Enceladus, Titan, and the Great Ammonia Mystery Options

[Guest_BruceMoomaw_*]

As promised: apart from the potentially gigantic biological significance of the Enceladus find, there are two big mysteries associated with it: what's driving such extraordinarily intense cryovolcanism, and why does Titan seem to have ammonia while Enceladus doesn't?

There seem to be four schools of thought about the latter, three of which look to me like serious possibilities.

(1) "Enceladus DOES have ammonia, but Cassini didn't see it because solar UV destroys it quickly." This is the theory of Baragiola et al in http://www.cosis.net/abstracts/EGU06/03150/EGU06-J-03150.pdf : "We will present results from laboratory studies on the radiation effects on ammonia–water mixtures pertaining to the environment of Saturn’s icy moon Enceladus. We show that ion irradiation destroys ammonia efficiently, and produces N2 that could be the source of N+ that has been detected in the exosphere. Warming the irradiated mixtures we observe outbursts of water and ice grains at temperatures much lower than those needed for sublimation of water ice. These radiation processes may explain the plume of water vapor and grains observed by Cassini at Enceladus." This was my own sentimental favorite when I went to the Europa meeting, and during F.J. Crary's presentation on the findings of Cassini's INMS at Enceladus I asked him about it -- but I think he disposed of it soundly in very short order by pointing out that, while this could explain the lack of frozen NH3 on the sufaces of Enceladus and the other icy moons, the gas analyzed by Cassini had been emitted from the vents only about 15 minutes earlier, which isn't time for most of the ammonia to be destroyed no matter how quickly the latter occurs.

Carolyn Porco's new "Science" article, I think hammers the last nail in the coffin by pointing out that -- since NH3 has a far lower boiling point than water -- if the liquid coming out of the vents was an NH3-water mix, most of the vapor detected by Cassini should have been NH3. Instead the article confirms that Cassini found none at all, confirming that it must at most comprise 0.5% of the plume gas. Enceladus really does lack ammonia -- which also means that the plumes really must be above 0 deg C, with all the biological implications thereof.

(2) "Contrary to initial appearances, Titan -- like Enceladus -- never had any ammonia either. instead, they both got their nitrogen, from the very start, as plain nitrogen." This actually fits better with the analyses of the heavier elements in Jupiter's air by the Galileo entry probe, whose ratios seem to indicate that the icy planetesimals that made up Jupiter were either much colder than expected, or were made of clathrates that were much more efficient than water ice at holding large amounts of molecular nitrogen and noble gases. ( http://www.lpi.usra.edu/opag/oct_05_meeting/probes.pdf , pg. 15-16)

Cassini's evidence that Titan does have large amounts of ammonia is twofold. Huygens found virtually no noble gases in the air (except for the Ar-40 produced by the decay of radioactive potassium in its rocky core), indicating that Titan was not made out of very cold ices or clathrates that could hold either those gases or N2; and Cassini's SAR images suggest large-scale cryovolcanism, which is definitely easiest to explain if it's due to a water-NH3 mixture erupting out of the ground (since this both has a very low melting point, and is less dense than solid water ice and thus capable of easily rising up through it). But Bar-Nun et al propose an alternative explanation for the lack of noble gases in Titan's air -- they claim that the rain of organic smog particles is capable of having long since scavenged all atoms of those gases out of the air and trapped them in the ground smog deposits ( http://www.aas.org/publications/baas/v37n3/dps2005/227.htm ; http://www.agu.org/cgi-bin/SFgate/SFgate?&...t;P34A-06" ). And, at the Europa meeting, Mikhail Zolotov told me that he seriously doubts Cassini's radar evidence of large-scale cryovolcanism on Titan, although he didn't say just why.

(3) "Titan, Enceladus, and Saturn's other moons acquired most of their nitrogen in organic compounds." This is Zolotov's belief. I presume he means mostly cyanide compounds, which Cassini apparently has sensed on the surfaces of the icy moons. ( http://www.aas.org/publications/baas/v37n3/dps2005/672.htm . See also Emily's log on Clark's DPS talk: http://redrover.planetary.org/news/2005/09...og_Special.html , Sept. 8 entry.) I haven't the slightest ability to judge the likelihood of this theory -- although I do note that Cassini didn't find any HCN in Enceladus' plume gas, either.

