Back to: Roko's Basilisk wants YOU | Forward to: Dragged kicking and screaming into the 19th century

Evidence exists that a large natural nuclear reactor formed and operated on Mars in the northern Mare Acidalium region of Mars. However, unlike its terrestrial analogs this natural nuclear reactor was apparently much larger, bred 233U off of thorium, and apparently underwent explosive disassembly, ejecting large amounts of radioactive material over Mars' surface.

Source (PDF).

(See also: Natural nuclear fission reactor. Only, on Mars, all geological features seem to be supersized ... I don't care if this is implausible, it's bloody going in a novel. OK?)



Note: I make the explosive yield (per page 2) about 36 billion megatons.

In other words, around a billion times more powerful than the entire combined Cold War nuclear arsenals; somewhere in the dinosaur-killer asteroid league table.

And now I have an idea for how to work this sort of thing into the new Merchant Princes books (because that series goes better with a really big bang :) ...


Yeah, I noted that too.

Whether it holds up to peer review or not, it is definite gosh-wow territory. The planets are getting interesting again.


Oh wait, yeah, your retweet in the first place.


I can't find the name of the short story but it was in Robert L. Forward's book Indistinguishable from Magic, last chapter. Starting from the true fact of the natural fission reactor, he worked out a novel way an alien race might solve the wealth concentration problem inherent in most economies. Money was made out of uranium and stored in money pits. The elders always advised not hoarding more money than you needed. A greedy and miserly alien accumulated more money than was good for him and his financial assets went critical. I always liked that story.


Strugatsky Brothers, The Land of Crimson Clouds, Uranium Golkonda. Only it was on Venus, not Mars.


Would this explosion have blown away Mars' paleo atmosphere (instead of slowly dribbling out into space)?

Would this explosion explain Mars' lack of a magnetic field?

Could Phobos and Deimos be ejecta from the explosion instead of captured asteroids?


It's dated to roughly 1Gy ago. I don't know enough about Areology to guess at the answer to those questions, but they don't sound obviously implausible.


Charlie - presumably on one of the alternate Earths?

gmuir77 @ 4 Also called "going Deeble" as in "A Cautious Collector" by R. Zelazny!

And, of course we now know where the Martians went - up in smoke ....


Charlie - presumably on one of the alternate Earths?



How could the fissile energy of 36 billion MT be held together before the boom absent a stellar or quasistellar mass? (I've not read the PDF, I admit)

Making a bomb go boom is a race against physics and requires a lot of just-in-time programming. A natural reactor with substantial amounts of ongoing fission will tend to tick along for a few hundred thousand years like the Oklo reactors did; anything more energetic will disassemble itself due to melting, steam explosions etc. well before any explosive fission cascade can occur. Colour me dubious.


I was a little sceptical about this paper when I read it. Even for a short paper it seemed rather thin on geology* and for that matter rather arm-wavy about the nuclear physics. It took one piece of evidence and extrapolated a remarkable conclusion from it without doing much to consider the implications of that conclusion and testing them. I would expect the first thing to do on coming up with the hypothesis of a 10^7 - 10^8 megaton explosion on Mars would be to ask someone familiar with martian geology what this would do to the landscape and then see if those predictions are borne out. Instead, we get the comment that there's a large, shallow depression in the candidate area.

At this point I put "John Brandenburg Mars" into Google. I commend this exercise to anyone else interested in following up this paper.

*Areology for the pedantic, I suppose.


All that yield, and nobody heard.


And another paper by John E Brandenburg, from a little while back: The Cydonian Hypothesis.

He also has a Unified Field Theory.


Blowing the atmosphere of any planet away isn't that easy, even on Mars. The explosion may have made a small contribution but a slow trickle fits the areology as currently understood.

The Martian magnetic field has been gone for much longer than a billion years, and the asteroid impacts that formed the Hellas and Argyre basins (along with a couple of others around the same time) currently have fingers pointed at them.

