Exomoons: A Direct Imaging Possibility

by Paul Gilster on November 9, 2012

It’s good to see that David Kipping’s work on exomoons is back in the popular press in the form of A Harvest of New Moons, an article in The Economist. Based at the Harvard-Smithsonian Center for Astrophysics, Kipping’s Hunt for Exomoons with Kepler (HEK) culls Kepler data and massages the information, looking for the tug of large moons on transiting exoplanets. The basic method will by now be familiar to Centauri Dreams readers:

Dr Kipping’s technique relies on the fact that moons do not simply revolve around their host planets; planets also revolve around their moons—or, rather, the two bodies both revolve around their common centre of mass. If a planet is large and its moon small the distinction is trivial. But if the planet is small and the moon is large, it is not. In the case of Earth and its moon, for example, the common centre lies only around 1,700km (1,100 miles) beneath the Earth’s surface. Someone looking from afar at the movement of Earth would thus be able to deduce the moon’s existence without having to see it directly.

And as we’ve discussed in previous articles, the need for a large moon is significant. Kipping has recently reckoned that a moon about one-fifth as massive as the Earth should be in range of detection, but Ganymede, the biggest moon in our Solar System, is only 1/40th as massive. The first exomoon detected, then, will likely be a very large object, big enough that its signal won’t be masked by the presence of other planets in the same system. I’ve always had a fascination with exomoon studies and thus am looking forward to the first presentation and data analysis by the HEK team, slated to occur at the American Astronomical Society meeting in January.

Image: Exomoon hunter David Kipping. Credit: CfA.

But I want to focus in on something else in this article, namely the work of Mary Anne Peters and Edwin Turner, who have asked in a recent paper whether a large enough exomoon orbiting close enough to its planet (and far enough from its star) might produce an infrared signature detectable from Earth. Think Io, and ponder how tidal heating churns the insides of such a moon, creating heat-generating friction and, in the case of Io, active volcanoes.

Peters and Turner, both at Princeton, produce an acronym I had never encountered before: THEMs, or Tidally Heated Exomoons. In terms of direct imaging, a THEM is quite interesting. For one thing, a tidally heated moon may remain hot and bright for the lifetime of its star, making it visible in solar systems both old and young. For another, such a moon may orbit its planet far away from the primary star while remaining hot because of tidal heating. It is thus a luminous target at a large separation from the star whose light would otherwise drown out its signal.

The researchers calculate that tidal heating could produce terrestrial-planet-sized moons with effective temperatures as high as 1000 K, moons as much as 0.1% as bright as the system’s primary (if the central star were low in mass). This is interesting stuff: Io has the highest measured temperatures of any body in the outer Solar System because of the tidal heating effect. Now imagine the system of Galilean moons orbiting Jupiter were scaled down to orbit Neptune in more or less the same configuration. If this were the case, Io would be more luminous than Neptune itself, and if it were as massive and dense as the Earth, the authors say it would be the brightest Solar System object beyond 5 AU, outshining even Jupiter at some wavelengths.

Given all this, there is a case to be made that tidally heated exomoons may actually be easier targets for direct imaging than the kind of hot, young gas giants at large separations from their star that are most likely to be found by current imaging efforts:

Direct imaging detection of physically plausible, tidally heated exomoons is possible with existing telescopes and instrumentation. If tidally heated exomoons are common, for example if typical gas giant exoplanets are orbited by satellite systems broadly similar to those found in the Solar System, we are likely to be able to image them around nearby Sun-like stars in the midst of their main sequence lifetimes with current or near future facilities.

The paper suggests that existing instruments can detect exomoons at temperatures of 600 K or above and radii of Earth-size or larger, while future mid-infrared space telescopes like the James Webb Space Telescope will be able to directly image heated exomoons with temperatures greater than 300 K and radii of Earth size and larger as long as they are orbiting at 12 AU or more from their star. Perhaps we’ve imaged our first exomoon already: The authors think it is possible that Fomalhaut b, whose identity as a planet remains controversial, is actually a tidally heated exomoon, or a blend of the emissions from a hot, young gas giant and a heated moon.

