Re: Terrain-following radar for airliners?

From:         rdd@cactus.org (Robert Dorsett)
Organization: Capital Area Central Texas UNIX Society, Austin, Tx
Date:         25 Aug 93 02:53:40 PDT
References:   1
Next article
View raw article
  or MIME structure

In article <airliners.1993.569@ohare.Chicago.COM> you write:
>There seem to have been many approach accidents due to the terrain
>being higher than expected by the pilot. These incidents include
>running into hills because the wrong VOR was tuned in, or the aircraft 
>descending more quickly than planned due to distraction from other
>things going on in the cockpit.
>
>I was wondering if civil aircraft designers have ever considered
>using terrain-following radar like some military aircraft do. 
>I understand that these point forward as well as down, thus causing
>the aircraft to pull-up in time if there is an obstruction. Apparently
>very high-speed fighters can fly safely at very low altitudes using
>this radar. 

Airliners can use their weather radar for terrain mapping near ocean bound-
aries, as cross-checks to navigation.  

However, flying into hills is just one aspect of the ground proximity problem.
The crash which inspired the mandating of ground proximity warning systems in
the United States involved an L-1011 which simply flew into a very level 
swamp in Florida, as the crew was distracted by a burnt-out lightbulb.

GPWS's tend to have five operational modes, called, intuitively, Modes 1-5.
These tend to have associated aural alerts, ranging from (in older airplanes)
"Whoop whoop, pull up!" to (in modern avionics suites) distinct configuration 
and terrain call-outs for every mode.

Mode 1 is a simple function of radio altimeter altitude and vertical climb
rate.  If you're closing too fast, an alert goes off.

Mode 2 is "excessive closure."  This is integrated over time: for instance,
suppose you're at zero ROC and the GPWS computer shows a rapidly deteriorat-
ing radio altitude: you get an alert.

Mode 3 is "altitude loss after takeoff."  

Mode 4 is a complex combination of configuration warnings: if, for instance,
your gear is up, and you're under a threshold altitude, you'll get a warning.

Mode 5 is designed for use with an ILS: if you drop beneath glide slope,
it'll tell you.

This is a pretty comprehensive suite, but unfortunately, what happens is that
the crew tends to get these signals in a high-workload environment.  And
the signals aren't always accurate: excess spurious signals was apparently
the rationale of Air Inter's decision to remove the systems from its air-
planes: due to their operations over mountainous terrain, they were getting
more spurious warnings than real warnings.  Some pundits claim that the lack
of a GPWS "killed" the Air Inter A320 which crashed near Strasbourg a few
years ago; I don't think it would have made any difference.  Even loud,
obnoxious warnings get filtered out if you hear them often enough.

At any rate, if a crew is CERTAIN that it is where it should be, it might
just choose to disregard the warning.  This is their prerogative.  They
might be wrong, though.

Would a terrain-tracking interface help?  I tend to doubt it: it'd have
to be part of the overall user interface.  And if we were to go to that
degree of trouble, it would certainly be possible to improve other aspects
of the interface, to ensure that "180 degree" reversals (a more common type
of CFIT problem) don't happen.

A terrain-following interface would need to be a monitoring interface, which
would add a great deal to its complexity.  GPWS is just a warning system
(aural and lights).  In the modern operational context, terrain-following
would also need to be a warning system, and I'm not convinced it would neces-
sarily be any more accurate than GPWS.





---
Robert Dorsett
rdd@cactus.org
...cs.utexas.edu!cactus.org!rdd