Phobos and Deimos have nearly circular orbits which are almost exactly over the Martian equator, if they were formed from ejecta the orbits would be much more elliptical and at varying inclinations. Spectroscopically their composition matches that of carbonaceous chondrite asteroids much more closely than it does the Martian crust.


How could the fissile energy of 36 billion MT be held together before the boom absent a stellar or quasistellar mass?

(It's a two-page paper, not hard to digest.) The hypothetical natural reactor is under immense pressure -- beneath at least a kilometre of solid rock. Water supplied by hydrated magma from below, not streams from above. It's also unlike Oklo in that it's breeding both 239Pu (from 238U) and 233U (from Thorium).


sjbradshaw @ 13 Oh dear, what a pity, he's a kook. "Cydonian hypothesis" indeed!


Never let kookery get in the way of a good plot element!


It is a cool idea, even if the purveyor lacks scientific rigor. Even more interesting would be the possibility that we might be able to locate additional asteroids that had that sort of concentration of fissionables in their makeup. Now there's something that might (at some point) make asteroid mining very interesting...


Not that I'm an areologist, but there are some good general guides to Mars out there. My personal favorite is Hartmanns' A Traveller's Guide to Mars.

First, a bit of Areological history. These eras are named after where the best evidence for them exists.

Noachian Era: 4.5-3.5 Gyr, (give or take 100 Myr). This is an era of active erosion, heavy volcanic activity, and possibly lakes or oceans. This is the time of liquid water, and probably the Earth and Mars were quite similar at this time.

Hesperian Era: 3.5-2.5 Gyr (give or take 500 Myr). This is when the planet dried out. Over a 1,000,000,000 years, which is not fast.

Amazonian Era: 2.5 GY-Present (give or take 500 myr). This is the Red Mars of Today. Note that the hypothetical reactor is well within this era. Water's deep underground and/or frozen.

This is to help get you oriented in deep time. Most of us don't spend much time thinking about that range. Remember that paleontology on Earth mostly cares about the last 400 million years ago, for comparison, and prior to a billion years ago, it was all bacteria. 4.5-1 GYr is a very, very long time before that.

A bunch of things are bugging me in this paper. For example: I found it fascinating, when I hauled out my topographic maps of Mars, the evidence for K and U concentration are at the two lowest points on the planet. If you dissolved salts into an ocean and freeze-dried said ocean, where would all the salts concentrate, I wonder? The lowest points of said bodies, perhaps? Certainly a couple billion years of wind would distribute stuff a bit, but still. Things roll downhill.

So I'm not putting a high probability on this one, based on the evidence presented. I'll suggest an alternate hypothesis, along with a way to test it:

Plate tectonics never really got started on Mars, but it almost got started along the Valles Marineris, which extends out into, guess where? The Acidalia Planitia. Acidalia's got a lot of darkish lava that looks like it came up through (ground)water, as does the floor of Hellas. I'd suggest simply that Mars' core almost boiled long enough to get plate tectonics going, but cooled too quickly to get continental plates forming (and note that continental plates are great elemental concentrators, contrary to what the article in question says. Absence of continents isn't a good thing if you want to mine veins of ore). The lowest places were the last places to stay wet, and as a result, they became salt concentrators, as well as areas where materials came up from the crust. The resulting evidence looks a bit like a nuclear reactor blew, but then again, most volcanoes kind of look like nuclear reactors blew.

Finally, I'd suggest a simple way to disprove the Martian reactor hypothesis: check ALL the elements that are concentrated in Acidalia and to a lesser extent at Hellas. If the only concentrated elements are products of a putative reactor, then we've got evidence. If everything that could dissolve in water is concentrated in those areas, then you're looking at dried oceans, nothing more, possibly with a stillborn mid-ocean ridge and subduction complex sitting in the middle of it.

