the cost of a right of way can be a serious challenge. Besides obtaining land rights, cable owners must also secure permits to cross existing undersea pipelines and cables, if any, and ensure those owners that no harm will come to their systems. The Minerals Manage- ment Service (MMS) of the Department of the Interior governs all pipe- lines and cables laid in U.S. waters, and each state has a department that controls the shallow ends to the high tide mark. An additional state agency may be legally responsible above the high tide mark and may share con- trol with county and city governments. This entire permitting process can take a year or longer.

96 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY Submarine Cable Protection and Route Design Submarine cables are vulnerable to damage from bottom fishing, an- chor drag, currents, and other hazards that scour the bottom in shallow water (i.e., depths up to about 1,000 m). In addition, rough topography can result in cable "spans" where the cable is suspended above the ocean floor and vulnerable to damage. Such spans are a serious problem in volcanic terrain, where sharp outcrops could easily damage cables. Since many observatories will be located in shallow water or in volcanic set- tings, protection is a critical issue. Protection can take two forms, cable burial and cable armoring. The cable industry uses both burial and armoring to protect cables in shallow water but no such protection is used in deep water, where careful surveying is done prior to installation to avoid rough bathymetry and volcanic terrain. Burial of available commer- cial telecommunications cables in deep water is not possible for logistical reasons, and armoring would result in serious weight problems at depths greater than about 2,000 m. For observatories in deep-ocean volcanic ar- eas, cable design must include armoring in appropriate sections and must account for adequate slack to prevent long spans. The most critical factor in determining adequate cable protection is knowledge of seafloor bottom characteristics. Such knowledge requires high-resolution near-bottom seismic, swath bathymetry, and sidescan data along the prospective cable route as well as knowledge of pipelines and other cables that will be crossed, resulting in a protection plan. Before any actual work is performed or contracts awarded, environ- mental, geopolitical, meteorological, oceanographic, geophysical, indus- trial, and regulatory factors affecting installation and long-term security of the system must be investigated in a desktop study. The desktop study should include the following activities: · A risk assessment must be performed to ensure that the most se- cure and environmentally friendly route is selected. While the general cable route will be driven by scientific objectives and required observa- tory node locations, the specific route must take risk into account. The risk of external damage to the cable or to components existing along the route, from seabed users (e.g., fishing, dredging, oil production) or natu- ral threats (e.g., turbidity currents, mobile seabed, storms, volcanic or tectonic activity) should be evaluated as part of this assessment. · Preliminary route design and route mapping should be produced. · Burial specification and armor design for protection of the system should be specified. · Concerns regarding pipeline crossings should be handled on a case-by-case basis. Some pipeline owners may require sand bags or con-

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 97 crete mats placed between the cable and their pipeline, while others may require only a plastic duct system. · The initial cable and installation design should be proofed against existing industry capability. Following the desktop studies, a physical route survey should be ordered and should include bathymetry, sidescan, sub-bottom (for buried systems), current, and other physical measurements. The purpose of this survey is to verify the initial route design and complete a threat analysis. Geotechnical measurements should be taken and analyzed on any por- tion of the route along which the cable will be buried. System Production Items that require a long lead time to acquire like the cable, terminal equipment, sub-sea components, and software will need to be ordered well in advance of installation. Contract management systems and proce- dures should be established to ensure system quality and timely delivery of all components. Any high risk or new design items that must be proof- tested well in advance of installation (in order to provide time to rebuild components that fail) should be given special attention at this time. Installation Planning and Execution Installation of a cabled observatory system is expected to be a signifi- cant fraction of its capital cost. Maintenance of a cabled system infrastruc- ture should be less than that of an equivalent buoyed system since cabled systems do not contain consumables needing replacement on a regular schedule. Installation of long cable observatories will require the use of at least one cable ship, and may require two where there are branches in the topology. Connections to shore will require cable burial, construction of shore stations, and connection to the land-side power and communica- tions infrastructure. Observatories utilizing less than about 100 km of cable could be installed from research vessels (temporarily) equipped with cable storage and handling equipment (provided that burial is not necessary), although use of a commercial cable-laying ship may be more cost-effective. An important component of the installation of a cabled system is selecting the right contractor for the job. In commercial operations this selection usually occurs through a process that involves pre-qualifying contractors and then "going out for bid." Pre-qualifying a contractor spe- cifically includes ensuring that the cable-lay ship and burial tools are suited for the task. It also means ensuring that adequate resources are

98 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY available for the duration of the job so that weather and other operational problems do not prevent timely completion. Once the cable lay contractor is selected, the cable owner should pro- duce system installation procedures. Additional permits, different from the long-term permits, must be obtained from all relevant authorities for the installation phase, including Notice to Mariners. Quite often a third- party installation contract management team of experts is hired to ensure proper installation. These experts would also work with the installation contractor on the production of "as-built" drawings and survey data for the entire sub-sea system and components. System Maintenance Planning To minimize costs, observatory system design must address reliabil- ity and maintenance issues. The greatest reliability concerns arise in those parts of the system that are the most difficult to recover, single-point failure locations, and locations that are the most vulnerable to failure. In a cabled system, such locations include the backbone cable system, junction boxes (Figure 4-4), and near-shore and rough topographic cable sections. Failure in the cable backbone will require an expensive repair by a cable ship. Considerable effort should be expended on identifying weak com- ponents, improving reliability by minimizing electronics in the backbone, using redundant systems, and protecting against faults. The most impor- tant hedges against failure will be the use of proven technologies and adequate testing of completed systems on land, at testbeds, and in the ocean during installation. Where possible, complex systems should be installed in stages to ensure that design goals are verified before the total system is in place. Managers and technicians will need to be trained to operate the sys- tem. The establishment of a usable database to archive and test technical materials is important. At this point system restoration agreements and arrangements for land and ship-based maintenance and repair facilities should be in place. Operational Planning Once the cable is in place, long-term concerns will become a decisive factor. It is important, therefore, to determine in advance the method by which the system will be operated. In many cases, commercial systems share terminal facilities with a local telecom company. The H20, for ex- ample, leases space in the Makaha Cable Station on Oahu, Hawaii. Build- ing a facility dedicated specifically for cable landing is usually done for

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 99 FIGURE 4-4 Deployment of the junction box developed for the H20 observato- ry. The junction box provides an interface for plugging instruments into the ob- servatory network, providing power and two-data communication to the instru- ments. The OOI will provide support for junction boxes for both cabled and moored buoy observatories. Figure courtesy of Fred Duennebier, University of Hawaii. security reasons or in cases where no facility exists in the area. The cable owners therefore must determine in advance whether they will build a dedicated facility or lease terminal facilities in an existing structure. Once the physical facility is established, the Network Operations Cen- ter may be run and maintained by an employee or through a third party contract. Many commercial cables and terminal facilities are maintained through such long-term maintenance contracts. A major factor in commercial cable systems is the "backhaul" issue. The backhaul route is that portion, or link, that can be shared with exist- ing trunk lines (large cable systems). For example, if a company were to lay a cable to connect platforms in the Gulf of Mexico, originating in Venice, Louisiana, running through the outer continental shelf, and ter- minating in Galveston, Texas, the company would probably contract with existing terrestrial cable providers to furnish the continuous ring or loop required for redundancy, rather than owning and maintaining the entire cable, including all terrestrial legs.

100 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY Opportunities for Experimentation Observatory design should maximize the opportunities for science experimentation and participation. Both the connection interface at the observatory and structures for data retrieval and command capability should be kept as simple as possible to encourage participation by groups lacking highly-skilled engineering support. It should be possible to have several connection levels, ranging from very capable, high-power, high data-rate interfaces for complex experiments to simple connections for analog sensors. An important management decision involves selecting the point at which the observatory infrastructure ends and user experi- ments begin. The observatory operator may control core and community sensors, but at some level experimenters should be allowed to experi- ment without observatory management oversight of as long as opera- tions stay within power and bandwidth constraints. Testing of innovative ideas, prototype sensors, and even high school experiments should be encouraged. It is strongly advised, however, that new or experimental hardware be evaluated at a system testbed prior to installation on an observatory. Cabled Systems Reused Cables There are three ways to re-use retired cables: (1) in-place use, (2) partial relocation, and (3) complete relocation. In-P/ace Re-Use and Partia/ Relocation In-place use utilizes the cable in its original location. The cable is cut and science nodes are placed on the cut end, and possibly along the cable, as was done on the H20 (Petitt et al., 2002~. Partial relocation cuts the cable and hauls some cable and repeaters aboard a ship until enough is on board to allow placement of an observatory at a location away from the original cable path as with ALOHA (University of Hawaii, 2002~. Both of these scenarios have the advantage of using the original cable station infrastructure, eliminating the cost of bringing the cable to shore. In-line observatory nodes can be placed on the cable by performing an operation similar to a cable repair. Since telecommunications cables come ashore at two different locations, at least two observatories can be established from each cable system. Re-use by Complete Relocation It should be possible to separate long sections of cables and repeaters and move them to other locations. Moving large sections of cable has been

