Originally Posted by
Pilot DAR
What remains unclear to me, is the "groundspeed" being reported, two dimensional, parallel to the surface of the earth? Or three dimensional over the earth?
I know very little about ADS-B, but I do have some training in geodesy, so may be able to add something useful.
There are multiple ways of measuring speed from a GNSS. One is by comparing two positions and another is by looking at the shift in the GNSS signal itself, which will give a velocity relative to the satellite (and therefore will include speed in a vertical dive).
Assuming the first, the difference in position can be calculated using spherical geometry to give a line along the spheroid (i.e. WGS84 for GPS) and converted to a speed either by using the spherical distance or the straight line, depending on which approximation you consider appropriate.
Or you can convert the positions to Cartesian coordinates and work out the distance using 3D Pythagoras.
Only if you decide to ignore height and use 2D Pythagoras will get the "horizontal" component only.
It is unlikely that you'll get speed parallel to the ground (as opposed to a mathematical approximation of the ground) just because that requires a surface elevation model.
I don't know how real world ADS-B transponders do it.