Ian,

I and many others appreciate your long, dedicated service and
willingness to interact on this forum. Please take these questions as
genuine requests to understand better the issues involved, not as an
attempt to undermine your position:

I am afraid that GPS altitude as recorded in IGC flight data files,
has been shown to be unreliable for accurate measurement purposes,
compared to traditional pressure altitude calibrated to the ICAO ISA.

How are "reliable" and "unreliable" defined here? Given that there is
no independent verification of either method of determining altitude
(pressure and GPS)--at least in IGC files--I assume you use some
measure of the number and severity anomalies. Could you elaborate?

Pressure altitude is extremely reliable in IGC files and has a
negligible anomaly rate. *In comparison, a significant proportion of
IGC files have GPS altitude anomalies, some small but some large. *

Could you give be more specific? Does reliability involve any
comparison of the two methods to cross check each other? That is, if
there's a sudden change in pressure altitude, is that presumed to be
in error solely because of the severity of the change or also because
the GPS altitude didn't change in a corresponding manner (and vice
versa)?

In addition there are quite a few examples of major GPS
altitude anomalies in IGC files, involving differences from pressure
altitude not just of tens or hundreds of feet, but occasionally
thousands.

Same question: is an anomaloy defined as a discontinuity in the trace
of a GPS or pressure altitude record, or a difference between the two
traces, or both? The above statement seems to indicate that the
difference between the two is taken into account. If so, are you
comfortable with judging the accuracy of one method by assuming the
other is correct?

Another factor is that, even in ideal conditions, due to the geometry
of SatNav position lines making up a fix, altitude accuracy will
always be poorer than Lat/Long accuracy by a factor of about two.

What does this translate to in feet or meters, or in percentage terms?
And how does that compare with known accuracy of pressure altitudes?

A report on this was made to IGC in the year after the Selective
Availability accuracy degradation was removed on 1 May 2000. *This
analysed many IGC files and came to the conclusion above. *It was
posted on the IGC web site and may still be accessible if you look.
Many thousands of IGC files have been analysed since then and the
conclusion is still the same, anomalies in GPS altitude in IGC files
continue to occur. *I say "in IGC files" because that is where the
data is from, and more expensive GPS receivers with more sophisticated
processing probably would not show these anomalies.

Do these "more expensive GPS receivers" refer to some or all current
COTS receivers, or to something else?

If comparisons of pressure and GPS altitudes are made to determine
anomalies, are they based solely on the records in approved flight
recorders or also on comparisons of traces made with other GPS
receivers. For example, I carry a Garmin GPSMAP 76 as a backup and
have downloaded and compared its trace with that of my Cambridge Model
20 for a number of flights. There are differences, as you say, but as
a layman I have no way of knowing which is closer to being correct.
And I have not compared the GPS altitude from the Cambridge to the GPS
altitude from the Garmin. Do the studies to which you refer do this?

Fortunately, lat/long accuracy is not affected, the anomalies are
confined to GPS altitude. *You might think that GPS receivers would
process one 3D position and then extract Lat/Long and altitude from
the same process. *However, this appears not to be the way it is done
in the low-cost receivers used in IGC-approved recorders. *Lat/Long is
processed separately from altitude and perhaps the manufacturer uses
more sophisticated processing for Lat/Long because that is what the
majority of customers want.

This is an interesting speculation. How likely is it that if the
flight recorder manufacturers wished to reduce anomalies in their GPS
altitude records, they could do so? And (this is purely speculative on
my part) would you agree there has been a disincentive for them to do
so since the more accurate that GPS altitude is proven to be, the more
likely that they might lose their exclusivity in the gliding
market?

There is more, but the above summarises the reasons why IGC has, so
far, not added GPS altitude to the Sporting Code where accurate
measurements are required.

I know that years ago there was a strong, justifiable sense of
appreciation for the work that Dave Ellis and Cambridge did to make
flight recorders possible and affordable for many pilots. Without
implying that GFAC or the IGC felt any moral obligation in the past to
"reward" this dedication by excluding other types of flight recorders
or verification methods, do you believe the flight recorder market is
well enough established now that new entrants, whether they be COTS
manufacturers or soaring equipment builders, are viewed without any
consideration for the early days of flight recorders or the difficulty
of making a profit in the relatively small world of gliding? To be
more provocative, is there any concern over the fact that opening up
the market to COTS equipment now would almost certainly harm flight
recorder manufacturers?

Chip Bearden
ASW 24 "JB"
USA