I suspect, but cannot prove, the wikipedia number is more likely to be correct.
[Edit 1]
Actually, if you consider the indirect effects of the use of such a weapon - particularly the large number of refugees fleeing contaminated areas, huge numbers of injured etc. while such a detonation wouldn't destroy an entire state it would probably reduce it to complete chaos.
e.g. The UK government estimated in the 1950s that three H-bombs would be all it would take to destroy the UK as a functioning society - the UK ambassador in Moscow even had a drunken debate with Kruschev on this point - suggesting that the hundreds of weapons the Soviets had pointing at the UK were complete overkill...
[Edit 2]
The UK government was always pretty gloomy/realistic about the likely impact of a nuclear war on the UK - basically even in fairly optimistic scenarios such as the Square Leg exercise used as the basis for Threads were absolutely awful. A real attack would have been far worse that the one shown in Threads, if anyone can conceive of that.
As Schlosser says in his book - because of the lack of coordination between the Air Force, Army, US Navy and Missile Forces, at one point there was something like 400(sorry if I remembered the number wrong,but it was in hundreds) nuclear rockets aimed at Moscow, many of them with multi-megaton payload. If they were launched, Moscow wouldn't be just destroyed - the entire area would be turned into the biggest crater on earth.
NB An interesting article by Charlie Stross on the relationship between the US and the UK's "independent" deterrent:
http://www.antipope.org/charlie/blog-static/2013/04/on-the-u...
http://en.wikipedia.org/wiki/Sprint_%28missile%29
Now picture the skies over Moscow with ~400 warheads coming in, most of them having been launched in a coordinated wave by Minuteman, Titan, or MX missiles within the space of a few minutes: it turns out that the prompt neutron pulse, blast, EMP, or thermal effects from one incoming warhead can disrupt another if it arrives within 10 seconds of the first. This is termed nuclear fratricide:
http://en.wikipedia.org/wiki/Nuclear_fratricide
... And there were enough weapons in play during the Cold War that this was a thing war planners worried about.
Apparently, the US military were tasked with destroying 50% of the socialist world's industrial capacity - which you could do by destroying 50% of the factories completely or 100% of the factories by at least 50%. However, if you attack say a shoe factory (the example given) with an H-bomb aiming to achieve at least 50% destruction you do actually get 100% destruction. So all factories were targeted.
NB I used the phrase "socialist world" as, at least until well into the 60s, the SIOP aimed to destroy everyone even vaguely left of center whether they were actually allies of the Soviets or not (e.g. China, Albania, Yugoslavia etc.).
Pretty sure I read this in "The Bomb - A Life" by Gerard DeGroot, but I could well be wrong!
Air defence systems back them were really pretty poor - e.g. the infamous case where the entire US air defence system couldn't stop UK strategic bombers taking out major us cities in an exercise (which was kept quiet for decades due to the embarrassment it caused).
"It appears that two weapons targeted on a silo must arrive at least 10 s apart to avoid fratricidal fireball effects, and less than 1 min or more than 1 h apart to avoid fratricidal nuclear dust cloud effects."
Anyone for Missile Defense with a 50-100Mt Tsar bomba?
If you can land a nuke, or even a conventional warhead, directly on whatever it is you're trying to make a was, your required explosive yield falls dramatically.
Blast radius and damage increases with the cube of yield, so if you're half as accurate, you need 8x the blast, generally, for effective damage.
Turns out that with extremely high precision warhead delivery, the US can dispense with nuclear weapons even for tasks such as destroying deeply buried and reinforced bunkers (command and control, or missile launch).
On the one hand this makes nuclear Armageddon that much less likely. On the other it makes use of highly advanced weaponry less objectionable.
Which way the risk needle ultimately swings given that calculus is an interesting question.
My (wild) guess is that you'd get a blob of extremely hot rock melting its way down through the crust and enough fallout to sterilise the Northern hemisphere.
So kind of a dumb idea from a strategic point of view.
http://en.wikipedia.org/wiki/Single_Integrated_Operational_P...
So, although it is possible for the fire to ignite the conventional explosives, it would almost certainly have not done so in precisely the way needed to activate the atomic components of the bomb -- at least that is what I have read somewhere. (Sorry, I do not remember where -- maybe Carey Sublette's Nuclear Weapons FAQ.)
The worst thing that could realistically have happened is that the fire and a conventional explosion would have spread the radioactive components of the bomb (particularly, plutonium) into the environment, in the manner of a dirty bomb.
Suppose the trigger device providing the precise timing could misbehave at high temperatures. Or, in the same source I mentioned earlier, it was disclosed that there were wiring paths which could have melted together in a fire and cause the weapon to fire.
[1] https://en.wikipedia.org/wiki/Operation_Castle [2] https://en.wikipedia.org/wiki/Castle_Bravo [3] https://goo.gl/maps/thGe6
Bikini Atoll was also worse than this hypothetical Arkansas scenario in two ways:
1) The soil there is potassium poor, causing local plants to more readily take in radioactive cesium.
2) It was also hit by 22 additional bombs.
Here's a link with the numbers plugged in: http://nuclearsecrecy.com/nukemap/?&kt=9000&lat=34.7464809&l...
Arkansas would not be "wiped out".
edit: for starters consider that in a state of three million people, 725k are in Little Rock, which is only 116 square miles. Then read the specs on that bomb you linked to again.
The Wikipedia number is not correct, because even a single blast has long term consequences, and in a full scale war, the consequences are more significant than the initial damage.
Even if the blast was just one a of a handful, survivors would have to contend with fallout from other detonations, followed by nuclear winter, followed by years of famine, mutation, cancer, and other diseases.
In a full scale war, nukes are targeted at nuke power plants, so there would be even more fallout - enough to sterilise many parts of the US.
Initial casualties would be a small percentage of the eventual total.
Even a limited exchange in a distant part of the world would have global consequences. See e.g.
http://phys.org/news/2013-12-nuclear-war-civilization-famine...
The specific point being discussed was a local disaster -- a 9 megaton bomb which was damaged by somebody dropping a tool on it -- not an all-out war scenario.
Which is why I responded the way I did. In that scenario, I don't see any realistic case in which the entire state of Arkansas would be "wiped out". Possibly fallout would have rendered a fraction of the state uninhabitable for several weeks, and possibly some long-term effects would render a smaller fraction of the state uninhabitable for much longer, but the whole state "wiped out" is just sloppy journalism.
http://nuclearsecrecy.com/nukemap/?&kt=9000&lat=34.7464809&l...
Also be aware that fallout from an underground blast (as in a silo) is going to spread much differently from fallout from an airburst.
Thus, I am unconvinced by the argument that the longer-range fallout would "wipe out" something the size of Arkansas.
Disappointing to read needless hyperbole in the New Yorker.