(4) "Titan and Enceladus really did start out with ammonia, but it was all broken down inside both of them by intense internal heat resulting from an early high concentration of short-lived radioiotopes inside their rocky cores." This is the view of Jonathan Lunine, Dennis Matson and J.L. Castillo ( http://www.cosis.net/abstracts/EGU06/05276/EGU06-J-05276.pdf , http://www.cosis.net/abstracts/EGU06/09655/EGU06-J-09655.pdf , http://www.lpi.usra.edu/meetings/lpsc2006/pdf/2219.pdf ), and they mesh it with their belief that the shape of Iapetus provides evidence that it too started out with a high concentration of such isotopes-- specifically, Al-26 -- in its rock ( http://www.aas.org/publications/baas/v37n3/dps2005/274.htm ; http://www.agu.org/cgi-bin/SFgate/SFgate?&...t;P21F-02" ). Once again, however, we have to assume that there is currently no ammonia-water volcanism occurring on Titan's surface, because "Preliminary calculation indicates that all of Titan’s initial NH3 can be processed [into nitrogen] within a few million years after formation."

As for Atreya's observations indicating that the outer planets themselves were made out of materials that never had any ammonia in them to begin with, Alibert and Mousis try to reconcile this with ammonia in Saturn's moons as follows ( http://www.lpi.usra.edu/meetings/lpsc2006/pdf/1141.pdf ): "We...favor an alternative scenario where Titan may be formed from satellitesimals that would have suffered a partial vaporization during their formation and/or migration in Saturn’s subnebula. The migration of satellitesimals in a balmy subnebula (as our model shows at intermediary epochs) could allow a partial or total vaporization of most volatile species (CO, N2, Kr, Xe) whereas CH4, CO2, NH3 would remain trapped in water ice."

So: does Titan really have cryovolcanism indicating that ammonia still exists in its interior, or not? That is the question, and Cassini may not be able to answer it by itself. ( http://www.lpi.usra.edu/meetings/lpsc2006/pdf/2136.pdf , unfortunately, indicates that Cassini's VIMS just does not have the ability to identify ammonia ice through the limited spectral windows allowed by Titan's atmosphere.)

Dominic Fortes ( http://www.lpi.usra.edu/meetings/lpsc2006/pdf/1293.pdf ) suggests an interesting twist: since it's likely that large amounts of sulfur has been leached out of Titan's carbonaceous-chondrite core rock by the water initially trapped inside it (as with the other moons of the outer planets), it may well have reacted with the ammonia dissolved in Titan's subsurface ocean to turn it into ammonium sulfate. He also suggests that ammonium sulfate "is a credible candidate" both for the mystery substance detected on Titan's surface by Huygens' near-IR spectrometer, and for the 5-micron "bright spot" Hotei Arcus. But such a solution would have about the same melting point as plain water, and a much higher density than ice, making it a lot harder for volcanic processes to drive it to the surface. Fortes suggests some mechanisms that might conceivably do so, at least in limited amounts -- and presumably these could work for plain water, too, to explain how Titan might have at least some cryovolcanism without any ammonia.

I wonder, vaguely, about a fifth possibility: it it possible that Titan still has large amounts of ammonia inside it but Enceladus doesn't, because Enceladus has recycled its inner liquid? Porco's "Science" paper concludes that "only ~1% of the upward-moving [ice] particles [in the plumes] escape to supply the E Ring" -- the rest falls back onto the surface as snow. If so, then the mounting weight of that snow on top will force the local ice crust back down into Enceladus' warm interior to remelt its lower layers into water. This explains why Enceladus hasn't dried up completely from these eruptions over the eons (although its high density suggests that the eruptions have stripped it of much of its initial ice).

But if so, then when the eruptions started, Enceladus might have had a layer of liquid water-NH3 mixture inside it -- which, since it could stay ductile at low temperatures, would have made it easier for some kind of heating to start up the eruptions in the first place -- and then, when this mixture was expelled onto the surface, the ammonia was quickly destroyed, leaving just water snow behind on the surface. Thus, after a relatively short period of ammonia-water volcanism, Enceladus would have depleted most of its ammonia (leaving only the small amount which may have been broken down by its internal heat into the reservoir of internal nitrogen which accounts for the 4% N2 in its current vapor plumes). But tidal convection -- once some kind of heat source has started it in the crust of an icy moon -- strongly tends to amplify itself and keep itself going, because the friction of ductile material churning convectively in the core (whether it's warm ice or some liquid) generates additional heat and thus warms the material further. And so the very frictional heat of the initial eruptive venting could produce enough additional warmth to keep Enceladus' eruptions going even after the ammonia was gone and the only liquid left was plain old easily freezable water. By contrast, there aren't any processes on Titan's surface that destroy ammonia erupted onto the surface -- its air pressure keeps any frozen NH3 from vaporizing, and its organic haze is a good shield against solar UV -- so its remaining ammonia keeps on getting recycled back down into the subsurface ocean instead of being destroyed as Enceladus' ammonia was, making it possible for cryovolcanism to continue on Titan even if it isn't as powerful and thus as frictionally self-heating as Enceladus' volcanism is.

