Know Thy Planet Know Thy Starspots

Talk given at e-Workshop in Boulder. How starspots confound exoplanet transmission spectra and derived exoplanet radii.

Transcript

1. Know thy planet, know thy starspots Michael Gully-Santiago Kepler/K2 Guest Observer Office e-workshop on exoplanet transit spectroscopy October 2019 sTARRy Luger et al. 2019
2. Key idea. Starspots confound transit depth. Intuitive explanation. Planets look bigger if they occult brighter-on-average flux.
3. Unocculted starspots confound the mapping of measured transit depth to estimated exoplanet radius. depth error ~ 8%
4. Spot-induced radius bias is wavelength dependent: you measure different transit depths at different wavelengths. Spot free
5. Transit Light Source Effect™ Spot-induced radius bias is wavelength dependent: you measure different transit depths at different wavelengths. Spot free e.g. Rackham et al. 2018
6. Spot-induced radius bias is wavelength dependent: you measure different transit depths at different wavelengths.
7. Spot-induced radius bias is wavelength dependent: you measure different transit depths at different wavelengths. Ben's talk tomorrow (9:00 AM MDT) for more about TLSE
8. How do you measure starspot coverage and contrast? Also: many other ways-- ask to see a list. Spot modulation from lightcurves: lower limit on coverage fraction Probabilistic spectral decomposition from spectroscopy: posterior constraint on starspot coverage and contrast
9. Spot modulation yields a lower limit for total spot coverage.
10. Spot modulation yields a lower limit for total spot coverage. Where should you normalize this lightcurve?
11. Spot modulation yields a lower limit for total spot coverage. Where should you normalize this lightcurve?
12. Spot modulation yields a lower limit for total spot coverage. A = 1.7% The starspots cover >1.7% of the surface. In Real Data™, you don't know where the normalization is.
13. V827 Tau, a heavily spotted young K7 in Taurus A = 23.5% phase-folded K2 lightcurve The starspots cover >23.5% of the surface.
14. V827 Tau, a heavily spotted young K7 in Taurus phase-folded ground-based lightcurves A = 48% The starspots cover >48% of the surface. Grankin et al. 2008; ASASSN; AAVSO
15. How do you measure starspot coverage and contrast? Spot modulation from lightcurves: lower limit on coverage fraction Probabilistic spectral decomposition from spectroscopy: posterior constraint on starspot coverage and contrast Also: many other ways-- ask to see a list.
16. Probabilistic spectral decomposition from spectroscopy: posterior constraint on starspot coverage and contrast Observed spectrum = Photosphere Starspot spectrum + Simultaneously vary photospheric, starspot, and spectral calibration parameters in a big MCMC to get joint constraints. Czekala et al. 2015, Gully-Santiago et al. 2017
17. Starspot emission is most amenable to detection in the infrared.
18. Starspot emission is most amenable to detection in the infrared. IGRINS R = 45,000 spectrograph All of H & K bands
19. V827 Tau, a heavily spotted young K7 in Taurus 17/48 available IGRINS spectral orders The starspots cover 79 ± 5% of the surface.
20. V827 Tau, a heavily spotted young K7 in Taurus 17/48 available IGRINS spectral orders The starspots cover 72 ± 8% of the surface.
21. V827 Tau, a heavily spotted young K7 in Taurus Constraints from all available data The starspots cover 65-80% of the surface, with a characteristic temperature of 2700-3000 K
22. I've shown examples on a heavily spotted young star. To what extent is it possible to apply spectral decomposition to exoplanet host stars? Ben's talk tomorrow (9:00 AM MDT)
23. What do you need for accurate characterization of transiting exoplanet atmospheres? 1. Precise knowledge of total spot coverage and contrast (this workshop / this talk) 2. Precise knowledge of the planet mass (Natasha Batalha et al. 2019) 3. Precision instrumentation (e.g. JWST) 4. Precise ephemerides (e.g. Benneke et al. 2017) 5. Critically evaluated exoplanet atmosphere models (last slides of this talk)
24. What do you need for accurate characterization of transiting exoplanet atmospheres? 1. Precise knowledge of total spot coverage and contrast (this workshop / this talk) 2. Precise knowledge of the planet mass (Natasha Batalha et al. 2019) 3. Precision instrumentation (e.g. JWST) 4. Precise ephemerides (e.g. Benneke et al. 2017) 5. Critically evaluated exoplanet atmosphere models (last slides of this talk)
25. "Yeah, but the model spectra are imperfect" -every astronomer ever when I tell them I'm using model spectra.
26. "Yeah, but the model spectra are imperfect" -every astronomer ever when I tell them I'm using model spectra. "You need to solve all of astrophysics just to fit a model to a spectrum." -David W. Hogg (NYU/CDS/CCA)
27. Mark Marley (NASA Ames) Has been solving all of astrophysics for decades. (insofar as it relates to modeling spectra of exoplanets)
28. https://youtu.be/HtWoexgYRsA?t=781 Mark Marley (NASA Ames)
29. *Caroline Morley (UTexas), Mark Marley, and me. **ultracool L/T dwarfs We* are critically evaluating exoplanet atmosphere models using high resolution spectra of exoplanet analogs**.
30. We* are critically evaluating exoplanet atmosphere models using high resolution spectra of exoplanet analogs**. *Caroline Morley (UTexas), Mark Marley, and me. **ultracool L/T dwarfs
31. Conclusions You need to know total spot coverage and contrast for accurate characterization of transiting exoplanet atmospheres. Caroline Morley, Mark Marley, and I are beginning an effort towards semi-empirical models of ultracool dwarfs, which will inform exoplanet atmosphere modeling. High resolution near infrared spectroscopy (e.g. IGRINS) can help with both of these goals.