Adding a Wave Buoy Under a Short-Term Grant

The first step is to turn the grant idea into an agreed scope with named decision rights. The cooperative agreement already identifies the port as non-federal sponsor and NOAA CO-OPS as operator, but it does not necessarily cover a new buoy, a second profiler, or a public dashboard. An addendum should state who owns each asset, who operates and maintains it, who is the data steward, who approves QC procedures and metadata, who authorizes public release, and who pays for operations after the 18-month grant ends. Because the grant has no operations funding, a sustainment plan and a named budget owner should be conditions of procurement, not afterthoughts. The 90-day deadline is a schedule risk: permitting, site selection, telemetry, and QC take longer than a direct dashboard feed. A phased release, with the pilots' products first and the public dashboard later, is more defensible than bypassing QC. The addendum should also define change control, so the port IT team cannot unilaterally alter the publication path.

Use one authoritative data portal with access controls and distinct product tiers rather than divergent copies. For the pilots, publish 30-minute significant wave height, dominant period, and direction after automated QC and with a clear status of preliminary or verified. For the public dashboard, publish only QC-flagged operational products with plain-language status and metadata; a direct unflagged feed would violate the data-management plan and invite users to treat provisional values as final. For the university, provide raw 1 Hz spectra and bottom temperature under a data-use agreement that identifies the data as provisional, documents calibration and processing metadata, and requires acknowledgement that operational QC has not been applied. Keep raw data immutable and retain the QC flags. If bandwidth or labor is limited, the pilots' tier should take priority because it supports navigation safety; the research tier can be delivered as delayed files. This tiering satisfies the different needs without creating conflicting public values.

The main risk is that the grant buys hardware but not the annual costs of ship time, sensor calibration, telemetry, QC labor, and replacements. Estimate those costs before deployment and identify a continuing sponsor, such as the port, pilots, or tenants who benefit from the data. If no sponsor exists, plan for decommissioning and notify users in advance rather than letting the feed fail silently. A second risk is stakeholder misalignment: the pilots want a narrow operational product, the university wants raw data, and the port wants a public dashboard. A short written product specification and a data-management plan can resolve this. A third risk is branding: if the data are presented as PORTS, they should meet PORTS QC and metadata expectations. Finally, confirm site access and permits for the buoy and profiler. Governance succeeds here only if it is written down before the 90-day clock creates pressure to cut corners.