@JasonC, Aeroplus is right about possible downdrafts, and Michel’s explanation is also correct. However, it may help to elaborate:
When we land an airplane, we’re supposed to flare and fly almost parallel to the surface for a short distance. Well, that’s what I try to do instead of my usual sack-of-spuds arrival. 
When the runway is horizontal, it’s easy, because even my old Maule is capable of sustained horizontal flight in ground effect.
When the runway slope is, say, 12.5% up (i.e. the first “flat” 130 metres at Courchevel), that may be two or three times steeper than we can climb with our little O-360 at altitude. So, just as we carry a bit of extra speed in a C150 Aerobat to enter a loop, we fly the approach to LFLJ a bit faster than we would for a short flat runway so as to have some kinetic energy to convert into potential energy as we touch down. We also maintain or add engine power during the flare.
It is all pretty easy, but the touchdown zone is small. If we land a bit short, we probably die. A late attempt to go-around has the same result as landing short, albeit a few seconds later. If we land a bit long, we may just bend the airplane. People do make errors of judgment on approach, and bending the airplane, especially a Piper, would be my choice.
The PA46 pilot did well to continue his misjudged approach. I take my hat off to him.
There is also the question whether it is sensible to try go around late (and risk a crash at climb speed, which the occupants won’t survive), or just accepting the risk to go off the end of the runway at 10 knots with some damage and perhaps some bruises.
So unless the pilot knows the aircraft can achieve a climb gradient which is a few degrees above the runway slope (or the obstacle surface), it is safer not to try.
The same applies to single engine go-arounds in most twins, where a crash landing may be the safest option below a certain height.
Cobalt wrote:
There is also the question whether it is sensible to try go around late (and risk a crash at climb speed, which the occupants won’t survive), or just accepting the risk to go off the end of the runway at 10 knots with some damage and perhaps some bruises.
Yep. The same applies with go arounds in jets. More people lost by attempting go around after touch down without enough time to stow spoilers etc and spool up than by running off the end at slow speed.
Why not get a topo map and plan the approach and the point of no return?
In the mountains you can get illusions as to where the horizon lies, etc.
The conventional wisdom is that we are committed to land when we intercept our final approach slope. After that, it’s usual to concentrate on maintaining that slope and speed to a selected aiming point some distance short of our intended touchdown point. If we screw it up, we land anyway, run into the berm at the top, and take a bus home.
But conventional wisdom may be a load of cobblers peddled by a bunch of leathery old instructors – or it could be one reason why they got to be old and leathery. Let’s do a quick check for my airplane.
The AFM says best angle of climb is at 70 KIAS, say 80 knots true at 10k DA. Suppose we feel lucky. No tailwind or downdraft, so we are sure we can climb at more than 300 ft/min at that altitude. I make that a gradient of 3.7%, give or take a bit for Jesus.
But wait, as AeroPlus pointed out, the bottom part of the runway at Courchevel looks horizontal – and it is actually 12.5%. Do we still feel lucky? Or do we follow the old goats?
Jacko wrote:
the bottom part of the runway at Courchevel looks horizontal – and it is actually 12.5%
This pic is from another forum, not sure how accurate it is.
What GS / rate of climb would you need to outclimb this and the mountains behind if going around from 200ft ?
JasonC wrote:
Cobalt wrote:There is also the question whether it is sensible to try go around late (and risk a crash at climb speed, which the occupants won’t survive), or just accepting the risk to go off the end of the runway at 10 knots with some damage and perhaps some bruises.
Yep. The same applies with go arounds in jets. More people lost by attempting go around after touch down without enough time to stow spoilers etc and spool up than by running off the end at slow speed.
Same apply even for SEPs on go around with full flaps/gear down, go around in turbulent air, go around in gusty crosswinds a stable approach is probably your first protection, runway length is the second one…
The above drawing is nice, but not accurate. The SIA VAC is here. Local copy
As for going around in a conventional GA airplane, I don’t really want to think about it. To clear ground less than 1 mile ahead on runway heading you would have to climb 2,000 ft. In case of engine failure, that might be tricky.
I have watched a couple of visitors to my airport going around. They survived, but I think they found it quite a thrill.
In a CAP 232, or even a Pitts S1 it should be possible. Fly the approach at suitable speed, pull up to vertical at the threshold and roll out on departure heading. Simples!
I have watched a couple of visitors to my airport going around. They survived, but I think they found it quite a thrill
I can attest to that – the surrounding tree covered hillsides provide a nice illusion of a false climb attitude, and a go around needs an early decision – and this is for an airfield not much above sea level with only hills (not mountains) around it.
