On Fri, 23 Jan 2009, Recliner wrote:

"David A Stocks" wrote in message

"Graculus" wrote in message
...
(No, GPS isn't accurate enough to say if its on the up fast/down
slow/etc track.)
Should be - I get accuracy down to about a metre on a very old
hand-held GPS box.

How can you be sure that it's so accurate? I took my Garmin wrist GPS
to Greenwich Observatory, and was crestfallen to discover that my GPS
wasn't nearly as accurate as it claimed. As I recall, it claimed to be
accurate to something like 15', but I had to move about 100' east from
the brass strip before it thought it was at logitude 0 degrees exactly.

That might be because GPS is giving you a position on the WGS84 ellipsoid,
and the strip at Greenwich is zero longitude on some other datum (Airy
1830?), and the two don't coincide there.

Might be - i don't know the details of the Greenwich strip, or whether
ellipsoids are all defined so as to align at zero longitude, or whether
any difference might be 100 feet.

In any case, if GPS consistently puts the meridian exactly 100 feet east
of the strip, that would be accurate enough for trains, provided that the
coordinates they work with were as measured by GPS, rather than by eg
brass strips and Victorian gentlemen.

I also have doubts about the vertical elevation it reports, though
that's not an issue for trains.

Yes, i believe the altitude is typically much less accurate than the
horizontal position. Something to do with geometry.

tom

--
sh(1) was the first MOO