How to read the brief
Aerodrome brief
The brief answers one question for the next 5 days: is there a window at this field worth a closer look? It scores the weather against a plane's limits, VFR by day. It does not replace a real brief, the TAF or a look at the sky.
How an hour becomes a colour
Every hour is judged on 8 constraints, one at a time, and the worst one decides the hour. Its letter is what you see on a strip cell from marginal upwards. If no model can judge a constraint for that hour (a gust figure nobody trusts, a model without visibility), the brief leaves it out instead of guessing and puts a corner mark on the cell.
- go
- marginal
- poor
- no-go
- dark
- already past
- D daylight
- X crosswind
- T tailwind
- W wind
- G gust
- C ceiling
- P precipitation
- V visibility
- dashed outline = models disagree
- corner mark = no information on some constraint
- letter = what decided the hour
Bands at the default limits
| Code | Constraint | Default | go | marginal | poor | no-go | From the limit | A plane may set |
|---|---|---|---|---|---|---|---|---|
| X | crosswind | 10 kt | ≤ 6 kt | ≤ 8 kt | ≤ 10 kt | > 10 kt | go 0.6 x, marginal 0.8 x the limit | 3-30 kt |
| T | tailwind | 2 kt | ≤ 1.4 kt | ≤ 2 kt | - | > 2 kt | go 0.7 x the limit | 0-15 kt |
| W | wind | 15 kt | ≤ 10.5 kt | ≤ 15 kt | - | > 15 kt | go 0.7 x the limit | 5-50 kt |
| G | gust crosswind | 10 kt | ≤ 8 kt | ≤ 10 kt | - | > 10 kt | go 0.8 x the crosswind limit | 3-30 kt |
| G | gust spread | 10 kt | ≤ 7 kt | ≤ 10 kt | - | > 10 kt | go 0.7 x the limit | 3-30 kt |
| C | ceiling | 2000 ft | ≥ 2000 ft | ≥ 1200 ft | ≥ 700 ft | < 700 ft | marginal 0.6 x, poor 0.35 x the limit | 500-5000 ft |
| V | visibility | 10 km good, 5 km floor | ≥ 10 km | ≥ 7 km | ≥ 5 km | < 5 km | marginal at floor + 0.4 x (good - floor) | 1-50 km good, 0.8-20 km floor |
| P | precipitation | - | ≤ 0.1 mm/h | ≤ 0.5 mm/h | ≤ 2 mm/h | > 2 mm/h | fixed, not a plane limit | - |
The bands are not set by hand. Each one comes from a limit, so a plane with its own limits moves every band with them. Pick the plane in the Limits popup on any airfield page. Without a plane you see the defaults above.
Crosswind and tailwind are taken on the favoured runway, the one with the most headwind. Model winds are degrees true and runway numbers are magnetic, so the runway is converted before any component is worked out. Wind (W) is the mean wind from any direction.
The gust is scored twice: its crosswind component, and how far it rises above the mean. Either one failing is a no-go.
Cloud base is an estimate: (temperature - dew point) x 400 ft, read beside the low-cloud cover. With 50% or more low cloud the base counts as a ceiling. Under 25% it is ignored. In between it is scattered and scores one category better than a deck would.
Precipitation is the rain rate. CAPE (on its own bands: 200, 500 and 1000 J/kg) can raise it only while rain is falling, because that is when showers turn convective. On an hour with little or no rain, CAPE above 500 J/kg nudges it to marginal and no further.
Visibility between the floor and the good value is a judgement call, not a no-go.
Daylight (D) closes the hour at the landing deadline: sunset minus the margin, 60 minutes by default. An hour cut by sunrise or by the deadline is scored on the part of it that is usable. Hours closed by darkness alone are drawn grey, not red. Darkness is not a weather problem.
Windows and day labels
A window is a run of at least 2 hours in a row, none worse than marginal, inside daylight. Each day is labelled by the best window it has: ALL DAY for 7 hours or more (or all of a short day's daylight bar one hour), MORNING when the window is over by 13:00, AFTERNOON when it opens at 12:00 or later, MIDDAY for the rest, and NO WINDOW. The hours are local time at the field. A day with a clear morning and a thundery afternoon averages out to a mediocre day, which helps nobody, so the brief names the window instead.
Temperature sits beside the window. It says whether the day is pleasant, and it is never scored.
Three models, never averaged
ECMWF, ICON-EU and GFS are scored separately and the hour takes the median, so a single outlier cannot close a day on its own. The headline crosswind, cloud base and visibility are the median model's, and the day page lists each model's own figure, so a disagreement stays visible.
When the models are two or more categories apart the hour is disputed and gets a dashed outline. A day where enough hours are disputed says so on its card: read its timing as provisional.
Why days further out look faded
They are less certain, and a grid that draws D+4 as boldly as D+1 would be claiming otherwise. From the first lead without hour detail, the cards drop the hour-level facts and keep the character of the day.
| Lead | Good to | Hour detail | Crosswind good to about | Drawn as |
|---|---|---|---|---|
| D+0 | today | yes | 2 kt | |
| D+1 | hour | yes | 2 kt | |
| D+2 | half-day | yes | 3 kt | |
| D+3 | part of day | no | 4 kt | |
| D+4 | character only | no | 5 kt |
These confidence figures were measured at EVAD and are used for EVAD, EVJA, EVLA and EVVA until the other fields have enough verified forecasts of their own. Each field's Accuracy page carries its current numbers.
Gusts treated as no information
Model gust fields fall apart in light wind, at any lead. A model gust is not trusted when it is more than 2 x the mean (on a mean of 3 kt or more), when it claims more than 10 kt over a mean below 3 kt, or when it is below the mean. Such an hour has no gust information. It gets the corner mark and a TAF check, and the gust never closes the hour on its own.
A TAF that covers the hour without a gust in any of its groups overrules the model: a forecaster looked at the same air and did not call it gusty. From D+4 the model gusts count as no information at all.
A METAR reports a gust only when it beats the mean by 10 kt or more, so a report without one means the gust was somewhere below that, not zero.
These gust rules come from what was measured at EVAD and apply to EVAD, EVJA, EVLA and EVVA for now.
What it cannot tell you
The ceiling, where nothing measures it. EVAD has no cloud report of its own, and the nearest METAR does not stand in for the field. Every cloud figure in the brief is a model estimate, and each page that shows one says so.
NOTAM. Military areas and other restrictions stay a manual check before every flight. The brief also scores the field you are looking at, not your destination.
Sunrise and sunset
Sunrise, sunset and the landing deadline are computed with the NOAA solar algorithm for each field and date. They are not taken from the AIP tables, so allow for a minute or two of difference.
When this updates
Model data, METAR and TAF are collected every 3 hours, at 02:40, 05:40, 08:40, 11:40, 14:40, 17:40, 20:40 and 23:40 (UTC).
The runs sit at an odd minute on purpose. A TAF can take half an hour to reach the wire after issue, and the models land hours after the cycle they are named for, so a run on the hour would keep missing them by minutes.
If the newest run is more than 7 hours old, the longest gap between runs plus some slack, the page turns red and says it is stale. It keeps the last good forecast rather than showing nothing.
Scoring rules version 2.0.0. All airfields