Search for: All records

Retrieval methods are a powerful analysis technique for modelling exoplanetary atmospheres by estimating the bulk physical and chemical properties that combine in a forward model to best fit an observed spectrum, and they are increasingly being applied to observations of directly imaged exoplanets. We have adapted taurex3, the Bayesian retrieval suite, for the analysis of near-infrared spectrophotometry from directly imaged gas giant exoplanets and brown dwarfs. We demonstrate taurex3’s applicability to sub-stellar atmospheres by presenting results for brown dwarf benchmark GJ 570D which are consistent with previous retrieval studies, whilst also exhibiting systematic biases associated with the presence of alkali lines. We also present results for the cool exoplanet 51 Eri b, the first application of a free chemistry retrieval analysis to this object, using spectroscopic observations from GPI and SPHERE. While our retrieval analysis is able to explain spectroscopic and photometric observations without employing cloud extinction, we conclude this may be a result of employing a flexible temperature-pressure profile which is able to mimic the presence of clouds. We present Bayesian evidence for an ammonia detection with a 2.7σ confidence, the first indication of ammonia in a directly imaged exoplanetary atmosphere. This is consistent with this molecule being present in brown dwarfs of a similar spectral type. We demonstrate the chemical similarities between 51 Eri b and GJ 570D in relation to their retrieved molecular abundances. Finally, we show that overall retrieval conclusions for 51 Eri b can vary when employing different spectral data and modelling components, such as temperature–pressure and cloud structures.

 

We present an atmospheric retrieval analysis of a pair of highly variable, ∼200 Myr old, early T type planetary-mass exoplanet analogs SIMP J01365662+0933473 and 2MASS J21392676+0220226 using the Brewster retrieval framework. Our analysis, which makes use of archival 1–15μm spectra, finds almost identical atmospheres for both objects. For both targets, we find that the data is best described by a patchy, high-altitude forsterite (Mg₂SiO₄) cloud above a deeper, optically thick iron (Fe) cloud. Our model constrains the cloud properties well, including the cloud locations and cloud particle sizes. We find that the patchy forsterite slab cloud inferred from our retrieval may be responsible for the spectral behavior of the observed variability. Our retrieved cloud structure is consistent with the atmospheric structure previously inferred from spectroscopic variability measurements, but clarifies this picture significantly. We find consistent C/O ratios for both objects, which supports their formation within the same molecular cloud in the Carina-Near moving group. Finally, we note some differences in the constrained abundances of H₂O and CO, which may be caused by data quality and/or astrophysical processes such as auroral activity and their differing rotation rates. The results presented in this work provide a promising preview of the detail with which we will characterize extrasolar atmospheres with JWST, which will yield higher-quality spectra across a wider wavelength range.

 

We present results from an atmospheric retrieval analysis of Gl 229B using the Brewster retrieval code. We find the best fit model to be cloud-free, consistent with the T dwarf retrieval work of Line et al.; Zalesky et al. and Gonzales et al. Fundamental parameters (mass, radius, log(L_Bol/L_Sun), log(g)) determined from our model agree within 1σto SED-derived values, except forT_effwhere our retrievedT_effis approximately 100 K cooler than the evolutionary model-based SED value. We find a retrieved mass of${50}_{-9}^{+12}$M_Jup, however, we also find that the observables of Gl 229B can be explained by a cloud-free model with a prior on mass at the dynamical value, 70M_Jup. We are able to constrain abundances for H₂O, CO, CH₄, NH₃, Na and K and find a supersolar C/O ratio as compared to its primary, Gl 229A. We report an overall subsolar metallicity due to atmospheric oxygen depletion, but find a solar [C/H], which matches that of the primary. We find that this work contributes to a growing trend in retrieval-based studies, particularly for brown dwarfs, toward supersolar C/O ratios and discuss the implications of this result on formation mechanisms and internal physical processes, as well as model biases.

 

Comparisons of atmospheric retrievals can reveal powerful insights on the strengths and limitations of our data and modeling tools. In this paper, we examine a sample of five L dwarfs of similar effective temperature (T_eff) or spectral type to compare their pressure–temperature (P-T) profiles. Additionally, we explore the impact of an object’s metallicity and the signal-to-noise ratio (S/N) of the observations on the parameters we can retrieve. We present the first atmospheric retrievals: 2MASS J15261405+2043414, 2MASS J05395200−0059019, 2MASS J15394189−0520428, and GD 165B increasing the small but growing number of L dwarfs retrieved. When compared to the atmospheric retrievals of SDSS J141624.08+134826.7, a low-metallicity d/sdL7 primary in a wide L+T binary, we find that similarT_effsources have similar P-T profiles with metallicity differences impacting the relative offset between their P-T profiles in the photosphere. We also find that for near-infrared spectra, when the S/N is ≳80 we are in a regime where model uncertainties dominate over data measurement uncertainties. As such, S/N does not play a role in the retrieval’s ability to distinguish between a cloud-free and cloudless model, but may impact the confidence of the retrieved parameters. Lastly, we also discuss how to break cloud model degeneracies and the impact of extraneous gases in a retrieval model.

 

ABSTRACT We present the most detailed data-driven exploration of cloud opacity in a substellar object to-date. We have tested over 60 combinations of cloud composition and structure, particle-size distribution, scattering model, and gas phase composition assumptions against archival 1–15 μm spectroscopy for the unusually red L4.5 dwarf 2MASSW J2224438-015852 using the Brewster retrieval framework. We find that, within our framework, a model that includes enstatite and quartz cloud layers at shallow pressures, combined with a deep iron cloud deck fits the data best. This model assumes a Hansen distribution for particle sizes for each cloud, and Mie scattering. We retrieved particle effective radii of $\log _{10} a {\rm (\mu m)} = -1.41^{+0.18}_{-0.17}$ for enstatite, $-0.44^{+0.04}_{-0.20}$ for quartz, and $-0.77^{+0.05}_{-0.06}$ for iron. Our inferred cloud column densities suggest ${\rm (Mg/Si)} = 0.69^{+0.06}_{-0.08}$ if there are no other sinks for magnesium or silicon. Models that include forsterite alongside, or in place of, these cloud species are strongly rejected in favour of the above combination. We estimate a radius of 0.75 ± 0.02 RJup, which is considerably smaller than predicted by evolutionary models for a field age object with the luminosity of 2M2224-0158. Models which assume vertically constant gas fractions are consistently preferred over models that assume thermochemical equilibrium. From our retrieved gas fractions, we infer ${\rm [M/H]} = +0.38^{+0.07}_{-0.06}$ and ${\rm C/O} = 0.83^{+0.06}_{-0.07}$. Both these values are towards the upper end of the stellar distribution in the Solar neighbourhood, and are mutually consistent in this context. A composition towards the extremes of the local distribution is consistent with this target being an outlier in the ultracool dwarf population.