Cold weather altimetry........this one is in pilot speak.

[Parvaz]

Cold weather altimetry

Up until this point the flight had been progressing normally however
the temperature at the airport located in the valley below was
miserably cold, thirty some odd degrees below zero miserably cold!
The instrument approach called for the crew to fly over the VOR then
track outbound on a radial to a DME fix, and then arc around keeping
the same DME distance while arching until reaching the localizer.
Once on the localizer they would proceed inbound to the airport.
Different altitudes were assigned to each segment of this approach.
All the onboard equipment was functioning correctly, and properly set,
the crew possessed the current weather including the correct altimeter
setting. As they turned onto the arc they continued descending in IMC
conditions to the published altitudes. It was at this point that the
ground proximity warning system went off alerting them to a collision
threat. Shaken by the unexpected warning, the crew instinctively
executed the well rehearsed escape maneuver climbing away from the
threat below. A later analysis of this incident would reveal that
this flight had come perilously close to the rugged mountainous
terrain surrounding the destination airport during their attempted
approach.

A dozen or so years ago I was a first officer on an MD-80. This
particular flight was headed to a city located just below the Arctic
Circle. It was late December, dark, foggy and cold. In fact it was
the coldest I had seen up until that point in my career. The surface
temperature at the airport was 46 degrees below zero Celsius (-46C).
The airport was reporting less than a quarter mile visibility. So to
land tonight we would be flying a Category 111(a) autoland approach
right down to the minimum permissible visibility with the approach end
right at 600 RVR.

As this approach progressed we were right on the localizer and the
glideslope. At the outer marker I was expected to make the mandatory
call: "Outer marker, altitude checks, no flags". Tonight when we got
to the outer marker something was wrong with our altitude. We were
not at the published glide slope crossing altitude; we weren't even
close; the altimeter was indicating that we were some 445' high. What
was going on here? This was the wrong place and time to be confused
about our position! Everything had seemed right with our execution of
this CAT 111(a) approach up until this point.

We had the airport's current weather including an accurate altimeter
setting. Our instruments were correctly set up. The navigation
instruments located on the airport were functioning properly. Our
flight deck preparation for this approach had been done accurately and
double checked. The localizer and glideslope needles were nailed
right on-everything was perfectly centered. Yet here we were at the
outer marker having done everything "right" but the altimeter show us
445' higher than we should be compared to the published glide slope
crossing height on the current approach chart. Why? What in the
world was going on?

There is a shared thread running through the two examples above.
What's in common is the unusually cold temperature being reported at
the destination airport. Cold weather has a significant influence on
aircraft altimeters, and this affect is often not fully understood or
accounted for. Early in my flight training I had heard the phrase:
"high to low look out below, altimeter is high, aircraft is low-same
goes from hot to cold". I had know just enough about this phenomenon
to get the occasional question answered correctly that came up on the
various written exams but I had never operated in extreme cold
conditions nor had I seen its effects personally up until that point
in my career. It was that Category 111(a) approach done during a cold
December night a dozen years ago that motivated me to unravel the
mystery of what had happened to me. This article is intended for
pilots with a working knowledge of basic altimetry and common aviation
vocabulary. Any questions regarding definitions should be taken to
the FAR Part 1 and/or the AIM for clarification. My purpose is to
present the topic of cold weather altimetry accurately in a manner
that simple to understand, remember and apply so that you can be safer
when you are flying without visual reference to the terrain and it is
really cold outside. Ignorance is not bliss, and what you don't know
or understand can hurt you in this business of ours!

Let's say for the sake of furthering our understanding of cold
weather altimetry that whatever airplane you are currently flying was
transported to an alien planet with a breathable life sustaining
atmosphere. If you were to fly your airplane in that alien atmosphere
to where your barometric altimeter indicated 2,000' how far above the
'surface' would you be? The answer is that you wouldn't have the
faintest idea using only your barometric altimeter. That instrument
was not designed nor calibrated for an alien atmosphere. So what is
my point? When the temperature is 30/40/50C degrees below zero you
are dealing with a very different atmosphere from the one your
uncompensated barometric altimeter and related pitot static system was
designed for. When the OAT is that far below zero at the airport it
is in many ways an alien atmosphere. If you have ever had to conduct
an exterior preflight under those conditions I'm quite sure you will
agree.

You remember from your early flight training that the altimeter takes
a sample of the atmospheric pressure and compares it to a setting and
then measures the difference. For every inch of mercury difference in
pressure the altimeter will read a thousand feet in one direction or
another. At ISA conditions at sea level you would have a setting of
29.92 in the Kollsman window of your barometric altimeter. If you
were to ascend straight up a thousand feet and sample the pressure it
would read precisely an inch of mercury lower at 28.92. This pattern
would repeat itself as you continued to ascend. Your altimeter simply
samples the ambient pressure and translates that to a height above or
below your initial setting (in this case sea lever & ISA). For every
inch of mercury pressure less it senses that is translated into a
thousand feet of altitude gained.

