Edit3: Links to the Cover Page and Table of Contents:
"86th Congress, 1st Session; House Document No. 86"
SPACE HANDBOOK: ASTRONAUTICS AND ITS APPLICATIONS
STAFF REPORT OF THE SELECT COMMITTEE ON ASTRONAUTICS AND SPACE EXPLORATION
http://history.nasa.gov/conghand/spcover.htm
http://history.nasa.gov/conghand/contents.htm
---
There is this:
http://history.nasa.gov/conghand/nuclear.htm#REF17-1
When a nuclear weapon is detonated close to the Earth's surface the density of the air is sufficient to attenuate nuclear radiation (neutrons and gamma rays) to such a degree that the effects of these radiations are generally less important than the effects of blast and thermal radiation. The relative magnitudes of blast, thermal and nuclear radiation effects are shown in figure 1 for a nominal fission weapon (20 kilotons) at sea level.1
And:
If a nuclear weapon is exploded in a vacuum-i. e., in space-the complexion of weapon effects changes drastically:
First, in the absence of an atmosphere, blast disappears completely.
Second, thermal radiation, as usually defined, also disappears. There is no longer any air for the blast wave to heat and much higher frequency radiation is emitted from the weapon itself.
Third, in the absence of the atmosphere, nuclear radiation will suffer no physical attenuation and the only degradation in intensity will arise from reduction with distance. As a result the range of significant dosages will be many times greater than is the case at sea level.
Figure 2 shows the dosage-distance relationship for a 20-kiloton explosion when the burst takes place at sea level and when the burst takes place in space. We see that in the range 500 to 5,000 roentgens the space radii are of the order of 8 to 17 times as large as the sea-level radii. At lower dosages the difference between the two cases becomes even larger.
A important consideration which is pointed out in some of the notes relating to nuclear weapons in space is a comment about the extra vulnerability to people: That is, because of radiation concerns, as pointed out above, in manned spacecraft, people can be a relatively vulnerable component.
Edit: To put some numbers on this, the charts show (approximately; the charts are hard to read in detail):
20KT Explosion
Air:
1000 RAD at about 0.7 miles
500 RAD at about 0.75 miles
100 RAD at about 0.8 miles
Vacuum:
1000 RAD at about 8.5 miles
500 RAD at about 13 miles
100 RAD at about 28 miles
The question then turns on the effectiveness of shielding, and on the relative vulnerability of electronics (I'm guessing the electronics of today are a lot more vulnerable than those of 1957, which is the date given for the referenced document.)
Edit2: Wow, some nice charts in that report, some very useful for game designers:
http://history.nasa.gov/conghand/mannedev.htm
Thx!
TomB