Illuminating the Hidden Planet: The Future of Seafloor Observatory Science (2000)

Given the wide range of disciplines that will be involved in a sustained observational presence on the seafloor and the investment required to establish seafloor observatory facilities, there is an unprecedented need for multidisciplinary coordination of scientific objectives and infrastructure needs. This scientific program and the resulting observatory infrastructure must be driven by high-quality, proposal-driven science. It is also imperative that each observatory site be located, configured, and used for multidisciplinary studies of ocean and earth processes occurring at various time and space scales. These objectives cannot be achieved without effective program and project management.

Management requirements for a seafloor observatory program will change significantly during the evolution from the planning phase into the initial development, design, and implementation phase, and then into the operational phase with subsequent development, design, and implementations. To properly administer a seafloor observatory program, several levels of management will be required, ranging from scientific oversight and systems engineering of distinct observatories, to management of a national observatory infrastructure co-supported by several agencies, to international coordination of global observatory programs. Independent scientific and engineering structures should also be established to determine the scientific direction of the observatory program and to ensure strong project risk management and systems integration.

It is beyond the scope of this report to outline a management structure for a national seafloor observatory initiative. The Committee does recommend

that a necessary first step in establishing a seafloor observatory initiative is the development of an implementation program that includes plans for a management structure. During the development of this implementation plan, other relevant agencies should be consulted to ensure cooperation on areas of common interest, such as data management. Interagency discussions will also help identify areas of potential conflict. The Committee suggests the following principles to guide the makeup of a program and project management structure:

• In all phases of a seafloor observatory program, management must be as responsive as possible to the scientific community it serves. As such, the scientific community must be fully involved in the development of an initial program implementation plan and must play a critical oversight role in the establishment and operation of a seafloor observatory network. This role should include the solicitation and coordination of scientific proposals that will drive the development and expansion of the seafloor observatory network.

• Likewise, management must be responsive to the funding agencies supporting seafloor observatories. If co-funding from multiple agencies is involved, as appears will be likely, then interagency agreements must be established as part of the development of a program implementation plan.

The initial development, design, and implementation phase for a seafloor observatories network should include engineering development, prototyping, manufacture, and installation of basic observatory infrastructure. Management of this phase should focus on the following:

• identification and prioritization of specific observatory projects and the development of timetables for their implementation;

• regular review of engineering development, systems integration, prototype efforts, and engineering risks by independent engineers and potential scientific users;

• the completion of previously identified milestones prior to proceeding further with program implementation;

• standardizing of design features common to both cabled and mooredbuoy observatories wherever possible, as well as common development of observatory components;

• the development of international interface standards to enable deployment of scientific instrumentation on any type of seafloor observatory;

• the development of a certified testing capability and standardized operational parameters for scientific equipment to minimize impacts on other observatory experiments;

• early identification of common instrumentation needs to advance the development of scientific measurement capabilities in parallel with infrastructure development;

• the establishment of successful demonstration projects before committing to major installation efforts;

• ensuring sufficient flexibility in the program management structure to evolve as development proceeds and scientific goals and opportunities are further elucidated.

A possible program model for the initial development, design, and implementation phase might involve the establishment of a scientific oversight committee and program office with distinct project efforts for observatory programs prioritized for initial development and deployment. Once the initial seafloor infrastructure is successfully installed, program management will shift toward more routine operational, administrative, and management functions, which should be guided by the following principles:

• As seafloor observatories will require tremendous investments in infrastructure, a project management plan must be established to both oversee infrastructure operation and maintenance and to facilitate proposals by independent investigators wishing to use that infrastructure.

• One of the highest priorities of project management should be to ensure fair and equitable access to observatory infrastructure by all funded investigators.

• Project management should encourage the use of seafloor observatory infrastructure by all potential scientific investigators by providing free access to information concerning observatory infrastructure (such as protocols for attaching instruments to observatory nodes) and to all data collected by core observatory instrumentation.

• Given the potential for synoptic measurements with seafloor observatories, it will be particularly important to facilitate truly interdisciplinary projects.

• Access by educators, students, policymakers, and the general public via the Internet must be encouraged.

Currently, other large, coordinated scientific programs in earth, ocean, and planetary sciences (e.g., University-National Oceanographic Laboratory System [UNOLS] ship operations, Incorporated Research Institutions for Seismology [IRIS], Joint Oceanographic Institutions for Deep Earth Sampling [JOIDES] and the international Ocean Drilling Program [ODP], and National Aeronautic and Space Administration [NASA] mission planning) share similar objectives to those listed above. Thus, the management systems of these

programs provide useful models for a multi-agency, international seafloor observatory program, and many of the most successful features of these models should be considered, such as the following:

• establishment of a prime management organization by an incorporated consortium of research institutions, universities, agencies, and partnerships that is under contract to the National Science Foundation (NSF) and cooperating agencies (e.g., JOIDES, IRIS, and CORE);

• establishment of a governing body for the incorporated consortium that is composed of experts in science, technology, engineering, and management pertinent to seafloor observatory science (e.g., IRIS);

• oversight of scientific planning by committees of experts appointed by the governing consortium following recommendations from member institutions (e.g., JOIDES);

• awarding of subcontracts to separate operators for distinct observatory projects to encourage competition and the establishment of high-quality efficient services (e.g., UNOLS ship operations);

• awarding of subcontracts to industry when it is best qualified to deliver cost-effective services, such as for engineering development, cable maintenance, large buoy construction, etc. (e.g., the ODP drill ship model);

• allocation of significant resources to data archiving and dissemination from the outset of the program (e.g., IRIS);

• allocation of significant resources for establishment of a high-quality public outreach program (e.g., NASA);

• dedication of significant resources to engineering development of sensors and measurement systems, either in academia or industry as appropriate (e.g., as with the development of ODP third-party downhole logging tools);

• the guarantee of major, sustained agency funding commitments for observatory science based on independent peer review of the observatory management structure and the scientific justifications and outcomes of the observatory program (e.g., ODP).

