The Stellar Variability Noise Floor – ISSI 2021

ISSI Workshop on “Getting Ultra-Precise Planetary Radii with PLATO: The Impact of Limb Darkening and Stellar Activity on Transit Light Curves,” Bern, Switzerland, September 2021

Transcript

1. The Stellar Variability Noise Floor Brett Morris (⭐/🌎) with Monica Bobra (☀), Eric Agol (🌎), Yu Jin Lee (🎓) and Suzanne Hawley (⭐) Sept 21, 2021 1
2. Ongoing and Recent Projects • Olin Wilson Legacy Survey: Decade-scale CaII H & K monitoring of planet-hosting stars • Starspot coverage of FGKM stars ◦ High resolution spectroscopy [Direct: 1; CCF: 2], ◦ Rotational modulation via photometry [ABC: 3] • Joint inference of phase curve parameters in presence of (possible?) stellar activity ◦ 55 Cnc e with CHEOPS [4] ◦ Kepler-7 b with Kepler [5] • Planet hunts for objects orbiting white dwarfs [6] 2
3. APO/ARCES Spectra
4. APO/ARCES Spectra S-Index
5. Case Study: 55 Cnc • Slowly rotating K dwarf host star to five planets (39 d) • Planet e is a transiting super-Earth with a density consistent with rock (6.6 g/cm3) • Phase curve of planet cannot be explained strictly by reflection from the planet’s surface • Is stellar magnetic activity to blame? Morris+ (2021) Raw Detrend
6. Case Study: 55 Cnc Activity maximum Activity minimum OWLS Keck/HIRES (In prep) OWLS observations confirm CHEOPS observations occurred near activity maximum Activity maximum
7. 7 https://ui.adsabs.harvard.edu/abs/2020MNRAS.493.5489M/abstract
8. Planetary radius Three forms of variability during transits may affect transit radius measurements 1. Variability caused by non-uniformity in the surface brightness of the star which is being occulted by the planet 2. Variability due to temporal variations in the total flux emitted by the unocculted star. 3. Limb-darkening affects the depth of the transit depending on impact parameter These three effects limit how well we can measure the planet-star radius ratio. 8
9. SDO/HMI • Observes six wavelengths centered on Fe I spectral line at 6173 Å • All observations available at http://jsoc.stanford.edu • Focus on: ◦ Observations near active minimum (2018) ◦ b=0 -> maximal granule contrast, minimal spot contamination 9
10. Primary objectives 1. Effects of granulation and super-granulation on transit photometry of Earth a. Simulated “without” photon noise or time variations 2. Effects of solar oscillations, granulation, and activity (all together!) a. Observed directly with photon noise 10
11. Fifty frames of the SDO/HMI images after PSF deconvolution 11
12. Granules: 0.5 Mm Supergranules: 16 Mm Granulation processes are stochastic but have typical length and turnover timescales which stay roughly constant -> model them with a Gaussian process Types of granulation Granule: 0.5 Mm Super- Granule: 16 Mm Earth: 6 Mm 12
13. Static sun, moving Earth • Simulate transits across 282 deconvolved images of the inactive Sun • Fit each transit with a Mandel & Agol transit model • Residuals typically have standard deviations of 0.5 ppm, and a typical range is 2 - 4 ppm 13
14. Static sun, moving Earth 14
15. Static sun, moving Earth • Typical uncertainty on the exoplanet radius is 0.02% R p . We consider this a noise floor, or a lower limit, since we modelled the granulation pattern from a synthetic transit across a single, static image. • Autocorrelation timescale in residuals: 18 minutes ◦ Given orbital velocity, 18 minute transit time corresponds to 16 Mm occulted region -> consistent with supergranule size ◦ Supergranulation turnover timescale is 2 days, so supergranules are ~constant during an Earth transit 15
16. Static sun, moving Earth Expectations for other spectral types: • Granule size -> (Surface gravity)-1 ◦ Smaller star = smaller granules = smaller residual signal • Supergranulation: ? ◦ Amplitude is small, horizontal scales are large -> hard to simulate 16
17. • Take four days of 45 second SDO/HMI continuum intensity images • Disk-integrate to compute photometry • Inject an earth transit (full duration ~ 13 hrs; ingress/egress duration ~ 15 min) Dynamic sun, moving Earth 17
18. • Oscillations+granulation+activity are similar in scale to the transit depth • Radius uncertainty is about 3.6% Dynamic sun, moving Earth 18
19. Three-way degeneracy Impact parameter, planet radius, limb-darkening R p /R s Inc. a/R s 19
20. Three-way degeneracy Impact parameter, planet radius, limb-darkening 20
21. Three-way degeneracy Impact parameter, planet radius, limb-darkening Ingress/egress duration becomes too long and can be excluded by the observations Long tail for larger impact parameter solutions 21
22. Mitigation strategies for PLATO • Careful propagation of the stellar density measured via asteroseismology into the measurement of the exoplanet's orbital parameters • Auxiliary IR observations of PLATO targets where stellar variability and limb-darkening are weaker to measure precision exoplanet radii and orbital parameters • A longer extended mission to improve upon the radius measurement uncertainties 22
23. Conclusions • Transit residuals show p-mode oscillations and granulation, of order 100 ppm in amplitude. • We find that the brightness variations due to (super/)granulation only is of order 2-4 ppm, to the in-transit photometry • Radius uncertainties of 3.6% are possible with photon noise-limited photometry with a Gaussian process 23