Whichever of these explanations for the Great Ammonia Mystery is true, we still have the other huge Enceladan mystery: what in the world started up such a violent eruption in the first place? Porco's "Science" article says that it's unlikely that Enceladus' orbit was ever eccentric enough to produce ice-melting heating by itself, but they propose an alternative: a chance giant impact. They conclude that Cassini's data on Enceladus' shape suggest that at one point it was librating four times during each orbit -- perhaps quite violently -- and that it could have been put into that resonant cycle by a giant impact, although it has since settled back down. "An initial [libration period to spin period ratio of] 0.26, with an associated libration amplitude of 22 degrees, yields a heating rate 100 times that due to the forced eccentricity today for typical fully elastic models and the same temperature-independent Q=20." As for what's kept the eruption going since that violent libration was finally damped out and disappeared: "...[I]t is plausible that Enceladus may be sufficiently heated today by some combination of the above mechanisms [tidal heating from its current slightly eccentric orbit, in which it is kept by its 2:1 resonance with Dione; and heat from the remaining long-lived radioisotopes in its rocky core] to explain the observed south polar venting, provided it underwent an early epoch of intense heating and, once heated, retained a low Q up through the present in some portion of the interior."

Lunine, Matson and Castillo instead blame the eruption's start on a concentration of Al-26 in Enceladus' rocky core. They also think that there's no way that tidal flexing or convection in the ice alone can produce enough frictional heat to keep the ice melted, and that instead what keeps the eruption going is frictional self-heating from a reservoir of magma in the rocky core itself, whose viscosity does create enough tidally generated frictional heat in Enceladus' current orbit to keep the magma molten. "The [tidal] dissipation factor in the core [for such magma] for tidal frequencies is Q~1-10, and this can provide up to 100 gigawatts, compared to 0.5 GW provided by the decay of long-lived radionuclides." But -- in their abstracts so far -- they don't seem to explain clearly why such an initial high concentration of short-lived hot radioisopes in the moon's rock didn't also happen to Saturn's other icy moons. In fact, in http://www.lpi.usra.edu/meetings/lpsc2006/pdf/2351.pdf , they puzzle over the fact that this hasn't happened to Mimas, since its orbit is more eccentric: "Enceladus’ present eccentricity is six times smaller than Enceladus’ and the tidal dissipation for a given Q and k2 is 40 times less at Enceladus...However, the fact that intense endogenic activity is observed at Enceladus, which contains twice as much rock as Mimas, might indicate that the difference in evolution is the result of a thermal event in the history of Enceladus that created the conditions suitable for significant tidal dissipation to start and be sustained." If I understand them correctly, they're saying that it might be simply because Enceladus has more rock in its core than the other small icy moons. ( http://www.lpi.usra.edu/meetings/lpsc2006/pdf/2200.pdf : "Rock mass fraction ranges between 3% in Tethys to 57% in Enceladus... Whether conditions are intense enough that endogenic activity can start and be sustained until the present is a function of several different factors, depending on the differences in size and rock mass fraction between the satellites and their orbital evolution (especially eccentricity damping)." And http://www.agu.org/cgi-bin/SFgate/SFgate?&...t;P32A-01" : "We note that in Enceladus and Titan conditions might have been such that the boiling point of water was reached and water might have been lost very early in the history of these satellites.") But this leaves us -- once again -- with the possibility that it might instead be a chance giant impact at some point in Enceladus' history that first created its eruptive heated region, which has since sustained itself through continuing positive-feedback tidal-friction heating in its ice, its interior rocky core, or both.