As for natural reactors on Earth, Oklo's the only one we know of, but given that radioactive elements were much more common a few billion years ago when Oklo formed, it's a straightforward prediction that there were quite a few others. Plate tectonics simply did away with the others over the intervening 1-2 billion years. Remember that vanishingly little surface rock survives from that period. Any other reactor evidence is over a kilometer underground.


The 4x10^24J of energy he postulates in the explosion is a LOT of energy and a mere kilometre of rock isn't going to contain or constrain it until the entire mass fissions simultaneously. For comparison the Sun emits 3x10^26 J per second so this purported supernuclear explosion is equivalent to about 10mS of the sun's output. Very little of Mars would be left if such an event had happened, not just a large crater.


Speaking of big bangs on Mars...

Could a Comet Hit Mars in 2014? (via /.)


To imply that the whole radioactive mass exploded at once is a bit dubious. One idea could instead be to consider the idea that the natural nuclear reactor is in equilibrium with the water reserves and is the equivalent of a gigantic mine fire, like Centralia PA, or ultimately, at Burning Mountain in eastern Australia.

That one has been running for 6000 years.

That would evenly distribute radioactive isotopes into the atmosphere, and weirdly, would also provide a constant, steady heat source around which a civilization might build itself as the atmosphere faded.


So, when are we getting our act together and seriously doing Spaceguard?

Nature keeps giving us these warning shots...

I don't care if this is implausible, it's bloody going in a novel. OK?



I like the part about 0.14 cubic kilometres of pure (U, Th)O2 -- a 1.4-billion-tonne deposit, on a world with not much running water -- justified by reference to a 140000-tonne deposit on Earth.


Sorry, the reference was to the Oklo deposit, said to be 70 percent pure (I don't know if that is right). I was mixing it up with the very large unconformity deposits in Northern Canada, one of which is about 140000 tonnes 20 percent pure.


Now that I'm aware of this, we've started bouncing it around in a subset of the open-source nuclear people.

First guesses - he got something wrong, as the reaction cross sections for all the fissile and fissionable isotopes involved should (waving hand, without resorting to a long dive into T2 and a pseudo-infinite physics model) ... should I think ... have higher cross sections with moderated reactions (i.e., wet) than fast fission-spectrum neutrons. I.e., anything that's just barely critical with some water, generating heat, should become less critical under any circumstances as the water's driven out.

I think. As does one of the compatriots. But I think we need to model some stuff.

Modeling including the oxygen, isotope mixes, etc. and potential water-loss scenarios might turn up a counterintuitive result.


Sounds a bit like a Larry Niven essay, "The Roentgen Standard". Which is up on his website.


This raises an interesting question. Was there anything in Mister Brandenburg's writings about DC-8 lookalike aircraft?

My coat? It's the grey one on the hook over there ...


foreign aid delivered by ICBM!


Slightly off topic perhaps but it occurs to me (and surely not only to me) that IF there ever was a martian civilization it would have existed about the time when the Red Planet had an atmosphere and liquid water, which apparently was WAY before this event supposedly took place.

Assuming that this civilization flourished and then became extinct, would any evidence of it still remain today, several billion years later?

(and no, I don't believe it existed, but it is a funny little idea to play with in one's head)

Also, if it really is the result of plate tectonics that has kept Earth from experiencing a similar cataclysmic event, perhaps we can put up with the occasional disastrous earthquake now and then?


Funny thing, I just stumbled across the same link at a more reputable source.


The duke himself figured it was the 6 pack a day habit that did for him, the cancer statistics for that movie crew correspond roughly to average statistics


I wonder if an explosion that large (assuming it happened) could have caused some axial tilt? In adition, wouldn't there be evidence of a lot of local bombardment at rock fell back to Mars? Certainly there are some interesting implications.


We have a few posters here talking about disassembly: wrong order of magnitude. Kilograms of fissionables fly apart; tons are a little more reluctant, and a hundred-thousand ton mass of (say) 70% fissile and fissionable fuel is going to stay put until an expanding sphere of severe compression and neutron flux has propagated outward to a surface which can throw off energy by mass ablation and radiation.

