FAA Ground Delay Programs as an Event Contract Signal

When the FAA slows an airport down, it does not do it quietly. It publishes a program — with a name, a rate, a scope, a start time, and an average delay — hours before most of the affected flights push back from the gate. A Ground Delay Program (GDP) is one of the most explicit delay signals in world aviation, and it is entirely public. For anyone trading flight delay event contracts on US routes, learning to read that program is learning to read tomorrow’s settlement data today.

This guide covers the US machinery specifically: what separates a GDP from a Ground Stop and an Airspace Flow Program, where the live feed sits, which advisory fields carry information, and the ways the signal decays after publication. The European equivalent — EUROCONTROL’s ATFM regulation system — runs on different plumbing and is covered separately in our EUROCONTROL ATFM delay trading guide.

GDP, Ground Stop, or AFP: Three Programs, Three Different Signals

The FAA’s Air Traffic Control System Command Center (ATCSCC) manages congestion with several traffic management initiatives; three matter most to a delay trader, and they are not interchangeable.

A Ground Delay Program is, in the FAA’s own definition, “a traffic management process administered by the ATCSCC, when aircraft are held on the ground. The purpose of the program is to support the TM mission and limit airborne holding” (FAA Pilot/Controller Glossary, accessed August 6, 2026). FAA Order JO 7210.3 adds the mechanism: aircraft are held “in order to manage capacity and demand at a specific location, by assigning arrival slots” (JO 7210.3EE, para 18-10-2). A GDP is measured, quantified, and usually runs for hours. It tells you delay is not a possibility but an assignment.

A Ground Stop is blunter. Per the same glossary, it “requires aircraft that meet a specific criteria to remain on the ground,” and — critically — “GSs normally occur with little or no warning.” JO 7210.3 reserves ground stops for severe situations: capacity below arrival minimums, snow removal, saturation, facility outages (para 18-13-2). For a contract on a flight into the affected airport, a ground stop is a step-change in probability — but typically short, reassessed at a published update time.

An Airspace Flow Program targets airspace rather than an airport: aircraft “are assigned an Expect Departure Clearance Time (EDCT) in order to manage capacity and demand for a specific area of the National Airspace System… to mitigate the effects of en route constraints” (Pilot/Controller Glossary). AFPs typically appear on thunderstorm days. Because flights can reroute out of one, the per-flight signal is softer than a GDP’s — real, but with escape hatches.

Where the Live Feed Sits: nasstatus.faa.gov

The FAA Command Center publishes its Operational Information System (OIS) at nasstatus.faa.gov. Three views do the work: the airport status view lists every active GDP and ground stop with headline numbers; the advisories view carries the full ATCSCC advisory text; the operations plan — reissued through the day — lists what the Command Center expects to break next.

A live snapshot at the time of writing (August 6, 2026, 14:06 UTC): SEA under a GDP for low visibility, average delay 26 minutes, maximum 1 hour 14; SFO under a GDP for low ceilings, average 43 minutes, maximum 1 hour 32; SAN under a GDP for airport volume, average 51 minutes, maximum 1 hour 47. The same day’s operations plan flagged possible ground stops or delay programs for LGA/JFK and EWR/PHL after 19:00 UTC — a forward-looking line published five hours before the event window.

Patterns matter: morning marine-layer programs at SFO and SEA recur almost seasonally — see our fog versus thunderstorm delay profiles guide — and the operations plan is the earliest public hint, naming airports the Command Center is watching hours ahead.

Anatomy of a GDP Advisory: The Fields That Matter

Every GDP implementation, revision, compression, or cancellation goes out as an ATCSCC advisory with contents prescribed by JO 7210.3: Airport, Delay Assignment Mode, Aggregate Demand List (ADL) Time, Arrivals Estimated For, Program Rate, Flights Included, Scope, Exempt Facilities, Maximum Delay or Delay Limit, and Reason, with Average Delay optional (para 18-10-6, accessed August 6, 2026).

