gully
commented
1 year ago Here is the itemized list of each takeaway, with the goal of incorporating some text for each. We completed item 3 in a separate, linked Issue.
- [x] 1. Attributing the Helium signal to planet rather than activity #24
- [x] 2. Dayside heating and a headwind "pile up" as a cause for the leading tail
- [x] 3. Do the archival Keck HIRES spectra show e.g. H-alpha variability? #23
- [x] 4. F-stars, corona loss, radiative-convective boundary, and ideas for comparison studies of F-stars #25
- [x] 5. Constraining the planetary mass
- [x] 6. Impact of stellar evolution on incident radiation
- [x] 7. Vertical extent of the escaping atmosphere
- [x] 8. Prospects for followup
1. Attributing the Helium signal to planet rather than activity
Suvrath asked about attributing the signal to planetary mass loss and not stellar variability. At face value, many of the observations appear to be consistent with either interpretation: the Helium signal is primarily in the stellar rest frame, and has a width comparable to the large vsini~36 km/s. Why a preference for one interpretation over the other? Isn't a stellar interpretation somehow more parsimonious? Gummi pointed out that the vsini, stellar radius, inclination would imply a rotation period in the ~3-4 day range, comparable to the 4.8 day orbital period.
We discussed the relative merits of assigning each signal to one-or-the-other interpretation, and ultimately agreed that the most convincing argument may be our newly revised priors: The HAT-P-32 result shows that its extended Helium signal can be uniquely attributable to planetary atmosphere escape, and HAT-P-67b's low surface gravity and close-in orbit suggests the same mechanism could and should be at play here.
The exchange catalyzed a key takeaway: we should strengthen the case for planetary interpretation through a range of quantitative comparison studies (takeaway #4).
2. Dayside heating and a headwind "pile up" as a cause for the leading tail
Morgan MacLeod led a discussion on hydrodynamic simulations. He introduced a simplified heuristic for understanding the leading tail morphology (as opposed to solely trailing tail or two symmetric lobes) as follows: the material escaping the planet's dayside has an intial velocity towards the star, encountering a headwind and leading to a pile-up of over-dense material with slightly faster Keplerian velocity. The nightside of the planet may encounter a tailwind, exhibiting no such pileup and so an overall lower density for material with slightly slower Keplerian velocity.
The distinction of photoevaporation from dayside heating versus core-powered mass loss came up in this context, and led to strategies for simulating a conspicuous leading tail. Morgan stated that the key dimensionless control parameter is the ratio of orbital shear to the sound speed, splitting systems into "shear-dominated" versus "pressure-dominated", with HAT-P-32-like systems and WASP-107-like systems as archetypes.
3. Do the archival Keck HIRES spectra show e.g. H-alpha variability?
We noted that the Helium signal strength is so large (>10%) and extended that we might consider searching for unconventional indicators of atmospheric escape. Someone (maybe Gummi?) suggested H-alpha, and I reported that there are 22 Keck HIRES visits sampled across the orbital phase, amenable to a similar search for line-by-line variability. This archival investigation can also address questions of stellar activity.
4. F-stars, corona loss, radiative-convective boundary, and ideas for comparison studies of F-stars
Suvrath took a step back to wonder why these F-stars are showing such conspicuous signals, when our expectation going into this program was to focus on the K dwarfs.
Morgan discussed how this star's 6400 K Teff places its surface close to the radiative/convective borderline, which has a few consequences. First, hot stars lose their coronae and the associated high energy coronal radiation. Morgan said Antonija has thought closely about these themes. Second, conventional starspots may go away, but surface inhomogeneities leading to detectable lightcurve modulation are still conceivable from, say, star-planet interaction or other peculiarities.
We'd like to quantify this boundary to place limits on the star's high energy output and the prospect for stellar activity contamination. In short, we want to learn as much as we can about the star to understand the mechanisms driving the outflow. For example, we'd like to know what fraction of stars in this chunk of the HR diagram show intrinsic Helium variability, as a function of say rotation, to know how unusual this signal is from a stellar astrophysics perspective. Unfortunately the dearth of RV searches around hot stars means the data to answer this question are not likely to exist in archives. Alternatively we could look at TESS lightcurves to gain a coarse measure of whether this star's modulation stands out compared to stars similarly-situated in the HR diagram. These questions seem like research-grade mini-projects in their own right, so there's an economic tradeoff-- I'll see what's feasible.
5. Constraining the planetary mass
One main claim I repeated is that HAT-P-67b is "special": it has a low escape velocity for a moderately high instellation and equilibrium temperature. Folks agreed that a lot of that claim hinges on a relatively weak mass constraint (K<36 m/s at 1 sigma), leading to Mp~0.3 M_Jup +/- 0.2.
Last week Quang Tran conducted a un-restricted planet search experiment (i.e. not forcing the orbital period and phase known from transit measurements) on HAT-P-67 using the subset of spectra from before the recent HPF warm-up. Quang found a noise floor of about 230 m/s, owing to the large vsini, limited sampling, and relative dearth of lines in this hot star. Several telecon participants agreed that revising the mass constraint would help to make this "special" claim more robust and believable.
6. Impact of stellar evolution on incident radiation
Caroline asked how much this evolved star's flux has increased over time, and how that would induce planetary radius inflation at the current age. I did not have the number at the time, but now see that model tracks suggest a ~2x increase compared to the Zero-Age Main Sequence, and would therefore be consistent with significant planetary radius inflation.
7. Vertical extent of the escaping atmosphere
I had made the claim that the width of the Helium line width was consistent with the escaping atmosphere overfilling the entire rotating stellar disk, both in the direction of the orbit and perpendicular to the orbital plane. Morgan clarified that the velocity structure only strictly requires a thin sheet of material. We agreed, and settled on a preference for the disk-filling scenario due to the ~15% amplitude of flux loss-- a thin sheet would need to approach 100% flux loss across the projected transit chord, which seems implausible.
8. Prospects for followup
We briefly discussed prospects for following up this detection, since we may expect a large and measurable phase curve from the hot dayside and cool nightside. The ~5 day orbital period seemed prohibitive as a pitch to JWST point-and-stare. Secondary eclipse could be a cheaper and more feasible strategy.
We also wondered what is happening in the gap between phases -0.1 and 0.0, where the signal appears to peak. We wondered whether the signal could be even larger at this phase. Obtaining new HPF spectra at this phase is feasible next year.