≡ Menu

Further Work on TRAPPIST-1

A closely packed planetary system like the one we’ve found at TRAPPIST-1 offers intriguing SETI possibilities. Here a SETI search for directed radio transmissions aimed at the Earth gives way to an attempt to overhear ongoing activity within another stellar system. For it’s hard to conceive of any civilization developing technological skills that would turn away from the chance to make the comparatively short crossing from one of the TRAPPIST-1 worlds to another.

Our more spread out system is challenging for a species at our level of technological development, but a colony on Mars or an outpost on Titan would surely produce intense radio traffic as it went about daily operations and reported back to Earth. Could TRAPPIST-1 be home to similar activities? The SETI Institute has continued to investigate the prospect, starting with ‘eavesdropping’ observations at 2.84 and 8.2 GHz in early April.


Image: A size comparison of the planets of the TRAPPIST-1 system, lined up in order of increasing distance from their host star. The planetary surfaces are portrayed with an artist’s impression of their potential surface features, including water, ice, and atmospheres. Credit:
NASA/R. Hurt/T. Pyle.

The SETI Institute’s work demands that the Allen Telescope Array be used in what this Institute news release calls ‘camera mode,’ which creates a series of ‘snapshots’ of the field of view every 10 seconds. The method is sensitive to the kind of broadband signals that might be used in spacecraft propulsion (think of the power beaming we’ve discussed often in these pages for getting payloads quickly to Mars and other local targets). The Benfords’ ideas on such a beaming infrastructure are examined in Microwave Beaming: A Fast Sail to Mars, while power beaming as a SETI observable is examined in Seeing Alien Power Beaming and elsewhere.

If a civilization were to be communicating between two worlds in the TRAPPIST-1 system, the time to observe its activities would be when two of the planets align with our view from Earth. It was just such an opportunity, between planets e and f in this system, that the ATA examined on April 6, while another two conjunctions occurred just six days later and were also observed by the ATA. If powerful transmissions were directed from one planet to the other, it might be possible for us to detect the spillover as the beam points toward the destination planet.

Similar observations are planned for May, looking for wideband signals much different from the kind of narrowband beacons the SETI Institute normally hopes to detect. Communications signals would doubtless carry a high data rate, while the broadband signals used in power beaming should likewise be distinctive. The raw data from these observations is being searched by high speed cloud computers, a computationally intensive task that is too expensive for the ATA to perform daily, but one that is assisted by its collaboration with IBM.

Refining Mass Information

While we follow the SETI search here with interest, it’s also worth noting that we’re getting a much tighter set of parameters on the masses of the planets around TRAPPIST-1. These are small worlds, but their proximity means we see substantial transit timing variations in the system. The masses calculated by Michaël Gillon and colleagues in the discovery paper from transit timing variation analysis produced masses for the six inner planets with uncertainties that varied between 30 percent and almost 100 percent.

Tightening up those numbers is the subject of a new paper from Songhu Wang (Yale University) and colleagues, who analyze data from the Kepler spacecraft’s K2 mission to refine the transit timing measurements. Several planets are much better constrained:

Perhaps the most significant conclusion that emerges from our analysis is that the masses of the outer planets, d, e, f, and g all show noticeable decreases in comparison to the values reported by Gillon et al. (2017). For example, the masses of planets e, f, and g (which have equilibrium temperatures of 251 K, 219 K, and 199 K, respectively) have decreased from Me = 0.62 M, Mf = 0.68 M and Mg = 1.34 M to Me = 0.24 M, Mf = 0.36 M and Mg = 0.57 M.

Reading through this paper simply reinforces how useful the planets around TRAPPIST-1 are proving to be — the authors say that they “…arguably constitute the most important exoplanetary system yet discovered.” The reason: We have large transit depths given the small size of the host star, along with extensive transit timing variations, meaning our ability to delve into mass, density and planetary composition here is greatly enhanced. And note this, which is based on the paper’s Figure 5 (not Figure 4, as is mistakenly referenced in the preprint):

…to within the errors of our determinations – the four most distant planets are consistent with pure water compositions, and in any event, are substantially less dense [than] either Mars or Venus.

Screenshot from 2017-04-25 08-53-02

Image: Figure 5 from Songhu Wang et al.. Caption: Planetary masses and radii for Trappist-1 system. Transit timing-inferred masses from the discovery paper (yellow dots) and this work (blue dots) are plotted with 1 σ error bars. Venus, Mars and, Earth are shown as black dots. Theoretical mass-radius relationships for different planetary compositions from Zeng et al. (2016) are plotted as colored curves. Credit: Songhu Wang et al.

The paper is Songhu Wang et al., “Updated Masses for the TRAPPIST-1 Planets,” submitted to the Astrophysical Journal (preprint).


Comments on this entry are closed.

