Chris W wrote:
It seems as though those who actually read what I was really asking
didn't think it was important to find or know the answer so let me get a
little more specific.

Hard to give a definitive response since there are quite a few
variables that aren't known to us. There's already been a discussion
of the various sources of discrepancy, so I'd look at each and decide
how large a value to assign, then add them up to get an overall
estimate. Those included:

1) Measurement uncertainty in the pressure sensor. Presumably the
manufacturer has a specification for this.

2) Measurement uncertainty in the GPS. Does it have WAAS? Without it
I've found that altitude is generally within about 35' 95% or more of
the time. With WAAS that should go down to about 20'. Both figures
assume a good skyview which you should have in your remote plane
provided the antenna is located properly. But note that there's still
that 5% of the time when the error could be greater.

3) Variations due to ground level pressure changes (i.e. Kollman window
settings). You indicate checking the value from the nearest measuring
points along the route so this depends on how far apart those are. But
more importantly it depends on how rapidly the weather is changing.
The uncertainly here could be quite large if you're flying even when a
front of thunderstorms/tornadoes is passing through - but I'd guess you
wouldn't fly in those conditions. Look at weather charts for a few
days with somewhat worse conditions than the ones you'll fly in to get
an idea of the expected variations.

4) Variations due to lapse/temperature deviations from the standard
atmosphere model. This will again depend on what range of conditions
you'll be willing to fly under, but also depends on the altitude - i.e.
not a big issue at 1000' AGL but would be more significant at the 6000'
level. But unless you fly in bad weather I wouldn't expect this to
change very much over the few hundred mile range that you mentioned -
especially at these relatively low altitudes.

5) The geoid vs. ellipsoid model effects. Do you know if your GPS
corrects for this (most consumer stand-alone models do, but you should
check on your particular one which I presume is a minimal board type)?
Early SiRF-based models generally didn't do the correction but the
later SiRFii and I think all SiRFiii models do have the correction
capability. Even if not, this varies pretty slowly over such a short
distance and you can use the website previously given to see how it
changes over the specific route in question.