Presentation is loading. Please wait.

Presentation is loading. Please wait.

Exoplanet atmospheres with JWST: degeneracy, systematics and how to avoid them Dr Joanna Barstow With thanks to: Patrick Irwin, Suzanne Aigrain, Sarah.

Similar presentations


Presentation on theme: "Exoplanet atmospheres with JWST: degeneracy, systematics and how to avoid them Dr Joanna Barstow With thanks to: Patrick Irwin, Suzanne Aigrain, Sarah."— Presentation transcript:

1 Exoplanet atmospheres with JWST: degeneracy, systematics and how to avoid them Dr Joanna Barstow With thanks to: Patrick Irwin, Suzanne Aigrain, Sarah Kendrew, Leigh Fletcher and Tom Greene

2 Outline Retrieval: how it works/doesn’t work Current status with exoplanet retrieval Exoplanet retrieval with JWST – golden age? Systematics and starspots

3 Wavelength Transparent Opaque Primary Wavelength Secondary Flux Transit Spectroscopy Gas abundances, cloud top pressure/particle size, mean molecular weight Temperature structure, gas abundances, cloud composition (maybe) Optical IR Optical IR

4 Spectral retrieval

5 v v

6 Opaque deep atmosphere Transparent upper atmosphere Opaque deep atmosphere Cloudy upper atmosphere Effect of clouds

7 Priors What is a prior? Initial model atmospheric state Additional constraint to beat underconstrained issue EXOPLANET PRIORS ARE TRICKY! If the prior drives the retrieval you have a BAD PROBLEM Image credit: NASA/ESA

8 How can retrieval go wrong? Example: Venus nightside spectra from Venus Express/VIRTIS 3 models, indistinguishable Details of cloud model trade off with water vapour abundance in the deep atmosphere

9

10 Outline Retrieval: how it works/doesn’t work Current status with exoplanet retrieval Exoplanet retrieval with JWST – golden age? Systematics and starspots

11 Exoplanet Spectroscopy: Current Problem is underconstrained. HD 189733b GJ 1214b Kreidberg et al. 2014

12 Exoplanet Spectroscopy: HD 189733b

13 Exoplanet spectroscopy: GJ 1214b 0.942 1.110 0.993

14 Outline Retrieval: how it works/doesn’t work Current status with exoplanet retrieval Exoplanet retrieval with JWST – golden age? Systematics and starspots

15 What we want from JWST In an ideal world… GJ 1214b HD 189733b Cloudy planets require broad wavelength coverage

16 What we want from JWST HD 189733b Sensitive to T at 1mbar < p < 100mbar Greater wavelength coverage provides more information  breaks degeneracy

17 Retrieval methods Optimal estimation (e.g. NEMESIS, Irwin et al. 2008) Fast Assumes Gaussian errors, priors and posteriors Great for solar system MCMC/DEMC/other fully Bayesian methods Allows full exploration of posterior including multiple minima Tends to be a bit slower than OE Get same answer if spectrum information content sufficient (Line et al. 2013)

18 Retrieval methods Available retrieval algorithms for exoplanets: NEMESIS (Irwin et al. 2008) – Oxford, OE/MCMC* CHIMERA (Line et al. 2013) – UCSC, DEMC/OE Tau-REX (Waldmann et al. 2015 a,b) – UCL, MCMC Madhu – Cambridge, MCMC SCARLET (Benneke 2015) – Caltech, nested sampling Caitlin Griffith et al. – UofA, ? BART (Harrington) – UCF, MCMC Others exist under development –Remco de Kok & collaborators at SRON *under development

19 Synthetic JWST observations Add Gaussian random noise (photon limit) to synthetic. Suggestions for improved noise model welcome! Hot Jupiter orbiting a sun-like star: cloud free, H2-He dominated, trace amounts of H2O, CO2, CO, CH4 *updated! Now includes background + read noise + dark current – see Greene et al. in press

20 Synthetic JWST observations Add Gaussian random noise (photon limit) to synthetic. Suggestions for improved noise model welcome! Hot Jupiter orbiting a sun-like star at 250 pc, single eclipse observation *updated! Now includes background + read noise + dark current – see Greene et al. in press

