Transmission Spectra of Three-Dimensional Hot Jupiter Model Atmospheres

Abstract

We compute models of the transmission spectra of planets HD 209458b, HD 189733b, and generic hot Jupiters. We examine the effects of temperature, surface gravity, and metallicity for the generic planets as a guide to understanding transmission spectra in general. We find that carbon dioxide absorption at 4.4 and 15 μm is prominent at high metallicity, and is a clear metallicity indicator. For HD 209458b and HD 189733b, we compute spectra for both one-dimensional and three-dimensional model atmospheres and examine the differences between them. The differences are usually small, but can be large if atmospheric temperatures are near important chemical abundance boundaries. The calculations for the three-dimensional atmospheres, and their comparison with data, serve as constraints on these dynamical models that complement the secondary eclipse and light curve data sets. For HD 209458b, even if TiO and VO gases are abundant on the dayside, their abundances can be considerably reduced on the cooler planetary limb. However, given the predicted limb temperatures and TiO abundances, the model's optical opacity is too high. For HD 189733b we find a good match with some infrared data sets and constrain the altitude of a postulated haze layer. For this planet, substantial differences can exist between the transmission spectra of the leading and trailing hemispheres, which are an excellent probe of carbon chemistry. In thermochemical equilibrium, the cooler leading hemisphere is methane-dominated, and the hotter trailing hemisphere is COdominated, but these differences may be eliminated by non-equilibrium chemistry due to vertical mixing. It may be possible to constrain the carbon chemistry of this planet, and its spatial variation, with James Webb Space Telescope.

Ground-based near-UV observations of 15 transiting exoplanets: constraints on their atmospheres and no evidence for asymmetrical transits

“…For illustration, we compare the observed R p /R * differences with wavelength for each target (Table 16) to theoretical predictions by Fortney et al (2010) for a model planetary atmosphere (Figs 8-10). The models used were estimated for planets with a 1 M Jup , g p = 25m s −1 , base radius of 1.25 R Jup at 10 bar, solar metallicity, and T eq closest to the measured value for each exoplanet (with model choices of 500, 750, 1000, 1250, 1500, 1750, 2000, 2500 K).…”

“…Czesla et al 2009;Oshagh et al 2013;Oshagh et al 2014; Zellem et al Variation of R p /R * versus wavelength for CoRoT-1b, GJ 436b, HAT-P-1b, HAT-P-13b, HAT-P-16b, and TrES-2b. Overplotted in red are atmospheric models by Fortney et al (2010) for planets with a 1 M Jup , g p = 25m s −1 , base radius of 1.25 R Jup at 10 bar, and T eq (specified on plot). The scaleheight of the planet is also shown on each plot for reference.…”

Erratum: Physical properties, star-spot activity, orbital obliquity and transmission spectrum of the Qatar-2 planetary system from multicolour photometry

“…developed by Fortney et al (2010) for comparison. The green line has been calculated for Jupiter's gravity (25 m s −2 ) with a base radius of 1.25 R Jup at the 10 bar level and at 1250 K. The opacity of TiO and VO molecules is excluded from the model and the optical transmission spectrum is dominated by Rayleigh scattering in the blue, and pressure-broadened neutral atomic lines of sodium (Na) at 589 nm and potassium (K) at 770 nm.…”

3D Models of Exoplanet Atmospheres

“…Qualitatively, the overall agreement of the models with the data is quite good: (1) the transit depth is weakly dependent on the viscosity; (2) there are no discrepancies between the data and model greater than 2 − σ; (3) the observed transit spectrum with wavelength shows the expected features due to water and Rayleigh scattering. In general, the success of this model is comparable to transmission spectra calculated from other models which utilize GCM simulations (Fortney et al 2010, Burrows et al 2010.…”

“…Fortney et al (2010) and Burrows et al (2010) perform similar calculations utilizing the 3D GCM models of Showman et al (2009) and Rauscher et al (2010), respectively. As mentioned above, the dynamical calculations differ amongst these models and ours, but the method for calculating the resulting spectra from the results is largely equivalent.…”