The altimeter setting you are given at the airport is not a direct
reading of the static pressure at the airport; rather it is corrected
setting so that when you dial that corrected setting in your
altimeter's Kollsman window the altimeter will accurately display
station elevation. The following point is important for our
discussion; the altimeter setting you are given at the airport (ATIS,
ASOS) is a corrected altimeter setting. That is true for whatever
temperature the airport happens to be experiencing as well as non
standard pressure. There is nothing wrong with the altimeter setting
at the airport. And that fact actually adds to the confusion. The
error related to cold weather altimetry begins to manifest itself as
you gain altitude above the station elevation. The colder it is from
standard ISA temperatures and the higher you are above the airport the
greater the error induced in the uncompensated barometric altimeter.
We will soon see that at 30/40/50 below zero we are talking about some
large errors. Most aircraft altimeter systems are uncompensated. I
fly a brand new next generation Boeing 737 and its altimeter system is
uncompensated. I have yet to actually see a temperature compensating
altimeter system.

To help us visualize what is happening let's say we constructed a
tower a 1,000 feet high and three miles from our sea level ISA
standard condition airport (surrounded by flat terrain). If we set
our altimeter to 29.92 at the airport it would read sea level. Now if
we were to take that altimeter up to the top of the tower we had
constructed it would read 1,000' MSL, and because we were over flat
terrain at sea level we would also be at 1,000' AGL.

Why does the altimeter read 1,000' MSL at the top of the tower? It
reads 1,000' MSL because the instrument senses an inch of mercury less
pressure than it did at the sea level airport. Under standard
conditions at 1,000' above the airport the instrument would be sensing
28.92 (one inch less of pressure) displaying this difference as being
1,000' higher. It is what the altimeter is designed to do.

While we are still located at the top of our tower let's drop the
temperature at the airport to 50 below zero Celsius (-50C). The
temperature at the top of the tower now reflects the temperature at
the airport as well. What is going to happen to the ambient air
around us? Cold air is denser, very cold air is significantly
denser. The denser air will sink and settle. It may help to visualize
it as a pancaking effect with layers of air sinking and getting
pressed densely together near the surface. Prior to us dropping the
temperature the altimeter was sensing 28.92 and indicating 1,000 MSL,
where is that same pressure now? As the air around us got so much
colder and heavier it sank. The actual pressure of 28.92, the
pressure that would result in our barometric altimeter indicating
1,000 MSL will now be found closer to the ground, below us on the
tower.

If it is -50C at the airport, our barometric altimeter located at the
top of our 1,000' tower would now be indicating 1,300 MSL. We are
still the same 1,000' above the ground but because the non standard
and extremely cold heavy air is so much denser and because it is so
much denser it has sunk closer to the ground, the air located at
1,000' AGL (the height of our tower) is significantly less dense than
we would find under ISA conditions. The altimeter senses this less
dense air (lower pressure) and displays it as a higher altitude.

Physically we haven't moved the altimeter at all but it is now
reading 1,300 MSL. To get the altimeter to read 1,000' we would have
to descend about 300' lower on the tower. Let's say we do that and
descend down the tower when we reach approximately 700' above the
ground and the altimeter finds 28.92 inches of mercury pressure and
indicates 1,000 MSL. This is the danger of flying published altitudes
that are uncorrected for non standard temperatures when the
temperature is extremely cold. If you fly the same published
altitudes when the temperature is extremely cold then you are actually
lower, and closer to the ground, than you would be under ISA
conditions. For example a missed approach altitude of 2,000' (over
flat sea level terrain) will not give you the same terrain separation
at -50C as it would at +15C; you would in fact be 600' closer to the
ground when the temperature at the airport is -50C than you would be
if the temperature at the airport was +15C.

Jeppesen puts out a chart to help you determine the correct altitude
correction for non-standard temperatures. The chart in entitled Cold
Temperature Corrections and is found in the terminal section of your
Jeppesen airway manual as Chart CA-19, CA-20. To use the chart you
will need two critical pieces of information. The first is the air
temperature at the station. The second is the height above the
elevation of the altimeter source the published or assigned altitude
is. With those two pieces of information you may enter the chart to
determine how many feet needed to be added under the conditions
stated. Some interpolation will probably be necessary. The only way
to be proficient using the chart is to work the examples as many times
as necessary until you are comfortable with the data and process. You
will then be able to apply this process to your own cold weather
flying experiences.

It is interesting to note that in Canada vector altitudes are
compensated for non-standard temperatures. They are also compensated
in the United States at military airports. However, in the United
States, at civilian airports when being serviced by civilian ATC
specialists their vector altitudes are not corrected for you when the
temperature drops to extreme lows. You will need to determine the
corrected altitudes you need to fly for all the relevant altitudes
during your approach.

One important closing note Autoland approaches and other approaches
that have minima based on a radar altitude are not affected. 50' RA
is 50' RA whether the temperature is -50C or +15C. However the
barometric altitudes assigned with those approaches will be affected
by the colder non standard temperatures and will need to have those
altitudes corrected or they will not align with the radar altitudes.

[pavas]

Very interesting, and factual.

A note of caution though, If you correct, make sure everyone else has
too.

B737-200 was my fave, also flew 3/4/7/800s, on the 320/319 now

[Parvaz]

I know nothing of the 320/319 but Airbus makes an excellent jet! You
have the advantage of having flown both Airbus & Boeing. Also you
have a good point about making sure that whatever altitude you are
flying in response to the cold weather correction needs to be clearly
communicated to everybody involved. Especially in a non-radar
environment! Please feel free to add your own experiences and
thoughts here. Thanks for your thoughts!