"David W" wrote in message
ups.com...
And is it true that such aircraft's navigation systems use maps
'optimised' for GPS?

What does 'optimised' for GPS mean?

I'm not sure. Somebody with whom I am having a 'debate' has asserted
that there is an altitude-dependent error component (if I may call it
that) on positions determined by GPS (and I presume that this alleged
error component affects the horizontal component of a 3D position, as
well as vertical (altitude) component).

It is theoretically true that altitude can affect the geometry of the GPS
receiver and satellites being received in such a way as to increase error.
However, the orbital altitude (10,988 nautical miles) is so much higher than
flight altitudes (usually up to 9 or 10 nautical miles at the very most, and
for commercial airliners 6 nautical miles give or take is more typical),
it's hard to imagine any significant error being caused by that.

Even in the most extreme case, an airplane is only going to climb high
enough to change the distance to the satellite by 0.05-0.10%. And this
assumes the satellite is directly overhead; when it's not, the distances are
even greater and altitude even less significant.

It's important to note: the GPS system doesn't actually care about
elevation. It cares about distances and calculates a 3D position based on
measured distances (simplistically stated, anyway). If altitude was a
significant problem, then so would any variation in distance from the
receiver to the satellite, and the change in that distance due to the
relative angle in the sky of the satellite is MUCH greater than any change
in elevation possible by an airplane. If altitude caused a problem, the GPS
would be incredibly unreliable all the time, since the satellites are rarely
directly overhead.

Not that your debate partner will see this. He'll just say something about
the system being "optimised" for the effects of satellite orbits on the
distance between receiver and satellite. But it's true, nonetheless.

In his own words:

"GPS is optimised for sea level, Blanchefort [a mountaintop ruined
castle] is 467 metres above sea level, couple this with a slant range
to a satellite of several thousand miles and the curvature of the earth
and you have error. At least up to 100 metres..."

I still don't know what it means for GPS to be "optimised for sea level".
It would be entertaining (though probably not educational) to learn what the
guy believes was done to "optimise" GPS for sea level.

Certainly is boggles the mind to think that elevation gain results in a
1-to-5 error ratio (that is, 1 unit of error for every 5 units of elevation
gain). That's a remarkably fragile system he's describing, and it certainly
doesn't apply to GPS as it exists today.

[...]
His reply (with some non-essentials removed):

"Aircraft ... are using a map optimised to the GPS system and this is
the key point which seems to be passing you by."

More BS. "A map optimised to the GPS system"? Again, even pondering what
this could mean is bewildering. Is he saying that, while there's some large
error, it's always a known error and so can be compensated for with the map
system? Why wouldn't a GPS receiver simply be designed instead to use this
known error and correct the calculated position based on that?

Of course, that's all theoretical. In reality, there's no "known error"
(not in the sense that one knows exactly the magnitude AND direction of the
error), and no reason to "optimise" the map "to the GPS system". You get
the same map on the ground as you do aloft, and the GPS simply plots your
position on the map.

It's true that there are different mapping systems in use, but they have to
do with how one projects the geometrically perfect information from GPS onto
the geometrically imperfect planet we live on.

I hope that that is sufficiently answered above. I didn't want to
burden this group with extensive background information and endless
quotes from this slightly silly debate!

Yes, I'd agree it's at least "slightly silly". I usually just put
people who write stuff that ridiculous into my killfile.

Pete