For probability work, four fields dominate. Program Rate is the arrival flow the program is built to deliver. Compare it against the airport’s normal Airport Arrival Rate — “a dynamic input parameter specifying the number of arriving aircraft which an airport or airspace can accept from the ARTCC per hour” (Pilot/Controller Glossary) — and against scheduled demand for the affected hours. A program rate of 30 at an airport that schedules 40 arrivals per hour implies a deep, compounding queue; a rate trimmed by ten percent implies a shallow one. Arrivals Estimated For gives the program’s time window: flights landing after the end time may escape entirely. Maximum Delay bounds the worst case, and Reason tells you whether the constraint is self-resolving (a morning stratus deck burns off) or open-ended (equipment outages, volume).

Scope and Exemptions: Which Flights Actually Carry the Delay

A GDP does not hit every inbound flight equally. Delay is assigned to “aircraft departing within a defined geographical area,” which “may consist of one or more ARTCCs and one or more Canadian airports” (JO 7210.3EE, para 18-10-4). That tiering is the Scope field: departures within a distance ring or within specific en route centers, with the Exempt Facilities line spelling out who is excluded. GDPs apply to departures from the contiguous United States and select Canadian airports; Alaska is excluded outright (para 18-10-1), and international origins beyond Canada typically sit outside the program’s reach.

The trading consequence is direct: two flights arriving at the same congested airport in the same hour can carry very different program exposure. A domestic departure from inside the scope holds an assigned slot delay; a long-haul international arrival may be exempt and land close to schedule. Before pricing any contract against a GDP, the first question is not “how bad is the program” but “is this flight’s origin inside the scope.”

EDCT: The Per-Flight Number — and How It Differs from CTOT

Inside a program, each captured flight receives an EDCT — “the runway release time assigned to an aircraft in a traffic management program” (Pilot/Controller Glossary). JO 7210.3 explains the arithmetic: “The EDCT is calculated based on the estimated time en route and the arrival slot” (para 18-10-2). The EDCT is where an airport-level program becomes a flight-level fact: a specific later release time, included in the flight’s clearance.

Traders who came to GADUIN from European markets will recognize the family resemblance to EUROCONTROL’s CTOT — a calculated take-off time issued against a regulated sector or airport. The mechanics differ — tolerance windows, substitution rules, who may amend the time — and the European reading workflow is its own discipline, covered in the EUROCONTROL ATFM guide rather than restated here. The US-specific point is this: an EDCT is anchored to an arrival slot. It is the arrival airport’s shortage, expressed as a departure hold hundreds or thousands of miles away.

RBS, Substitutions, and Compression: Why the Delay You Saw Is Already Stale

The initial slot allocation follows RBS — “Ration by Schedule” in the FAA’s CDM vocabulary (CDM Acronym Sheet, tfmlearning.faa.gov, April 2024 edition, accessed August 6, 2026): slots are rationed in the order flights were originally scheduled to arrive. That first cut is equitable by design — JO 7210.3 states that “GDPs provide for equitable assignment of delays to all system users” — but it is only the opening position.

Then the airlines start moving. Under the FAA’s Collaborative Decision Making framework, “users are permitted to exchange and substitute CTAs congruent with CDM agreements concerning substitutions” (para 18-10-12). An operator facing a bank of connections will sacrifice a low-value flight — delay it further, or cancel it — to move a high-value flight into the freed slot. The Slot Credit Substitution (SCS) mechanism, referenced throughout JO 7210.3’s EDCT amendment procedures, formalizes the trade. Separately, the ATCSCC “may make revisions and compressions to the GDP as conditions at the airport or within the airspace change” (para 18-10-5) — compression sweeps unused slots forward, pulling flights earlier when cancellations open gaps.

The consequence: per-flight delay inside a GDP is a moving target after publication. The program’s average delay describes the pool at one ADL timestamp, not your flight. A hub carrier’s marquee evening departure is likely to be substituted out of a deep delay; a thin regional frequency is the natural donor. If your contract sits on a flight the operator has every incentive to protect, discount the average.

