—

Counting movements since you opened this panel — the hourly rate is an extrapolation and firms up the longer it runs.

Positions are advanced between fixes from each aircraft's own track and groundspeed, so the picture moves continuously instead of jumping every poll. Gold dashes are the extended centreline of the runway in use.

A standard ILS glidepath is 3° — about 318 ft per nautical mile from the threshold. Aircraft are plotted at their real altitude against that line.

Being below the path far out is normal, not an error: aircraft level off at an intercept altitude and join the glidepath from underneath. What matters is being high and close in — that's the one that ends in a go-around. Read the deviations near the threshold, not at 12 nm.

Field elevation isn't carried in this feed, so the path is drawn from sea level. At an airport like Denver the whole picture shifts up, and the deviations rather than the absolute heights are what to read.

In-trail spacing between consecutive aircraft in the same approach stream. Airports like LAX land two parallels at once, so aircraft are grouped by their offset from the centreline first (streams A, B…) — otherwise two jets side by side on parallel runways read as a 0.6 nm violation, which is nonsense. Minima come from the ADS-B wake category of the pair (leader → follower). Red is below the applicable minimum. Note we still can't resolve L from R within a pair of closely-spaced parallels.

Everything here is computed live from the ADS-B feed and the METAR — nothing is scheduled data.

Runway in use is inferred by averaging the ground tracks of aircraft established on final (3–12 nm out, where the track really is the localizer). ADS-B track is true north while runway numbers are magnetic, so a variation correction is applied — without it JFK reads "RWY 21" while it's actually landing 22. It's still an inference from a handful of aircraft, so it can land one number either side of the published designator, and L/C/R cannot be resolved from track alone. Treat it as "roughly which way they're landing", not as an ATIS.

Movements are counted from the moment you open this panel, so the hourly figure is an extrapolation that firms up the longer it runs. There is deliberately no delay estimate: real delay needs scheduled times, which this feed doesn't carry. Inbound count and holding aircraft are shown instead — measured facts rather than a guess dressed up as a number.

Wake spacing uses the ADS-B emitter category (plus a type list for Super/Heavy), and applies leader→follower radar minima. Aircraft not transmitting a category default to Medium.