[Stratoscope]

> Fun fact: Flight Levels are really only used at 18000 feet and above. That's where local barometric pressure is ignored and aircraft use the standard value. I believe it's a global standard to use local barometric pressure below 18000.

I don't know if there are any fun facts in this particular conversation. May the crash victims rest in peace, and may their families, friends, and colleagues somehow find peace after this tragedy.

In any case, what you're referring to is the Transition Altitude, where pilots switch from local altimeter settings to a standard setting of 29.92. In the US and Canada, this is 18,000 feet, but it varies in other countries. In Australia, for example, the transition altitude is 10,000 feet. The transition altitude in any country is supposed to be above the highest elevation of any point in the country.

That's the theory anyway: the highest elevation in Australia is 7,310 feet, but the highest elevation in North America is Denali (Mount McKinley) at 20,310 feet - which puts it up in the flight levels.

Leaving that aside, the idea is that when you're flying at lower altitudes, you not only want to know your altitude relative to other aircraft to avoid collisions, you also want to know if you will clear any obstacles on the ground. So you set your altimeter to reflect current local air pressure to get an accurate altitude above mean sea level, which also lets you determine your altitude above ground level.

But once you're well clear of any terrain, this isn't helpful. It simplifies things to have all aircraft use a common altimeter setting, even if it doesn't truly reflect local conditions. In the Flight Levels, you are no longer worried about the ground, you're only worried about other aircraft. (And Denali!)

https://en.wikipedia.org/wiki/Flight_level#Transition_altitu...

[grkvlt]

How does this work in practice, since I can imagine a situation where using the local air pressure value, altitude is measured at 18k feet - then, switching to the standard setting of 29.92, it turns out that this converts to an altitude of 17.5k feet - so, now the system switches back to using the local air pressure for calculations - and so on... Is there some form of hysteresis or rounding used, or would this never happen in practice?

[Stratoscope]

Interesting observation! You could easily imagine an automated system going into some kind of oscillation if it didn't handle this properly.

Here is why it doesn't happen: whether you are climbing, descending, or flying level, you always have a specific target altitude or flight level. This target determines the setting you use, and you know which kind it is.

For example, you may be flying level at an altitude of 8000 feet and Air Traffic Control instructs you to "climb and maintain flight level 240" (roughly 24,000 feet).

You start a gentle climb and then change your altimeter to 29.92 because you're going up to the flight levels. You don't need to fiddle with the altimeter when you cross the transition altitude, because the target altitude/flight level is the only one that counts.

[FabHK]

Excellent rejoinder. FWIW, GP reflects this tendency I've seen more than once to assume that specific US aviation rules or conventions apply everywhere.