That surface is a long, long way away.

Note the phrase 'expanding sphere' - this is not an expanding thin spherical shell, the interior is going to stay hot, and severely compressed, until the outer surface of heat escape and ablation propagates all the way in again: and, like the tamper of a thermonuclear weapon, the outward momentum of the vaporising surface will have an equal and opposite inward reaction - a force of compression that holds the critical mass together for more than long enough to ensure a complete 'burnup' of the fuel.

We are talking about mountain-sized masses, and 'disassembly' is irrelevant. The question is one of fuel enrichment and purity - and no natural 'control rods' of neutron absorbing isotopes - and of the volume of near-critical material around the initial criticality.

Admittedly, the assembly of such a structure is extremely unlikely - Oklo's hydrological cycle is instructive here - and I am skeptical that such a thing occurred on Mars, or anywhere else. But a mountain-sized mass of weapons-grade material will not disassemble until it has achieved a fuel burnup that would make a weapons physicist weep with envy.


Please feel free to contact me offline but...

It's not going to be a highly compressed mass of material; it's going to start out at standard pressure conditions and density. This is a bulk mass just sitting there.

That said, the two modes of expansion - bulk expansion in the material, and rarefaction waves moving down from the solid surface above the hotspot / reaction zone, will react significantly differently for something that approximates a pseudo-infinite body (mean diameter greater than 500 meters) where the tamper thickness is "a kilometer" rather than a few centimeters. Trivial review of the Serber equations will reveal that were you to assemble a 500 meter diameter supercritical sphere with km thick tamper the reaction is going to come close to completion.

The entire problem is that this seems to assume something backwards about relative reactivity of moderated and fast fission systems. As a general rule, for nearly all of the materials in play (U-235, Pu-239, other Pu isotopes, Th-232, U-233, U-234, etc) the cross section is maximized for thermal neutrons. The only exception is U-238 where the cross section is both much lower and starts to ramp up only at energies at / off the top of the significant fission neutron energy spectrum.

Essentially, without further evidence, what should happen as heating causes the water to withdraw / be driven out, is that reactivity should drop. The system should be self-correcting - if water intrudes and it starts to react, the water being driven out by the heat should reduce reactivity back to subcritical. If it ever reaches moderated supercritical, it would start a reaction pulse, but as the temperature rises the water should be driven out and reactivity drop due to the decreases in cross section at increasing temperature effects, the lower fast fission cross sections, etc.

The reaction time for the fast reactions is much faster, but the lower cross section is really the dominating factor. Once the hydrogen's driven out, reactions should cease.

This is a wild generalization, not backed up by a computer model, but even making the system pseudo-infinite shouldn't change the underlying physics. If a model can be constructed where driving out the water or increasing the temperature increase reactivity, then this could conceivably explode. I can't see how to make one do that under the circumstances, but that's not saying it's not possible. Someone who's arguing that this is a possible explosion mode (the planetary scientist in question) needs to fill in the details on what reaction model he's assuming, or someone else could.

If that matches an increasing-reactivity-by-withdrawl behavior, then that's valid. If not, then not...


And suddenly my Martian Ancient Astronaut novel just got way more plausible. And interesting. Science is so fucking cool!


We're talking about different things: you're discussing the steady-state or hydrological-cycling state of the natural reactor; I'm discussing the detonation of a prompt criticality. Hence the reference to the explosive 'disassembly' of small man-made warheads, and my assertion that a prompt criticality in a mountain-sized mass will not disassemble until long after a complete fuel 'burnup'.

You are discussing reactors, I am discussing a nuclear detonation.

I believe you are correct in your assertion that a natural reactor operates at or near Standard Temperature and Pressure; and, further, that increasing activity will eventually lead to expulsion of the moderator and reactor 'shutdown' which limits the progress of natural fuel 'enrichment', even over repeated cycles of reactor operation and moderator expulsion.

