Date: 03 May 2000 11:26:02 -0400
From: Derek Atkins <[EMAIL PROTECTED]>
Do you mean 13 feet radius or 13 feet diameter? I was seeing a
vertical error of approximately +/-10 feet (although I'd believe 13 ;)
Horizontally I was seeing approx +/- .0015 minutes in both lattitude
and longitude (which equates out to approximately +/-9 feet).
Considering the inacuracy of my measurements, I would mostly concur
with your observations, except that I'd extend it to a sphere (instead
of a circle) of radius 10'. I can't wait to get up flying again :)
Remember that the exact error you see is a function of the geometry of
the satellites in your line-of-sight at the instant that you compute
the fix. Most GPS's take this into account when showing you your EPE.
For example, in the limiting case, you may only be able to see 3-4 birds
which are all approximately colinear in the sky above you. In this case,
your position error normal to the "line" made by the birds will be quite
large, and so your EPE won't be very good. This happens surprisingly
often if you can only see a few satellites. Of course, given a more
reasonable geometry, your EPE should be pretty small.
Also, recall that altitude fixes are much harder to get right. An
intuitive way of visualizing this is that surface fixes can essentially
"see" 180 degrees of sky, from horizon to horizon, with you in the
middle of that, but an altitude fix only sees birds above you and never
below (since they'd be below your horizon). This makes things harder.
[Yes, this description will give anyone who actually knows the math
behind GPS fits, not to mention that all the "lines" I'm talking about
are embedded in sphere, etc. I'm being approximate; this is not the
list to go through all the math in detail...]
