Local pilot — real public weather, not a synthetic demonstration. Unlike the rest of this dashboard, this page makes a real, live connection to a public weather source for the airports listed below. Separate, clearly labelled sections also show a hypothetical flight compared against that same real weather, and fabricated test cases that never touch the network — each section states which one it is. Local pilot only: no live training, no automation, no telephony or admin system is touched, and nothing here is a validated forecast-accuracy result or an endorsement by any aviation authority.
Ceiling & Visibility Outlook
A bounded, local pilot: for 3 real public airports, whether the official TAF forecasts a worse ceiling/visibility flight-category within the next 60 minutes. Deterministic interpretation of NOAA's own decoded fields — no trained model, no invented probability. Back to Predictive Analysis · Review guide.
Local pilot, not a production feature. Real public weather from NOAA's Aviation Weather Center; no telephony, admin or training system is touched by this page. See "What this page cannot tell you" below before treating anything here as advice.
Real public weather — live outlook
From source reviewThe current TAF for each station, evaluated for the 60-minute window starting now. Refetched at most every 10 minutes; a reload sooner shows the same cached fixture.
New York/JFK (KJFK) — next 60 min from now
No crossing
Base forecast category in this window: MVFR.
TAF KJFK 291444Z 2915/3018 33007KT P6SM BKN015 BKN050 FM291600 31007KT P6SM SCT020 BKN030 TEMPO 2916/2920 BKN020 FM292100 24006KT P6SM BKN040 FM300100 29005KT P6SM BKN050 FM300400 33004KT P6SM BKN050 FM301100 33004KT P6SM SCT250 FM301600 22006KT P6SM FEW250
No outcome shown: either this window has not happened yet, or no observation was available after it.
Persistence baseline
Unavailable: this window has not happened yet.
Chicago/O'Hare (KORD) — next 60 min from now
No crossing
Base forecast category in this window: VFR.
TAF KORD 291454Z 2915/3018 20008KT P6SM SCT250 FM292300 17007KT P6SM SCT150 BKN200 FM300600 20007KT P6SM SCT060 BKN100 PROB30 3008/3014 6SM -SHRA BKN060 FM301500 20008KT 6SM -RA BR SCT020 BKN050
No outcome shown: either this window has not happened yet, or no observation was available after it.
Persistence baseline
Unavailable: this window has not happened yet.
Dallas/Fort Worth (KDFW) — next 60 min from now
No crossing
Base forecast category in this window: VFR.
TAF KDFW 291446Z 2915/3018 18008KT P6SM FEW020 BKN250 FM291600 16016KT P6SM FEW070 SCT250 FM300800 18015KT P6SM BKN015 BKN250 FM301700 16016KT P6SM SCT025 BKN180
No outcome shown: either this window has not happened yet, or no observation was available after it.
Persistence baseline
Unavailable: this window has not happened yet.
Real public weather — a recent example with a real outcome
From source reviewThe same evaluation, for a fixed recent date (updates once a day, UTC noon, so this stays a stable, cache-friendly example rather than a new lookup on every page load), so real later observations exist to compare against — a demonstration of the outcome comparison, not a live reading.
Chicago/O'Hare (KORD) — 60 min from 2026-09-28T12:00:00Z
Crossing possible (TEMPO/BECMG/PROB)
Base forecast category in this window: VFR. Possible (TEMPO/BECMG/PROB) category in this window: MVFR.
TAF KORD 281129Z 2812/2918 34006KT P6SM FEW100 SCT250 TEMPO 2812/2813 5SM BR FEW003 FM281500 05007KT P6SM BKN250 FM290600 VRB04KT 6SM BR FEW003 TEMPO 2909/2913 2SM BR SCT003 FM291500 20006KT P6SM SCT250
Outcome (real later observations, never fed back into the forecast above)
Worst observed category across the checks: VFR.
1 distinct observation(s) checked, up to every 20 minutes, strictly after the cutoff. This is NOT continuous coverage: a shorter excursion between two checks would not be captured. Only 1 distinct observation(s) were actually available for the 3 check(s) requested (METARs are not necessarily published every 20 minutes, so nearby requests can resolve to the same observation).
Persistence baseline (holds the last observed category unchanged — the simplest possible comparator, never a real forecast)
At the cutoff, the most recent observation showed VFR (2026-09-28T11:51:00Z UTC, 9 minute(s) old at the cutoff, receipt-verified).