21 Synthetic JWST observations Add Gaussian random noise (photon limit) to synthetic. Suggestions for improved noise model welcome! Hot Jupiter orbiting a sun-like star at 250 pc, single transit observation *updated! Now includes background + read noise + dark current – see Greene et al. in press

22 Retrieval tests add noise generate spectra retrieval NEMESIS compare results See Barstow et al. 2015 MNRAS

23 JWST spectroscopic pick ‘n’ mix NIRISS 0.6 – 2.5 micron slitless spectroscopy Low throughput NIRCam 2.5 – 5 micron slitless spectroscopy Two grisms, 2.5 – 3.9 microns and 3.9 – 5 microns (also 1-2 um?) NIRSpec 0.6 – 5 micron Single shot with prism but low R, bright target restrictions Or in 3 shots MIRI 5 – 28 micron 5 – 12 micron at R~100, one shot 5 – 28 at R~2500, 3 shots

24 Instrument Choices Barstow et al. 2015: NIRSpec + MIRI Greene et al. In press – NIRISS + NIRCam + MIRI Differences? NIRISS + NIRCam + MIRI LRS

25 Instrument Choices Barstow et al. 2015: NIRSpec + MIRI Greene et al. In press – NIRISS + NIRCam + MIRI Differences? NIRISS + NIRCam + MIRI LRS NIRSPEC + MIRI LRS

26 Instrument Choices Barstow et al. 2015: NIRSpec + MIRI Greene et al. In press – NIRISS + NIRCam + MIRI Differences? NIRISS + NIRCam + MIRI LRS NIRISS + NIRCam + MIRI MRS

27 Retrieval tests Hot Jupiter orbiting a sun-like star, 250 pc Hot Jupiter single eclipse provides good constraints with NIRSpec + MIRI

28 Outline Retrieval: how it works/doesn’t work Current status with exoplanet retrieval Exoplanet retrieval with JWST – golden age? Systematics and starspots

29 Stitching and systematics Can’t observe simultaneously with NIRISS, NIRCam, NIRSpec and MIRI Therefore need to observe at least two transits What happens if baseline varies between observations: 1)Due to instrumental effects? 2)Due to stellar activity variation?

30 Offsets Wavelength invariant baseline offsets between NIRSpec/MIRI observations of same planet can be easily corrected for No offset 1x σ offset 3x σ offset

31 Starspots Image taken from Sanchiz- Ojeda et al., Nature, 2011 (lightcurve shows Kepler- 30c)

32 Starspots

33 No spots 3 % spots out of transit Hot Jupiter orbiting sun-like star at 250 pc: H2-He dominated, trace amounts of H2O, CO2, CO, CH4

34 Starspots No spots 3 % spots out of transit Hot Jupiter orbiting sun-like star at 250 pc: H2-He dominated, trace amounts of H2O, CO2, CO, CH4

35 Starspots No spots 10 % spots out of transit Hot Neptune orbiting M5 star at 15 pc: H2-He dominated, trace amounts of H2O, CO2, CO, CH4

36 Starspots H2O Spot-free 10% spots (out of transit) T*=3000 K, T_spot=2300 K 10% spots, offset corrected

37 Starspots CO Spot-free 10% spots (out of transit) T*=3000 K, T_spot=2300 K 10% spots, offset corrected

38 Conclusions Degeneracy is our enemy – JWST provides us with the best tools yet to beat it NIRISS+NIRCam+MIRI LRS/NIRSpec + MIRI LRS valid NIRSpec not so good in bright limit, but need more exposures for NIRISS + NIRCam Good stitching is important Starspots! Can do a lot with R = 100 spectroscopy MIRI MRS only if high resolution/>12 microns a requirement, otherwise doesn’t add much

39 Earth analogs? 30x spectra

40 Venus analogs? 50x spectra


Download ppt "Exoplanet atmospheres with JWST: degeneracy, systematics and how to avoid them Dr Joanna Barstow With thanks to: Patrick Irwin, Suzanne Aigrain, Sarah."

Similar presentations


Ads by Google