A Large and Variable Leading Tail of Helium in a Hot Saturn Undergoing Runaway Inflation

My talk at Towards Other Earths III, in Porto Portugal, on Monday July 17, at 5:10 PM.
The talk is about our recently submitted paper here:
https://arxiv.org/abs/2307.08959

Note that the HAT-P-67 stellar and planet properties reported on slides 20 and 21 reflect the previously published values from Zhou et al. 2017. Our new paper (link above) slightly refines these values from revised Gaia DR3 distance, use of new isochrones, and more numerous RV samples with HPF (albeit still consistent with an RV non-detection). See the paper for more information.

Slide 38 shows an animation from Thorngren, Lee, & Lopez 2023 available on the HTML version of the ApJ website:
https://iopscience.iop.org/article/10.3847/2041-8213/acbd35

The top-down orbit on slides 28 - 31 is merely a cartoon to emphasize the leading nature ot the tail, and not its actual orbital location-- a genuine planetary wind consistent with the outflow will initially reside at orbital separations closer into the star, and not directly on top of the planet orbit.

Transcript

1. Michael Gully-Santiago
   The University of Texas at Austin
   Towards Other Earths III
   July 17, 2023
   Porto, Portugal
   A Large and Variable Leading Tail of Helium in a
   Hot Saturn Undergoing Runaway Inflation
   Caroline V. Morley, Jessica Luna, Morgan MacLeod,
   Antonija Oklopčić, Aishwarya Ganesh, Quang H. Tran,
   Zhoujian Zhang, Brendan P. Bowler, William D.
   Cochran, Daniel M. Krolikowski, Suvrath Mahadevan,
   Joe P. Ninan, Guðmundur Stefánsson, Andrew
   Vanderburg, Joseph A. Zalesky, Gregory R. Zeimann
   Gully-Santiago et al. submitted; on tonight’s arXiv posting
   

   View Slide

2. The Mass-Radius diagram
   

   View Slide

3. Why so few
   inflated sub-
   Saturns?
   The Inflated Sub-Saturn Cliff
   

   View Slide

4. Scenario 1: Nature does not make them.
   Scenario 2: Nature makes them, but they are unstable.
   Why so few
   inflated sub-
   Saturns?
   

   View Slide

5. Scenario 1: Migration prevents sub-Saturns from reaching high Teq
   .
   Scenario 2: They reach high Teq
   , but quickly undergo Runaway Inflation.
   Thorngren & Fortney 2018
   

   View Slide

6. Scenario 1: Migration prevents sub-Saturns from reaching high Teq
   .
   Scenario 2: They reach high Teq
   , but quickly undergo Runaway Inflation.
   Thorngren & Fortney 2018
   Key question:
   How efficiently does insolation couple into the interior?
   

   View Slide

7. Anomalous heating efficiency, ε
   Thorngren & Fortney 2018
   Observationally
   disfavored
   Favored
   Observationally
   disfavored
   Agnostic to the physical mechanism
   causing the anomalous heating
   Peaks at ~3% for Teq
   ~ 1600 K
   Governs the extent of radius inflation
   

   View Slide

8. The steady state Radius-Mass slope steepens with increasing Teq
   .
   Thorngren & Fortney 2018
   

   View Slide

9. Thorngren & Fortney 2018
   The steady state Radius-Mass slope steepens with increasing Teq
   .
   

   View Slide

10. Thorngren & Fortney 2018
    The steady state Radius-Mass slope steepens with increasing Teq
    .
    

    View Slide

11. Thorngren & Fortney 2018
    A positive feedback loop ensues.
    

    View Slide

12. Thorngren & Fortney 2018
    A positive feedback loop ensues.
    

    View Slide

13. Thorngren & Fortney 2018
    A positive feedback loop ensues.
    Losing mass makes you larger, which makes you lose mass faster.
    

    View Slide

14. Inflated sub-Saturns should exhibit profound mass loss rates.
    Thorngren & Fortney 2018
    

    View Slide

15. Inflated sub-Saturns are rare.
    Thorngren & Fortney 2018
    

    View Slide

16. This talk: Helium 10833 Å observations of HAT-P-67 b & HAT-P-32 b
    

    View Slide

17. Habitable Zone Planet Finder (HPF)
    Helium Exospheres Survey
    λ = 8100 – 12,800 Å
    R = 55,000
    Hobby Eberly Telescope (HET), Texas, USA
    HET has fixed-altitude design: not fully-steerable
    - Restricts available visit times
    - Restricts available visit durations to <1 hour
    We get abundant orbital phase coverage:
    Large orbital phase coverage
    Visits
    In–Transit Out-of-Transit
    HAT-P-32 b 3 18
    HAT-P-67 b 7 35
    

    View Slide

18. Large tails of Helium excess were previously missed due to
    limited out-of-transit phase coverage.
    (i.e. self-subtraction of genuine planetary signal)
    Zhang, Morley, Gully-Santiago et al. 2023 DOI: (10.1126/sciadv.adf8736)
    Giant tidal tails of helium escaping the hot Jupiter HAT-P-32 b
    We detect a 12-hour Helium transit.
    (the white-light transit lasts for 3.1 hours)
    The sky-projected length of the tails is 53 Rp
    

