Exoplanet Atmospheres

Our group is using different techniques to characterize large-separation exoplanets and brown dwarfs. These projects complement our studies of transiting planet atmospheres, which all represent heavily irradiated atmospheres.

My group’s papers on rotational mapping of brown dwarf and exoplanet atmospheres: ADS Library.

Daniel Apai at a conference in Bern, Switzerland.

Daniel Apai at a conference in Bern, Switzerland.

Rotational Mapping

Clouds play a critical role in the physics and chemistry of brown dwarfs and giant exoplanets. Clouds are intrinsically complex and multi-dimensional structures; however, no existing telescope can spatially resolve brown dwarfs. We use a new method for brown dwarfs, rotational phase mapping, to explore the properties of the cloud cover in these sources. By precisely measuring the changes in the brightness and color of these sources as they rotate, we can explore their surface brightness distributions, thus creating rough maps of their cloud cover and temperature distributions.

For these studies we use the high-precision time-resolved Hubble Space Telescope near-infrared spectroscopy and Spitzer Space Telescope photometry in multiple programs.

Clouds in M, L, and T atmospheres

The techniques we are developing for the brown dwarf characterization will, in the near future, be also used for mapping directly imaged exoplanets.

Recent press releases:

Spitzer PR: Stormy Stars
Spitzer PR: NASA Telescopes See Weather Pattern in Brown Dwarf
UA News: Scientists Peer Into a Brown Dwarf’s Stormy Atmosphere

If you would like a video summary of our results, check out my recent talk at the HST2020 Symposium: http://tinyurl.com/pjjbyv4

Some recent papers and results:

Two-dimensional Maps of Exoplanets and Brown Dwarfs from Lightcurves

A series of HST images of Jupiter are used to validate our MCMC-optimized exoplanet/brown dwarf mapping algorithm Aeolous.

A series of HST images of Jupiter are used to validate our MCMC-optimized exoplanet/brown dwarf mapping algorithm Aeolous.

How can we map exoplanets based on light curves? In this new paper led by Theodora Karalidi we show that with MCMC modeling of high-quality light curves the dominant cloud features of gaseous exoplanets and brown dwarfs can be mapped in 2D. The method is demonstrated on a really cool HST time series of Jupiter images and then applied to two brown dwarfs. http://arxiv.org/abs/1510.04251

Cloud structure at the L/T Transition
Apai et al. 2013 ApJ: HST Spectral Maps Reveal Cloud Thickness Variations in L/T Transition Brown Dwarfs

Our HST spectral maps demonstrated that L/T transition brown dwarfs have thin/thick cloud cover and not cloud/cloud-free patches, as predicted.

Our HST spectral maps demonstrated that L/T transition brown dwarfs have thin/thick cloud cover and not cloud/cloud-free patches, as predicted.

The L/T spectral type transition marks the transition from cloud-covered photospheres (L-type) to mostly cloud-free atmospheres (T-type). It was long thought that objects at the transition region would display photospheres with patches of clouds and cloud-free regions. In our 2013 paper we applied HST spectral mapping to L/T brown dwarfs for the first time and showed that these objects do not have large, cloud-free holes, but instead their photospheres is covered by patches of thin and warm clouds, and thick and cool clouds (see illustration).

Brown Dwarfs are Patchy
Buenzli et al. 2014: Brown Dwarf Photospheres are Patchy: A Hubble Space Telescope Near-infrared Spectroscopic Survey Finds Frequent Low-level Variability

We carried out an HST SNAP survey to assess the frequency of patchy cloud covers in brown dwarfs as a function of their temperature and spectral type. Our short (40 minutes), but very sensitive time-resolved observations identified temporal variations in 1/3 of our sample.

Once the limitations of our study (short coverage, limited precision, viewing angle) are considered, this frequency indicates that most (and probably all) brown dwarfs have intrinsically patchy cloud covers and display rotational variability when observed at sub-percent accuracy. These results are important because:  1) they demonstrate the variability can be used as a general tool to study ultracool atmospheres, including directly imaged planets; and 2) they show that the variable brown dwarfs we study in other studies are representative to brown dwarfs in general.

Multi-Layer Scans of Ultracool Atmospheres

Buenzli et al. 2012 ApJ: Vertical Atmospheric Structure in a Variable Brown DwarfPhase Shifts

Due to the presence of molecular and gas-phase absorbers in planetary atmospheres different wavelengths probe different pressures. In a pioneering study published in 2012 we used the Hubble and Spitzer Space Telescopes simultaneously to measure rotational phase curves (think vertical scans) of the surface brightness of an ultracool brown dwarf. Thanks to the combined wavelength coverage of HST and Spitzer we were able to scan six distinct layers in the atmosphere, probing pressures between 0.1 to ~20 bars.

Surprisingly, we found very clear rotational phase shifts in the different layers; furthermore, the offsets were monotonically changing with the pressure. In other words, we see similar brightness distribution in the atmosphere at different levels, but the deeper we look the more offset the pattern is.

This exciting study demonstrated how we can derive multi-layer maps for ultracool atmospheres; and the phase shift pattern provided the first evidence for a vertically-longitudinally organized brightness distribution in a brown dwarf.

High-Altitude Hazes in L-dwarfs and Water-band Variability as Cloud Probe
 Yang et al. 2015 ApJ: HST Rotational Spectral Mapping of Two L-type Brown Dwarfs: Variability in and out of Water Bands indicates High-altitude Haze Layers


In our 2015 study we again used powerful HST time-resolved spectroscopy to study clouds in brown dwarfs, but this time we targeted earlier type targets, L-dwarfs, which have thick silicate clouds in their photospheres. These objects have not been mapped previously. As part of our Extrasolar Storms program we observed two L-dwarfs and compared these to the L/T dwarfs we have observed previously.

The key discovery of this study was a difference in the wavelength-dependence of the variability. While the L/T transition dwarfs (early T) show reduced variability in the water band, the L dwarfs display the same amplitude in and out of the water band. We proposed that this is due to a difference in the depth at which the clouds are located: in L dwarfs the clouds are high up in the atmosphere, while in the L/T dwarfs they are deeper. This means that there is a larger gas column above the L/T dwarf clouds than the L clouds. As the opacity in the water band is higher than outside the water band, the larger gas column translate to a reduced flux in the water band. In contrast, the clouds in L dwarfs are so high up that the optical depth difference between the water-band and out-of-the-water (continuum) wavelengths is negligible, therefore there is no difference in the variability amplitude.

In an interesting twist in this finding is that our radiative transfer modeling revealed that the particles in L-dwarfs have to be so high up in the atmosphere that they cannot be part of a simple single cloud, which base is set by the condensation of silicates. Therefore, the particles have to be very small grains lofted in the upper atmosphere with a very long settling time — in other words, haze particles.

This study presented the first L-dwarf spectral mapping and discovered evidence for unexpected high-altitude hazes in these objects.