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 101 accomplished by the military (D. Gunderson, AT&T [retired], personal communication, 2003~. Such relocation is the most costly way to re-use cables, since it involves the use of a cable ship, the shoring of the cable, and construction of new cable station infrastructure. It also requires that the cable section to be moved is not crossed by a newer cable. However, relocation can also support observatories where they could not be sup- ported by other methods. OPERATION AND MAINTENANCE The operation and maintenance (O&M) of ocean observatories will require a significant, long-term commitment of resources and facilities. Observatory operation will also require a high level of interaction be- tween the scientific community and observatory operators since observa- tories are still an evolving research tool. While power and bandwidth might initially seem unlimited, demands on the facility infrastructure could exceed capabilities. For example, an event could occur (e.g., a sea- floor eruption or a harmful algal bloom) that would create simultaneous calls on resources by multiple observatory users. Therefore policies devel- oped to apportion available power and bandwidth under standard opera- tional conditions must also provide flexibility for intensified redeploy- ment of resources during such episodic events. In general, prioritization of use of community assets, including community instruments and ROVs, will likely be an important, on-going operational issue over the lifetime of an ocean observatory. Interference between experiments may also be a significant concern: light, chemical, or radiated acoustic noise might be a requirement for some users, but a source of interference for others. Moored Buoy Observatories Open-ocean moorings should be designed for an operational life of 10 years (Figure 4-5~. They will require annual service, with a three to five year refurbishment interval. Coastal moorings will require more frequent servicing, probably at least every quarter. Refurbishment of moored ob- servatories could involve replacing mooring components or sub-systems on the buoy (e.g., the satellite antenna system, power generators, acoustic modems). Due to the hostile environment in which these moored obser- vatories are likely to operate, these maintenance and refurbishment costs are expected to be relatively high. Annual maintenance and refurbish- ment costs are estimated to be 20% of the capital cost of the buoy and mooring (DEOS Moored Buoy Observatory Working Group, 2000) and higher costs can be expected in remote regions.

102 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY FIGURE 4-5 Surface moorings such as this one will need to be serviced annually, requiring up to one month of ship time (including transits) for each of the widely spaced moorings of the OOI global observatory network. Figure courtesy of Woods Hole Oceanographic Institution. Operationally, the most important environmental limitation for open- ocean moorings will be the effect of sea state on the satellite telemetry system. The DEOS specification calls for operation at pitch, roll, and yaw rates of less than 10 degrees per second. However, recent experience with a C-Band antenna on a large UNOLS vessel as part of the Real-time Ob- servatories, Applications, and Data Management Network (ROADNet) high seas project (see Chapter 3) suggests this requirement is too stringent and that these systems may be able to operate in much higher sea states than previously thought (beyond sea state 6~. Tri-moored spar buoys at high-latitude sites and other areas that frequently experience bad weather are likely to experience smaller motions and deliver higher communica- tion efficiency than a discus buoy. There will obviously be a strong sea- sonal variation in sea state at some sites, and there may be sustained periods where no data can be transmitted to shore. However, these data would be recorded on the buoy and can be either transmitted later or retrieved during annual servicing.

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 103 Biofouling is also a persistent problem for coastal moorings and fre- quent servicing will be required to keep instruments operational. Other major hazards for coastal moorings are fishing activity and vandalism. Small- or intermediate-class UNOLS or comparable commercial ves- sels can be used to maintain coastal moorings. Open-ocean mooring maintenance will require large UNOLS or commercial vessels. Typical maintenance tasks will include replacing or repairing sensors and com- munication systems, removing biofouling, and, in the case of high-band- width systems, refueling. Mechanical components, electronics, and even diesel generators for the high-bandwidth buoys should be fairly easily exchanged utilizing standard UNOLS ship equipment. The modular na- ture of the spar topside unit allows it to be replaced separately from the rest of the spar buoy a UNOLS vessel could conduct these operations. Recovering the mooring for a spar buoy, however, would require a com- mercial workboat-class vessel. In the case of a three-point spar mooring, the spar buoy can be recovered for repair or replacement without recov- ery of the mooring. Long-term operation of cable-linked moored observatories will re- quire the use of ROVs. In some cases an ROV may be needed even for acoustically-linked observatories to ensure proper placement of sensors, although in many cases this will not be necessary. If a buoy is lost, Argo's transmitters would allow the GPS position of the buoy to be tracked and recovered by a vessel chartered for this purpose. Cabled Observatories Many of the operational issues associated with cabled observatories have been described in the SCOTS report (Dickey and Glenn, 2003~. Main- tenance costs and logistics will vary widely for different cabled obser- vatories. The cables associated with existing cabled observatories, for example, H20, LEO-15, FRF, and MVCO have been relatively robust, requiring minimal maintenance. However, it should be noted that if a cable is damaged or cut, repair would represent a significant cost, thus O&M budgets will need to include a contingency fund for such occur- rences. A stand-by maintenance contract with a cable company would most likely be used for cable repair, as these repairs cannot be completed by a standard UNOLS vessel. The cost of such a contract would depend on the required response time. Due to the fact that repairs for a research observatory will not be as time-critical as for a commercial telecommuni- cations cable, the observatory operator should be able to negotiate a com- petitive price for this contract. The majority of cabled observatory maintenance costs are likely to be associated with nodes and instruments. Basic node maintenance (e.g.,

104 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY repair or replacement of a junction box) may not be required every year. Shallow water nodes such as those at LEO-15, FRF, and MVCO are rou- tinely serviced by divers and small coastal vessels. Node maintenance and replacement at open ocean sites, however, will require ship and ROV assets. Routine node maintenance may either be contracted or conducted from a UNOLS vessel, although node replacement may require a vessel with a heavy lift capacity (see Chapter 5~. A sufficient spare inventory should be maintained so that nodes can be replaced, rather than repaired, at sea. However, the level of spares that can be maintained will be depen- dent on the overall operations budget, which will need to be balanced against research needs. The installation and servicing of instrumentation and experiments at observatories will require the largest amount of ship and ROV days each year. It would be prudent to plan on annual instrument servicing at nodes at open-ocean sites and more frequent servicing at coastal sites. Since much of this work is unlikely to be routine, observatory instrumentation work is probably best conducted by UNOLS vessels outfitted with ROVs. Given the expense of instrument maintenance, reliability and cali- bration standards will need to be developed for core science and com- munity instruments before deployment. Operations and maintenance plans should include a commitment to rigorous testing and calibration of core and community instruments to ensure the quality of observatory data. Sensors should be calibrated in the lab prior to deployment, and again after recovery, and should be as close as possible to their condition when recovered (i.e., with weathering and biofouling intact). An effort should also be made to make in situ checks and calibrations, using stan- dards and new sensors brought to the observatory sites by ships. Stan- dardization of these calibration procedures should be pursued across observatory sites within the coastal, regional and global networks and across various groups and nations to ensure comparability across all ob- servatory systems. Where biofouling and sensor aging are serious prob- lems, it may be possible to periodically switch sensors to new units. Shallow coastal sites will likely require more frequent servicing, but in- strument revisits in this setting are less expensive and more easily ac- complished. Maintenance of deep-water instruments will necessarily be less frequent; therefore instrument design and calibration should take these factors into account. Observatory Operation and Maintenance Costs The operation and maintenance of the observatory infrastructure ac- quired as part of the OOI will require a substantial, long-term financial commitment on the part of the NSF. Experience demonstrates that these

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 105 maintenance costs are often initially significantly underestimated (Chap- ter 2~. The DEOS Steering Committee provided information for this report on projected O&M costs for the OOI. While these projections are rough, as would be expected at this stage in program development, they do provide a useful estimate of the potential costs. The DEOS projections are in- cluded in this report to provide some guidance to the NSF concerning the level of resources that may be required. Table 4-1 summarizes projected annual costs for the three main OOI components: a global observatory network, a regional-scale cabled obser- vatory, and coastal observatories. The global network is assumed to con- sist of 20 nodes at widely separated locations in the world's ocean. Half of the sites are assumed to be occupied by low-bandwidth moorings (acous- tically-linked or cabled) with the remainder comprised of either high- bandwidth, cabled moorings or sites using retired telecommunications cables. The O&M cost estimates are based on the DEOS Moored Buoy TABLE 4-1 Estimates of Ocean Observatory Operations and Maintenance Costs Observatory Type Approximate Annual O&M Costs Global Network Observatorya O&M (20 nodes) $7M Ship time (20 months/yr; 10 with ROV; 10 without) $15M Contingency $2M Regional-Scale Cabled Observatoryb O&M (30 nodes) Ship time including ROV (4-8 months/yr)C Contingency Coastal Observatoryd O&M Ship time (3 months/yr) Contingency Total $11M $3.6 to $6.3M $1.5M $4M $1.5M $0.5M $46 to $49M aCosts based on data from DEOS Moored Buoy Observatory Working Group (2000~. bCosts based on data from NEPTUNE Phase 1 Partners (2000) and updated figures pro- vided by NEPTUNE Office. CNEPTUNE cost estimates assume 3 days/node for sensor maintenance and 1 day/node for node maintenance annually. The NRC Committee recommended budgeting one week/ node for nodes and sensor maintenance. dVery crude estimate based on operation of existing coastal observatories.

