The Air New Zealand incident, like many others, proves more than one point:
Despite our mental calculations- times 3 arithmetic, etc., people still bust the correct altitudes during the final approach, because they don't do a proper glide check.
Capturing a false glide slope, or experiencing wrong indications on our glideslope indicators, is a rare event (never had one, personally). But it's still possible, and it is the rarity of the event which makes it so dangerous.
Distance indications are not always available, so it's not always
that easy to make a X3 mental check.
Let's have a look at the Faleolo incident, then at the Alitalia crash in Zurich, 1990.
Both of them did not check the glide properly: the 767 crew went around at ~400ft, (!), after the whole approach screwed up!
Had they called,
loudly ,"At 2500 we must be at 7.5 DME" and "At 2000 we must be at 6 DME" etc, they would have been totally safe, instead of waiting for the DME to check their altitude: they happened to be at 900ft. at 8miles, and at ~400ft. at 5.8 miles, very close to Apolina island!
I'm aware that such callouts seem so obvious, so basic, so natural... but if you make a habit of thinking that way, and declaring loudly "2500, 7,5 DME,
Glide Checked ", then you are safe on
every approach, where the numbers are not so straightforward.
In 1990 an Alitalia DC9 crashed short of rwy 14 in Zurich: they did not check the glide, and flew the approach 1400ft parallel and below the glide.
They were waiting for the outer marker to do the check ("Il marker é passato?"), and they didn't realise they were low.
It's easy to say, from behind a keyboard, they missed the obvious, that they should have made a constant X3 mental check and so on.
Truth is the conditions (from a human factor point of view) were bad, the elevation was high, and so on... and they died. In the cockpit, when things go to worms, our brain decreases its abilities to 80, 50, 30%..... even if we don't like to admit it.
So
we must adopt a fool proof method, wich GUARANTEES no mistakes can be done, even by retarded people! Only then we won't experience anymore accidents which seem so stupid, but unfortunately fill our statistics, like the Faleolo or Zurich or Guam, and the list is looong.
Please don't get me wrong when I say you must check the DME or marker at the altitide instead of the opposite.
It's just a way of thinking the check which saves your sweet a§§.
Actually, when you make sure your glide is correct, you
simultaneously check the altitude accuracy.
And you can make or confirm all the corrections you like due to temperature and whatever.
And now we have come to the second interesting point: do we correct or not for the temperature?
As C. Chukle has so rightly pointed out, if we fly in extremely cold conditions,
and at altitude, like Calgary, then we must correct our minima.
The 4XaltXISA deviation rule of thumb is well known and still valid if we don't have tables or more accurate means of correction.
In Calgary the error could be very significant.
Here are a few interesting articles on the subject:
http://www.bluecoat.org/reports/Long_98_Cold.pdf
http://www.fmcguide.com/media/calgary.pdf Academically, we should always correct, even at sea level, but there don't seem to happen many crashes because we decide at 184 feet instead of 200.
Anyway, this is really a big black hole in our culture, as it seems only Canadian ATCO make allowances for very cold temperature when radar vectoring, and even the FAA is dragging behind trying to establish new safer rules in North America.
Also, if we consider that at despatch an altimeter is considered acceptable with an error as big as 75ft (if my textbooks are still correct), and on a B733 the max allowable difference between altimeters at sea level is 50ft, then we realise the Gods really love us!
Is the average pilot really safe, or just a bit safe and a lot lucky?
I suggest a new rule for establishing when to decide at the minima: At the
corrected barometric decision altitude/height,
or at 200 ft radioaltimeter, whichever comes first.