Bennett, Matt wrote:

> I think EPE (from Jim's message below) stand for estimated position error,
> which, if my logic holds, is related to the GDOP- Geometric Dilution of
> Precision.

That's correct (Jim)

> The more SVs in good position, the better your solution would be
> (lower EPE and GDOP).  Horizontal error is usually about half of the
> vertical error, due to the geometry of the position solution.

Right again.  Most of the excursions were directly associated with satellite
switch-outs. The gentle 1-2 foot oscillations that were going on continuously
were probably associated with atmospheric travel time due to turbulence,
temperature, and pressure changes along the paths.

> Is that enough acronyms for ya?

I apologise for using the term EPE without explaining it.  I thought it would be
obvious from context.

> The error number you see is not related to the presence or absence of SA,
> that error number is related to the accuracy of your solution from the
> geometry
> of the SVs (the smaller the included angle of all the SVs in your solution,
> the larger the error will be).  The excursions you see are probably due to
> SVs being switched in and out of your position solution.

Your second and third statements are true. Your first statement doesn't appear
to be quite inclusive.  EPE is calculated as you say, but it is larger when SA
is turned on, I think because the timing errors throw additional positional and
calculated distance errors into the path length from each satellite.  Haven't
given it much thought though.  As I said, I was making an idle observation.