I wonder this every time I land.
Well if this is to be believed
140 knots - 30% (42) = 98 knots (flaps up) should be your answer.
This cite below seems to have a slightly different calculation for flaps down
UPSVAC Boeing 747 Pilot Operation Handbook
General Information: Boeing 747-400
If 150 = 1.3 x stall then flaps down stall speed = approx 115 knots
This probably isn’t what you meant, but in “course made good” or speed over the ground, it sounds like a 747 could hover or go backwards if it used a hurricane or the jet stream.
I don’t think that is exactly what the OP means but it is still a good point. Planes fly by aerodynamic principles which depends on air speed and not ground speed. Air speed is what the pilots read on their instruments and that is all the plane cares about. If there is a perfect 15 knot sustained headwind while the plane is approaching the runway, then 15 knots of groundspeed are subtracted from the landing. You, as a passenger could not see that there was this headwind, you would just see that the groundspeed is slower than normal. Small planes such as Piper Cubs that can fly very slow can sometimes remain still or even fly backwards relative to the ground because of this.
I was a master helmsman for two years so I am no stranger to this concept 
astro, you second, more official source looks about right. The first one is obviously off because flaps up stall speed should be higher, not lower than the flaps down stall speed. Landing flaps provide additional lift and create drag so that the plane can land more slower than when it is configured for cruising flight.
The only thing that they leave out is how much fuel they are allowing for. I read a startling statistic in a flying magazine last night that said that 90% of a 747’s payload weight is used for fuel and only 10% for bags and passengers. I will try to find a cite but I may need to go and buy the magazine. A nearly empty 747 will have a lower stall speed.
A fully loaded 747 can weigh up to 875,000 lbs. and 500,000 lbs. of that can be fuel. I doubt it can land at all with a full load however. Fuel would have to be jetisoned.
Airline pilot here.
To answer your real question: “How fast are we going on approach?”, the WAG I always tell people is 160mph. It’s close enough for passenger info purposes.
The long answer is that the “slowest you can fly” is not a fixed number. I don’t fly the 747 so I don’t have access to the performance data for it, but as a reasonable proxy we can talk about the 767-300, which is also a very long range aircraft. And I do have the performance charts for it.
The airplane weighs about 175,000 lbs empty. If you were ferrying it with just 2 pilots, no flight attendants, no passengers and no cargo, and you were down to the bare minimum legal fuel on board of 7,000 lbs ~= 1,000 gallons, you could weigh as little as 182,000#. By our standards, 90 tons is a very, very light aircraft and it’ll handle like crap, bouncing around like a chip in a windstorm.
At that point you’d have about 45 minutes of fuel remaining on board. If you were within a couple of minutes of running out of gas, you’d weigh more like 176,000#.
Under those conditions you could get approach speeds as low as about 110 knots ~= 125 mph.
OTOH, right after takeoff with a full load of pax & fuel for a 14-hour jaunt halfway around the world, the same airplane would weight 420,000# and need an approach speed of 180 knots ~= 207mph.
As you can see, weight due to fuel & payload makes a big difference.
Now those speeds, 125 mph to 207mph, are what we call V[sub]ref[/sub], or the planned minimum speed to fly when approaching for a landing. The airplane can be slowed another ~25% before it quits flying and falls from the sky.
So that gets us to a bare-minimum flying speed that varies from a low of 96mph to a high of 160mph. We’d never actually operate that slowly; it gets pretty dangerous. But the airplane will do it.
There are a couple of other arcane factors that work to raise the bare minimum speed on some aircraft types by 5-15 knots at light weights, but I’ll leave that explanation out.
As well, you can add or subtract for wind. Wind doesn’t affect how fast you fly through the air, but it does affect how fast you and the moving air slide over the ground. If you were landing into a strong but steady wind, you could have a speed over the ground that was 20 to 30 mph less than your speed through the air. I’ve actually seen ground speeds on approach as low as 92 knots on those days when you’ve got a strong frontal wind howling at ground level. It seems to take forever to get to the runway. It’s really a pretty uncomfortable feeling.
If the wind isn’t steady, we’ll speed up a little so we won’t get caught short if the wind suddenly quits. So we’ll offset some / most of the headwind’s effect on ground speed with additional speed through the air.
Hey, LSL Guy, you’ve started to beat me to these threads! Great answer, as always.
I’d just like to point out a couple of things from astro’s post. (Disclaimer: I don’t fly the 747, and never have. My heaviest airplane was 340,000 lbs and the one I currently fly is around 140,000 lbs).
First, the info about the Evergeen tanker includes this:
This means that the flaps are down. The landing gear is up, but the flaps are at least partially down to reduce the stall speed. Depending on the amount of fuel on board (and the amount of water about to be rapidly offloaded), 140 knots seems reasonable. (This is how I flew airdrops in the Air Force - flaps at some intermediate setting and gear up. The speed depended on the weight, but was usually around 130-140 KIAS).
Also, the info from the 747 POH said this:
Note that “MLW” means Max Landing Weight, which as LSL Guy pointed out is usually much higher than normal landing weight.
The point is that these two sources are not necessarily contradicting themselves. Approach speed for a MLW 747 is very high, but if you bring down the weight a speed of 140 is entirely reasonable.
Again, approach speed is (depending on the airplane and operator) usually 1.25 or 1.3 Vref (that is 25-30% above Vref).
The lowest Vref will occur at the lightest weights. Just a guess, but I would not be surprised to see an empty 747 with 10,000 pounds of gas have a Vref of 110 or even 105 knots.
As noted, weight is important.
In theory, the stall airspeeds for two identical aircraft in the same configuration will be related as the square root of their weights. So if an aircraft is twice as heavy, its stall speed has increased by 41%.
This is true for many passenger aircraft, if they have problems during take-off for a long flight they must off-load the fuel down to their landing weight. It’s another reason why they take as little fuel with them as they can safely.
When landing a large aircraft, at what altitudes would ground effect come into play and would GE effect landing speed?
As high as a zombie can reach. A nine-year old zombie.
I worked on one aircraft where the max flying weight was greater than the max take-off weight. WTF you might say? The answer was in-flight refueling. The aircraft was the F-111.
Of course the landing weight of the F-111 was much less than the flying or take-off weight. Besides dumping fuel, the pilot could also drop things like external fuel tanks and pylons full of bombs before an emergency landing.
Oh, a zombie. I was wondering why nobody mentioned treadmills.
The standard answer is that ground effect is quite small until the aircraft is within one wingspan of a hard surface (e.g. level ground or water), and becomes significant only when the height is less than half a wingspan.
Speed matters as well. Ground effect is attributable to the ground interfering with the formation of wingtip vortices, which is another way of saying that it represents a reduction in induced drag. Induced drag is inversely proportional to the square of the indicated airspeed - so a reasonably slow aircraft (e.g. one that’s landing) would see much more ground effect that a fast one (e.g. one making a fast low pass down the runway).
Slight hijack, but is there some sort of “bombsight” on those fire-fighting planes, or do they just do their best without any specific targeting system?