This is a 7.5 miles distance from the airport.
This seems quite a big positional error (IMHO);Even with MLAT. Is there a description somewhere of the precision available with MLAT ?
MLAT estimates distance by measuring the time difference of a radio signal received at different locations. It requires precise time-stamping, usually derived from GPS signals.
Radio waves travel at 300 000 km (300 000 000 metres) per second. An error of just one microsecond (1 * 10^-6 seconds) in the time measurement could equate to an error of [typo edited] 0.3 km (300 metres) in the estimated position.
Reflections, multipath reception and other effects can have a great influence on the reception of radio signals as could the system used for processing the signals.
Usually positions which are substantially wrong like this are a result of poor receiver geometry more than anything (there are some areas of the solution space where there are two equally plausible solutions and which gets picked is very sensitive to small errors in measurement; this is mostly only an issue when there are few degrees of freedom in the solution, e.g. few receivers and no altitude data - as with an aircraft on the ground)
It also occurs to me that what you may be seeing (I can’t easily tell from the screenshot) is an old mlat position, coupled with a recently seen direct message after landing. In this case skyview will show the old position with the new altitude. In the tracklog above we lost mlat coverage a couple of minutes before landing.
Yeah you got it ! There are not enough mlat sites to track the plane down to the ground. So the last known position is used to report plane on the ground.
I just checked a mlat flight (GLG71TG) which just landed on either LFLL or LFLY, but is shown somewhere in the countryside :).
The point is you only notice this, if you happen to check the airplane position within 1 minute of its landing, since it stops reporting, and just vanishes from the map afterward.