GPS altitude

[rit]

GPS altitude shown in MCDU of A320 - is it measured from ground or MSL?

[beerdrinker]

Surely if it is showing "Altitude" it must be AMSL. If it was AGL it would read "Height"

[Dan Winterland]

Altitude above the WGS84 sea level datum.

[Tim Zukas]

Slight clarification-- no such thing as WGS84 sea level. WGS84 uses a spheroid that approximates the sea level surface; here in California the spheroid is 30+ meters above sea level.

[beerdrinker]

So Tim, are you saying that if your altimeter in CA is set to the correct local Altimeter Setting (QNH to the rest of the world) the altimeter will read approimately 100 feet different from the Altitude reading on your GPS?

[Tim Zukas]

For all I know maybe GPS's nowadays include a worldwide map that gives the altitude difference between spheroid and sea level, wherever you are. If so, they could give a good approximation to height above sea level. I'm just saying such an addition to WGS84 (based on thousands? millions? of additional measurements) would not be WGS84 itself.

Or put it this way-- once you've determined sea level, it's the same sea level whether your GPS is set to WGS84 or NAD27 or OSGB36. The WGS84 spheroid won't be at the same height as the NAD27 or the OSGB36, so the map of altitude corrections will be different for each-- but they're all aiming at the same result.

[ATCast]

The GPS system uses WGS84 (http://en.wikipedia.org/wiki/WGS84) internally. WGS84 is an ellipsoid earth model. The GPS height with respect to that ellipsoid is called Height Above Ellipse (HAE). Unfortunately the Mean Sea Level does not coincide with this ellipsoid. In fact there are quite a few bumps and dents in the sea caused by local deviations in Earth's gravity field. The difference between mean sea level and WGS84 is modeled by another model : EGM96 (http://en.wikipedia.org/wiki/EGM96). When this model is available, it is easy to transform the HAE in height above MSL. I am not sure whether this is done in the Airbus, but I would expect so.

Now if you are on the ground, and your altimeter setting is local (QNH) of course your altimeter should read the local elevation, and so should your GPS if it's using the EGM96.

However, in the air it's a different story. Your altimeter derives altitude from ambient pressure using formula's from the ICAO Standard Atmosphere. When the atmosphere is different from the standard, and it usually is, then the altitude shown by your altimeter is different from the GPS height above MSL.
For example, on a hot day air expands, and so the "barometric foot" is longer than the "GPS foot", in other words on a hot day the GPS will read a higher number than your QNH tuned altimeter.

I hope this explains the different readings.

ATCast

[malcolmf]

GPS altitude measurement is also 1.5 times less accurate than horizontal, so if you have 5m horizontal accuracy it could be out by 7.5m. Altogether not good for RVSM.

[Bushfiva]

I'm not au fait with the aeronautical GPS world, but wasn't WAAS designed with aviation, and particularly approaches, in mind? I believe over the US, WAAS-equipped GPS can be accurate to within 2 meters vertically and 1 meter horizontally.

[kijangnim]

Greetings
GPS Altitude is true altitude, however it needs to be corrected from Geodic error.

[Microburst2002]

ATCast

Very interesting post. Never heard of that model before. I wonder if EGPWS systems use these models. And I guess it will be used in GPS precision approaches.

[kijangnim]

Greetings,
Scientists have redefined level Zero as being the point where Gravity is equal, at sea level it works, but it gets complicated on mountaneous area, because gravity is measured perpendicularly to the surface, and unfortunately a mountain has a slope....So it offsets the level by sometimes 700ft or more.
Geodic correction software are available on the net just google it:ok:

[stilton]

So why not replace Barometric altimeters with GPS data derived ones ?


I'm probably missing something but if this could be done you would eliminate pressure data derived errors and there would be no need for transition altitudes / levels.


You could retain the standby pneumatic altimeter as a back up.

[ATCast]

Stilton:
I'm probably missing something but if this could be done you would eliminate pressure data derived errors and there would be no need for transition altitudes / levels.


You could retain the standby pneumatic altimeter as a back up. Since GPS derived altitude and pressure derived altitude give different values it is not possible to replace the pressure altimeter with GPS. The GPS altitude and pressure altitude are just not compatible. If your GPS would fail, you can't fall back on the barometric altimeter, because you would end up at a different altitude. That is simply not acceptable in a crowded airspace.

Consider the following:
The barometric altimeter is very sensitive to vertical deviations (think RVSM). Unaugmented GPS is far less accurate in the vertical plane. For vertical separation the barometric altimeter is the preferred solution.
The barometric altimeter does not require a power source. It helps to have passive instruments you can rely on in case you loose electrical power. (again think RVSM).
GPS is dependent on satellites, the barometric altimeter on the presence of the atmosphere. Atmosphere has proven to be more reliable than satellites.
GPS measures absolute altitude, the barometric altimeter assumes Standard Atmosphere. For your height above the hills, obstacles, runway etc, a reliable GPS can give a solid indication. Barometric altimeters depend on the correct setting (QNH), and suffer from temperature induced errors, so can be less accurate when it comes to absolute height.So when it comes to vertical separation between aircraft, the barometric altimeter is and will always be the preferred solution. For operations where the vertical distance between the aircraft and ground is of the essence, (augmented) GPS may become an alternative. But still you will always need the altimeter as a backup.


ATCast

[MD83FO]

what temperature limits are built into procedure design?
is this info on 8168?
thanks.