Runway Intelligence Methodology

• Active — a runway the tower states is in use: from the tower's TFDM surface system (structured data, via FAA SWIM) where the airport has it, otherwise from the D-ATIS broadcast text. Active is never inferred from wind.
• Favored — the end of a runway with more headwind in the current METAR. Pure arithmetic, recomputed every minute; no knowledge of traffic, noise abatement or what ATC is doing.
• Predicted — the ATIS Alerts estimate of the runway most likely in use at future hours, from the TAF wind and runway geometry. Labelled with a confidence band; never an assignment.
• Recorded use — hours during which the facility stated each end, from every change of the TFDM configuration or the D-ATIS since we started listening.
• Wind-favored history — at airports without a stated-runway source, the share of hours each end would have had the most headwind. Always labelled as wind-inferred.

• Tower runway configuration, closures, demand and traffic-management restrictions: FAA SWIM Cloud Distribution Service, TFDM (Terminal Flight Data Manager) Airport Information and Traffic Management Restrictions services, at the towers that publish them (about 36 in September 2026). Configuration lists come from per-runway rates; a runway with a zero rate for a role is not used for that role.
• Runway visual range and runway-light settings: FAA SWIM STDDS Airport Data (APDS), about 130 airports. Values are in hundreds of feet with 6,000 ft the reportable maximum; a reading of zero means the sensor is off, not zero visibility.
• D-ATIS text: the community relay at atis.info of the FAA digital ATIS, polled every five minutes and every minute on the page; parsed with an open, tested role-aware parser; the broadcast text is always shown.
• METAR and TAF: aviationweather.gov. Wind components use TRUE runway headings against TRUE METAR wind; magnetic headings are shown for orientation only.
• Runway record: OurAirports / FAA runway data (length, width, surface, lighting, headings). Declared distances, lighting systems and ILS types are not yet included.
• Climatology: Iowa Environmental Mesonet ASOS archive, 2005–2025, one observation per clock hour, the same archive and hourly method as the Flying Weather Index.
• Instrument approaches: the FAA d-TPP chart index for the current charting cycle; plates link to aeronav.faa.gov and no minimums are reproduced.

Every change of the tower's configuration (TFDM) or of the D-ATIS opens a new configuration interval and closes the previous one. Use over the last 30 days is the duration-weighted share of hours in which each end was stated for arrivals and for departures, split by wind sector (eight 45° sectors from the wind reported when the interval began) and by airport-local time of day. TFDM intervals start at the tower's own start time, so the history begins before we started listening; D-ATIS intervals start at the broadcast time. A station whose broadcast goes silent or stale is credited only one ATIS cycle past its last broadcast. Where a tower has TFDM the page uses the TFDM stream; the two streams are never mixed.

For each forecast hour the TAF wind is resolved against every open runway end. Each end is scored by headwind, minus a crosswind penalty, minus a penalty for any tailwind above five knots, with the longest runway breaking ties. Confidence starts at 50, rises six points per knot of headwind margin over the runner-up, and falls with forecast distance, gusts, TEMPO groups and a TAF older than six hours, clamped between 20 and 95. When the tower has stated a configuration, the “Now” row shows that instead and says so. The model does not know about noise abatement, flow programs or construction unless the D-ATIS states them.

For every runway end, each hourly observation with a steady wind direction is resolved into headwind and crosswind components against the end's true heading. The page reports the share of hours with a headwind, tailwind above 5, 10 and 15 kt, crosswind above 5 to 25 kt, mean, median, 90th and 95th percentiles, the strongest and calmest months, and the same by month, time of day and wind sector. Calm and variable-direction hours are excluded from the wind denominators and reported separately. Today's runway geometry is applied to all years, so a runway renumbered or built during the period carries the current heading throughout. A headwind requires at least 0.01 kt of component, so an exactly perpendicular wind counts for neither end.

• Current operational status and the FAA published baseline are separate: the status card shows what the D-ATIS, the tower and current NOTAMs state now; “FAA NASR baseline: Open” in the overview is the published record and never overrides a current closure. Precedence: a baseline-closed record; a current NOTAM full closure; a tower TFDM closure; a D-ATIS explicit closure (which stands even when no matching NOTAM is in our dataset); NOTAM restrictions; D-ATIS restrictions and outages; then OPEN only with a complete, fresh NOTAM picture. “No NOTAM” never means open.
• The FAA runway record marking a runway closed (a permanently closed runway) wins over everything and is labelled CLOSED (FAA BASELINE).
• A tower TFDM closure record for a runway the current configuration does not use reads CLOSED (TOWER TFDM); one for a runway the configuration still uses reads CLOSURE ON RECORD, because the closure may be partial or the record may lag.
• A D-ATIS closure reads CLOSED (D-ATIS); a broadcast restriction or equipment outage naming the runway reads RESTRICTED or EQUIPMENT OUTAGE with the clause quoted verbatim.
• NO CLOSURE REPORTED requires a tower feed or a D-ATIS broadcast fresh within three hours; a stale feed reads STATUS UNKNOWN, and an airport with neither source reads NOT MONITORED. No page ever reads “open” without a live source, and runway NOTAMs are not monitored yet.