In one figure, the researchers look at the performance of JWST’s Mid-Infrared Instrument (MIRI), charting exomoons with temperatures similar to those found on Earth. The results:

…it is plausible that some of the exomoons JWST is capable of detecting, could potentially be habitable, in the sense of having surface temperatures that would allow liquid water to be present. Some of these exomoons have comparable irradiance to the gas giants in our solar system. At ~ 14µm and 300K, an Earth-radius exomoon would be as luminous as Jupiter. However, if [the] Jupiter were colder due to being less heated by its primary and/or being older, the Earth-like moon would be much brighter than the planet.

The paper is Peters and Turner, “On the Direct Imaging of Tidally Heated Exomoons” (preprint).



jkittle November 9, 2012 at 11:13

excellent post – this is one of the reasons I follow CD…
Heated expmoons overcome lots of problems with detection. this is also possible if there is a large un- discovered planet in the oputer reaches of the solar system ( out beyond 1,000 AU). out there, a planet half the size of neptune would be hard to detect because of low abedo and low light, and it would be pretty low temperature ( though detectable to a meter class telescope operating at 12 microns for example) . However if an Io class ( or Encelatus-class geyser moon) were out there, it heat signature would be visible with a very distinct signature. The only question is .. how hot would it have to be. Io has some really extreme hot spots.
As for as exoplanets/Exomoons,.. what a great idea! the distance from the star now becomes an asset to detection , not a liability, since the radiated energy is not a function of the luminescence form the star. I am prediction success using the JWST! This could happen people… in the next decade!

Gerry November 9, 2012 at 15:23

Fascinating article. I wonder how much of an impact tidal heating could have on the surface temperature of an Earthish-size moon existing outside the nominal habitability zone of a star system? If Io were more Earthlike in mass, gravitationally capable of holding onto a thick atmosphere, how much heat could be trapped? Might the surface be warm enough for water to remain liquid, even at Jupiter’s distance from the sun?

Mike November 9, 2012 at 16:31

Regarding the effort to detect exomoons by the HEK team, detecting even a large moon orbiting a Jovian size planet will be a challenge. What maybe easier to detect are double planets, which could be defined as two planets orbiting each other which are similar enough in mass that their common centre of mass is above the surface of the larger member of the pair. Pluto and Charon would be an example in our Solar system even though they are dwarf planets.
If such pairs exist elsewhere and are planetary in size they could be detectable examining the TTVs and TDVs in Kepler transit data. Searches using the Kepler data to hunt for exomoons may also detect double planets.
We know dwarf double planets exist, I think it would be wonderful to discover full-size double planets. And ofcourse discoveries of exomoons would be wonderful too!

TTV Transit timing variations. TDV Transit duration variations.

Kappy November 9, 2012 at 21:22

“Direct imaging detection of physically plausible, tidally heated exomoons is possible with existing telescopes and instrumentation.”

Is it possible what we see at Fomalhaut b (http://en.wikipedia.org/wiki/Fomalhaut_b) is just such an exomoon? The article states that “Further observations from Subaru Telescope confirmed the existence of the planet; concluding that the light from the planet originates from the dust cloud around the planet, not Fomalhaut b itself.” But an exomoon around the planet might also glow in the IR enough for our instruments to detect.


coolstar November 10, 2012 at 5:32

Mike makes a very good point about double planets: it could be that the odds of finding a terrestrial sized double planet in the Kepler archive are larger than that of finding a terrestrial sized exomoon. This is particularly true if present theory holds and exomoons are unlikely to be more than about 10e-3 the mass of their primary. The HEK team is doing good work, but their real contribution will likely be in finding more subtle transit timing variations (and thus “unseen” planets and possibly more good determinations of planetary masses).
Earth sized double planets in habitable zones would certainly be exciting, but whether they’d truly be habitable is uncertain. They’d certainly be tidally locked (to each other), which would slow their rotational periods and thus likely decrease the strength of their magnetic fields (and increase the day and night side temperature differences). They can’t be TOO close, since tides go as one over the distance cubed. Too large a mass and too close an orbit and tidally induced heating could turn both worlds into something like Io…..

