Major Research Equipment and Facilities Construction
In its statement of task (see Box 1-1), the committee was asked “to assess whether the Science Plan makes a compelling case for establishing the WATERS Network with Major Research Equipment and Facilities Construction (MREFC) funding.” The committee addressed this question in its letter report (NRC, 2009; see also Appendix A), and its findings are described below.
DEFINING A FACILITY UNDER MREFC
The National Science Foundation (NSF) defines a facility under MREFC as an essential part of the science and engineering enterprise that will advance science in ways that would not be possible otherwise. The facility can either be centralized or consist of distributed installations. The project “should offer the possibility of transformative knowledge and the potential to shift existing paradigms in scientific understanding and engineering processes and/or infrastructure technology and should serve an urgent contemporary research and education need that will persist for years” (NSF, 2007).
MREFC projects are so large that the total construction costs would be >10 percent of the budget for the sponsoring directorate or office. Thus, funding of the facility would distort the base program of funding in that discipline(s) without MREFC funding. Investments in computing resources and for supporting cyberinfrastructure can be included in the design plan and the construction costs. In most MREFC projects, construction occurs over a relatively limited period of time (approximately 5 years). Within this short construction time line, there is little time for learning and adaptation, given the magnitude of the project and the advanced planning requirements.
Although many funded MREFC facilities are large items such as particle accelerators, telescopes, research vessels, and polar aircraft, several geographically distributed but networked MREFC facilities have been approved and are in various stages of development. Earthscope, a continental-scale seismic and magnetotelluric observatory designed to provide a foundation for integrated studies of continental lithosphere and deep earth structure over a wide range of scales, has completed the construction phase and continues to collect data and address fundamental science questions. The Ocean Observatories Initiative (OOI), a set of ocean observatories designed to encompass nearly every area of ocean science at global, regional, and coastal scales, is a project approved for construction. The National Ecological Observatory Network (NEON), a national network of observatories for continental-scale observations and experiments on ecological systems, has passed the preliminary design review phase of the MREFC process.
CASE FOR ESTABLISHING THE WATERS NETWORK WITH MREFC FUNDING
MREFC is a possible mechanism to fund the infrastructure envisioned for the WATERS Network. One of the major strengths of the WATERS Network is that hydrologic sciences, engineering disciplines, and the social sciences are cooperatively developing the WATERS Network plan with the full support of the three NSF directorates. This would be the first MREFC project to span natural sciences, social sciences, and engineering.
The NSF guidelines (NSF, 2005) state several conditions that a project must meet to qualify for MREFC funding. Among the conditions are the following: “To qualify for MREFC investment, networked infrastructure must exhibit systems characteristics greater than inferred simply by the connectivity of its parts.” The understanding also is that the facility would “require large investments for construction/acquisition, over a limited period of time, such that the project cannot be supported within one or more NSF Directorate(s)/Office(s) without severe distortion to the funding of its portfolio of activities.” The committee understands that NSF intends these guidelines to mean that a facility must satisfy the condition that addressing the proposed science questions would require construction of the network in its entirety over a short period and that pieces of the network (e.g., one or a few of the observatories) could not effectively meet the science objectives. That is, the committee as-
sumes that a proposed network would have to satisfy both the “systems characteristic” and the “large investments over a relatively short period” conditions to qualify for MREFC funding. The Long-Term Ecological Research (LTER) Network and the more recent Critical Zone Observatories, funded under directorate research programs, are examples where individual observatory sites conduct transformative science without meeting the MREFC criteria, even though additional comparative insights are gained by having multiple observatories.
The Science Plan makes a convincing case that the WATERS Network will likely lead to strong, transformative science in its individual pieces. The committee, however, was not clearly convinced by the Science Plan that a collection of such pieces will meet the MREFC criterion that the WATERS Network will “exhibit systems characteristics greater than inferred simply by the connectivity of its parts.” Each of the three hypothetical examples of regional, theme-based science (snow hydrology, eutrophication of estuaries, and urban water systems) illustrates how our understanding of particular issues could be significantly advanced. However, as noted in Chapter 2, there do not appear to be clearly articulated compelling questions or hypotheses in the Science Plan that require integration across individual observatories at the same time. Rather, the proposed network of observatories appears to be a collection of many strong pieces. Some of the components are new, while others would consist of existing sensors or observatories, operated by mission agencies, that could be shared or repurposed to meet objectives of the WATERS Network.
