Yuhan Yao SPF @ University of Michigan Jan 16 2018 “An HST/STIS Optical Transmission Spectrum of Warm Neptune GJ 436b” (Lothringer et al. 2018) arXiv: 1801.00412 

GJ 436 ~3400K 0.41 ±0.05M⊙ GJ 436b 700-800K, 21.4 M⊕ , 4.2 R⊕ 2.6 d orbital period a = 0.029AU Mass loss from photo evaporative escape. Ehrenreich et al. (2015) 

GJ 436b 700-800K, 21.4 M⊕ , 4.2 R⊕ 2.6 d orbital period a = 0.029AU GJ 436 ~3400K 0.41 ±0.05M⊙ 

HST/WFC3 obs. (Knutson et al. 2014) Rule out H/He-rich cloud free High μ vs. Aerosols (Clouds & Hazes) 

HST/WFC3 obs. (Knutson et al. 2014) Rule out H/He-rich cloud free High μ vs. Aerosols (Clouds & Hazes) Spitzer obs. (Stevenson et al. 2010) Low CH4, High CO or CO2 

Modelling (Morley et al. 2017) Photochemical hazes does not fit obs. Best fit self-consistent model require high metallicity, tidal heating, and disequilibrium chemistry via quenching. 

Why optical? Sing et al. (2011) HST/STIS, hot Jupiter HD 189733b slope Spot occultations were trimmed for this measurement. Rayleigh scattering by small aerosols model Cloud free model (Fortney et al. 2010) 

Light curve: STIS/HST, 530-1040nm Visit 1 Visit 2 Raw white light curve Systematics-corrected Limb darkening coefficients are derived from interferometrically determined stellar parameters; transit properties are from literature. 

Marginalization 1. HST orbital phase 2. Time 3. Slope of the spectral trace 4. Position of the spectral trace in spatial direction 5. Position of the spectral trace in dispersion direction Covariates: 

Marginalization Dots: Raw data (7721-8210 Å) Solid line: the most complex model. Dashed line: the least complex model. 1. HST orbital phase 2. Time 3. Slope of the spectral trace 4. Position of the spectral trace in spatial direction 5. Position of the spectral trace in dispersion direction Covariates: 

No slope shortward of 0.6µm. Both visits are in general agreement (with a flat line). Perhaps an increase of transit depth at 0.8µm. Results: Transmission spectra of GJ 436b 

No slope shortward of 0.6µm. Both visits are in general agreement (with a flat line). Perhaps an increase of transit depth at 0.8µm. Results: Transmission spectra of GJ 436b No Na & K absorption 

Other effects? ~7.4yr, ∆max ~10mmag Folded by 44.1d, ∆max ~3mmag Periodograms of raw (gray) and de-trended (black) photometry. 44.1d APT’s 14-year stellar monitoring of GJ 436: 1. Star Spots 

Other effects? 1. Star Spots Interplay between star spots and plages (faculaes). 

Other effects? 1. Star Spots Un-occulted spot makes planet appear larger in shorter wavelength. Interplay between star spots and plages (faculaes). 

Other effects? 1. Star Spots Any difference between visit 1 and visit 2 is not due to stellar activity. ∆max ~(10+3)=13mmag ∆flux ~1.4% Up to 150ppm. Trend (the can be) introduced by star spots. Trend (that can be) introduced by plages (faculaes). 

Other effects? 2. Different Orbital Solutions Lanotte et al. (2014) Morello et al. (2015) Knutson et al. (2014) Use parameters from: For each orbital solution, there is a uniform offset. à Orbital solution does not affect non-detection of a scattering slope and K/Na. 

Other effects? 3. Model Comparisons Models from Morley et al. (2017) fsed <0.3: thick cloud fhaze : haze efficiency (~haze mass) Rule out low fhaze . Disfavor small particle radius hazes. 

GJ 436b HAT-P-26b GJ 1214b Stevenson et al. (2016) Rackham et al. (2017) Wakeford et al. (2017) This work 

GJ 1214b Rackham et al. (2017) GJ 1214b Stellar Plages? 

GJ 1214b Rackham et al. (2017) GJ 1214b Stellar Plages? GJ 436b HAT-P-26b Model with stellar plages Biases from LDCs? Additional opacity source? 

A JWST GTO target ! Simulated JWST data (NIRSpec, MIRI), 100*Metallicity, no quenching, with internal heating. 

Thank you to: Michael Meyer, Ke (Coco) Zhang, Larissa Markwardt Next steps? 1. Need JWST to distinguish between high metallicity and moderate metallicity with clouds scenarios. 2. Why 0.8 micron increase of transit depth ? 3. Possible interplay between star spots and plages ?