  • Robin Datta April 25, 2017, 14:24

    If we find ourselves eavesdropping on an intelligent but unintelligible conversation, or even one direction of it, that would bring with it a re-focusing of humanity’s efforts in all fields of endeavor.

  • Robert Feyerharm April 25, 2017, 14:28

    If the above plot is using 1 sigma error bars, then the true mass of TRAPPIST-1 e, f, and h is still up in the air – especially with TRAPPIST-1 e.

    But if we are looking at water worlds, my money is on the outer worlds in the TRAPPIST system being former ice giants such as Uranus and Neptune which have migrated inward and somehow lost much of their hydrogen/helium atmosphere, leaving a small exposed rocky core surrounded by a water envelope.

    Having said this, a Neptune sized ice giant ought to still retain a significant portion of its hydrogen/helium atmosphere at temperatures of 200 K-300 K. Perhaps solar wind/flares from TRAPPIST-1 would be sufficient to strip away the lighter gases?

    • Michael April 25, 2017, 14:40

      I wonder if sychronised gas stripping is occurring or has occurred i.e every nearest planetary approach the atmospheres will be held on to less because of change of gravitation strength each planet has on each other.

      • Harry R Ray April 25, 2017, 20:09

        PRECISELY! Although you finally convinced me that planet h is relatively spherical with respect to its core and any possible ocean, the atmosphere itself(ESPECIALLY a larger than normal STRATOSPHERE) is a whole other horse. The big question here would be whether any cirrus clouds in such a stratosphere STILL make the planet appear to be egg-shaped.

  • FrankH April 25, 2017, 15:07

    @Robert Feyerharm – it would depend on the age of the system; none of these planets are massive enough to retain hydrogen and helium for more than a few million years at their current temperatures and masses. TRAPPIST-1g might just be able to hold on to some helium with some generous assumptions about its density (I guessed at a density of around 2gm/cm^3). Solar flares and wind wouldn’t help with retention, either.

    From the ever useful Gas Retention Plot page:

    • Robert Feyerharm April 26, 2017, 10:59

      @FrankH Thank you for the Gas Retention Plot link. I can see that even at 250 K Uranus and Neptune’s position on the plot would still lie above, but certainly “close”, to the 10 x V_avg gas retention threshold for H2. But several hundred million years of flares and solar wind might do the job. There have been papers researching atmosphere loss to terrestrial planets orbiting red dwarfs, even after taking strong planetary magnetic fields into account.

  • Harry R Ray April 25, 2017, 16:47

    Planet g is the ONLY ONE with a mass CONSTRAINED ENOUGH to to make ANY KIND OF DETERMINATION as to what KIND of a planet it is. What it is is UNIQUE to the exoplanet zoo! It can be called either two things: Gas Midget OR Micro-Neptune. Here’s why. Unlike “water worlds”(which planets e and f MOST LIKELY ARE, and planet h, which COULD BE, but in a VERY PECULIAR WAY), Planet g’s(PRESUMABLY)H2O envelope is NEVER in a truly LIQUID form as we know it. Here’s why; The pressure exerted by the UPPER LAYERS on the LOWER LAYERS is so great that the TEMPERATURE of the lower layers would be WAY TOO HIGH for Ice-7 to form, and provide a definite BOUNDARY between the super-hot, super-dense core and a truly liquid ocean. Therefore, convection would PREVENT a liquid ocean from ever forming. As a result, Andrew LePage should SERIOUSLY CONSIDER STRIKING this planet from his “potentially habitable list, unless there is a REALISTIC CHANCE life could exist in the CLOUDS of this planet. In September, the Spitzer Space Telescope will observe the TRAPPIST-1 system CONTINUOUSLY for 45 days. The ensuing data should be able to constrain the masses of ALL THE REMAINING PLANETS with the exception of planet h to the same narrow parameters which the mass of planet g is NOW constrained, allowing us to definitively know what king of system we are dealing with here.

    • EricRoboto May 14, 2017, 0:28

      Can you please not post randomized capital words? That’s hard to read and just not necessary. Please don’t, thanks.

  • Dave Moore April 26, 2017, 1:27

    I’m a little dubious of these results. Planets e,f, and h are shown as having the density of water. There is one other planet with this sort of density, Gliese 1214 b. It is thought to consist of 25% rock and iron, 75% water with a thick H and He atmosphere. But it has a mass 6.6 times Earth’s.

    From these results, we are supposed to assume that planets with masses of 0.24, 0.36 and 0.57 of Earth’s can hold onto a large H and He envelopes, which would be the only way a planet can achieve this sort of density.

    • Harry R Ray April 26, 2017, 13:08

      BELIEVE IT! A recent paper has claimed to constrain the masses of Kepler 444d and Kepler 444e to even a greater extent than TRAPPIST-1g’s has been. They are BOTH roughly 1/3 Me with radii similar to MARS’S. They both orbit their parent star at a distance SIMILAR to that of the TRAPPIST-1 e,f,and g planets, but their parent star is MANY TIMES MORE LUMINOUS than TRAPPIST-1, so their Teff’s approach 1000 degrees Farenheit! They both orbit in a PERFECT RESONANCE with one another, and have been doing so for ELEVEN BILLION YEARS! Kepler gathered four years of near-continuous data on this star, so the evidence is UNIMPEACHABLE!