There is some space for doubt in the possibility that our models for Oklo's hydrology are not applicable to Mars, with added points of difference in the sheer scale of Martian 'geological' structures, and the extraordinarily long timescales available in a stable non-tectonic environment. Some space, but not much.


As far as I can see, there are two different things that have to happen:

1: Assembly

A series of processes which produce a large and dense volume of fissionable material, deep below the Martian surface, without actually going bang.

2: Disassembly

This is where things are supposed to go bang, distributing traces of particular radio-isotopes across Mars to give the claimed pattern.

Problem 1: the pattern may have other causes. I am also not sure of the basis for some of the assertions. For instance, it is only this week that a human instrument on Mars has managed to take a sample from inside a piece of rock.

Problem 2: can we really assume a bang, when the event may be under a kilometre of rock. The original paper just talks about a very hypothetical total energy released. What would it look like if we had an assembly of low-grade bang-stuff, and a thousand year fizzle?

Problem 3: I still don't get how so much fission-stuff could get into one place to start the whole thing off.


P2.3 - Neither do I; it takes some reasonably clever engineering to keep 2 just sub-critical masses of, say, U235 sufficiently far apart that they don't go critical until you want them to, and stay close enough together to develop the critical chain reaction when yuo do want them to.

Which is pretty much the sum total of my knowledge on "how to make a fission bomb".


hairyears: We're talking about different things: you're discussing the steady-state or hydrological-cycling state of the natural reactor; I'm discussing the detonation of a prompt criticality. Hence the reference to the explosive 'disassembly' of small man-made warheads, and my assertion that a prompt criticality in a mountain-sized mass will not disassemble until long after a complete fuel 'burnup'. You are discussing reactors, I am discussing a nuclear detonation.

I am probably one of the 3-4 people in the world who will willingly publicly have this conversation with you who know that difference best. This conversation is taking me away from (in my spare time, not day job) modeling of Iran's R265 and possible North Korean weapons based on their test explosions evolutionary sequence and very very sketchy public statements on their designs.

My point is this, in short:

Anything like this that could fast-fission detonate would seem to moderated-fission react first. Not just because of the water; see graphite reactors, moderated assemblies using only UO2, etc.

I cannot see any situation where such a natural system could be more reactive in a fast-fission reaction than in the moderated fission reaction.

In other words, anything that started to assemble towards a configuration that could fast-fission detonate would first start reacting in a moderated fashion, which is almost certain to 'defuse' further concentration. I see no way for something that's just barely moderated critical to become supercritical for fast reactions.

Unless you believe you can assemble a non-moderated-fission-critical mass in a way that can reach prompt supercritical for fast fission, using just natural geology, the explosion you posit requires starting conditions that are naturally impossible.

What we do inside nuclear bombs is not vaguely natural, of course.

If you can posit the configuration and assembly method to avoid the intermediate moderated reaction and self-disassembling tendency of the ore body at moderated criticality, I am all ears.


[quote=Charlie Stross] Note: I make the explosive yield (per page 2) about 36 billion megatons.

In other words, around a billion times more powerful than the entire combined Cold War nuclear arsenals; somewhere in the dinosaur-killer asteroid league table. [/quote] Not to be picky, but I'm sure the cold-war arsenal was bigger than 36 megatonnes. Did you mean a million times greater?


How to get a manned mission to Mars before the end of the decade:

1: Devise a "natural" sequence of nuclear reactions that depends on the presence of large quantities of crude oil. and climaxes in a long-lasting "fizzle".

2: Elect Dick Cheney.

The details are left as an exercise for the reader.


In what must be an element of extreme irony...

There's now a noticeable computed probability (experts quoting 0.01 to 0.03% chances in the last day) that comet C/2013 A1 will hit Mars on Oct 19/20, 2014. Error bar is significant, but the planet is clearly near the center of the estimated trajectory.