106 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY Observatory Design Study (2000~. NEPTUNE has been used as an example of a regional cabled observatory with 30 nodes over a 500 x 1000 km area. The O&M costs are based on the NEPTUNE feasibility study (NEPTUNE Phase 1 Partners, 2000) and updated figures have been provided by the NEPTUNE Office. Coastal observatories are assumed to be a mix of moor- ings and cabled observatories. Since the required coastal observatory in- frastructure is not well-defined at this point (see Chapter 3), the associ- ated O&M costs for this OOI component are of necessity the most speculative. A number of important assumptions have been considered in these cost projections. The O&M costs apply to observatories after initial instal- lation and commissioning as operational systems and they include labor and project management. Nodes for both moorings and cabled observato- ries are assumed to be serviced annually. Ship costs for a Class I UNOLS vessel are assumed to be $20,000/day with an additional $10,000/day for ROV costs. Although vessel costs for servicing coastal observatories can vary widely depending on the type of vessel; $10,000/day has been as- sumed for this cost projection. Commercial charter rates for ships and ROVs are market-driven and vary significantly from year to year, how- ever, these figures should be representative of longer-term average costs. Contingency costs are included for unscheduled maintenance and other unforeseen costs (estimated here at ~10% of annual O&M plus ship costs). Based on 2003 fiscal calculations, the figures shown in Table 4-1 indi- cate that O&M costs, not including ship time, for the OOI could run about $25M annually. If ship time is included, these costs approximately double, approaching $50M annually. By comparison, the FY 2002 budget of the ODP was approximately $46M for operating the JOIDES Resolution, a drilling vessel and associated program activities (drilling and science sup- port services, information services, publications, administration). The OOI O&M costs are thus not out of line with other major geoscience initiatives. Nonetheless, the oceanographic community has expressed concerns that the costs associated with operating and maintaining ocean observa- tories will drain resources from other areas of the ocean sciences (e.g., funding, ship and ROV assets, intellectual resources) and negatively im- pact non-observatory ocean science. This fear is based to a significant degree on concerns that funding levels will not be adequate to support this new facility or that these costs will initially be significantly underes- timated. In order to allay these concerns, the NSF needs to take steps to ensure that observatory program costs and infrastructure needs for ships and ROVs are accurately estimated early on, that budgets are augmented by amounts sufficient to operate whatever is acquired, and that manage- ment oversight and fiscal controls be implemented to ensure that the observatory program operates within budget.

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 107 The funding profile for the OOI shown in Figure 1-5 indicates the NSF is projecting O&M costs of about $10M/yr in 2011 when the infra- structure is fully installed. The figures shown in Table 4-1 suggest a more realistic estimate is $25M/yr, not including ship time, and twice that figure if ship time costs are included (it is not clear if the O&M costs shown in Figure 1-5 include ship costs or not). None of these estimates include the cost of funding the scientific research that this new infrastruc- ture will enable. These costs are difficult to estimate, but certainly could amount to a significant fraction of the annual O&M costs. A successful observatory program will require sufficient funding for both maintenance and operation of the observatory infrastructure and the science and in- strumentation that this infrastructure will enable. The NSF needs to take appropriate steps now to ensure that sufficient resources are in place to meet these needs by the time the observatory infrastructure is in place. NATIONAL SECURITY ISSUES Ocean observatories have the potential to provide public access to technology or data that could raise significant national security concerns. The most obvious issues relate to the U.S. Navy submarine fleet and hydrophore and geophone arrays that may be used with seafloor obser- vatories, although other sensors could raise security concerns as well. The issue of submarine security and ocean observatories has occurred several times in the past; however, the new capabilities of the proposed observa- tories combined with near-continuous operations raise issues well be- yond those of the past. In the past, oceanographers have had access to small arrays of hydrophores, geophones and seismometers (e.g., the Wake Island and Ascension Island systems), but these arrays had a very limited capability to detect and track a low signal-to-noise source such as a sub- marine unless it was very close. The towed arrays currently used for seismic reflection and refraction profiling have a greater capability. The arrays are several kilometers in length, have a large number of sensors, and operate in frequency bands (5~00 Hz) relevant to anti-submarine warfare (ASW), but since these arrays operate in an active mode with large air gun and/or explosive acoustic sources, only record in limited time segments, and are not stationary, they do not raise concerns for submarine security. Consequently, oceanographic research has not been a problem for submarines; indeed they have benefited immensely from the basic science accomplished with these systems. The scientific community has had access to the Navy SOSUS arrays for seismicity, marine mammal, acoustic tomography, and a few other applications. This access, however, was allowed only to investigators with security clearances; time-series data could not be published; large uncer-

108 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY tainties existed regarding receiver location; fated publication was required. All of these restrictions are not generally acceptable to either the scientific community or its many referred jour- nals, so use of these data has not been widespread. Shortly after the end of the Cold War and the demise of Russian submarine operations, there was a period when the Navy desired scien- tific usage of the SOSUS arrays. The reasons for this usage are uncertain, though it could possibly have been used to justify 'dual usage' or to transfer the cost burden for operating and maintaining these systems to another entity. Since then no efforts have been made to encourage 'dual usage' except within these same security constraints. Nevertheless, a panel of scientists convened to document the possibilities of using SOSUS for scientific purposes (Ioint Oceanographic Institutions, Inc., 1994~. When the approval process reached senior levels of the Navy, the policy of restricting use to cleared individuals operating under the above protocols remained in place. All submarines radiate acoustic signals (a signature) from a variety of sources, including machinery, structural resonances, propellers, and tur- bulence. Measured submarine signatures are highly classified since they can be readily exploited for submarine detection, classification, and track- ing. Arrays of hydrophores and/or geophones with sufficient array gain and state-of-the-art signal processing are used to detect, track, and clas- sify a surface ship. The capability for such measurements involves many factors; however, several are relevant to deep-sea observatories and are discussed below. and prior review of any re- Location The proximity of an observatory to a submarine operating area is important since signatures are louder and at shorter ranges. The pro- posed NEPTUNE site is in a sensitive location since the proximity of a Trident SSBN station (Submarine Ship Ballistic Nuclear) located in Wash- ington State raises concerns as submarines leave and enter port. A highly capable hydrophore or geophone array at a node with large horizontal line array (HLA) and/or large vertical line array (VLA) could acquire the acoustic signature of an SSBN. There are also nonacoustic "signatures" used in ASW; however, classification of such information restricts a com- plete listing. Such signatures are highly classified because they can be exploited for detection and compromise the stealth of these submarines. There are also other modalities by which such arrays could be used for detection. In addition, some of the proposed OOI global network sites raise concerns because they may be near regions of SSBN operations.

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 109 Array Capability Hydrophone or geophone arrays with many sensors and wide aper- tures lead to large array gains and high resolutions, and both of these capabilities are critical in detecting and tracking a quiet submarine. NEP- TUNE nodes with a capability similar to the existing SOSUS arrays cer- tainly could prove a concern. Furthermore, the distributed nature of the entire NEPTUNE system will provide a significant enhancement capabil- ity in terms of improved tracking by triangulation and increased signal to noise levels. NEPTUNE's array configuration (the number and location of sensors) at each node, as well as the rest of the sensor suite, is not speci- fied at present, so a determination of the threat to submarine operations cannot be made. A single sensor at each node may not pose much of a threat, but a SOSUS level array at each node certainly would attract the Navy's attention. Continuous Observations Continuous data acquisition by an observatory raises operational con- straints for submarines by potentially eliminating any opportunity to es- cape observation. Potential solutions might be to turn off, delay, or de- grade observations by pre-arrangement with the U.S. Navy; however, even knowledge of such scheduling is valuable information regarding the transit or location of submarines. Another alternative employs extensive noise masking, which entails its own set of operational problems and may degrade scientific observations. There are other security issues related to system configuration and control. The junction boxes used with ocean observatories would be de- signed to accept a wide variety of instrumentation operating with the system protocols for power and communication to shore. The openness and flexibility of these junction boxes are important for the scientific com- munity as new and improved instruments become available during the proposed lifetime of the observatories. However, this openness raises issues over network configuration control. Procedures need to be put in place so access to the junction boxes is controlled and the Navy has full advance disclosure about the capabilities of the instruments connected to the boxes. Moreover, the Navy will not be comfortable if data streams are encrypted or other actions are taken to limit its availability, and will want real-time access to the data. The Navy may want to perform some data processing for ASW purposes solely to assure that sensitive information is not compromised. Implementing such requirements represents signifi- cant costs that are yet to be determined.

