When the Governance Instrument Exists: ICOS, NEON, and the Two Paths to

SCENE: The Clause That Is Not There

The most instructive documents in policy work are not the ones that tell you what exists. They are the ones that reveal, by their specificity, what is missing everywhere else.

To understand the gap in how we govern scientific data, you have to look at the few institutions that get it right. Both the National Ecological Observatory Network (NEON) in the United States and the Integrated Carbon Observation System (ICOS) in Europe are massive, continental-scale monitoring networks. They track multi-generational climate and ecological baselines. Because they operate at this scale, they cannot rely on the goodwill or “culture” of individual scientists to keep their sensors running. They must guarantee continuity.

I came to the NEON cooperative agreement the way most policy analysts come to contracts: looking for something specific rather than reading for comprehension. I was trying to figure out how federal science infrastructure handles the transition between operators, and the NSF-Battelle arrangement kept coming up as a reference case. What I found was a data continuity obligation written into the operative terms of the agreement itself. Not a best-efforts commitment. Not a programmatic goal. A requirement, with named parties and a defined scope.

I have read a fair number of federal contracts and cooperative agreements in the course of work at the Forest Service Washington Office. Most of them handle continuity the way most organizations handle continuity: by assuming it. The NEON language was different because it named the obligation explicitly and assigned it to a specific institutional party. In a contract, the difference between an assumed obligation and a named one is the difference between something you can enforce and something you can only complain about after it fails.

The question that followed was whether this was a one-off innovation or a pattern. ICOS answered that question.

BREAK: Two Things Worth Holding Separately

The first is what ICOS and NEON have in common. Both are large-scale ecological and atmospheric observation networks. Both include long-duration data collection that depends on continuity of physical infrastructure and named stewards at individual sites. Both arrived, through entirely different institutional pathways, at the same governance solution: write named stewardship obligations into the operative governance structure before the need becomes acute.

That convergence is not a coincidence. It is a pattern. When sophisticated network designers working independently arrive at the same structural solution, they are solving the same underlying problem: distributed networks cannot maintain stewardship continuity without a separately governed accountability layer. Other global observing systems (like the GCOS Reference Upper-Air Network aka GRUAN) solved this by building deliberately non-federated, centrally controlled reference networks. ICOS and NEON solved it by writing the obligation into the governance instrument at the node level.

The second thing worth holding is that ICOS and NEON are not the same solution in different packaging. They represent genuinely different governance architectures that happen to address the same structural gap. Understanding both pathways matters because the appropriate instrument depends on institutional context.

SCHEMA: The Principle and the Two Pathways

The stewardship obligation in a long-duration observing network must be named, assigned, and embedded in the operative governance structure before the need becomes acute. The specific instrument depends on institutional context. Two documented pathways exist.

The Intergovernmental Pathway: ICOS

ICOS, the Integrated Carbon Observation System, is a European Research Infrastructure Consortium with 13 member countries. It measures greenhouse gas concentrations and fluxes across terrestrial and atmospheric stations distributed across Europe. Each station has a named Principal Investigator who carries a formal stewardship obligation as a condition of the station’s membership in the network. That obligation includes succession planning. If a PI leaves, the obligation does not leave with them. The station must name a successor before the transition, not after.

This governance structure did not emerge from a general principle. It emerged from hard-won understanding of what happens to long-term observational records when station-level stewardship is treated as a local management problem rather than a network governance requirement.

Voluntary stewardship at the node level is not stewardship. It is hope.

ICOS formalized what the community had learned the expensive way.

The ICOS model is relevant to the experimental forest context because the institutional structure is more similar than it might appear. ICOS coordinates member institutions that retain sovereignty over their own stations. No central authority can compel a member university or national institute to do anything. The stewardship obligation is enforced through the membership agreement rather than through hierarchical authority. That is exactly the governance mechanism available in a domestic federal context through cooperative agreements and interagency instruments. The USDA does not need to build a new institution. It needs to write the obligation into the instruments it already uses.

The Contractual Pathway: NEON

The National Ecological Observatory Network operates differently. NSF funds it as a large facility program and Battelle Memorial Institute operates it under a cooperative agreement that includes explicit data continuity obligations.