In any case, the fact that the eruption site is now at the south pole is easier to explain; Bob Papplardo points out ( http://www.lpi.usra.edu/meetings/lpsc2006/pdf/2113.pdf ) that, wherever the eruption started on Enceladus, the lower density of the moon's internal material in that region would likely make plain old centrifugal force and Saturn's tidal tuggings cause Enceladus' polar axis to wander until the lowest-density region was at one of the poles. Moreover, if an internal layer of liquid separates its outer crust from its mantle, the crust would tend to reorient in this way without the moon's interior doing so -- which, since Enceladus is slightly oblate, would cause the crust around the eruption region to get somewhat squeezed together as it moved to the pole, producing compressional tectonic features of exactly the sort seen around Enceladus' current south pole ( http://www.lpi.usra.edu/meetings/lpsc2006/pdf/2182.pdf ).

[Bob Shaw]

Bruce:

Thanks for all that! Very interesting overview!

As regards Enceladus Vs Mimas, and the big impact explanation for the 'kick-start' of the geysers, another aspect is, of course, Herschel on Mimas. That crater is 130 kilometers (80 miles) wide, one-third the diameter of Mimas, and the impact is said to have nearly disintegrated Mimas. So, with a 'kick-start' like that, *and* more tidal energy, why oh why isn't Mimas displaying geysers too?

Very strange!

Bob Shaw

[Guest_BruceMoomaw_*]

You WOULD ask the major question I hadn't thought about, wouldn't you? Hmmph. Well, we can perhaps assume that the Herschel impact didn't happen to throw Mimas into the kind of rapid rocking-horse libration that Porco's team thinks may have played a major role in initially stirring up Enceladus' big-time tidal dissipation. Alternatively, maybe it took a combination of both a giant impact AND the extra heating from Enceladus' rocky core (which makes up 57% of its mass, versus 27% for Mimas' rocky core) to set off the process.

Or maybe this really does serve as evidence that Lunine & Co. are correct in implying that it was really just the larger size of Enceladus' rocky core, and thus the larger amount of Al-26 initially stored in it (which has a half-life of only 20 million years and generates a lot of heat accordingly) that determined Enceladus' fate. (On this point there's still another caveat, though: do we know that Enceladus' rock initially constituted such a big share of its total mass, or has that moon hemorrhaged away a big part of its initial ice over the eons? Questions, questions -- and proably no firm answers to them for a long time.)

[dvandorn]

That was a really fine summary of the current thinking on Enceladus, Bruce. I knew there was a good reason for wanting you around here...

As you say, though, pinning this issue on the relative ratio of rock to ice in Enceladus vs. the other Saturnian moons may be a McGuffin, in that Enceladus has been ejecting significant amounts of mass for an unknown period of time. And it has been *preferentially* ejecting mass -- you don't see any of that rocky core getting ejected in these plumes. Just the ices that cover the rocky core.

I'm positive that one could develop boundaries from currently available data for the rate of mass ejection that's happening right now, and from that come up with a time scale that tells us when the ratio of rock to ice in Enceladus will rise from 57% to 58%, and thence on to 60%, etc. Assuming, of course, a steady rate of mass ejection. One could then work that backwards and determine how long, at present mass ejection rates, it would take for a Mimas-sized moon (for example) to dissipate itself down to the size and rock/ice ratio we see at Enceladus today.

My gut feeling is this -- we see several rather similar Saturnian moons, with relatively small rock cores and large ice mantles, and then we have this much smaller moon with a much higher ratio of rock to ice than the others. And this much smaller moon is, right now, actively ejecting mass -- something the other moons aren't doing, and don't show any major signs of having done in the past. That tells me that Enceladus started out very much like Rhea, Tethys and Mimas, and has dissipated itself down to its present size over billions of years due to this continuing mass ejection. It explains why Enceladus is unique in its size, mass and rock/ice ratio when compare to the other icy moons of Saturn.

-the other Doug

[ugordan]

(On this point there's still another caveat, though: do we know that Enceladus' rock initially constituted such a big share of its total mass, or has that moon hemorrhaged away a big part of its initial ice over the eons?

As Porco et al. estimate that only 1% of the ejected material does escape Enceladus, that would imply a really long time these eruptions were going on in order to lose that much mass. Which then adds another point to the story -- if Enceladus was once much more massive, then the loss rate would probably have been even lower than 1% due to its much larger mass (assuming the venting wasn't also much more violent in the past, which is yet another question).
If the eruptions indeed have been going on for aeons, that again raises interesting biological implications, but that's beside the point here. If Enceladus did indeed turn itself inside-out in the process, here we have a good look at the majority of the interior of an icy moon, apart from the rocky core. One could then conclude that the ice either is very pure in the whole outer shell in all moons or the heavier organic and inorganic compounds just sink lower to the core and never get expelled.
We really do need gravity passes to at least get a hint what's actually going on below the surface!

