Hubble Cloud Atlas

Cloud Atlas is a large Hubble Space Telescope Treasury program that is exploring the atmospheres and condensate clouds in directly imaged exoplanets, planetary-mass and more massive brown dwarfs, with typical temperatures between 800 and 1,700 K.

The program is using high-precision, time-resolved photometry and spectroscopy of these rotating objects to derive course maps of their cloud structures. By comparing the cloud maps between different objects we can determine how the cloud properties (covering fraction, cloud thickness, vertical structure) depend on the upper atmosphere’s temperature and the surface gravity of the objects.

Observations: Cloud Atlas consists of two components: First, in the VAAS (Variability Amplitude Assessment Survey) we scout for the most likely high-amplitude variable brown dwarfs. We do this by obtaining 2-orbit-long quick looks: long enough to determine with high fidelity whether the brown dwarfs show modulations and rank-order them, but short enough to allow surveying about 20 objects economically. Second, we carry out a detailed study of the most interesting objects identified in the VAAS: our Deep Look Observations (DLO) are between 6 and 12 orbits long. These long observations allow us to cover a complete rotation (or a large fraction of it) with high-cadence, very high quality spectrally resolved data.

Targets: When completed, Cloud Atlas will provide time-resolved photometry/spectroscopy on about twenty brown dwarfs and directly imaged exoplanets. We are also utilizing all similar archival datasets on brown dwarfs, directly imaged exoplanets, and hot jupiters.

The Hubble Cloud Atlas program is led by Daniel Apai and includes an international team of 14 researchers: Yifan Zhou, Ben W. P. Lew, Glenn Schneider, Elena Manjavacas (all from University of Arizona), Theodora Karalidi (UCSC), Mark Marley (NASA Ames), Stan Metchev and Paolo Miles (UWO), Nicolas Cowan (McGill University), Adam Burgasser (UCSD), and Patrick Lowrance (Caltech/IPAC).


Cloud Atlas results

Ramp Effect Correction

Our Cloud Atlas survey is obtaining uniquely large set of very precise time-resolved near-infrared spectroscopy and photometry of rotating ultracool atmospheres. Our observations have been limited by the infamous HST/WFC3 “ramp effect” (a hook-like/exponential/linear shape in lightcurves) that has also plagued the transiting exoplanet observations.

As a spin-off of Cloud Atlas, our small team led by Yifan Zhou has developed the first solid state-physics based correction for the charge traps and their delayed release, which was the source of the ramp effect.

Even with constant illumination HST’s WFC3 detector does not show a flat line but a hook-like pattern known as the “ramp effect”. A major challenge for precise time-resolved observations required by exoplanet observations.

This correction has a major impact on HST observations: it increases observing efficiency by about 25%, provides more reliable and as precise (and often even more precise) data than the traditional empirical ramp effect fits — plus, it does not require a flat baseline, therefore it can correct exoplanet phase curves and increase orbital phase coverage.

Our work has been published in Zhou et al. 2017 and the correction method is available on a collaborative basis to the community. It has been also covered in a blog post.


Large-amplitude Rotational Modulations in an Extremely Red Brown Dwarf

Cloud Atlas had an exciting start: among the first brown dwarfs we targeted was WISE0047, a well-known L-type object among brown dwarf aficionados for its extremely red near-infrared color. Our initial HST/WFC3 observations obtained in our short, scouting VAAS observations indicated a very high likelihood for WISE0047 also being one of the few high-amplitude rotators. And indeed: our follow-up observations (Deep Look Observations) provided another six HST orbits showed a beautiful, high-amplitude modulation.

We found that WISE0047, an exceptionally red brown dwarf, is also one of the most variable brown dwarfs – and not only it changes its brightness, but also changes its color. It is likely that all three of its unusual properties stem from a single property affecting its clouds.

In a nice, focused study led by Ben Lew we presented the discovery and showed that this very red, very high amplitude object also shows the second most chromatic modulations yet seen – in other words, it changes more in the blue than it does in the red. This behavior suggests that the cloud particles that introduce the changes are small, comparable to the wavelength of the light we used to observe the object. Our findings also raise the question: Is it just a very lucky a coincidence that this very red brown dwarf is also the highest amplitude L-type variable brown dwarf and also the one with the yet strongest color modulations? It is quite possible that all three of these unusual facts stem from a single same property or cause – with the larger sample provided by Cloud Atlas we will be able to figure this out.


A Surprising Lightcurve in a Stellar Companion Brown Dwarf

In our new paper, led by Elena Manjavacas, we present the discovery of surprisingly complex lightcurve in the L6-type brown dwarf LP261-75B. This object is the first brown dwarf companion for which time-resolved spectral mapping is available and it offers an exciting clue on whether brown dwarfs and large-separation giant exoplanet that formed as companion to stars have different cloud composition and structure from single brown dwarfs and unbound planetary-mass objects.

LP261B is a late-L type companion to an M-dwarf star. Our Cloud Atlas data showed an exceptionally complex lightcurve: six peaks within 9.5 hours!

Elena’s study shows that the spectral modulations are quite gray, i.e, similar in the red and blue edges of our wavelength range. This suggests larger grain sizes than seen in our study for WISE0047 (Lew et al. 2016, see above); it is also more similar to two other unbound (non-companion) brown dwarfs we described in our 2015 paper led by Hao Yang.

LP261B is, however, quite unusual in the complexity of its lightcurve: while most other sources we studied display only roughly sine-like modulations in few hours worth of data, LP261B shows at least six peaks in just 9.5 hours. This complex behavior is reminiscent of the lightcurve complexity we have seen in our Extrasolar Storms program, where we monitored brown dwarfs for tens or hundreds of hours – that study (Apai et al. 2017 Science, blog) revealed that planetary scale waves are common in the high-amplitude brown dwarfs. Similarly, LP261B may be a rapidly rotating brown dwarf with quickly evolving lightcurve due to the beating of the waves – if that is the case, it will be the first non-L/T transition brown dwarf with such behavior!