≡ Menu

New Work on Planetary Inflation

Once in space in 2018, the Transiting Exoplanet Survey Satellite (TESS) will be observing, among many other things, hundreds of thousands of red giant stars across the entire sky. Planets around red giants are an interesting topic, because such stars point to an evolutionary outcome our own Sun will share, and we’d like to know more about what happens to existing planets in such systems as the host star swells and reddens, engulfing inner worlds.

New work out of the University of Hawaii Institute for Astronomy now examines two ‘hot Jupiters’ around red giants, stellar systems where we see the gas giants swelling up as the result of processes that remain controversial. The inflated size of planets like these can be explained in at least two ways, one of which involves a slowing of the cooling in the planet’s atmosphere, which causes the planet to inflate soon after formation. But the data presented here, drawn from NASA’s K2 mission, tend to corroborate the thinking of co-author Eric Lopez (NASA GSFC) that direct energy input from the host star is the dominant cause of this planetary inflation.

Image: Upper left: Schematic of the K2-132 system on the main sequence.
Lower left: Schematic of the K2-132 system now. The host star has become redder and larger, irradiating the planet more and thus causing it to expand. Sizes not to scale. Main panel: Gas giant planet K2-132b expands as its host star evolves into a red giant. The energy from the host star is transferred from the planet’s surface to its deep interior, causing turbulence and deep mixing in the planetary atmosphere. The planet orbits its star every 9 days and is located about 2000 light years away from us in the constellation Virgo. Credit: Karen Teramura, UH IfA.

The work is now available in The Astronomical Journal, where lead author Samuel Grunblatt and team show that each of the two planets is about 30 percent larger than Jupiter, though in each case only about half as massive. The two planets — K2-132b and K2-97b — are similar in orbital periods, radii and masses. Each orbits its red giant star in about nine days, with planetary radii being calculated at 1.31 ± 0.11 RJ and 1.30 ± 0.07 RJ respectively.

The researchers used models to analyze the evolution of planets like these over time, determining that their radii are typical for planets receiving their current level of radiation, but calculating back to main sequence values of radiation, they find the gas giants would have been considerably smaller. Stellar flux flowing to the planets’ deep convective interiors could therefore explain their current size, an indication that planet ‘inflation’ is directly tied to stellar irradiation rather than delayed atmospheric cooling after the planets’ formation.

But other factors remain to be tested, metallicity in particular. From the paper:

Further studies of planets around evolved stars are essential to confirm the planet re-inflation hypothesis. Planets may be inflated by methods that are more strongly dependent on other factors such as atmospheric metallicity than incident flux. An inflated planet on a 20 day orbit around a giant star would have been definitively outside the inflated planet regime when its host star was on the main sequence, and thus finding such a planet could more definitively test the re-inflation hypothesis. Similarly, a similar planet at a similar orbital period around a more evolved star will be inflated to a higher degree (assuming a constant heating efficiency for all planets). Thus, discovering such a planet would provide more conclusive evidence regarding these phenomena.

Also in play is the issue of heating efficiency, which may well vary between planets depending on their composition. Back to TESS, whose investigations should complement these results. Grunblatt and team point out that TESS should be able to observe additional planets in roughly 10 day orbits around more evolved stars, including oscillating red giants. The data should allow us to distinguish between the delayed cooling possibility and stellar irradiation scenarios.

The paper is Grunblatt et al., “Seeing Double with K2: Testing Re-inflation with Two Remarkably Similar Planets around Red Giant Branch Stars,” Astronomical Journal Vol. 154, No. 6 (27 November 2017). Abstract / preprint.


{ 10 comments… add one }
  • Mike November 28, 2017, 23:08

    I’m having a real problem trying to figure out how a planet can orbit a red giant star in nine days. A star like our Sun is expected to expand past the orbit of Venus possibly as far as Earth’s orbit or a 1 AU radius. Even a K-dwarf star is thought when it becomes a red giant to expand out to at least .5 AU radius.