During the initial stages of this study, the committee was charged by NSF to provide an “order of magnitude” estimate of the costs needed for the establishment and initial phase of a seafloor observatory program. To this end, there was much discussion in the technical breakout sessions at the symposium to attempt to gauge these costs. The outcome of these discussions was that the initial cost of this program could approach several hundred million dollars, with annual operation and maintenance costs approaching tens of millions of dollars annually.

As with program and project management, it is also beyond the scope of this report to detail specific information management protocols. Despite this, from discussions at the symposium and Committee deliberations, certain guiding principles can be suggested.

The three components of a visionary data and information management system that could be incorporated as part of a seafloor observatory program include Internet services, data processing and management, and long-term stewardship. Internet services should include a quick-look, versatile graphics interface with discovery, browse, subsetting, ordering and delivery, interactive visualization, data mining, analysis, and data-driven numerical modeling capabilities. Data processing and management functions should include such items as ingest processing, quality control, reprocessing, product generation, automated catalogs and inventories, browse image generation, and system navigation.

The data collected at seafloor observatories should be centrally managed, with a short time interval between receiving the data and making it available to end users. All data should be preprocessed and permanently archived, and should be non-proprietary unless collected from experimental sensors or individual investigator-initiated experiments. In the latter case, data will remain proprietary for an agreed duration but not to exceed two years, as is consistent with present federal ocean data policy. Raw data and metadata from such experiments should be archived at the earliest possible time to prevent loss. Core financial support should be made available to principal investigators for production of data products suitable for public distribution. Furthermore, support should be given to facilitate educational outreach using seafloor observatory infrastructure, where schools could have direct involvement in the planning and execution of the experiment and receive real-time data. It will also be important to ensure that the data collected from the research-based observatories proposed here can be easily integrated with the data management strategies proposed for the Global Ocean Observing System (GOOS).

Objectives for seafloor observatory information management:

• encourage public involvement by providing as much near real-time information as possible;

• support educational uses for seafloor observatory information;

• provide both the scientific community and the public an efficient method of obtaining information from seafloor observatories;

• support the flow of data from seafloor observatories to data processing nodes;

• assist researchers in the integration and management of their datasets

by providing support for the transition of experimental information into synthesized data products;

• provide the up-to-date status of observing elements and maintenance schedules.

Principles of seafloor observatory information management:

• Routine (facility-generated) information products and data should be made available without restriction as quickly as possible.

• Data from experimental sensors or individual investigator-initiated experiments should be made publicly available after quality control procedures have been applied, but within one to two years of retrieval.

• Data summaries (graphics, photographs, etc.) and metadata should be made available in near real-time whenever possible.

• To the extent possible, existing data management facilities, such as the National Oceanographic Data Center (NODC) should be used and enhanced as necessary to meet the needs of seafloor observatories.

A possible model for a “Data, Information, and Knowledge System” for a seafloor observatory program is provided by the Space Physics and Aeronomy Research Collaboratory at the University of Michigan (http://sparc-1.si.umich.edu/sparc/central). This system was designed to facilitate interdisciplinary science, while meeting the needs of disciplinary science and the general public. The Collaboratory consists of a comprehensive and visionary data management system; a set of display, analysis and data mining or information-extraction tools; access to current data; and timely access to older data. These capabilities could facilitate widespread use of seafloor observatory data and data products.

Seafloor observatories will offer an excellent opportunity for educational outreach, as they have the potential to bring the excitement and discovery of ocean science to the public. Virtually all aspects of observatory research are amenable for inclusion in outreach programs, but the most unique are the capabilities for real-time data transmission and, particularly, real-time display of visual images, such as those produced by video, acoustic, and optical sensors.

Outreach efforts can be adapted from successful programs currently in use combined with new approaches designed specifically for observatories. Examples of possible outreach efforts include the following:

• incorporation of real-time image and data feeds into museum exhibits via the Internet;

• incorporation of real-time image and data feeds into K-12 curricula (presented either through schools or museums, as with the Jason program);

• development of curriculum modules for K-12 students that will incorporate research results and scientist profiles;

• establishment of a summer research sabbatical program for K-12 teachers to facilitate interaction with scientists using observatory data;

• development of public World Wide Web sites, publicized through libraries, lay science journals, etc., containing real-time images and data;

• use of public television to broadcast the exciting science coming out of seafloor observatories.

Currently, numerous museum and curriculum outreach programs are being established, but for ocean sciences most of these partnerships are related to coastal studies. An important objective of observatory outreach should be to ensure a broad geographic scale for seafloor observatory public outreach programs.

Involving students and the lay public in the excitement of ocean science needs to be a primary objective of an observatory effort. This effort requires dedicated resources and formal inclusion in project and data management plans. Nationally funded science initiatives that have demonstrated successful outreach programs include NASA, Sea Grant, and IRIS. Other approaches include individual efforts through NSF education programs (the Directorate for Education and Human Resources), local partnerships between individual K-12 schools and nearby marine research institutions or universities, and the privately funded JASON project. Seafloor observatory outreach plans should build on the successes of these programs by incorporating their best and most appropriate aspects. Also, it would be advisable to have a centralized outreach office to coordinate individual efforts and provide information and support. Funding for educational and public outreach efforts should be sought not only from observatory program funds but also from other governmental, private, and industry sources.