      • Harry R Ray April 27, 2017, 9:27

        OOPS! I meant …roughly 1/30 me…, or roughly one third the mass of MARS! Here’s the COMPLETE Kepler 444 system in its most updated form: Planet b(radius: 0.403 Re, orbital period: 3.6001053 days), Planet c(radius: 0.497 Re, orbital period: 4.5458841 days), Planet d(mass: 0.036 Me, radius: 0.53 Re, orbital period: 6.189392 days, DENSITY!!!: 1.127gcm3), Planet e(mass: 0.034 Me, radius: 0.546 Re, orbital period: 9.740486 days, DENSITY!!!: 1.108gcm3), Planet f(radius: 0.741 Re, orbital period: 9.740486 days). By comparison, the Moon has a mass of 0.0123 Me and a radius of 0.273 Re. Enceladus has a density of 1.609 gcm3, GREATER than the densities of Planets d and e!!!

    • Holger May 9, 2017, 10:03

      “From these results, we are supposed to assume that planets with masses of 0.24, 0.36 and 0.57 of Earth’s can hold onto a large H and He envelopes, which would be the only way a planet can achieve this sort of density.”

      Not at all, you’re ignoring the error bars. e can still even have an Earth-like density at 1 sigma, and f and h can be 50% rock, 50% ice at about 1 sigma each.

  • Adam April 26, 2017, 5:11

    Hi Paul,
    The outer planets, with those densities, are probably *failed* Gas Giant cores. Just didn’t get big enough to pull in nebula gas before it blew away. I very much doubt they’d be able to have support indigenous technological life (no solid surface and likely no free O2), but they might be astrobiologically interesting.

  • RobFlores April 26, 2017, 11:07

    The age of the system is still not known for certain,
    it appears to be older than 3×10^9 years. It cannot be ruled out that it could be three times that age. If this is 9 billion old system, does this help the chances of multi cellular life arising here? or do the planets begin to become inert/sterile at that advanced age?

  • hiro April 26, 2017, 16:47

    Who could have thought that the mass of planet g would drop down to around 1/2 Me, this is more like a snowball super-Mars than a (potential) habitable planet.

    …….So WTF Star is still on the top huh?

  • EricSECT April 27, 2017, 7:57

    I am curious as to why they chose to monitor the Trappist system at 2.84 and 8.2 GHz frequencies, way outside the “water hole” AFAIK. What do these frequencies correlate to? Radar trying to detect asteroids? Airports scanning for thunderstorms?

  • Harry R Ray April 28, 2017, 9:30

    arXiv: 1704.08449; “The full spectral radiative properties of Proxima Centauri” by Ignasi Ribas, Michael D. Gregg, Tabitha S Boyajian, Emeline Bolmont. XUV irradiance on Proxima b would ONLY be 60 times that of Earth’s, instead of 1200 times suggested in previous literature. Warm dust(disk?)orbiting Proxima Centauri. Could dust be blocking MOST of the XUV flux? Could the SAME THING be happining at TRAPPIST_1?

  • Michael Fidler April 30, 2017, 11:01

    Several interesting articles on arXiv.org > astro-ph, dealing with material we have been discussing:
    The first is a very good up to date PhD Dissertation on; Habitability of Exoplanetary Systems:

    High-precision stellar limb-darkening in exoplanetary transits;

    On the feasibility of studying the exospheres of
    Earth-like exoplanets by Lyman-α monitoring.
    Detectability constraints for nearby M stars.

    And finally: Mapping Exoplanets;

    Surprising what you can see thru a pinhole looking glass!

  • Harry R Ray May 4, 2017, 10:31

    Dr Andreas Faisst just tweeted this: “#GISS 17 starts tomorrow@caltechipac! Featuring updates on #TRAPPIST1! giss.ipac.caltech.edu//meetings/2017/”. thirteen hours ago. If any reader of this website CAN attend this meeting, PLEASE DO SO AND REPORT BACK ASAP!

  • Harry R Ray May 17, 2017, 9:35

    Good news on BOTH TRAPPIST-1 AND Proxima b arXiv:1705.05535; “Atmospheric escape from the TRAPPIST-1 planets and implications for habitability”. by Chuanfei Deng, Menf Jin, Manasvi Lingham, Vladimir S. Airapetian, Yingjuan Ma, Bart Van Der Holst. ALSO: The UK MET paper on Proxima b has finally been published.

  • ljk May 22, 2017, 13:10
  • ljk May 22, 2017, 13:12

    Talk about the music of the spheres – more evidence that art and science working together give us a better picture of the Cosmos: