How fake-able is a nuclear test?

If North Korea piled 50,000 tons of TNT (don’t ask how they could acquire that much TNT) into an underground chamber, then detonated it, and then released Xenon-131, Xenon-133, Xenon-135 and the other element byproducts associated with a nuclear test so that those elements could be sniffed by American spy planes that sniff the atmosphere for this sort of thing, would it be able to pass for a 50-kiloton “test?”

If North Korea pump Umpteen tons of Liquefied Natural Gas, from it’s own sources, into an underground chamber, say a abandoned underground mine, then detonated it, and then released Xenon-131, Xenon-133, Xenon-135 and the other element byproducts associated with a nuclear test so that those elements could be sniffed by American spy planes that sniff the atmosphere for this sort of thing, would it be able to pass for a 50-kiloton “test?”

Was that supposed to be a link?

I am not an expert, but as I understand it, the seismic signature of a nuclear explosion is unique. It can be differentiated from a conventional explosion. I will try to find a cite.

Here is one easy cite:

it shows that I overstated the case. The signatures are different but not unique. There is actually more concern that by setting off conventional explosives one might mask an actual nuclear test.

ETA: mostly ninja’d by rbroome.
IMO the seismic signature of a large LNG explosion is going to be very different from that of a nuke. A TNT explosion will be more similar to a nuke. Whether it’s similar enough to fool our (and Russia’s) detectors is a secret I sure don’t know.

One thing I thought was that the number of byproduct isotopes from a nuclear explosion is so broad that they could never all be collected beforehand and released. It’s possible that you might know which ones are most likely to be looked for–many of them would never last long enough to be found anyway.


Natural gas doesn’t detonate, and there probably isn’t enough oxygen in an underground chamber to burn kilotons of it.

So, can North Korea pass the Nuclear Turing Test?

A nitpick: If you mix in enough oxygen you can get it to detonate, in a “fuel-air explosion”.

It would probably be much harder to pull off on such a scale than just piling up lots of TNT, though.

I think it would be difficult. A nuclear explosion is not just large, it is extremely prompt. Nuclear reactions in a weapon are over in microseconds, orders of magnitude faster than the fastest possible chemical detonation. I imagine that gives rise to an extremely rapid rise in initial pressure that is pretty unfakeable. Only a meteor impact would seem likely to have the same kind of very prompt onset.

Detonate, or just combustively explode?

Detonation is when explosives decompose rapidly, usually without additional oxygen.

You can get a detonation or deflagration depending on conditions. FAS (Federation of American Scientists) has a good summary. —
"…There are dramatic differences between explosions involving vapor clouds and high explosives at close distances. For the same amount of energy, the high explosive blast overpressure is much higher and the blast impulse is much lower than that from a vapor cloud explosion. The shock wave from a TNT explosion is of relatively short duration, while the blast wave produced by an explosion of hydrocarbon material displays a relatively long duration. The duration of the positive phase of a shock wave is an important parameter in the response of structures to a blast.

Although the detonation combustion mode produces the most severe damage, fast deflagrations of the cloud can result from flame acceleration under confined and congested conditions. Flame propagation speed has a significant influence on the blast parameters both inside and outside the source volume. The blast effects from vapor cloud explosions are determined not only by the amount of fuel, but more importantly by the combustion mode of the cloud. Significant overpressures can be generated by both detonations and deflagrations. Most vapor cloud explosions are deflagrations, not detonations. Flame speed of a deflagration is subsonic, with flame speed increasing in restricted areas and decreasing in open areas. Significantly, a detonation is supersonic, and will proceed through almost all of the available flammable vapor at the detonation reaction rate. This creates far more severe peak over-pressures and much higher amounts of blast energy."

Explosives contain their own oxidizers.

The cite earlier from rbroome is interesting as it suggests that it would be quite possible to deliberately fake the signature of a below ground nuke test. They note that the staggered explosions in a mine lead to a characteristic scalloped response in frequency space. One that with care can be discerned. What isn’t at issue is that the power of the blast is too small. Only that its seismic signature has the right profile. They provide the clue that a carefully crafted set of explosions staggered over time can be used to create different signatures. (Not really a surprise, this is Fourier theory in action.) The question is then of the amplitude of the blast. This comes down to how well the energy of the nuke couples to the Earth, and over what frequency range. The question is probably simply one of how much rock is vaporised, and over what timespan.

My gut feeling is that an air blast is vastly more efficient in coupling the energy to the air into a proper destructive blast that then propagates through the atmosphere. I suspect that a nuke underground expends a lot more energy melting rock than vaporising it, and that the impedance mismatch from the vaporised rock to the surrounds is sufficiently worse to not couple a lot of the higher frequency energy at all well. Basically meaning that you can approximate the seismic effects of the nuke with TNT in a hole in the ground with a lot less TNT than the equivalent rating of the nuke. Still probably a goodly amount of TNT, but not tens of thousands of tons.

Faking the radioactive trace gasses may be a lot herder. It isn’t as if you can go to your local chemical supply company and buy a bottle of [sup]135[/sup]Xenon. With a half life of 9.5 hours, harvesting enough from a running nuclear reactor is going to be a challenge. If you knew the precise location of the sensing systems looking for the isotope signatures you might be able to fake things with a limited supply. Not easy to get right however. You need to get the ratios right and concentrations plausible.

I know. That’s what I meant by’ without additional oxygen’

NB: there are also some explosives that contain no oxidizer at all and work by decomposing without oxidation.

Whooo, yes.
Favorite blogs are by Lowe. “Things I won’t work with…” are some exciting reading.
ex. "But remember, N-amino azidotetrazole (I can’t even type that name without wincing) is the starting material for the work I’m talking about today. It’s a base camp, familiar territory, merely a jumping-off point in the quest for still more energetic compounds. The most alarming of them has two carbons, fourteen nitrogens, and no hydrogens at all, a formula that even Klapötke himself, who clearly has refined sensibilities when it comes to hellishly unstable chemicals, calls “exciting”. Trust me, you don’t want to be around when someone who works with azidotetrazoles comes across something “exciting”.

Yeah I think that’s about it. My dad used to work as a physicist in the area of nuclear testing treaty verification, in the 90’s and this was exactly the sort of question he was trying to answer. He couldn’t share too much with me, all his work was classified, but my impression was that there were some people who said they could do it but he was skeptical.