[Gsnorgathon]

Bruce - thank you for that wonderfully detailed post.

Rather than follow Bruce's example of well-thought and organized prose, I'll just spew (Enceladus-like?) a few random bits out and see if any of y'all think they're worth noting.

How old is the oldest terrain on Enceladus? If it has turned itself inside out, there's a limit on the last time it might have done so. Those relatively densely cratered northern plains have long had me scratching my head; the big impact scenario does a lot to soothe that particular itch.

I'm wondering if perhaps one reason Mimas might not be active is because it got smacked hard enough to knock it completely apart. Just how differentiated is Mimas's interior? Disrupting a body could cause it to lose a good bit of its interior heat, and prevent its "convective engine" from starting back up again when it re-accreted.

[Guest_Richard Trigaux_*]

Thank you Bruce for your detailed thoughs about icy moons. When I saw such a long post, my first idea was that it would be too tedious to read completelly. But I found myself reading it entirely.



There is a problem about the idea of Enceladus having lost most of its mass in the past:

if so, the surface would have shrinken too, and the ancient terrains would show shrinken craters, for instance become oval or elongated. It don't seem to be the case. So it is clear that there was not a tremendous diminution in size, Enceladus was never Mimas-sized. Of course a close studdy would perhaps reveal some shrinking faults, but not many. For instance if Enceladus had shrinked half of diametre, its surface would habe be divided by four, and we would see many craters cut by faults, half-moon shaped, or their middle parts absorbed underground, with only the opposite edges remaining. There are indeed many faults on Enceladus, but I don't see things like this. (except for the margin between ancient and recent terrain, but in this case it don't account as shrinking faults)


So we have to find other explanations to Enceladus heating. On another thread it was said that small objects could be easily tossed away. If we consider a couple like Mimas-Enceladus, Mimas would toss Enceladus, but Enceladus would have little effect on Mimas, this explaining that Enceladus produces gravitational flexing heat, and not Mimas. Relevant too seems the idea that flexing heat would occur in the rocky core and not in ice, although we can imagine heating of ice, below zero, but enough to create diapirs of rising ice. Relevant too is the idea that the Enceladus heating we see today would not be permanent, but a relatively short episode of heating.

[The Messenger]

Thanks Bruce, this was worth the wait.

It is good to see so many alternatives explanations for the plumes - this is the way it should be when the physical model is so ambiguous. - be sure to check out the art in the art/geyser thread.

And the evidence of ammonia in Titan is still circumstantial, and will likely remain that way for the balance of the Cassini mission. Wouldn't it be fun if we could scrap all this manned moon mission crap and dedicate an equal amount of spending on planetary missions? We could be scrambling right now to build a heavy lift rocket to saturn.

[Guest_AlexBlackwell_*]

I know this is going to shock you, Bruce, but I have to agree everyone else in this thread: that was a very nice summary.

[Guest_BruceMoomaw_*]

Oh, I'll convert you all to recognition of my genius at some point.

Now, a question: what's Enceladus' total mass? I haven't found that yet on the Web, but I do know that 57% of its mass is rock -- so this should allow us to calculate just how much of Enceladus' total quantity of ice it has lost to the E Ring over the eons, assuming that the geysers have continued erupting at their current rate. (Or at least set an upper limit to it, since some of the E Ring particles must eventually recollide with it. Given how far they disperse to form the E Ring, though, and the fact that radiation and cosmic dust will quickly vaporize them, I imagine the share of the E Ring that recollides with Enceladus isn't all that large. Obviously there's a scientific paper just waiting to be done on this topic alone.)

Indeed, the assumption that the geysers' current eruption rate is their long-term average is risky -- it's now clear that the huge surge in E Ring content that Cassini's UVIS saw when it was approaching the planet in Feb. 2005 was probably Enceladus giving a major burp. (It might have been a big meteoroid impact against Enceladus, but the odds are very much against it -- among other things, that would have been just as likely to have happened to any of Saturn's other inner moons instead.) But it's also conceivable that the geysers shut down well below their current activity level for long periods. (On that point, however, note that the huge Feb. 2005 increase in E Ring material had totally vanished again within just a couple of months, which gives us some idea of how very fast E Ring material DOES vaporize and disappear -- surely only a small part of it recollided with Enceladus. Since no observers have yet noted the E Ring vanishing or shrinking in a major way for any period, the geysers have probably not shut down during the last few decades.)