Approach lighting, visual glide slope, edge / centerline / touchdown-zone lights, REIL, runway markings, RVR equipment, ILS components (localizer, glideslope, DME, marker beacons with their FAA status), LAHSO hold-short points with available landing distance, arresting systems and facility frequencies come from the FAA National Airspace System Resources (NASR) 28-day subscription, imported every cycle. A blank FAA field is shown as “Not available” and never inferred. Equipment is then annotated from two live sources: FAA NOTAMs received through SWIM that name the runway end, and outage clauses in the current D-ATIS. An annotation is a reading of the official text shown beside it (“OUT OF SERVICE” for OTS / unserviceable, “DEGRADED” for a portion, unmonitored or restricted); the text itself is always kept. A frequency is called runway-specific only when NASR sectorizes it by runway (“RWY 08L/26R”); every other row is labelled airport-wide or by its published sector text. LAHSO availability on record never means LAHSO is active; “LAHSO currently reported in D-ATIS” appears only for an explicit “in effect” statement. Declared distances (TORA/TODA/ASDA/LDA) are stored but not shown in this phase.

The D-ATIS collector stores one row per changedbroadcast text (an unchanged broadcast only refreshes its last-seen time) and opens a new configuration event only when the stated arrival or departure set changes. Letter-only re-issues therefore never create a configuration row; on the “Today's runway configuration” timeline they collapse onto the row they belong to (“ATIS: Bravo → Charlie, configuration unchanged”). Each row shows the arrivals, departures, flow, the LAHSO statement and what changed against the previous row. The tower's TFDM stream is recorded the same way, from the tower's own start time.

A flow is the compass direction aircraft are landing toward, read from the true headings of the ends the facility named for arrivals (or departures when no arrivals are stated). Eight compass words are used. A flow is named only when every arrival end points within 30° of the others and the mean heading sits at least 5° inside its 45° sector; crossing runways in use together are reported as “Mixed configuration”, and a heading on a sector boundary is left unclassified rather than named inconsistently. Where the tower's own TFDM configuration name carries a direction word (“SOUTH_WEST_FLOW”), that word is used as the facility's statement. Flow is stored separately from the runway sets and is never an assignment.

• Observation — one stored broadcast change (a D-ATIS text that differs from the previous one) or one tower TFDM configuration message. Quoted as “broadcasts recorded”.
• Configuration — the pair (arrival runway set, departure runway set) the facility states. A new information letter with the same sets is the same configuration.
• Configuration change — a recorded interval whose arrival set OR departure set differs from the previous interval. Counted once, at the start of the later interval. Today's timeline, the 24-hour / 7-day / 30-day windows and the rollup all use this one definition; an interval re-opened after a feed gap with identical sets is not a change.
• Coverage hour / day — an hour (airport-local day) during which at least one recorded configuration interval was in effect. Coverage, not calendar time, gates every statistic, and a window is called “Last 30 days” only when coverage spans it.
• Closure event — kept per source with its provenance: a D-ATIS-observed closure (broadcast days and first / last seen), a NOTAM-observed closure (the NOTAM number, official text, effective times, first / last seen, cancellation), and the FAA NASR baseline. They are shown side by side and never merged into one anonymous event.

The page prefers a stored rollup only when it was computed after the newest recorded interval started or ended; otherwise it computes from the same intervals live, so a configuration change can never show in one module and not another.

Configuration history is computed over the last 24 hours, 7 days and 30 days from the recorded events (refreshed every 3 hours by a rollup; the page computes live when the rollup is older than 4 hours). A “meaningful observation” is one recorded configuration interval weighted by its duration. Nothing is shown below these bars; the module says “Insufficient history” with the hours collected so far instead:

• Usage shares: 12 recorded hours (24-hour window), 48 hours (7 days), 168 hours (30 days).
• Use by wind direction: 24 hours in that wind sector, and at least 5 covered days at the airport.
• Use by time of day: 7 covered days and 20 hours per daypart; the hour-of-day chart follows the same rule.
• Flow shares: 48 recorded hours. Configuration changes per day: 3 covered days.
• The runway prediction cites recorded use under similar wind only when that wind sector has 24 hours on record.
• Seasonal and monthly patterns: 180 covered days (not yet shown anywhere).

Every history block is labelled observed runway usage and names the source (tower TFDM or D-ATIS), the hours behind it and the date it runs through. Wind sectors use the wind reported when each configuration began. Patterns by time of day are described, never explained, unless an official source establishes the cause.

Official PIREPs from aviationweather.gov (past 6 hours, within 100 NM) are attached to a runway page only when the raw report names the runway explicitly (“FINAL RWY 8L”) and the report is tied to the airport by its identifier in the text or by a position within 15 NM. The original report is always shown. The layout diagram is drawn from the FAA NASR runway-end coordinates: the selected runway is highlighted, the ends the facility states in use are marked at their own threshold (never the reciprocal), closed runways are hatched, LAHSO hold-short points are dotted and the current METAR wind is drawn. It is a reading aid, not a chart; the official FAA airport diagram is linked beside it.

SWIM content is received under the FAA SWIM Cloud Distribution Service terms for non-operational use; ATIS Alerts is not an FAA product and nothing shown here is FAA data or a clearance. Recorded use begins the day we began listening (September 18, 2026 for D-ATIS; earlier for TFDM where the tower's start time predates it). Runway NOTAMs have streamed from the FAA NOTAM Management Service since September 19, 2026 with no initial load, so a NOTAM issued before that and never re-issued may be missing and no page reads “open” from the feed. FAA NASR data is public domain and refreshed every 28 days. Declared distances are scheduled for a later phase. Always verify runway assignments, closures, equipment status and weather with official sources before flight.

Questions or corrections: the contact page. See also the about page for how the rest of the site is built.