OK, for my National Novel Writing Month entry, I got WWII B-29s flying at 48,000 feet (don’t ask) over Europe (don’t ask).
So, can you hit anything from 48K? From that height, can you see the darkness of ‘space?’ The strange curvature of the horizon? I am just looking for some color to add. What sort of WWII-ish technology could be used to hit (say) a city from such a height?
The novel is here
Well, you can be pretty sure of hitting the ground. 
Seriously though. From 48000 feet using the Norden bombsight you could hit a city without much problem. A casual Gooble has turned up no data on Norden sight dispersion. However, a lot of the accuracy of the sight depends upon being able to fly the airplane on the proper course. Remember, the sight isn’t aiming the bomb, what is gong on is trying to fly the plane to the proper release point so that a bomb with nominal aerodynamics will hit the bombsight aim point within Y feet X percent of the time.
You might stress the difficulty of maneuvering the plane at 48000 ft. (especially since the B29 operational ceiling was about 35-36000 ft.). Probably the pilot would need to fly the plane on the bomb run since I suspect the autopilot-bombsight loop would be pretty jittery at that altitude.
As an absolute WAG I think you would do well to get within 1000 ft. of the aimpoint. That’s an error of 27 milliradians which is pretty good considering that the dispersion of a machine gun is about 9 milliradians. Maybe 2000 ft. for even half a mile would be a better WAG.
That’s about half a degree, for those of us who don’t normally think in radians.
I’d call that an ideal, best-case estimate. I think the typical error could be many times that.
I know a guy who flew bombers (mostly B-26) in WW2. He says that to render a bridge unusable (which really required just a single good hit), they figured on many dozens of bombloads. And they were operating at something like 24,000’.
By the end of the war in Europe, the USAAF was using radar to bomb through clouds. I don’t think they actually used that practice over Japan because LeMay brought Harris’s firestorm policy to the wood-and-paper Japanese cities, but the B-29 did mount the AN/APQ-7 radar bombsight (from Bell labs). (There was also an AN/APG-7 bombsight that was used to guide glide bombs that might have been more accurate.)
You might track down information on those devices to see how accurate they could be.
I have not found anything definitive, but my memory is that the curvature of the Earth’'s horizon and a steadily deepening darkness in the overhead sky begins somewhere between 60,000 feet and 100,000 feet.
Former USAF fighter-bomber pilot …
48,000 ft is high altitude, but nothing special in terms of the appearance of the Earth or sky. I’ve been there & seen it myself & it looks just like normal airliner altitudes.
It’s also waay too high for B-29s, but if you’re stuck with it, go with it.
Any bombsight for dumb bombs, from the Norden of WWII to the modern CCIP (continuously computed impact point) systems of current fighter-bombers / attack airplanes (F-16, F/A-18, Typhoon, Rafale, SU27) , can only compensate for the wind drift at the altitude of the bomber.
If the wind field varies widely in the atmosphere below the release altitude, the bombsight won’t know about it & can’t compensate for it. The result is an uncontrolled error in the solution & a large miss distance. And the more atmosphere below you, the more likely the wind on the way down is not homogenous.
Note that the unknown lower wind field won’t increase the dispersion of any given salvo, it’ll just push it all, say, 500 meters Northwest of the aim point on this particular drop. Over many salvos on many missions the overal effect is to increase dispersion by a statistically significant & fairly reliable factor.
WWII era bombs were not highly streamlined & were more affected by wind during freefall than modern bombs. WWII bombs also were notorious for not very repeatable fin installations which made for a few fliers in each pattern. I’d add 70+ mils to the CEP to acount for lower altitude winds & crappy WWII bombs.
I don’t have any direct sources for typical Norden accuracy , but Dave was there & his 27 mils doesn’t feel too far off; that’s about 500 ft at 24,000 ft release altitude which sounds pretty much like typical WWII numbers I’ve read elsewhere.
In terms of effectiveness, typical WWII bombs were either 250 or 500 lbs. A reasonable kill radius for soft buildings (factories, offices, houses, etc) is 100 ft, with damage radius out to 250 ft. Clearly dropping single bombs with a 100 ft kill radius and a 3000’ CEP is not going to be effective except as harassment. Dropping large quantities from a whole formation will give you the large shotgun pattern you need to probably hit something the enemy values.