    View Slide

19. Large tails of Helium excess arise naturally
    from Keplerian orbital shear:
    dayside mass loss precedes the planet
    nightside mass loss trails the planet
    Giant tidal tails of helium escaping the hot Jupiter HAT-P-32 b
    Zhang, Morley, Gully-Santiago et al. 2023 DOI: (10.1126/sciadv.adf8736)
    3D MHD simulations estimate Mass Loss at
    ̇
    𝑴 ~ 1 ×1012 g/s
    ~ 5 M
    ⨁
    / Gyr
    

    View Slide

20. HAT-P-67
    an F subgiant
    Zhou et al. 2017
    

    View Slide

21. HAT-P-67 b
    a very low density, inflated hot Saturn
    Zhou et al. 2017
    

    View Slide

22. HAT-P-67 b
    with HPF
    39 nights over 3 years
    13.8 hours of on-sky integration time
    152 individual exposures
    

    View Slide

23. HAT-P-67 b shows conspicuous variability in He I 10833 Å.
    

    View Slide

24. Up to 10% transit depths.
    

    View Slide

25. HAT-P-67 b also has an extended ingress.
    

    View Slide

26. View Slide

27. View Slide

28. View Slide

29. View Slide

30. View Slide

31. View Slide

32. Leading tail resides in the stellar rest frame.
    

    View Slide

33. Weak trailing tail blueshifts
    indicating acceleration away from the planet.
    

    View Slide

34. The leading tail
    is direct evidence for preferential dayside mass loss.
    

    View Slide

35. ̇
    𝑴 ~ 2 ×1013 g/s (105 M
    ⨁
    / Gyr )
    with 1D Parker Winds models † (p-winds)
    †Significant uncertainty:
    - XUV radiation
    - T0
    - 3D effects (streams)
    - self-shielding
    Dos Santos et al. 2022
    with Mp
    < 100 M
    ⨁
    implies inflationary timescale
    𝜏infl
    < 1 Gyr
    

    View Slide

36. What physical mechanisms
    cause runaway inflation?
    

    View Slide

37. Ohmic Dissipation and XUV irradiation
    both predict runaway inflation for hot Saturns
    Batygin, Stevenson & Bodenheimer 2011 Thorngren, Lee & Lopez 2023
    

    View Slide

38. Thorngren, Lee & Lopez 2023
    XUV irradiation removes hot Saturns from the mass-radius plane.
    Mass loss is a positive feedback loop near the 0.1 g/cm3 threshold.
    

    View Slide

39. Ohmic Dissipation and XUV irradiation
    make different quantitative predictions for inflation timescales.
    HAT-P-67 b
    Theory:
    𝜏infl
    ~ 5-50 Myr
    Observed:
    𝜏infl
    < 1000 Myr
    

    View Slide

40. XUV irradiation
    better matches the population of hot Saturns
    0.1 g cm-3 threshold
    divides observed
    planet sample from
    sub-Saturn cliff.
    

    View Slide

41. Conclusions
    We have detected up to 10% transit depth of He I 10833 Å from HPF spectra of HAT-P-67 b.
    The excess absorption preceeds the transit by up to 130 planetary radii in a large leading tail.
    The prominence of this leading tail is direct evidence for preferential dayside mass loss.
    We estimate a mass loss rate of 2 x 1013 g/s, and lifetime less than a Gyr.
    This pattern broadly agrees with theoretical predictions and explains the lack of inflated sub-Saturns.
    

    View Slide

42. blasé
    interpretable machine learning for high-resolution stellar spectra
    Gully-Santiago & Morley 2022 github.com/gully/blase
    - Forward models an entire high-bandwidth échelle spectrum
    - Treats properties of all 10,000 spectral lines free parameters
    - Transfer learns from precomputed synthetic spectra
    à Evaluable semi-empirical templates, extensible to EPRV & activity mitigation
    

    View Slide

43. blasé
    interpretable machine learning for high-resolution stellar spectra
    Gully-Santiago & Morley 2022 github.com/gully/blase
    - Forward models an entire high-bandwidth échelle spectrum
    - Treats properties of all 10,000 spectral lines free parameters
    - Transfer learns from precomputed synthetic spectra
    à Evaluable semi-empirical templates, extensible to EPRV & activity mitigation
    

    View Slide

44. We have detected up to 10% transit depth of He I 10833 Å from HPF spectra of HAT-P-67 b.
    The excess absorption preceeds the transit by up to 130 planetary radii in a large leading tail.
    The prominence of this leading tail is direct evidence for preferential dayside mass loss.
    We estimate a mass loss rate of 2 x 1013 g/s, and lifetime less than a Gyr.
    This pattern broadly agrees with theoretical predictions and explains the lack of inflated sub-Saturns.
    Conclusions
    

    View Slide