Rob Henry November 10, 2012 at 5:59

Gerry asks a very common type of question with “ If Io were more Earthlike in mass, gravitationally capable of holding onto a thick atmosphere, how much heat could be trapped? Might the surface be warm enough for water to remain liquid, even at Jupiter’s distance from the sun?”

The simple answer is no. The surface of your hypothetical warm planet with identical atmospheric structure to us needs to be two orders of magnitude more active per unit area than Io for that to occur. Furthermore, volcanic activity allows only a fraction of the potential for life to build high energy chemicals that photosynthesis does, so the biosphere in that hell should be four or five orders of magnitude less powerful than our own.

Of cause you could have a super dense atmosphere such a Venus’, and tone down the necessary volcanism, but then the biosphere would be even more anaemic.

Rob Henry November 10, 2012 at 19:52

Coolstar writes “They can’t be TOO close, since tides go as one over the distance cubed. Too large a mass and too close an orbit and tidally induced heating could turn both worlds into something like Io…..” and I can’t help wondering if this is correct.

If a double planet has all its surfaces tidally locked, and their eccentricities w.r.t. their c.o.g. is zero there should be no cause for heating, and I can’t see perturbations from other planets in the system providing anything like those in the tight Laplace resonance orbits around Jupiter. If however, they had a large mutual moon, I could see problems here.

Of cause one partner can’t orbit within the Roche sphere of the other, but this is such a complex construct that it is even possible to construct Rocheworlds (binary planets sharing a common atmosphere) – given sufficiently tight parameters.

Rob Henry November 10, 2012 at 20:22

Oops Coolstar, I left out the most relevant bit. In that linked Rocheworld commentary above, a scheme is given in which an environment that experiences about 40 times Io’s tides is invoked yet no mention is made of extraordinary tidal heating. I was wondering if you disagree with that assessment.

Rob Henry November 11, 2012 at 17:32

I am starting to wonder if any object radiating at 1000K could be stable for more than a few years. Here is my problem…

Sure, I can believe that we can come up with a theoretical object that could suffer such tidal forces and not be ripped apart, but a (black body) radiating at that temperature puts out 56,000W/m^2. Now Io has mass/surface of 2.14 million tons per sqm, so this power is equivalent to it being accelerated by 800m/s per annum. Even for an Earth sized object it would equate to 200m/s per year. So, the orbital parameters must be changing like crazy unless rotational momentum from the primary is being continuously transferred to that “Io” with incredible precision to offset this potential.

andy November 11, 2012 at 18:13

The authors think it is possible that Fomalhaut b, whose identity as a planet remains controversial, is actually a tidally heated exomoon, or a blend of the emissions from a hot, young gas giant and a heated moon.

I am not convinced that Fomalhaut b is consistent with being an exomoon. For starters, it is observed in the optical not the infrared: this is not what would be expected from thermal radiation from a tidally-heated moon.

The assertion that Fomalhaut b is an exomoon needs to be backed up with better analysis, at present they just appear to state that it could be the case without attempting to back up this idea by relating it to the evidence at hand.