From a purely scientific view, the Science Plan does not clearly articulate a rationale for why WATERS as a facility is required to address the key science questions. That is, the Science Plan does not present a convincing case explaining why the simultaneous construction of the entire infrastructure is essential to answer the science questions, as opposed to phased construction of a few observatories at a time. The document also does not explain clearly why any of the three major questions cannot be approached regionally and, in fact, why some current efforts are not addressing the science questions, at least in part. For example, the first major WATERS question is “how is fresh water availability changing and how can we understand and predict such changes?” (Dozier et al., 2009). The U.S. Climate Change Science Program (CCSP) is carrying out work described as follows:
FY 2008 activities will focus on a few regional case studies in which both models and measurements will be used to develop
closure in the terrestrial water cycle budget for those regions. This multiagency CCSP project will use existing regional sites to improve observational capabilities (surface, subsurface, and remote sensing). A range of climate zones will be considered to provide a suitable research framework that concurrently addresses climate/water cycle science and water resource management issues. (USCCSP, 2007)
The WATERS Science Plan does not describe, even conceptually, why efforts such as these are insufficient to address the first grand challenge identified, and thus, why the network as a facility (in the sense of NSF’s MREFC program) is needed.
The Science Plan does argue that the accelerating pace of human-induced changes to the environment calls for enhanced knowledge for policy makers and water users to cope with potentially disastrous effects. The committee envisions that policy-relevant social sciences questions could be addressed at a national scale through the WATERS Network, at least for the specific issues developed in the Science Plan. For example, how do incentive- and regulatory-based management approaches differ in their ability to control nutrient loading and hypoxia in rural, suburban, and urban watersheds, and how consistent are these differences across different hydrologic conditions? Important policy-relevant discoveries from one observatory could be cross-checked rapidly at others, providing a national-level knowledge base for policy makers. Thus, if implemented as a facility, the WATERS Network could provide an integrating source of knowledge needed for rapid adaptation in the very fragmented governance system. This is a powerful argument in support of a national network supporting interdisciplinary research in water science, with a strong social sciences component, that is not clearly articulated in the Science Plan. However, the persuasiveness of the argument for WATERS as a unified facility also requires a strong case for the scientific and engineering knowledge to be gained from a national network.
As the WATERS team goes forward, there are two possible options that the committee sees. First, the WATERS team could bolster its case that a national network of observatories is required to address the science questions that are posed. Second, an alternative funding procedure under the MREFC or some other mechanism within NSF might be considered, if feasible, for establishing a phased network of observatories such as envisioned in the WATERS Science Plan. It may not be best for advancing water science to demand that spatially distributed and temporally extensive measurements at a set of observatories pass a “facility” test.
Rather than emplace a network of sensors that is based on a fixed initial design (or an initial set of hypotheses), it may be more efficient to build out a field design as one learns how the hydrologic–human systems operate at a site, provided that there is a long-term commitment to the program. If the entire network is not simultaneously required to address the science questions, it is probably more sensible to build the network incrementally and let the questions and experiments evolve in an adaptive framework. This approach, which is not constrained by MREFC time lines for design and construction phases, could take better advantage of advances in technology over time, such as for sensors and components of the cyberinfrastructure. Also, capital costs would be lower initially and would be spread out over a longer period of time. Downsides of this phased approach are that the delivery of data from the network as a whole would be substantially delayed, and it is possible that observatories in the network would be based on different technology and even different science questions. Given the current vision for WATERS as outlined in the Science Plan, the potential benefits of a phased approach appear to outweigh these drawbacks, assuming that long-term funding support for phased implementation can be found within NSF.
CONCLUSIONS AND RECOMMENDATIONS
To qualify for MREFC funding, NSF requires that facilities “must exhibit systems characteristics greater than inferred simply by the connectivity of its parts” (NSF, 2005). The proposed network of observatories appears to be a collection of many strong pieces, but the Science Plan does not explain clearly why any of the three major questions cannot be approached regionally and, in fact, why some current efforts are not addressing the science questions, at least in part. As the WATERS team goes forward, it should bolster its case that a national network of observatories is required to address the science questions that are posed. The committee believes that such a case can be made, especially with a strong social sciences component as part of the interdisciplinary water science network. However, the persuasiveness of the argument for WATERS as a unified facility also requires a strong case for the scientific and engineering knowledge to be gained from a national network. Alternatively, a different funding mechanism within NSF might be considered, if feasible, for establishing a phased network of observatories that could address the questions posed in the WATERS Sci-