NSF built this language into the agreement deliberately. The history behind that decision is specific. In December 2015, following an $80 million cost overrun and congressional oversight hearings, NSF terminated NEON Inc. as operator after writing that it had minimal confidence in the contractor’s ability to complete the project. The House Committee on Science, Space, and Technology had substantiated allegations against NEON Inc., resulting in one of the largest Federal cooperative agreement terminations for cause in history. Battelle was selected as successor operator in 2016. The continuity obligation written into the Battelle cooperative agreement is not a best practice borrowed from elsewhere. It is a direct institutional response to watching a major operator transition nearly collapse a continental-scale scientific infrastructure program. The continuity obligation was a proactive measure to ensure the scientific record would survive the institutional change, regardless of who held the operating contract.

The specific significance of this arrangement is that continuity is treated as a requirement that survives operator transition, not as a cultural commitment of the current operator. When NSF issued a notice of intent to re-compete the management contract, the data continuity obligation did not dissolve. It transferred to the competitive requirements for the successor operator. An incoming operator who cannot demonstrate the capacity to maintain continuity of the physical infrastructure and data record does not qualify for the contract. The obligation is not a promise made by Battelle. It is a condition of operating the network, regardless of who operates it.

To be clear, no governance architecture survives a total collapse of funding. The re-competition clause is a shield against administrative indifference and sloppy operator transitions, not a magic wand against budget elimination. But it forces the abandonment of the archive to be an explicit funding decision, rather than a quiet administrative accident.

This is the governance architecture the experimental forest system lacks. The Forest Service Experimental Forests and Ranges network operates under no equivalent instrument. If a facility reduces operations or closes, there is no cooperative agreement clause, no interagency instrument, and no designated party with a named obligation to maintain the physical archive and its environmental controls. The transition happens inside the administrative process with no external governance requirement triggered.

Why the Contractual Pathway Matters Most for Domestic Federal Science Infrastructure

The ICOS intergovernmental model requires building or joining a consortium governance structure. For individual experimental forests and ranges, that path exists but it is slow and requires willing institutional partners. The contractual pathway requires only that the right language appear in the right instruments.

Federal research facilities already operate under authorizing statutes, cooperative agreements, and budget justifications. While the specific contracting mechanisms and statutory authorities differ between an independent agency like NSF and a cabinet department like the USDA, the underlying governance principle is identical: the stewardship obligation must be written into the binding agreements those specific agencies are legally authorized to use. Adding a stewardship continuity requirement to those instruments is a policy decision, not a structural innovation. The NEON cooperative agreement demonstrates that NSF already knows how to write this language. The gap is not technical. It is that nobody has specified the requirement for the experimental forest system. The experimental forest system does not need a new idea. It needs an existing one applied.

The most instructive detail in the NEON arrangement is not the continuity language itself but the re-competition clause. By requiring incoming operators to demonstrate continuity capacity as a qualification, NSF embedded the governance requirement into the procurement process. An operator who cannot maintain the physical baseline cannot win the contract. That is an enforcement mechanism that does not depend on anyone’s goodwill, institutional memory, or cultural commitment to the science.

Goodwill, institutional memory, and cultural commitment are what the experimental forest system currently runs on. They are not nothing. But they are not architecture.

When culture changes, architecture holds. When culture changes and architecture is absent, the record is at risk.

The NEON re-competition clause is architecture. The experimental forest system needs an equivalent.

Featured Image: A field technician services the instrument array atop a NEON flux tower at the Bonanza Creek site in Alaska. Each of NEON’s 81 field sites operates under a named stewardship obligation embedded in the cooperative agreement between NSF and Battelle. The obligation is not a promise made by the current operator. It is a condition of operating the network, regardless of who holds the contract. Photo: National Ecological Observatory Network (NEON), operated by Battelle. Retrieved from NEON Flickr archive

Doctrine Note

The pattern across GRUAN, ICOS, and NEON is not three separate governance innovations. It is one governance principle implemented through three different instruments matched to three different institutional contexts. The principle is the same in each case: name the steward, assign the obligation, and embed the requirement in the operative governance structure before the need becomes acute. The instrument varies. The principle does not.

This field note is part of the Ground Truth, Federation, and the Anchor Point series. Read the companion piece: Federation Cannot Anchor Itself: GRUAN and the Hidden Stewardship Layer in Global Observing Systems.

Last Updated on May 5, 2026