It has what is estimated now at an 8 km diameter nucleus and would be moving at 56 km/s relative to Mars.

It WOULD have energies in the range we're talking about now, and would approximate the Chixulhub impact energy, on a much smaller planet.

At the very least, it looks like Mars will get saturated by comet debris and gas cloud. Not clear yet what that will mean. For example, for our rovers there. At the very least we'd expect they'll get a lightshow. It might be worse.

Not clear yet what side of the planet will be facing the direction the comet is coming from. Trajectory variance in time as well as space is at factor, though we will hopefully get some data soon.



Before anyone says "Areforming! Elebentee!!", consider these numbers:- Mars'area ~145000000km^2 Comet'volume ~8^3km^3 for 512km^3.

If we presume 100% efficient convertion of comet to gas, and that the comet is 100% H20 and 02, this gives an extra gas layer some 0.003mm thick.

It looks like we're going to need a bigger comet!


Either you or my calculator are misplacing a decimal point, because I get approximately 3 * 10^-6 km thick, or 3 millimetres.

If that's solid water ice, and it were to evaporate, then vapour thickness at STP (yeah, right) would be about a thousand times as much, or three meters.

So yes, we do need a bigger comet. But not as much of a bigger comet as it might have appeared.


In the 1800s, William Thompson calculated the age of the Earth to be in the millions of years,

based on the fact that the core is still hot. In fact, it is much older and the core is only molten because of decaying Uranium. Thus Mars failed to maintain a magnetic field and thus an atmosphere, and to develop plate tectonics, mainly because it has less Uranium rather than because it is smaller. Though its probably more complex, and it's smaller because it has less Uranium or something.

So, it's probably unlikely that Mars would really have enough Uranium for all that.


Err, there are several problem with this argumentation; first of, AFAIK there is no relation between relative content of uranium and size in planets; to go for a gross example, gas giants are the biggest, but I guess their relative uranium content is quite low compared to hydrogen and like. For absolute content, well, that depends...

Second of, Mars has about half the radius of Earth, so it has only (1/2)^3=12.5% of Earth's volume. Which would translate to 12.5% of the radioactive decay energy. OTOH, it has (1/2)^2=25% of Earth's surface area.

So there is much less heat energy for Mars, but not that much less surface area the heat energy is radiated from.

Err, hope I got that one right, biologist and physics, never the twain...–Boltzmann_law

Not just because of the water; see graphite reactors, moderated assemblies using only UO2, etc.

What about a natural RBMK reactor? At least there'd be a positive void coefficient. While graphite is often of biological origin, lump graphite seems to be more or less abiogen, e.g. hydrothermal. Where this could also deposit uranium and thorium, especially if geochemistry and/or a deep biosphere help with it. Since we already have water, this'd be quite similar to a Chernobyl type reactor. Compared to the smaller man-made variant, the high pressure would mean it takes some time to get to the "positive void" part. One could also posit that first there is some other neutron absorbant slowly removed from the system.

As power output increases, we're closing in on the critical point, where small changes of temperature or pressure can lead to big changes in density:

Which might translate to a sudden decline in neutron absorption.

For the specific scenario, breeding materials like thorium, nuclear poisons like xenon created by the reactions and the exact timing might matter.

For the latter, imagine the reaction goes critical in the periphery, where pressure is low. Positive void coefficient means we get more energy, but not much because the assembly is destroyed in some small bang. The energy ledas to the material somewhat more inside becoming supercritical, too, this assembly is destroyed, too, but pressure etc. from the first reaction means there is a little bit more energy released and so on. In the end we get a shock wave going in, just like a implosion type nuclear weapon.

Just some musing, as already said, biologist and physics, never the twain...


On another note, U-238 might stabilize the system thanks to higher temperatures leading to higher neutron absorption, and when it's spent, we have some plutonium in the fire.