Revisions, Extensions, and Cancellations: Trading the Program’s Life Cycle

A GDP is not a single announcement but a life cycle, and every stage generates a fresh advisory: implementation, revision, compression, cancellation (para 18-10-6). Reading the sequence beats reading any single message.

A revision that cuts the program rate or pushes the end time later is the airport admitting the constraint is worse than modeled — probability of threshold-length delay rises across the affected window. A revision that raises the rate, or an early cancellation — for which the ATCSCC conferences facilities “to develop an operational plan for release of ground delayed traffic into the system” (para 18-10-10) — collapses the tail risk fast. Ground stop advisories even carry an explicit forward-looking field: JO 7210.3 requires national ground stop advisories to state a “Probability of Extension” (para 18-13-4). The Command Center is literally publishing its own probability estimate — and the market rarely prices it within minutes.

The practical cadence: note the advisory number when a program starts, and treat every subsequent advisory for that airport as a repricing event. Programs that get revised once tend to get revised again — weather models miss in trends, not in single steps.

From Program Parameters to a GADUIN Price

Turning an advisory into a position update is arithmetic, not art. Take the program’s average and maximum delay and set them against the contract’s threshold. A GDP averaging 43 minutes against a 15-plus-minute delay threshold on an arrival inside scope means the base outcome has flipped: on-time arrival is now the minority case. The same program against a 3-hour threshold barely moves the needle — the maximum delay field tells you whether the tail even reaches that far. Matching threshold to signal depth is its own topic, covered in the threshold selection guide.

Anchor the timing too: a flight arriving after the program’s window may carry little risk; one landing mid-program carries close to the full assignment. And remember what settles the contract: GADUIN resolves against verified actual flight data, not FAA program parameters — the process is documented in how GADUIN verifies flight delay outcomes. The GDP is your forecast input; the aircraft’s actual movement is the settlement fact. In BTS delay-cause statistics this signal family lands in the “National Aviation System Delay” category (transtats.bts.gov, accessed August 6, 2026), which is one reason airports with chronic flow programs top the delay tables in our airport delay hotspots review.

When the GDP Signal Lies

Four failure modes account for most losses on this signal.

The early cancellation. Programs built on forecast weather get cancelled when the forecast busts. A marine-layer GDP scrubbed two hours early releases held traffic quickly, and afternoon arrivals recover. A position that needs the program to run its full published length is exposed to that advisory — and cancellations, unlike implementations, come with no operations-plan warning.

The average is not your flight. In the live snapshot above, SEA’s program averaged 26 minutes but maxed at 74. RBS ordering, substitutions, and compression spread delay unevenly; a specific flight can sit at either end. Where possible, check the flight itself — an assigned EDCT, a gate hold, prior-leg lateness — rather than trading the pool statistic. The aircraft’s inbound rotation is often the larger factor, and that mechanism has its own guide.

Arrival program, departure contract. A GDP delays flights into the constrained airport. Departures out of that airport are not assigned EDCTs by that program. A contract on an SFO departure during an SFO arrival GDP is exposed only indirectly — through late-arriving aircraft and gate congestion — a real but weaker, laggier link. Do not price a departure-side contract as if it held an arrival slot.

Outside the scope. Exempt facilities, international origins, and post-window arrivals ride out the program largely untouched. The advisory’s scope and exemption lines are one paragraph long. Read them before assuming a flight is captured.

The FAA Ground Delay Program is the rare case where the referee announces the delay before the game: published rates, defined scope, per-flight times, and revision history, all in the open. The signal decays — airlines substitute, programs compress, forecasts bust — so its value sits in the first minutes after each advisory, and in knowing precisely which flights it does and does not touch. Read the program, then verify the flight.

This content is provided for informational and educational purposes only and does not constitute financial, investment, or trading advice. Trading event contracts involves risk of capital loss. GADUIN services are not available to US persons. Please review the Terms of Service and User Agreement before opening any position.