1 10 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY Since the Trident SSBN fleet is a primary component for U.S. national security, the Navy will not accept siting sensors (e.g., NEPTUNE) where the data derived compromise the fleet's stealth and operational capabil- ity. Resolving this possible conflict between the desire of the research community to use the most capable instruments available in the best possible locations and national security concerns could lead to an impasse among the Navy, the new Department of Homeland Security, and the NSF. The NOPP National Ocean Research Leadership Council (NORLC) already has a security subcommittee, an inter-governmental infrastruc- ture which may be capable of working out these security problems. There are two particular concerns: (1) these difficult problems will certainly require resolution especially if the OOI expects to maintain the schedule described elsewhere in this report; and (2) submarine security concerns the operational Navy whereas NOPP has more of a research focus, mak- ing operational Navy participation in these discussions imperative. Discussions will need to take place to establish observatory security policies well in advance of any observatory installations, and will need to involve senior officials at the NSF, the NOPP-NORLC, the U.S. Navy, and the U.S. Department of Homeland Security. As observatories become op- erational, an on-going process will be needed in order to address security concerns as they arise and to regularly review security policies and proce- dures. It is imperative for the NSF and other supporting agencies to initiate this dialogue with the Office of the Secretary of the Navy (SECNAV) as soon as possible. The SECNAV is the appropriate office because it can represent the concerns of the entire Navy in these discussions. Any obser- vatory systems compromising submarine security or other Navy capabili- ties will represent a threat to national security and could be challenged at the highest levels by the Navy. With the present schedule, installation of observatory systems is not terribly far off, making it critical that the poli- cies impacting national security are put in place soon, lest their absence potentially delays deployments. DATA MANAGEMENT There have been a number of planning activities over the past few years to develop requirements and policies for an ocean observatory data management system. Although there are differences in the requirements and policies recommended by participants in these various workshops and reports, there are a number of overarching themes that address key issues for an ocean observatory data management system. The require-

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 1 1 1 meets and policies adopted in this report are based on the following four documents: · Illuminating the Hidden Planet (National Research Council, 2000~; · Ocean.US IOOS-Data and Communications (DAC) sub-system (Appendix V.8 in Ocean.US, 2002a) ~ Please note that the IOOS Data Man- agement Plan was released for formal review on April 16, 2003 after this report had been sent to review, therefore conclusions drawn in that report are not available here.~; · The Argo Data Management Handbook (Argo Data Management Committee, 2002~; and · NEPTUNE planning documents (NEPTUNE Canada, 2000; NEP- TUNE Phase I Partners, 2000; NEPTUNE Data Communications Team, 2002~. Recommendations for an OOI data management implementation plan have been developed based on those requirements. Further, the OOI ini- tiative will take advantage of the work that these listed activities have accomplished. The plan provided in this report provides guidance for developing an integrated, cost-effective data management strategy for global, regional, and coastal ocean observatories (see Appendix B for defi- nitions of any unfamiliar terms). There are several challenges that an ocean observatory data manage- ment system must address: · Heterogeneity of data sets: Data products are generated from vari- ous instruments and have different characteristics in format, metadata, resolution, validation of data, and other similar characteristics. · Absence of existing infrastructure: Data archive centers exist for some data types that will be collected by ocean observatories but do not exist for others. Moreover, the responsibility for supporting oceano- graphic data management centers is divided among several different agencies. · Integration of data productsfrom observatories: In order to access data products from multiple sources, data need to be quality-controlled in a uniform manner and coordinated by data processing centers for all observatories. At present, there is a lack of coordination between data providers and data users. · Data volumes: The observatories being planned as part of the OOI will potentially generate huge volumes of data (on the order of 1 petabyte/year). The system needs to be scalable to accommodate data volume growth.

1 12 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY Due to the variety of disciplines involved in ocean observatory re- search and the different requirements of each observatory type, a distrib- uted data management architecture should be adopted in order to be cost effective and manageable. The observatory systems proposed for the OOI share common needs to address metadata and data standards, data pro- cessing strategies, and information sharing. In order to develop a highly automated and cost-effective data handling system, however, the com- mand and control and data management models will need to be im- plemented differently for each observatory system due to differences in instrument configurations, data volumes, and real-time data delivery re- quirements. Table 4-2 summarizes the data management system require- ments for the OOI and their impact on the software and hardware data management system design. Ocean Observatories Initiative Data Management Architecture The software system architecture defines the performance and rela- tionship of various data management services including data processing, data archiving, data mining, operational and science user interfaces, and data distribution services. Each service will be described by a set of com- ponents allowing for the construction of the software system. Addition- ally, the system should be designed to leverage a distributed systems architecture to allow for scalability across multiple servers, both locally and geographically. These design guidelines have been proven to support large data management systems in both commercial and research envi- ronments. The data management sub-system (DMS) interfaces with sensor net- work operations, science users, archive centers, program management, algorithm developers and calibration engineers, and ancillary data sources. The recommended OOI-DMS design priorities are: · high speed of data product processing and delivery to science us- ers; · advanced level of full process automation to enhance speed and to minimize labor cost; · high adaptability of system to shift between processing priorities in the case of an unscheduled event; · scalability of the system to consistently support rapidly growing data collection and increased processing demands; and · ability to extend the system to incorporate new algorithms, proces- sors, delivery medium, processing centers, and command stations, as well as new observatory nodes.

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 1 13 Architecture The purpose of the software architecture is to define a reference DMS that includes components that can be integrated to serve a variety of ocean observatory requirements. It should allow for any number of data types to be integrated into the system. A key goal of the architecture should be to achieve the scalability and performance requirements neces- sary to support the observatory system. The architecture should include components that facilitate the capture, location, interpretation, and distri- bution of science products. The final data management framework archi- tecture should have the following goals: · Scalability and reusability, to accommodate different observatory requirements and data sizes; · Hardware independence, to provide hardware configuration that is not driven by the framework; · Database independence, to provide a database that is not driven by the framework; · System adaptability, to allow for observatory or project specific features that can be plugged into the framework; · Location independence, to allow data sharing from multiple dis- tributed repositories for analysis, decision-support, and knowledge dis- covery; · Ease of use, to provide data management framework software that is portable and easy to install and manage; · Autonomy, to enable a rule-based management to support autono- mous operations; and · Interface efficiency, to provide an Application Programming Inter- face (API) to interface with user analysis tools. A key principle in implementing these goals is to separate the data architecture from the technology architecture. The data architecture speci- fies standard models for describing and exchanging data that can evolve over time. For example, XML is a part of data architecture that specifies data interchange format. On the other hand, JAXR is a Java™ technology architecture that provides API for handling XML registries in lava™. If a new technology emerges, it can still take an advantage of XML data archi- tecture. However, if application specific data architecture is used (e.g., Microsoft Word format), it is hard to take advantage of new technology without affecting data architecture. The technology architecture specifies basic communication middleware between geographically distributed data systems, a common software component framework, and methods

1 14 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY TABLE 4-2 Requirements and Design Impacts of Data Management Systems Requirements Design Impact Data from ocean observatories must be readily available to the scientific community and the general public. The OOI-DMS should be integrated with the data management strategies proposed for IOOS/GOOS. The OOI-DMS should provide the up- to-date status of observing elements and maintenance schedules. The OOI-DMS should provide data in real-time or near-real-time to end users. Data summaries and metadata should be available in near real-time. The OOI-DMS should provide reliable continuous deliveries of real-time data streams from observing system Requires professionally managed DMS and data repositories. Develop a middleware framework for interoperability among heterogeneous, cooperating systems. Develop API to interface with various data access methods. Develop a "free market" of globally accessible ocean science information, including officially sanctioned OOI data and products, as well as products and analyses from other sources. Requires fault tolerance with real-time data delivery. Support special orders of various level products. Interface to health of DMS, system monitoring, data acquisition requests, and planning. Develop NSF and community driven data and metadata archive standards for the three observatory types. Develop capabilities for automatic generation of descriptive metadata associated with data. Need data discovery services that allow rapid location and access of data based upon queries formulated by machines or humans. Provide unambiguous citations and linkages, assuring users of the version of data provided. Develop highly automated and cost- effective systems, with built-in feedback mechanisms for rapid detection and downlink sites to operational modeling suggestions for repairing problems. centers and the scientific community. Recognize needs of all users and coordinate and/or implement required actions. Use Spiral Model (IEEE Computer 21, May 1988) for development of systems and technology insertion for stability of the system. continues

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 1 15 TABLE 4-2 Continued Requirements Design Impact Data products must be preprocessed and permanently archived with standard data formats. Data or data products from different observatory systems must be readably available to users. Data from core and community instruments should be available without restriction to any interested user or the general public. Data products from core instruments should be supported by basic observatory operating costs. Data products from community instruments and "investigator-owned" experiments should be supported by responsible funding agency. Develop comprehensive, well-documented and supported standards and protocols to guarantee interoperable delivery of all observations and numerical products. Develop a hardware resource management plan must also be developed. Possess sufficient data transport capability to handle large-volume exchanges of raw data and model outputs between modeling centers and high volume users. Develop robust network capability. Standardize interfaces. Develop data management plan for data from all three types of instruments: core instruments, community instruments, and "investigator-owned" experiments. The OOI-DMS should assist researchers Develop a scalable data management in the integration and management of framework. Develop a componentized their datasets by providing support for architecture for Plug-n-Play capability of the transition of experimental data processing and assimilation information into synthesized data procedures. products. The OOI-DMS should facilitate educational outreach using seafloor observatory infrastructure and information. Existing data management systems COTS software should be used if possible. Enable diverse communities to easily utilize the varied and distributed forms of marine data in a wide range of current and future computer applications. Provide opportunities to engage the private sector as a powerful development engine in creation of value-added products targeting the needs of specific user groups. SOURCES: Data from National Research Council, 2000; NEPTUNE Phase I Partners, 2000 Appendix V.8, Ocean.US, 2002a.