Distinct from this card's own "Base forecast category" above: one is what the forecaster stated for the period, this is what was actually observed at the cutoff — they can disagree, and are never merged into one figure.
Simulated flights, real weather
SimulatedNever a real recorded call. A labelled hypothetical flight, evaluated against the same real, fetched weather the panel above uses. Simulated cases are never counted as real calls or real predictive accuracy anywhere on this dashboard.
Hypothetical: departing KJFK now
Hypothetical flight: KJFK to KORD, ETD now — never a real recorded call.
No crossing
Base forecast category in this window: MVFR.
TAF KJFK 291444Z 2915/3018 33007KT P6SM BKN015 BKN050 FM291600 31007KT P6SM SCT020 BKN030 TEMPO 2916/2920 BKN020 FM292100 24006KT P6SM BKN040 FM300100 29005KT P6SM BKN050 FM300400 33004KT P6SM BKN050 FM301100 33004KT P6SM SCT250 FM301600 22006KT P6SM FEW250
No outcome shown: either this window has not happened yet, or no observation was available after it.
Persistence baseline
Unavailable: this window has not happened yet.
Hypothetical: KDFW pattern work planned ~20h ahead (near the edge of current TAF validity — may legitimately show "no forecast covers this window")
Hypothetical flight: KDFW local pattern work, ETD in 20 hours — never a real recorded call.
No crossing
Base forecast category in this window: MVFR.
TAF KDFW 291446Z 2915/3018 18008KT P6SM FEW020 BKN250 FM291600 16016KT P6SM FEW070 SCT250 FM300800 18015KT P6SM BKN015 BKN250 FM301700 16016KT P6SM SCT025 BKN180
No outcome shown: either this window has not happened yet, or no observation was available after it.
Persistence baseline
Unavailable: this window has not happened yet.
Synthetic stress tests
Simulated Not measuredWholly fabricated — never touches the network, never a real airport reading. Hand-built edge cases (a TEMPO at the window edge, a TAF validity gap, a period with unreported visibility) used only to exercise the interpretation logic's own edge cases, kept in tests/wx-outlook-php.php too. Never presented as real weather anywhere else on this page.
TEMPO straddling the window boundary
Crossing possible (TEMPO/BECMG/PROB)
Base forecast category in this window: VFR. Possible (TEMPO/BECMG/PROB) category in this window: LIFR.
No outcome shown: either this window has not happened yet, or no observation was available after it.
No TAF period covers the window at all (a validity gap)
No forecast covers this window
Base forecast category in this window: unknown.
No outcome shown: either this window has not happened yet, or no observation was available after it.
A BECMG period with missing visibility (unknown, not assumed clear)
Category unknown or coverage incomplete
Base forecast category in this window: unknown.
Part of this window has no TAF period covering it at all. A BECMG (becoming) group overlaps this window. This pilot does not read NOAA's own timeBec (transition-completion) field, so whether/when that transition had actually completed relative to the window is unresolved — full coverage cannot be certified for this window even when a base period also nominally spans the same time.
No outcome shown: either this window has not happened yet, or no observation was available after it.
What this page cannot tell you
Not measuredWhether to fly Not measured
This is a forecast-category comparison, not an operational go/no-go decision, and carries no endorsement from the FAA or any other authority. Every real weight of judgement about a flight belongs to the pilot, using official sources directly.
What would measure it: A licensed dispatcher or the pilot's own review of the current official products.
Forecast accuracy over time Not measured
A historical observation (from NOAA's METAR archive, the same family of data as NCEI's Integrated Surface Database) shows what the weather WAS — it is an outcome, not proof that a matching forecast was actually available to read at decision time. This pilot only evaluates the TAF it can itself confirm was already issued by the cutoff (see the "Issued" row on every real card above); it does not claim a full retrospective verification study.
What would measure it: A dedicated forecast archive with confirmed publication timestamps for every historical product, which this pilot does not build.
Anything beyond ceiling and visibility Not measured
Wind, icing, turbulence, convective activity and every other TAF element are visible in the raw text shown on each card, but are not scored for a threshold crossing here.
What would measure it: A dedicated hazard-crossing outlook, not built in this pilot.