Originally Posted by 737Jock
I can also counter that by saying you would not lower the nose close to the ground to recover the path if you get an updraft above the runway, nor would you raise the nose dramatically to stop a sink close to the ground. You either reduce or increase thrust, or change the rate at which you are changing the thrust (slower/faster).
If you get an updraft "above the runway" (I assume you have started to flare - by easing back the stick back with power coming off) you don't just pull the power back and wait! You'll either land very long or very slow/hard. You decide either to lower the nose again if you have runway space and height (add power if getting slow before touchdown) or go around.
If you get a bad sink close to the ground you would most certainly raise the nose (depending on type/tail clearance, of course); put the power up as well if it's a significant pitch change. By the time these donks spin up to meaningful thrust you'd have hit the deck hard if all you did was maintain the pitch attitude you had before the sink hit you. Obviously, the lower it hits, the less options you have, but we're moving away from the main argument which is what you use to control what during a somewhat normal approach.
Whether deliberate or not, you are continually mixing up the scenarios:
So lets start with BUSS, BUSS tells us (or at least you do) to maintain pitch and use thrust to stay in the green.
So what happens when I'm descending and I maintain pitch at say 5 degrees nose-up and I add thrust:
We're not talking (never have been) about general descents where pitch is used to control the speed (fixed power ie idle) and obviously if you add a bit of power, the rate of descent will reduce (secondary effect of controls); we're talking in this thread about being stabilised coming down final and flare with a known vertical path that you must stay on.
A computer can't anticipate, it can only react after the fact. Humans can anticipate.
So how does autoland work then?