1 16 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY for using the data architecture. Allowing the data and technology archi- tecture to evolve independently will extend the life of the OOI-DMS. Components Although there will be differences in managing data from the three types of observatory systems, some common features apply, such as: de- velopment strategy, metadata and data standards; and software architec- ture to allow data management and data archival. These components are described by addressing the importance of software development strat- egy, standardized metadata and data formats, archive policy, and educa- tional outreach. Figure 4-6 illustrates the components that would be re- quired to support DMS in a typical observatory network (Hughes et al., 2001~. Several recommendations in this report are adopted from the IOOS DAC components definition (Appendix V.8 in Ocean.US, 2002a). Further- more, any existing components that have been developed by the research community which may include metadata format, data format, interchange protocols, and API, should be adopted for this effort if feasible. Metadata Management A distributed DMS similar to that required for ocean observatories should develop metadata standards to specify contents and formats of the metadata. A Data Management Committee should exist at the OOI pro- gram level (Figure 4-1) to establish simple guidelines and extensible stan- dards for metadata and to develop a data and metadata search and re- trieval framework to enable searches across multiple data repositories established by the observatory program. This committee should also de- velop well-documented and reliable standards and protocols to guaran- tee interoperability among all data centers. Development of standards and protocols should be coordinated with other national and interna- tional programs. Based on these guidelines, the primary activities to be conducted by the data centers are: · implementing and maintaining a comprehensive OOI data discov- ery service; · ensuring that linkages between data and metadata are maintained with reliability; · developing data mining techniques and an analysis framework to recognize features and find clusters in large datasets (i.e., capability for event notification and automatic science discovery);

117 o tL — ~7 c~ 07 c5 ~ ~7 c5 {~7 O ~ Q ~7 tL ~ CD C 7 ~ — CD ~ ~ ~ CD tL `~ 07 cay C5 c5 c.7 -1 ~ ~0 I~\ #~. ~~ ~ ! ~0 ~o0~l ~~(,) ~1 0~! ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ (~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ TIC ~ ~ ~ ~ ~ 1 ~~ ~~ l |~ ~~ l ~~)~ ~ 1~ ~0~ l ~~.~ l 1~ ~~;,,~ ~ ~~o~ 1 l ~0~ r ~~ l 1 ~~ ~~ ~ I 1 . ~~ ~~ . ~~ ~~ ~ ~ ~ ~~ l ~~ ~~ 1 ~~t~ 1 ~~'>~ 1 ~~ . ~~ ~~ r ~ 1 - 07 07 c5 ~ c5 .' ~7 ~ ~7 ° Cal tL 07 07 cay ~ ._ tLO CD r Hi ~ ~ ~ 5 1 ~ 1~1 ~ ~ it_ I ' 7 1~1 ~ .-~ ~ ~ 7 . ' 5 1~1 0 -0 cay o ~ c5 l ~ ; 1~L ~ i= ~ QO 07 l - ,~ ,:~ ~~_~" B |~W~fflJ~I 1~;5~::;'~ ~1 H~°~1 B~*~' ~1 1 l B~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~ 1 1~.C~ l B~c~ l B~:~:~:~:~:~:~-—~:~:~:~:~:~:~:~:~ 1 l~:~ 1 l~l Observatories D 1 ~ c5 ~ ,~ o ~ CD ~ Q 1 ~ CD C~ c5 v ° ~ c5 Q— Q~ ~ c c5 c~ c5 c5 -_ ~ ~ C, ~ ~ ~ 1 ~ ~ 1 CD ~ c~ ~ c5 `~ ~ c5 Q~ c5 ~ + c5 ._ , . ~ c5 o CD ~ Q ~ ~ CD cn r 1 5 ~ c5 ~ c~ ·— c5 ~ ~ CD 1~1 ~ ''''''''''''''''! r r o o au V) o au o o o au 5- o V) au o o au p V) ·_I

1 18 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY · developing data translation middleware to geographically project, register, and subsample data products (i.e., co-registration). To sustain the large amount of data to be handled by the OOI, a special focus should be given to providing education, training, and tools to increase OOI effectiveness in metadata generation and management. Data Archive The ocean sciences community has not yet widely accepted that data should be made freely available to the rest of the community in a timely manner. For example, the history of WOCE demonstrates how difficult it has been to make WOCE data acquired over the past ten years available to the public, and the struggles the data processing centers have undergone attempting to make these data available in a uniform fashion (Appendix V.8, Ocean.US, 2002a). The research community also lacks a centralized or coordinated system for archiving and distributing oceanographic data. While some data types are managed through NOAA's National Oceano- graphic Data Center (NODC), not all investigators submit their data to NODC, nor does the NODC archive all of the many kinds of data col- lected by ocean researchers today. In many cases, data archiving and distribution has thus become the responsibility of individual investiga- tors, institutions, or programs. As a result, valuable data are generally not widely available and may be lost over time. Experience shows that the ocean observatory program cannot rely on individual investigators to manage, archive, or disseminate observatory data. Data must be professionally managed and distributed through es- tablished data centers according to nationally and internationally agreed upon standards. Due to the interdisciplinary nature of the data to be collected at ocean observatories, they will not be archived in a single central archiving center for these data, but rather in a network of distrib- uted centers dedicated to particular data types (e.g., seismological, oceano- graphic, biological, or geodetic). The program will need to provide tools for scientists to search and retrieve data across this distributed network of data centers (a "virtual" data center or "data grid" concept). The OOI program should have an open data policy with all data from core instrumentation and community experiments available in as near real-time as possible. Data archive centers will therefore require sustained funding to support data archival and distribution even beyond the end of the program. It is also important that hardware upgrades, maintenance, and media migration are conducted and funded as a part of the operation of data archive centers.

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 1 19 User Outreach, Applications, and Products A main concern of the OOI-DMS is providing data products in an efficient, reliable, and timely fashion. To achieve this ambitious goal, the OOI needs to establish and maintain contact with outreach program de- velopers. The OOI must continually determine the adequacy of the qual- ity and timing of release of OOI products in relation to user needs. In particular, these products should include the following: · quality-controlled collections of observations maintained by OOI- DMS and archive centers according to commonly agreed upon quality control procedures; · a minimal, guaranteed geo- and time-referenced data visualization capability accessible through a standard Web-browser; and data. · information on products produced by outside groups utilizing OOI To accomplish this goal, two key user interfaces will need to be cre- ated: a science user interface and an operational user interface. The sci- ence user interface allows scientists to enter product requests. Requests for previously captured data will be processed and products staged for download by the data distribution function; requests for products that have not been acquired will be recorded by the system, which will sched- ule the user for notification and distribution once acquired. An opera- tional user interface will allow system operators to manage the data sys- tem. As the size and complexity (in temporal and spatial resolution) of data sets grow, so will the inadequacy of an information discovery pro- cess that can descend only to the level of a data set. Data mining tech- niques should be included to allow the user the ability to identify and retrieve "features" such as "a decrease in transport of the Gulf Stream" or "abnormal seismic activity along the fuan de Fuca Ridge." Data System Security Data should be protected based on policies adopted by the global, regional, and coastal observatory community. Some data products might be restricted to the U.S. due to International Traffic in Arms Regulations (ITAR). To protect network infrastructure, communication between sen- sor network and commanding centers should include the capability to authenticate users and encrypt data. The following methods are available for adhering to various security measures:

120 cation; ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY public key infrastructure-based data encryption and user authenti- · a secure network and firewall; · authentication based on user profiles and data profiles; · a log and monitor function; · security plans for data management systems; and · risk assessment and countermeasures. A security plan should cover data management system security re- quirements, hardware and software architecture, network interconnec- tions and internal architecture, and system components. It should also cover non-technical security features such as personnel, user training, physical security during the development and operational phases, and daily cycle of operation. Finally, risk assessment and countermeasures should be discussed to prevent operational failure, to increase data sensi- tivity, and to protect command and control components. Such an assess- ment will be extremely important for regional and coastal cabled observa- tories due to concerns over homeland security. Administration and Operation From an administrative and operational standpoint, the OOI-DMS should: · guarantee day-to-day system operations (e.g., for data acquisition, data and metadata portals, monitoring, and evaluation of system perfor- mance); · guarantee liaison with product and archival facilities; · ensure "help desk" functions and software development; and · establish and publicize policies. Implementing Observatory-Specific Components The three OOI components (global, regional, and coastal) will each have different data management needs. A regional cabled network will produce huge volumes of data from a diverse set of sensors on a single telemetry link, while the global network will generate smaller data vol- umes, but from a globally distributed network of nodes, each of which uses a different telemetry link to shore. The coastal network is likely to be a mix of these two scenarios. Data management capabilities and opera- tional scenarios must therefore differ in order to efficiently manage the flow of data generated by the networks.