All in all, WWII aerial bombardment was a fine way to attack cities, a marginally effective tool for attacking even large factories, and wholly useless for pinpoint strikes. As long as you don’t care about collateral damage & have the time, bombers, and bombs to spend on repeat attacks, eventually you can hit any desired shack in the county. The question is whether you’re willing or able to turn the rest of the county into the surface of the Moon just to get that one key shack.
Barring the lucky hit:
In Martin Caidin’s 1957 photo history of the Air Force, Air Force, there is one photograph of a very long bridge over a Chinese river or swamp that was struck by a single bomb from a single B-24. I no longer have the book, so I can’t reproduce it on the web, and I have not been able to find the photo online. My memory of the drop altitude was that it was in the high teens. (I’d really love to find the box that held that book when we moved.)
It will be really cold in the airplane, especially if/when bomb bays are open. And the guys on board will be getting goofy from the altitude even with oxygen available. So… hypothermic and hypoxic. Every task will be harder and take longer.
There were two main operational blind bombing systems in WWII, Oboe and Gee H. Oboe planes were referred to in the 9th AF as Pathfinders and were part of specialized Pathfinder groups. In such bombing, the pathfinder would lead the group as a formation, usually 36 planes but sometimes 54, and drop its bombs at the proper point. When the group bombardiers saw the pathfinder’s bombs they dropped theirs. Crude but maybe better than nothing. The main problem was, rapid damage assessment was impossible and a separate recon plane had to be sent out after the weather cleared. This meant, of course, that you never really knew whether or not the technique was all that effective.
At least it was a big improvement on the RAF’s early performance before these methods were developed. They couldn’t always hit the right city, or any city at all, in night bombing.
In addition there was this radar system used by both the RAF and the USAF.
As to crew environmental conditions,the B-29 had heated and pressureized crew modules. The whole plane wasn’t pressureized, just the parts where the crew was. There was a crawlway to get from front to rear. The gun were remote controled so that the gunner could be in a pressurized space and the guns could be in unpressurized turrets.
B-29 were pressurized in the parts where the crew was stationed. Bomb bay was not pressurized.
From Wiki “full pressurized comfort. The nose and the cockpit were pressurized, but they had to have a large bomb bay that was not pressurized, or they would have had to de-pressurize to drop their loads. So the B-29 had a long tunnel over the two bomb bays so that crews could crawl back and forth between the front end and the back, with both areas and the tunnel pressurized.”
Also from that Wiki link it said that the B-29 could fly at 40,000 feet.
::: Shakes fist:::
Damn you and your flying fingers
Well, you had from 04:26 PM to 05:08 to get in your licks. Don’t blame me. 
The B-29 might be able to fly at 40000 ft. but the operational ceiling is around 35000, depending upon the model it might be a little higher.
As you go up in altitude the indicated airspeed, which is what counts, decreases for the same true airspeed. When you reach an altitude where the indicated airspeed equals the stalling spee you can’t go any higher.
The stalling speed in a turn is higher than in level flight. So it you want to be able to do any maneuvers, your ceiling is lower than the level-flight, absolute max.
[hijack]You often couldn’t even get WWII bomb configurations to leave the airplane. And even the modern bomb shapes are dropped from ejection racks that give the bomb a good shove to get it the hell away from the airflow around the plane as fast as possible.
We had a program here to drop some of the old, blue practice bombs from a North American A-3 which had a bomb bay. When they first tried to drop them, the bomb would hit the airstream and bounce off it back up into the bomb bay. So a co-worker of mine had the job of designing some bomb shapes that we could get to leave the plane. And even with more streamlined shapes the A-3 had to slow down some.[/hijack]
To steal a line from Jonathon Winters, “What the colonel meant was…” you couldn’t get the bomb to leave a high-speed jet plane.
I did not know that. But now I do.
Weapons separation is still problematic. Modern ejector racks give them a damn good shove, but even with modern computational fluid dynamics simulations, the first releases of test shapes or known shapes from new airplanes or racks is a bit fraught. Release envelope expansion all the way to the corners take brass cojones.
Where’s the rest of it? 
Hey, hold your horses, I’m still writing it! All I have to do is get Hitler to trip over some luggage in the last 9K words!
Thank you all for very serious answers to another of my very silly questions.