First, I wish to complement the IGC for it's work on GPS flight loggers.
They have made a huge and favorable impact on soaring performance and cross
country flying. The fact that the availible certified units aren't cheap
enough to satisfy all is just an artifact and not at all the intention of
the IGC. I'm sure everyone in the IGC hopes time will bring cheaper units.

For those pilot wishing to attempt a badge flight, just ask to borrow a
logger from a fellow pilot. My Volkslogger is available for the asking - as
long as I know you.

Now, some rambling thoughts.

Both pressure altitude and GPS altitude sensors suffer a "signal to noise"
ratio but of a different character.

In the case of pressure sensors, the main source of "noise" is the synoptic
distribution of highs and lows on the weather map. This noise signal
changes slowly over time scales of hours and distances of hundreds of miles.
There are errors due to differences in the temperature of the atmospheric
collumn and the Standard Atmosphere which change slowly on a daily cycle.
There may also be a shorter time scale noise source related to the static
source or cockpit pressure although this is small compared to synoptic
pressure changes.

A GPS receiver, by nature of being a radio, suffers noise of a different
type. 'Static' is short term noise on the scale of milliseconds. There is
also some noise from the slowly changing geometry of the Navstar satelites
but this is predictable, thus not really 'noise'. There's also the issue of
each system using a different sea level reference or "datum"

Both GPS altitude and pressure altitude are both "right" and "wrong" for
different reasons.

Sometimes, a clever engineer can fuse data from two different sources in a
way that uses the best of each and cancels out the worst. This is seen in
fusing data using a Kalman Filter from a GPS receiver and an Inertial
Measurement Unit. GPS, is a 'position finding system' and the IMU is a
'position keeping system'. GPS keeps the IMU honest and the IMU smooths out
the GPS signal. The result is basically a huge "signal to noise"
improvement.

Simularly, fusing pressure data with GPS altitude data could work the same
way. For example, a millisecond scale spike in GPS altitude, not confirmed
by presure data could be ignored with impunity. A slow drift of the
pressure signal from the smoothed GPS signal is almost certainly a synoptic
pressure change and can be canceled by the long term stability of the GPS
signal making the pressure signal more useful in detecting airspace
incursions. (Think automatic Kollsman settings.)

The differences between sea level references and temperatue differences from
the Standard Atmosphere can be subrtacted out producing an altitude signal
referenced to whatever datum you choose.

Of course, all this ignores the question of just what it is you want to
measure. If it is deemed desirable to have an altitude logging system for
future flights that is directly comparable to past flights that used a
barograph, then including GPS data may not be desirable.

Does any of this have something to do with reducing the cost of loggers?
Maybe. If two low cost sensors fused together can produce superior data to
one high cost sensor, the result might be a cheaper, more accurate logger.

Bill Daniels