/MPLEMENTAT/ON OFA NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 121 Data Transport Data management capabilities already exist for some components of the global and coastal observatory networks, (e.g., NODC, IRIS, Argo). Since some of these capabilities are not completely coordinated, however, issues may arise with regard to interchanging products and distributing data in a timely manner. A common interchange format and a data dictio- nary should be developed through working groups and joint research opportunities for the national and international science community. As specified in the IOOS recommendations, depending on the ocean observatory network the following topics need to be addressed and imple- mented by surveying existing implementations and adapting them to fit the ocean observatory network: · robust networking technology to transmit data between sensor sub- systems, assembly centers, modeling centers, product generating centers, archival centers, and data users in both real-time and delayed modes; · extensible data Codeless that assure interoperability of diverse classes of transmitted data, which should be created if an appropriate model does not exist; · software strategies that translate data, as accessed from diverse data management systems, into an interoperable data model, which should be created if none exist; and · software strategies that assure security, performance monitoring, and fault detection of the OOI data management network, which should be created if none exist. Global Observatory Although data processing will be implemented at a national level, the network to support a global observatory will provide data for several international research programs ranging from climate studies to global seismology. Data requirements may differ significantly from site to site both in the volume of data acquired and in the type of link to shore, depending on the type of mooring. Real-time or near-real-time data deliv- ery will be required for many instruments on this network, including data from meteorological, oceanographic, and seismological sensors. The diffi- culty of implementing a data management system for the global observa- tory network therefore lies in the distributed network of nodes shared internationally rather than the amount of data generated. As a result, it is important to agree on quality control procedures, standards for data dis- tribution, data-level processing, and version controlling among the net- work. As experience from the Argo program has indicated, techniques for

122 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY uniquely tagging and tracking the versions of data sets must be devel- oped. The global network data management system should include a dis- tributed processing framework among international data centers, a unique Web portal to access these data annotated with quality statements among the network, and standardized quality control procedures to be devel- oped for real-time and delayed mode data. Regiona/ Observatory A regional observatory based on a cabled seafloor network has simi- lar data processing requirements to a satellite system in that huge amounts of data (Tb/s or greater) will be acquired in real-time. The issue of imple- menting either software or hardware networking capabilities needs to be addressed in the future, and creative solutions to these problems will need to be devised. Solutions may include techniques to reduce the de- mand for large data transfers (e.g., more effective sub-setting, server-side analyses) as well as improved delivery (e.g., higher bandwidth, new com- pression schemes). Resource management of power consumption, instruments, net- works, and AUVs should be handled by autonomous capabilities by moni- toring product requests and data access and production, and through process monitoring and control. Although a distributed framework will be used for data management, centralized coordination of observation planning, data process requests, and access is required. Coastal Observatories Since coastal observatories may include both relocatable moorings and fixed cabled or moored observatory systems, they share requirements with both global and regional observatories. A scalable, componentized capability in the data management architecture should be developed to accommodate future observatory needs. Use of middleware is critical for merging existing capabilities with new capabilities to be inserted into the architecture. These functions include metadata profiling, data translation, APIs, and service-oriented components such as Common Object Request Broker Architecture or lava™ message service. Command and Control While it is not clear at this point whether instrument command and control at ocean observatories will be a data management function or an operations function, the DMS should provide interfaces to a command

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 123 and control system to provide information on scheduling, instrument control and monitoring, telemetry processing, network security, and maintenance. Constant feedback regarding data acquired versus observa- tions planned should be available for planning and scheduling purposes. Another important aspect of this component is predicting system over- loads and providing information for instrument and system monitoring in order to guarantee continuous operations. This component should also include autonomous control capability for network resources and data flow. EDUCATION AND PUBLIC OUTREACH Research directorates of funding agencies like NASA and the NSF are increasingly encouraging the integration of science and education and greater scientist involvement in education and public outreach (EPO). The NSF's Geosciences Directorate has developed new programs for geo- science education and outreach to teachers, students and the public, and the NASA strategic plan makes it the responsibility of its strategic enter- prises to "embed" education and outreach into its programs. In testimony before the Committee on Science, U.S. House of Repre- sentatives, 28 April 1999 Daniel S. Goldin, NASA Administrator, stated that: No longer is it an acceptable practice to say, "we are too busy." Re- search, knowledge generation and education are all equal components of the NASA mission. We must combine our traditional methods of in- volving the education community with new and innovative ways so that the impact NASA has on education is greater. (U.S. House, Com- mittee on Science, 1999, p. 123) On the same date Dr. Rita Colwell, Director of the NSF, testified be- fore the same committee that: All researchers whether at a university, a national lab or circling the Earth in a space station should link their inquiries with the education of the next generation. (U.S. Congress, 1999, p. 77) Seafloor observatories have the potential to provide unique opportu- nities for educational outreach by conveying the excitement of discovery of ocean sciences to the public (National Research Council, 2000~. Many aspects of observatory research are suitable for education and outreach programs, especially the capability of observatories to transmit real-time video and data from acoustic and optical sensors. Examples of possible outreach efforts presented in the Illuminating the Hidden Planet (National Research Council, 2000) report include:

124 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY · real-time image and data feeds into museum and aquarium exhib- its via the Internet; · incorporation of real-time image and data feeds in K-12 curricula (e.g., The JASON Project™ [see Appendix Bit; · development of curriculum modules for K-12 students that incor- porate research results and scientists' profiles; · establishment of summer research (sabbatical) programs for K-12 teachers using observatory data to facilitate interaction with scientists; and · development of public Web sites containing real-time images and data, and publicizing exciting science at observatories via public televi- sion. Large sectors of the American public are functionally illiterate in the sciences most cannot explain the mechanism responsible for the four seasons of the year (Goodstein, 2002~. The U.S. scientific research commu- nity has an obligation to aid American educators in improving the science literacy of its citizenry. The National Science Education Standards (NSES) provide a consensus on what educators and scientists nationwide believe students should know and subsequently be able to understand and apply at various K-12 grade levels (National Research Council, 1996~. The NSES also address the need for best teaching practices, preparation and profes- sional development programs for teachers, and implementing systemic reform of education. EPO should be an important objective of the observatory effort in order to involve K-12 students and the interested public in the excitement of ocean sciences research. The EPO program should use ocean research data to help students and teachers meet NSES, not just to entertain (Na- tional Research Council, 1996~. The ocean observatory EPO program should be implemented through a collaborative effort with the National Sea Grant Program and with the recently funded Centers for Ocean Sciences Education Excellence (COSEE). Sea Grant has long been involved with K-12 education and public outreach, coupling research marine scientists in universities with educators to promote the benefits of (primarily) Sea Grant-funded marine research. In general, Sea Grant EPO managers are science educators with expertise in marine and aquatic science content, educational pedagogy, and pre-college curriculum development. These highly qualified marine and aquatic educators work within an established infrastructure and have made valuable contacts in many pre-college schools, museums, science centers, and aquariums across the country. The eight COSEE are nationally coordinated programs for ocean sci- ences education in both formal and informal sectors. These regional cen-

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 125 ters and other similar facilities and programs will help facilitate the inte- gration of research into high-quality educational materials, promote a more ocean-literate society, provide opportunities for teacher preparation for pre-service teachers and professional development programs for in- service teachers, undergraduate faculty, and administrators, as well as other audiences. The most cost-efficient means of establishing a viable EPO program within the seafloor observatory network program is to avoid redundancy and take advantage of the existing Sea Grant and COSEE programs. Us- ing funds provided by the NSF, designated Sea Grant and COSEE offices could expand their EPO efforts to encompass the broader aspects of ocean science research associated with the national seafloor observatory net- work. This leverage of expertise would mean involved Sea Grant pro- grams would have to accommodate a more global perspective in order to ensure that all aspects of the global, regional, and coastal observatory networks are utilized in the EPO activities. It is recommended that the OOI program office contain an EPO coor- dinator to oversee and coordinate EPO activities going on at various lev- els within the OOI. One role of the EPO coordinator should be to assist observatory program principal investigators as they establish EPO col- laboratives and partnerships with appropriate Sea Grant, COSEE, and related EPO program managers. Principal investigators should be able to select the Sea Grant or COSEE program for their EPO activities based on individual inquiries, programmatic needs, or assessments that are not necessarily based exclusively on office location. Other factors that must be considered are the qualifications of individual prospective Sea Grant, COSEE, or other appropriate EPO managers who will be involved in the observatory program through possible matching funds or resources pro- vided by these agencies, professional organizations, academia, the pri- vate sector, or industry. The EPO coordinator should establish specific goals with the desig- nated EPO officer (whether Sea Grant or COSEE) associated with each observatory program. Goals should include the establishment of specific partnerships between the observatory research program and K-14 schools, museums, aquariums, and other public outreach institutions. There should be a clear expression of the manner in which the EPO programs address advancement of scientific literacy through the integration of iden- tified ocean sciences concepts within the NSES. One of the most promising potential activities mentioned in Illuminat- ing the Hidden Planet (2000) is the establishment of paid land or sea sum- mer internships for K-14 teachers. These internships will allow classroom teachers to gain hands-on, inquiry-based experiences in many aspects of ocean sciences research. These internships and pre- and in-service accred-