Another thing, condition like on Titan might mean no water, but plenty of hydrocarbons. Which might act as a neutron moderator:


Now we're getting somewhere...

CANDU reactors also can have a positive void coefficient in the coolant, for similar reasons (marginal moderation behavior doing spectrum shifting in the presence of a stronger main moderator; for RBMK the graphite, for CANDU the D2O.)

Those only work in carefully constrained geometries; the distance between fuel elements has to be far enough for the moderation to work, but not too far or capture and moderation below the peak cross section neutron energies happens. Homogenous solutions of D2O, H2O, and Uranium salts are entirely different than a CANDU reactor, for example.

I don't think you can naturally match those conditions, and certainly not across a roughly 500-meter across ore body.

Solution reactors - liquid salt critical assemblies and the like - happen. A lot of criticality accidents happen that way, during processing. The big one in Japan a few years ago for example. But they all have negative void coefficients.

CANDU and RBMK are only positive coeficients so long as the main moderator (the graphite and the D2O) remain more or less intact. If you disrupt those, their reactivity drops right quick.

H2O is an OK moderator; D2O is often nearly ideal; Li-7 is ok (Li-6 is an absorber, but turns to Li-7 when that happens); He is ok but gas at any relevant pressure and temperature; C is ok, and various C-H materials have been used. Be is not a bad choice. O has been considered here and there (and moderating effects of oxygen in fuel oxide pellets are noted). F is not great but as a chemical constituent in other moderators works ok.

Finding a material and geometry on Mars that could give you a background moderator, in combination with a secondary moderator you could then boil off / remove / etc leading to positive void coefficient, seems difficult. But not impossible. Unlikely is not impossible..


Numerous problems with this "paper", and it pains me to see it being spread around without a bit of research done first.

1) It's a two year old LPSC (Lunar and Planetary Science Conference) abstract. LPSC abstracts have not been peer-reviewed at all.

Now, I would strenuously like to point out that the vast majority of abstracts submitted to LPSC are perfectly good science written by legitimate scientists and students of science. Most of them are ideas in progress that may or may not be later developed as papers, and some are presented at the conference as posters or talks. Some of them get shot down, some pass scrutiny - that's how science works (I submitted and presented a few myself while I was at university - some worked out, some didn't).

However, a handful of kooky ones do slip through every now and then (because they're not vetted after they're submitted) - I remember reading one abstract a few years back suggesting that the sun had accreted around a neutron star - it was clearly nonsense written by someone who had no idea about star formation or anything. This nuclear reactor abstract is one of those kooky abstracts, and here's why:

2) The first reason to be suspicious is usually if the author isn't associated with a university. It's not always the case, but many of those who submit from a 'company' or independently are "on the fringe", so to speak. In this case, a bit of digging reveals that the author is (or was) a Cydonia nut - the only other thing I found by him was a "paper" written about how the Face on Mars was built by an alien civilisation. While this doesn't necessarily mean that the science presented here is bad, the author's credibility is already suspect.

3) Obviously, the proper way to proceed is to analyse the science. Unfortunately, even disregarding the author's credibility, the science in the paper is pretty terrible and full of holes. A few issues I found: a) he says that the "reactor" was "tamped" by the overlying rock but doesn't provide any calculations to support this (and for all I know forgot that Mars has lower gravity than Earth, so pressure is lower at a given depth). That's a fairly critical part of the scenario that we just have to take his word for. b) He also doesn't explain how uranium ore forms and concentrates on Mars in the first place. c) He doesn't provide any evidence for this supposed explosion beyond "it looks like the interesting stuff is concentrated around a depression" which could have been caused by a number of other means. Occam's Razor seems nowhere to be found. d) We know of precisely one natural nuclear reactor on Earth, which implies that they're somewhat unlikely to form... and it didn't blow up. And yet there was supposedly one on Mars that did? Seems like a lot of unlikely coincidences would have to line up to make that happen on the next planet over from us. e) And he spends a lot of time telling us his interpretation of the data, and not a lot of time just objectively describing the data and saying what other options could explain it. f) if this happened so long ago, why would there be evidence left on the surface after a billion years of aeolian deposition and erosion and redeposition?