coolstar November 12, 2012 at 0:13

@Rob Henry The eccentricities of the Galilean satellites are all actually very low, and yet Io’s tidal heating amounts to about 2 w/m^2 which is about 1/500 that of solar insolation at 1 AU and about 25 times that of earth’s internal heating, if memory serves for the latter. I wouldn’t expect a double planet to have eccentricities THAT low. Planet-star orbital eccentricities would likely keep the double planet’s orbital eccentricity pumped up as well as would stellar tides, if the semi-major axis is small (as required for the HZ around a K or M dwarf). Also, one would have to consider perturbations caused by giant planets in the system.
However, tidal heating is VERY complex (a good starting reference to the habitability of exomoons is a paper by Heller and Barnes at http://arxiv.org/ftp/arxiv/papers/1209/1209.5323.pdf and references therein), and I could be wrong (tidal heating seems to go as the mass of the planet cubed and is a very complicated function of the eccentricity but varies as the planet-satellite distance as 1/r^9, where r is the separation between the bodies). Naively plugging in some numbers leads me to think I might be VERY wrong (factor of 50 or so) for a double earth! Hard to get around that factor of 300 cubed! I’d originally just assumed (guessed) that the tidal heating would only go as the mass to the first power…..
Perhaps it’s just a lot easier to just consider the tidal FORCES which are proportional to the mass of the body and inversely proportional to the distance cubed. So even with a separation equal to that of the present day earth-moon separation, you’re starting with tides about 80 times higher than today and MUCH, MUCH higher than that for a long time after formation of the double planet system (since any reasonable capture or collision model starts out with a much smaller initial orbital radius). I’ve seen one paper (sorry, can’t find the reference easily), that derives much more heating for Europa’s ocean just due to the “sloshing” around rather than the tidal heating (though one presumes that there wouldn’t be any sloshing without the tidal heating!). My suspicion is that a double planet would not be a great place for life, given all this and the fact that the day would likely be quite long due to tidal locking, just for starters. Of course “not great” does not mean impossible……
Thanks for keeping me honest Rob, which led me to the very nice papers by Heller and collaborators.

coolstar November 12, 2012 at 0:42

@Rob One last thing, the same first order calcs done for the planets for the Roche world you referenced would likely be completely molten as they’re about 24x closer than the earth-moon distance. (24)^9 trumps 2.7×10^7!. As mentioned at the url, lots of other reasons they can’t exist (for long).

I remember reading Forward’s Rocheworld, long ago, and don’t recall even thinking about this objection. Forward was such a competent physicist I’m pretty surprised he didn’t think of it either. Or maybe he did and it was such a neat idea he ignored this part of the physics??

spaceman November 12, 2012 at 3:30

Last year I think there was an online article for laymen describing the search for exomoons. If I recall correctly, one of the HEK investigators interviewed said that they found several promising candidate moons. It will be interesting to see if they end up panning out after the data is examined even more thoroughly.

ljk November 12, 2012 at 12:27

coolstar said on November 12, 2012 at 0:42:

“I remember reading Forward’s Rocheworld, long ago, and don’t recall even thinking about this objection. Forward was such a competent physicist I’m pretty surprised he didn’t think of it either. Or maybe he did and it was such a neat idea he ignored this part of the physics??”

Probably in the same way that even “hard” science fiction authors ignore or otherwise get around the speed of light when its so-called limitations will get in the way of their galactic empires and such.

Even the 1968 film 2001: A Space Odyssey, which is often rightly hailed as having done its darndest to be as scientifically accurate as possible let some things slip when reality would have gotten in the way of the story.

I am not even talking about a thinking computer or hibernating astronauts here. I am referring to no radiator fins on the USS Discovery, even though they would have been desperately needed to get rid of all that heat from the spaceship’s nuclear engines. They were removed because it would have given the vessel “wings” and that was just not the aesthetic look the filmmakers wanted.

Another one was the look of the lunar surface. Even by the time of the film it was known that the lunar mountains were smooth and rounded, not jagged like in a Chesley Bonestell painting. Just take a gander at no less than Arthur C. Clarke’s 1964 book Man and Space from the Time-Line Science series if you can find a copy: He knew the true shapes of the mountains of the Moon and had illustrated examples of them. But in those pre-Apollo days, most people expected Earth’s natural satellite to have a rough face with pointy mountains, so that is what they got.

andy November 12, 2012 at 13:46

Regarding the Galilean satellites, the point to note is the Laplace resonance that forces the eccentricities of the satellites.