[edstrick]

Note that the e-ring material seen by the UV spectrometer is disassociated gas atoms (I believe) and not the few micrometer size ice particles seen at optical wavelengths. Careful analysis at different phase angles and with different filters over the course of the mission will be needed to detect any but the most massive variations in the ring's solid particle population.

[scalbers]

Oh, I'll convert you all to recognition of my genius at some point.

Now, a question: what's Enceladus' total mass? I haven't found that yet on the Web, but I do know that 57% of its mass is rock -- so this should allow us to calculate just how much of Enceladus' total quantity of ice it has lost to the E Ring over the eons, assuming that the geysers have continued erupting at their...

The total mass is 1.08x10^20kg according to the Wikipedia article at http://en.wikipedia.org/wiki/Enceladus_(moon)

[Guest_Myran_*]

If Enceladus have a quite higher ratio of rock than the other small moons at Saturn, we could assume the eruptions have been going on for a long time with quite some water lost to the ring system. So far so good.
But thats the problem I get with assuming the isotope Aluminium 26 are responsible, with a halflife of 20 million years its relatively shortlived, or if a lot of it would have been incorporated in the center of Enceladus from the moons formation, then the core of the moon would have been extremely hot at the time of formation ~4 billion years ago, wouldnt it?
Secondly, if it is true this moon actually have turned itself inside out more than once, then the volatiles like ammonia already would have been lost to space leaving us with the recycled water in the geysers.

[Guest_BruceMoomaw_*]

Hmmph. Then if it's been losing 100 kg of ice all along (as is the current estimated rate of loss), it would take fully 1.7 trillion years to lose its remaining supply of ice -- which means that, unless its rate was enormously higher at some point, the fraction of its original ice that it's lost during its entire 4.5 billion lifetime is infinitesimal -- so the relatively high mass of its rock core can't be explained by ice loss, but must have existed all along. This would seem to be another point in favor of Matson's and Lunine's view that the radioisotopes in the core itself might be responsible for initiating the eruption, without any assistance needed from a giant impact. So that question is still wide open.

As for the short lifetime of Al-26: Lunine's theory is that its emitted heat did initiate the eruption -- but that, as I said in my original message, it then sustained itself over the eons through the combination of the frictional heat produced by its initial convection, and the small continuing additional input of heat energy which relatively mild tidal forcing (such as Enceladus continues to undergo from its current orbit) can exert on viscous or ductile substances. This is the same reason why there was serious doubt before Galileo as to whether or not Europa had an ocean: calculations had indicated that, if Europa started out with an ocean in the first place, the tidal flexing of its icy crust due to that layer of liquid water sloshing around tidally underneath it would generate enough heat to SUSTAIN the ocean -- but if Europa didn't have an ocean to begin with, its rigid icy crust would never have flexed tidally enough to generate enough frictional heat to melt in the first place. And we had no idea whether or not it had started out with an ocean originally.

So tidal forces could very likely sustain Enceladus' eruption by themselves IF, but only if, something else started it -- so the debate is over what that something else is which originally jump-started it: a giant impact, a concentration of rocky-core radioisotopes, or conceivably some episode of a much more eccentric orbit that induced far stronger tidal flexing earlier in the moon's hstory. But the "Science" article suggests that long-term orbital analyses have ruled out the latter, leaving the other two possible initiators of the eruption.

[Guest_AlexBlackwell_*]

So tidal forces could very likely sustain Enceladus' eruption by themselves IF, but only if, something else started it -- so the debate is over what that something else is which originally jump-started it: a giant impact, a concentration of rocky-core radioisotopes, or conceivably some episode of a much more eccentric orbit that induced far stronger tidal flexing earlier in the moon's hstory. But the "Science" article suggests that long-term orbital analyses have ruled out the latter, leaving the other two possible initiators of the eruption.

The "giant impact" model could, theoretically, account for the large power output that CIRS is observing, though I would really like to see what modelling results indicate. However, in my opinion, this comes dangerously close to one of those dreaded "tooth fairy" explanations. Of course, that doesn't mean such an explanation is impossible, but the timing is fortuitous.

In any event, with respect to tidal forcing/libration, one of the papers that Porco et al. reference is Jack Wisdom's 2004 paper from The Astronomical Journal:

Spin-Orbit Secondary Resonance Dynamics of Enceladus
J. Wisdom
Astron. J. 128, 484–491, (2004).
716 Kb PDF reprint

It's a pretty good read.