    If the planet was in a nine day orbit before its star went off the main sequence and expanded it would be swallowed up as it would have been in a very close in orbit.

    If a planet was in an orbit more than .5 AU from its star I would think the orbital period would be around 200 days more or less.

    When the star becomes a red giant and expands so the star’s surface approaches the planet’s orbital path the planet would accelerate?

    I’m not well versed in orbital mechanics but this doesn’t make sense to me. I’m probably overlooking some basic factor because I don’t think the astronomers who wrote this paper overlooked this factor but I’m confused.

    And the light curve is showing a 9 day orbit? How can a planet travel that fast in a orbit that might be as much as 2 AU in circumference? What would be the planet’s escape velocity?

  • Alexander McLin November 29, 2017, 10:19

    I came here to post a comment asking the exact same thing Mike is asking. I’m quite confused by the 9-days orbit around a red giant star. I thought such stars would expand to at least several AUs in diameter.

    Initially I thought maybe it was possible for lower mass stars to expand into red giants on smaller scale than what we’d expect for stars with similar masses to Sol but even so I’m unable to understand how a 9-day orbit at least one AU wide would remain bound to the star.

    • Paul Gilster November 29, 2017, 10:28

      Thank you both for the question. I thought the best thing to do would be to forward the comments to Dr. Grunblatt and let him explain the orbital periods involved. I’ll post any response I get as soon as it’s in.

      • Paul Gilster November 29, 2017, 14:30

        OK, here is Dr. Grunblatt’s response:

        “Happy to clarify. The reason these planets could be discovered at all is because these stars are not fully evolved red giants. We restricted our search to low luminosity red giant stars, stars with radii 3 to 9 times the size of our Sun. In this regime, stars were large enough for us to detect stellar oscillations but still small enough for us to detect giant planet transits. In less than 100 million years (very soon from an astrophysical perspective), these stars will grow to ~AU sizes and swallow up their planets, but for now, the planets can continue to exist in their 9 day orbits.”

        Thanks very much to Dr. Grunblatt for the clarification!

  • andy November 29, 2017, 12:35

    The paper includes the stellar parameters. It does not give the semimajor axes of the planets, but these can be worked out from the stellar masses and orbital periods: 0.085 AU and 0.088 AU for K2-97b and EPIC 228754001b respectively. The paper gives the stellar radii: K2-97 has a radius 4.20 times that of the Sun, while EPIC 228754001 has 3.85 times the radius of the Sun, which translate to 0.020 and 0.017 AU respectively. As expected, the stars are smaller than the orbits of the planets.

    These stars are substantially expanded relative to main sequence stars of the same mass, they are red giant branch stars (notably the paper indicates they are low-luminosity red giant branch stars, which are at an early stage of their evolution as giant stars) but have not yet expanded to the full radii they have when they are on the asymptotic giant branch. The star does not snap instantly from a ~solar radius main sequence star to an asymptotic giant branch star the size of the Earth’s orbit, the expansion takes time!

  • Mike November 29, 2017, 14:57

    Thanks to Dr Grunblatt, Paul Gilster and Andy for their efforts in clarifying and explaining the findings and thus eliminating my confusion. I suspected I was overlooking something and it was a big help to have it pointed out to me.

    • Paul Gilster November 29, 2017, 16:05

      You bet, Mike. I learn something every day from the questions readers ask and the scientists who help to clarify them. Makes editing the site a huge pleasure.

  • Alka December 1, 2017, 4:06

    I came here to post a remark asking precisely the same Mike is inquiring. I’m very befuddled by the 9-days circle around a red mammoth star. I figured such stars would extend to no less than a few AUs in distance across.

    At first I thought perhaps it was feasible for bring down mass stars to venture into red monsters on littler scale than what we’d expect for stars with comparable masses to Sol yet even so I can’t see how a 9-day circle no less than one AU wide would stay bound to the star.

Leave a Comment