126 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY ited teacher workshops (two to three days) or institutes (three to four weeks) are perhaps the most effective ways to excite and engage teachers about science; to enhance their content knowledge, esteem and effective- ness in their schools; and to encourage them to develop new innovative teaching strategies using the opportunities provided by the seafloor ob- servatory network. An excellent example of this enhanced effectiveness is the NSF-spon- sored Research and Education: Volcanoes, Exploration and Life (REVEL) Project operated by the Department of Oceanography at the University of Washington. REVEL is described as a Program that allows: -- r--~- the interaction of highly-motivated science teachers hungry for opportu- nities to engage in science and innovative scientists pursuing cutting- edge research. The scope of this research encompasses a wide variety of scientific problems that range from the origin of life to new aspects of biotechnology. (REVEL Project, 2003, p. 1) Approximately 50 science teachers have been involved in the REVEL Project since its inception in 1996. Many of these teachers have partici- pated in research cruises aboard the R/V Thomas G. Thompson, and have become highly motivated in science education as a result of their oceano- graphic experiences. There are also other model EPO programs that can be emulated by the OOI. The NSF or, once it is established, the OOI Program Office, should solicit proposals for a workshop to address the EPO issues raised in this report and to develop a specific EPO implementation plan for ocean re- search observatories, including recommending a budget for EPO activity. This workshop should consist of 20-30 individuals, including educators with strong links to Sea Grant, COSEE, aquariums, science centers, and science museums. Some of the workshop participants should include re- search scientists who are leading the establishment of the national obser- vatory network in order to ensure an effective and mutually proactive dialog between the research and education communities is established from the outset. A similar Education Workshop, sponsored by the NSF, was conducted by the Integrated ODP in Narragansett, Rhode Island in May 2003, involving approximately 80 formal and informal teachers, edu- cators, and scientists. A report from this workshop should be available by summer 2003. PHASING OF OBSERVATORY CONSTRUCTION AND INSTALLATION The OOI is expected to provide approximately $200 million dollars over a five-year period, beginning in FY 2006, for the acquisition and installation of ocean observatory infrastructure (Figure 1-5) (National

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 127 Science Foundation, 2002~. This section discusses scientific planning, en- gineering development, and system testing needs over the next two to three years; describes the factors that might affect the phasing of imple- mentation of each of the three main OOI components; and recommends an implementation strategy for the five-year OOI-MREFC program. It was beyond the scope of this study to develop a detailed cost analy- sis for the construction and installation of each OOI component (coastal, regional, and global). Planning for each of these elements has developed largely independently of the others, and important decisions on the exact proposals have yet to be made. Although precise cost estimates are thus impossible to make at the present time, the development of a detailed and comprehensive project implementation plan and cost analysis for each of the three major components of the OOI, and the review of these plans by knowledgeable and independent experts is needed as soon as possible. These plans should be completed by the end of 2004 and reviewed in early 2005. If the total cost of the infrastructure envisioned for the OOI exceeds the resources available through the MREFC, the scope of each proposed component, as well as each component's relative priority will need reassessing. Pre-lnstallation Planning and Development Needs A program as large and complex as the OOI requires an extensive planning and development effort prior to installation of the MREFC- funded infrastructure. These activities have been under way for some time (Chapter 3) but much remains to be accomplished; planning and development work will need to be accelerated between now and the beginning of construction and installation. Significant levels of additional funding (several million dollars over the next two to three years) will be required to support these activities. An essential first step for pre-installation planning is the establish- ment of the OOI Program Office, described in the first section of this chapter, to oversee and coordinate these planning activities. It is recom- mended that this office be established by the end of 2003. The Program Office's first task should be the development of a detailed and compre- hensive project implementation plan, as outlined above. The Program Office also needs to oversee scientific and technical plan- ning to better define locations, scientific objectives, and core instrument and infrastructure requirements of specific observatory nodes and obser- vatory systems. Scientific planning will allow individuals, groups, and programs to compete through a peer-reviewed mechanism for research time, bandwidth, and power usage on observatory infrastructure. This

128 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY process should begin immediately and should involve the broadest pos- sible cross section of the ocean sciences community. The process may include planning workshops, solicitation, and review of proposals from individuals and community groups, and the establishment of science and technical advisory committees to the Program Office. A third major task of the OOI Program Office should be the develop- ment of both a comprehensive data management plan for the OOI and a strategy for an innovative and effective EPO program. In addition to scientific and program planning, development and test- ing of the more advanced observatory infrastructure components envi- sioned for the OOI is required prior to the beginning of the MREFC (these needs have been outlined in some detail in Chapter 3~. Funding has been secured for prototyping and testing of some critical sub-systems, and for the establishment of testbeds for cabled and moored buoy observatories (e.g., the Acoustically-linked Ocean Observing System [ALOOS], MARS, MOOS). However, additional funding will be required to complete this development and testing work prior to the construction and installation phase of the MREFC. Of particular importance is testing of the power and communications sub-systems for the multi-node, looped-network topol- ogy envisioned for the NEPTUNE-type, regional-scale cabled observa- tory, as well as the design, prototyping, and testing of critical sub-systems for the high-bandwidth, cable-linked moored buoys (i.e., EOM cable, C- Band antenna, and diesel power generation). The technical feasibility and cost-effectiveness of utilizing retired telecommunications cables for some global network observatory sites should also be thoroughly evaluated during this period. Ocean Observatories Initiative Implementation Strategy The proposed OOI MRFEC funding profile shown in Figure 1-5 will increase from $27 million dollars in FY 2006 to approximately $80 million dollars in FY 2008 and will decrease to approximately $43 or $44 million dollars in FY 2009 and FY 2010. While it is not clear how much flexibility there may be in changing these proposed year-to-year expenditures, the DEOS Steering Committee or, once it is established, the OOI Program Office should review this funding profile and determine if it is optimal for the OOI's specific requirements. In considering possible phasing of construction and installation of the ocean observatory infrastructure over this five-year period, a number of different criteria have been considered including scientific and technical readiness, risk, cost considerations, timely payoff, and leveraging oppor- tunities. Table 4-3 summarizes these criteria for each of the three OOI

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 129 components. The following sections discuss, these criteria and offer rec- ommended phasing strategies. As there is insufficient information avail- able to develop a five-year construction and installation budget, tasks have been assigned to one of three different phases (early, middle, late) for the five-year OOI MREFC program. Global Observatory Network Scientific planning for the global observatory network is mature and nas proceeded to the level of identifying specific sites and multidisci- plinary instrumentation requirements for each node (Chapter 3; Figure 3- 1~. Site selection is being coordinated at the international level and there are significant opportunities to leverage the investment the NSF makes with additional nodes funded by other nations (Appendix E). The low- bandwidth, oceanographic mooring design proposed for many low- and mid-latitude sites is already in use for oceanographic and meteorological applications and additional buoys can be built and deployed as soon as funds are available. Priority has been given to sites that fill gaps in the present time-series observatory system and that meet interdisciplinary research needs (DEOS Moored Buoy Observatory Working Group, 2003~. The opportunity for early scientific payoff for these sites is high, espe- cially for climate and oceanographic research applications. The high-latitude and high-bandwidth buoy systems proposed as part of the global program will, however, require additional prototyping and testing before large-scale construction and deployment of these systems can begin (Chapter 3~. The proposed high-latitude sites are characterized by severe weather, including high surface winds and seas, which will require new engineering approaches for buoy and mooring design to ensure survivability. The high-bandwidth buoy systems that have been proposed are also new, and will require validation and testing of critical sub-systems (e.g., EOM cable design and terminations, C-Band antenna performance, and reliability of unattended diesel generators), preferably by initial deployment of a prototype system at a low or mid-latitude location. These considerations suggest phasing for installation of global obser- vatories (Box 4-2), assuming that the OOI is divided into three phases of approximately one and a half to two years each. Regiona/-Sca/e Observatories Through the efforts of the U.S. and Canadian NEPTUNE groups, sci- entific and technical planning for a plate-scale cabled observatory in the