In contrast, the Saturnian system presents the Mimas-Enceladus paradox: Mimas is located closer to Saturn than Enceladus and has a higher eccentricity. Despite this, Mimas is apparently geologically dead, while Enceladus is active.

It turns out that while Mimas is in a 2:1 resonance with Tethys and Enceladus is in a 2:1 resonance with Dione, the Mimas-Tethys resonance is inclination-type while the Enceladus-Dione resonance is eccentricity-type, that is Enceladus has forced eccentricity as a result of the resonance which seems to be key to the greater activity on Enceladus. Here’s a quick summary.

As for Rocheworld-type scenarios I’ve never been particularly convinced that such a system is particularly plausible: I suspect the dissipation and eventual merger/disintegration would happen over a relatively short timescale compared to the lifetime of the system.

Rob Henry November 12, 2012 at 19:37

Thanks for the help on this unbelievably difficult problem. I am also surprised by that ninth power law, and suspect it could only hold for bodies that were semi-elastic throughout. Strangely though, you made a mistake on an easy one.

Solar insolation at 1AU might be about 1kW/m^2 if we are a solar collector, but if we are a spherical planet we need to multiply by (1-albedo)/4, so about 150W is more appropriate.

coolstar November 12, 2012 at 22:17

Rob, no mistake: insolation is the total amount of solar energy incident on a surface, an energy flux (joules/sec/m^2). And by definition it’s a flat surface. At the top of the atmosphere it’s closer to 1400 w/m^2. And it’s defined for the sun at the zenith, which it never is, most places. Handy to remember in that it’s a nice round number. It’s actually pretty easy to approximately measure using a telescope, insulated flask of water, and a thermometer (yeah, one would want to be careful doing this, and I doubt I’d be able to do it in an intro lab anymore!) . You defined the total energy absorbed averaged over a spherical surface, which is a different thing.
Oh, I make easy mistakes all the time! just not on this one.

@Andy I recently saw a paper which showed that the Laplace resonances of Io, Europa, and Ganymede are NOT a permanent thing. Surprised me, as I always thought they’d be locked that way forever. Sorry, don’t have the reference handy but Dr. Google or the NASA ADS site ought to pull it up pretty quickly.

Torbjörn Larsson, OM November 13, 2012 at 17:01

the need for a large moon is significant.

I can’t find those articles, but I have to assume they claim something like “the need for a large moon is significant for having life on a planet” as it is a common claim. That should be an incorrect claim.

- Surface habitability (defined as withing the zone of liquid water on average) is independent of moons.

- Orbital stability is ensured in periods larger than our average periods between mass extinctions.

The first simulations made erroneous predictions, but the mistake was identified recently.

- Atmospheric stability is threatened by a large impact generated moon.

Our own early atmosphere was nearly loosing its magnetosphere protection due to the strength of the close and still dynamo active Moon. In some models that could have triggered an hydrogen escape atmosphere, with drastic effects on the volatile supply.

In sum, we can safely predict that the presence of a moon has in general an insignificant effect on habitability. If anything, there is a small risk a large moon may be harmful.

Rob Henry November 13, 2012 at 17:06

Coolstar, I hate to quibble over a definition, but I must since insolation is a word that I have frequent cause to use. I thought that insolation could have either meaning depending on context, and in in your previous statement that should have given it the second one that I implied. The Wikipedia article on Insolation uses it in this way, but attaches the word “direct” in front of it when it is using it in the way you defined, implying that that is now seen as a secondary definition (at best?). I hope this is not another case of different sciences using the same technical word in a different way.

PS, when you wrote “You defined the total energy absorbed averaged over a spherical surface, which is a different thing” I finally realised how the human race will end. How you ever noticed how most physicists insist of building a Dyson sphere at 1AU, and comparatively few correctly at 2AU. It seems that, in a few millennia, we will all fry!

Comments on this entry are closed.

{ 2 trackbacks }