And this is before I even get to the nuclear physics side of it... (I'll leave others to do that)

So all in all I'd say the "evidence for a nuclear reactor on Mars" is actually pretty darn sketchy!


@ 45 - 47 Is there an easily accessible site giving updates on this close pass / possible impact of comet with Mars? I assume one of the NASA / JPL / Astronomical ones, but where? Incidentally, what happens if comet misses Mars, but impacts Phobos? Also, if major impact does occur, what is our orbital position relative to Mars at that point - are we in danger of splashback?


Cancel that!

"C/2013 A1" in google gets me all the info I need ....


The main problem I have with this happening on Mars is that the wikipedia article on natural nuclear reactors mentions they have two constrains:

a) uranium is only soluable when there is some oxygen around, which means it takes some time with developing biosphere and such.

b) there has to be lots of fissile uranium isotopes around, which means it has to be early in planetary history, else the isotopes decay.

Which, first of, explains why natural reactors are somewhat rare, and second of, means some problems for Mars with the oxygen. Now there are quite some salts with oxidation numbers of chlorine too high to be comfortable on Mars, and there might be oxygen thanks to geochemistry or chemolithoautotroph organisms deep below. Still, something of a bummer.


Whatever happens, 19/20 October next year is going to be very interesting, from an astronomical-observation viewpoint, isn't it?


Constantine wrote, in part: a) he says that the "reactor" was "tamped" by the overlying rock but doesn't provide any calculations to support this (and for all I know forgot that Mars has lower gravity than Earth, so pressure is lower at a given depth). That's a fairly critical part of the scenario that we just have to take his word for.

I think that I mentioned some expertise in this. In this case, tamping is the correct term of art for material intimately in contact with the fast fission "pit" or mass of reacting fissile and fissionable material, which does not significantly act to reflect neutrons back into the bulk material, but does act to inertially retard pit expansion and disassembly.

Materials which return (some) neutrons to the core are reflectors, those that do both efficiently are tamper/reflectors.

Tamping works via two mechanisms:

First, by being in intimate physical and thermal transfer contact with the pit, it moves the surface expansion wave start point further out from the critical radius.

Second, to some degree inertially retarding the direct bulk expansion within the pit.

See for example the Nuclear Weapons FAQ and tamper-effects adjusted Serber efficiency equation.

On this one point, the useage is correct ( or appears so; the papers' details on the model are sketchy ). Not sure if the effect would matter for a fast fission reaction in a hundreds of meter wide zone if the conditions were to set one up; time and geometric size will be in play mostly. But it is tamped.

That does not change any of the other criticisms leveled so far.


Thanks george - to clarify: I'm not doubting that that tamping is a real thing, it's more the claim that "Because of the size of the ore body, and its burial at kilometer depth, the reaction was inertially confined or “tamped” so that explosive disassembly was delayed until a high degree of fission burn-up was achieved."

If that's the case, I would expect to see some numbers and calculations to demonstrate (a) that this actually is at kilometre depth, (b) whether the depth is actually enough to "inertially confine" the reaction, and (c) why it's such a big explosion that it penetrates the surface from kilometres so that it can spread everything across the planet. There is none of that in the abstract.


Yeah, I'd slipped magnitudes when I tried to do a re-normalisation in my head.


That's the original paper's claim, but papers at conferences can be rather flimsy pieces of science.

This one doesn't hold up to informed examination.



About this Entry

This page contains a single entry by Charlie Stross published on February 25, 2013 2:25 PM.

Roko's Basilisk wants YOU was the previous entry in this blog.

Dragged kicking and screaming into the 19th century is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.

Search this blog