130 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY TABLE 4-3 Phasing Criteria for Seafloor Observatory Implementation Global Network Regional- Scientific Readiness Scientific objectives well-defined; scientific planning Scientific very mature with ~20 potential multidisciplinary scientific observatory sites identified. Good coordination at the system or international level. definition objective' requirem Technological Readiness Low-to-mid latitude, low-bandwidth acoustically- Major pro linked moorings feasible now; cable-linked and high- power ar latitude moorings need prototyping and testing of design d. critical sub-systems (EOM cable, C-Band antenna, Two testl power generation) before large scale deployment; for valid; re-use of telecom cables needs feasibility study. Need ful a multi-e Risk Low for low-to-mid latitude, low bandwidth systems; Moderate moderate for low-to-mid latitude high-bandwidth common systems; high for high-latitude systems; should because ~ consider cable re-use to minimize risk at some high node, me latitude sites. risk can l designs ~ installation Financial Considerations Unit cost is about $1 million dollars to several Desirable million dollars/node; total costs scalable by number installati' of moorings acquired. advantag market cat will inch Timely Payoff Depends on science, but opportunities for early Given sit payoff are high, particularly in remote regions. for route likely to . five-year Leveraging Opportunities High, with international collaboration with other High, wi nations (Japan, United Kingdom, Europe). and poss Northeast Pacific is well advanced (NEPTUNE Phase I Partners, 2000~. As described in Chapter 3, however, a large, plate-scale cabled observatory like NEPTUNE presents some major engineering challenges. The progress in developing and testing new technology to meet these engineering re- quirements and the long lead times required for many of the tasks in-

/MPLEMENTAT/ON OFA NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 131 tation Regional-Scale Coastal Planning nary on at the cally- nd high- ing of enna, Dent; dy. systems; width 1ld he high al number rly as. other Scientific objectives well-defined; scientific planning for NEPTUNE-like system mature, but need better definition of location, scientific objectives and infrastructure requirements of individual nodes. Major progress made in the design of power and telemetry systems but final design decisions have not been made. Two testbeds are under development for validation of major sub-systems. Need full system integration test using a multi-node, loop network topology. Moderate-to-high because power and communication systems are new and because of the complexity of multiple- node, multiple-loop network topology; risk can be minimized by validating designs using testbeds and a phased installation. Desirable to acquire cable and sign installation contracts early to take advantage of present depressed market conditions; phased installation will increase total costs. Given significant lead time required for route surveys and permitting, not likely to be operational until late in five-year period. High, with collaboration with Canada and possibly other nations. Scientific planning still in early stages; relative importance in OOI of mobile Pioneer Arrays, cabled observatories, and long time-series sites requires more community input. Relationship to IOOS coastal sites needs definition. Pioneer Arrays and Codar use standard "off-the-shelf" technology; use of simple cabled systems in coastal environment demonstrated; more complex cabled observatories with mesh topology or multiple nodes need development; issues with damage from fishing, corrosion, vandalism need to be addressed. Risk for coastal radar systems very low; risk for mooring arrays low-to-moderate; risk for cables low. Requires further definition of components of coastal OOI. Does not appear that phasing will have a major impact on costs. Early payoff possible with operation of first Pioneer Array or augmentation of existing coastal observatories. High, through leveraging of funding from state and other federal agencies, IOOS. valved in installing a NEPTUNE-like cabled observatory will place strong constraints on the funding timeline of the OOI MREFC. Power and communications systems for a system like NEPTUNE will be significantly different than systems used with conventional submarine telecommunications cables. The multi-node, multiple-loop network

132 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY topology of NEPTUNE is also unprecedented for a submarine cable sys- tem. These engineering and technical issues are being addressed on a number of fronts. Engineering design studies have been completed or are under way for both the power and telemetry sub-systems and two test- beds (VENUS and MARS) are under development for validation of these system designs. As presently funded, however, MARS will provide only a partial test of the key power and data telemetry sub-systems since its relatively short, single-cable, single-node design will not test the opera- tion of these sub-systems with the more complex multi-node, looped net- work topology of a NEPTUNE-like observatory. A full system integration test of all major sub-systems (power, telemetry, timing, and command and control) with a multi-node, looped network topology is recommended before the full deployment of such a network. This test could be accom- plished by augmenting the MARS testbed with an on-land, full-configu- ration test or with a phased deployment of the full network by initially

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 133 installing only one of the three planned loops. While a phased two-stage installation of a looped network will cost more than a single installation, the reduction in risk brought about by this approach may be worth any additional cost. Important logistical considerations will also affect the timing of in- stallation of a NEPTUNE-like system. Obtaining the necessary permits, especially near cable landfalls, can take up to two years. Cable routes need to be surveyed prior to installation in order to assess bottom charac- teristics and topography for hazards. Time is required to fabricate and test the cable. In addition, nodes must be designed, ordered, manufac- tured, and tested, both individually and in their final configuration. Given the present depressed state of the telecommunications industry, signifi- cant cost savings may be achieved by purchasing cable and contracting for installation sooner rather than later; the NEPTUNE system may not use standard submarine optical amplifiers or cable power systems, how- ever, so non-standard cables will likely be needed. The operation of the major sub-systems (power, telemetry, and timing), and the operation of the system as a whole, need to be simulated and validated through exten- sive computer modeling and physical testing throughout the construction and installation phase. These considerations suggest the phasing for installation of a NEP- TUNE-like, regional-scale cabled observatory over the five-year MREFC (Box 4-3~. Coastal Observatories As described in Chapter 3, scientific planning for coastal observato- ries in the context of the OOI began in 2002, but there is still no commu- nity consensus on the appropriate balance between spatial mapping and high-resolution time-series. Additionally, it is essential to establish a num- ber of long-term time-series sites in U.S. coastal waters, including the Great Lakes, using cables and buoys. There is, however, no agreement on whether this need can be met by the moorings that Ocean.US is planning to deploy as part of the coastal IOOS, or whether the OOI will require moorings specifically dedicated to coastal ocean research. The coastal community will need to develop a consensus on the appropriate balance of Pioneer Arrays, cabled observatories, and long-term measurement sites required to meet future coastal research needs. This consensus can be achieved by bringing together the diverse coastal community, including representatives from Ocean.US. Discussions should focus on implemen- tation with pragmatic consideration given to the appropriate mix of re- locatable and permanent observing systems. This planning effort should

134 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY identify the number and location of long-term time-series sites and the instrument requirements at these sites. While additional scientific planning is needed, the available technolo- gies for coastal observatories are relatively mature and installation of these systems is feasible within the five-year MREFC time frame. How- ever, some important technical challenges exist. Biofouling and corrosion remain significant problems for long-term observations in the coastal ocean and a major effort to mitigate their affects is required. Coastal moor- ings have not been outfitted with bistatic radar arrays and they rarely integrate the new generation big-optical sensors required for biogeo- chemically relevant measurements. For the coastal radar arrays, the de-

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 135 velopment of multi-static arrays will require further development of syn- chronous timing technology that allows different radars to use the same radio frequency. Coastal cables are largely limited by a lack of robust multi-sensor auto-profiling of the water column. Until a community consensus is reached on the infrastructure re- quired for coastal research observatories, any implementation plan will, of necessity, be rather notional. The following plan (Box 4-4) assumes construction of two Pioneer Arrays, the establishment of a coastal instru- ment testbed, a new coastal cabled observatory, and the augmentation of the IOOS national network of long-term coastal time-series moorings with additional instrumentation to make them suitable for interdisciplinary coastal research.

136 ENABLING OCEAN RESEARCH IN THE 2 lST CENTURY

/MPLEMENTAT/ON OF A NETWORK OF OCEAN OBSERVATORIES FOR RESEARCH 137 Data Management System Implementation Establishment of the OOI-DMS will need to be phased and coordi- nated with observatory installations (Box 4-5~. Even though the operators of each observatory system will manage data functions individually, co- ordination at the program level will be necessary to guarantee compat- ibility across observatory types. A Data Management advisory committee should be established by the OOI Program Office to oversee the imple- mentation strategy outlined in Box 4-5.

Next: 5 Related Facility Needs for an Ocean Observatories Network »
Enabling Ocean Research in the 21st Century: Implementation of a Network of Ocean Observatories Get This Book
×
Buy Paperback | $60.00 Buy Ebook | $47.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

As the importance of the oceans to society grows, so does the need to understand their variation on many temporal and spatial scales. This need to understand ocean change is compelling scientists to move beyond traditional expeditionary modes of investigation. Observing systems will enable the study of processes in the ocean basins over varying timescales and spatial scales, providing the scientific basis for addressing important societal concerns such as climate change, natural hazards, and the health and viability of living and non-living resources along our coasts and in the open ocean.

The book evaluates the scientific and technical readiness to move ahead with the establishment of a research-driven ocean observatory network, and highlights outstanding issues. These issues include the status of planning and development, factors that affect the timing of construction and installation, the cost and requirements for maintenance and operations, needs for sensor development and data management, the impact on availability of ships and deep submergence facilities, and the role of research-based observatories within national and international operational ocean observing systems being developed and implemented.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!