NavList:

   

Reply

----- Original Message -----
From: "Richard B. Emerson" <navsys@XXX.XXX>
To: <NAVIGATION-L@XXX.XXX>
Sent: Wednesday, May 03, 2000 11:07 AM
Subject: Re: GPS Accuracy Now.
> Roger M. Derby writes:
>  > Thus when aircraft travel into a low pressure
>  > area, all of them descend.  When leaving the low pressure area, they
all
>  > climb.  A similar deviation occurs flying into an area of cold air.
One
>  > wouldn't want to have a mixed system in use.
>
> Scuse me?  It's been a long time since I was a pilot in command but I
> learned very early on that he who doesn't keep his altimeter set for
> local pressure is begging for a "controlled descent into terrain".
The ups and downs occur at and above flight level 180 (18,000' MSL) where
all altimeters are set to 29.92 mm-Hg or equivalent.  To preserve
separation, FL 180 doesn't exist when the pressure is below 29.??  since the
guy at FL 180 might meet the VFR pilot at 17,500' who's using local
altimeter settings.
> Anyway, I leave it to all the "alphabets" to come up with a scheme for
> resolving GPS and pressure altitude compatibility and integration
> issues.  My guess is, however, pressure and radar altimeters will
> remain the standard with GPS altitude being applied only in
> approaches and departures.
Radar altimeters are only useful  in the last few hundred feet AGL and they
don't look ahead, sort of like depth finders.  Binghamton, NY, Hazard, KY,
and many other cities, site their airports on flattened hill tops.  One mile
out you're hundreds of feet above the terrain.  Then you're hauling back to
climb over the fence at the airport boundry.  Hazard also surrounded their
field with a net of high tension lines.  It was well named.  (I think its
been moved.)
> As to the geoid definitions, I beg to differ there, too.  There are a
> number of geoid description systems in use around the world because
> GPS is capable of resolving the differences in the Earth's shape
> (i.e., it's not regular in any direction) compared to a regular,
> mathematically defined shape.  WGS84, for example, works well for
> the continental US.
GPS reports your altitude relative to the mathematical geoid and local lumps
in the real one can make a big difference to your serenity.  Don't give me
RMS error, I need to know the peak deviation to have confidence.
Do you have any idea as to what precision is used in the internal
calculations?  I've seen some government data that included huge cliffs
where segments switched; e.g. moving from data derived from one transverse
Mercator chart to the next one  (NTDC, Orlando, FL, ca 1972).  I've also run
into staircase effects, particularly in trig functions, if there weren't
enough bits being used.  (The now defunct SEL shipped a 16 bit math library
with their 32 bit SEL 840.  It made for very strange graphics.)  I spent the
last seven years before retirement in a "software quality assurance" role
and I left with a bad feeling for the state of the art.  Schedule and cost
are paramount and there's no glamor for the programmer in designing an
exhaustive set of tests just to show that he or his buddy overlooked
something and has to start over.  On the F-22, the Software Verification and
Validation function was defined as complete by administrative fiat, long
before the production software flew.  It was having an adverse effect on
some managers' careers.  There was also a considerable misuse of "security
concerns" to insure people didn't ask embarrassing questions.  Let's hope
none of the overlooked bugs results in a bomb landing in Central Park or
another missile launch at an airliner.
That said, I have been moderately pleased with the vertical navigation
function on my Garmin GPS 90.  You slap it up the side of its head to make
its initial altitudude agree with your opinion (and altimeter), and it then
tells you when to start down and how far above or below the chosen descent
path your are.  Useful for preserving one's sinuses while getting under the
TCA layers in a timely fashion.  The number of menus involved and the ease
with which it gets turned off result in its getting used mainly on long
boring flights.
Roger

   

Reply