Predicting the Presence of Clouds in Exoplanet Atmospheres

Accepted for publication in ApJL  • 

The H2O – J index measures the strength of the 1.4 μm water feature using HST's WFC3 instrument in units of atmospheric scale height (H). The nine planets with strong measured water features (H2O – J > 1, blue squares) have Teq > 700 K and log g > 2.8 (dotted lines). The remaining five planets with H2O – J < 1 (red squares) have muted water features. The size of the squares is inversely proportional to the uncertainty in the H2O – J index, thus larger symbols carry more weight. Like the dotted lines, the dashed diagonal line also delineates these two "classes" of exoplanets. Within the left panel, the gray circles depict confirmed transiting exoplanets with known masses and radii. The coloring in the right panel depicts the best-fit 2D model segmented into four quadrants (yellow lines and numerals). Additional measurements are needed to more precisely determine the transition region for atmospheres with and without obscuring clouds/hazes at the depths probed by HST/WFC3.


Water in the Atmosphere of HAT-P-26b?

Accepted for publication in ApJ  • 

Left: The newly-commissioned LDSS-3C instrument on Magellan provides enhanced sensitivity and suppressed fringing in the red optical, thus advancing the search for the spectroscopic signature of water in exoplanetary atmospheres from the ground.

Right: Using data acquired by LDSS-3C, we search for evidence of water vapor in the transmission spectrum of the Neptune-mass planet HAT-P-26b. Our measured spectrum is best explained by the presence of water vapor, a lack of potassium, and either a high-metallicity, cloud-free atmosphere or a solar-metallicity atmosphere with a cloud deck at ~10 mbar.


Phase-resolved emission spectrum of WASP-43b

Published Oct 9, 2014 in Science  • 

Band-integrated "white" phase curve of WASP-43b using 61 orbits from the Hubble Space Telescope. The systematics-corrected flux values are binned in time, normalized to the stellar flux, and have 1σ error bars. Each color represents data acquired from a different HST visit. The phase curve depicts steadily increasing and decreasing observed flux which originates from different longitudes of the tidally-locked planet as it makes one complete rotation. Light from the planet is blocked near an orbital phase of 0.5 as it is eclipsed by its host star. The shape of the phase curve is noticeably asymmetric. Inset, for comparison, is the white light curve primary transit.

Phase-resolved emission spectrum of WASP-43b relative to the stellar flux. We illustrate the planet's emitted light in 15 spectrophotometric channels (colored points with 1σ error bars) and compare our measurements to best-fit atmospheric models (colored lines). White diamonds depict the models binned to the resolution of the data. The absorption feature from 1.35 to 1.6 μm confirms the presence of water vapor in WASP-43b's atmosphere. For clarity, we only display planet-to-star flux ratios at four planet phases: full (0.5, secondary eclipse), wanning gibbous (0.62), half (0.75), and wanning crescent (0.88). A time-lapse video of the WASP-43b's phase-resolved emission spectrum is available here.


C/O > 1 in the Dayside Atmosphere of WASP-12b

Published July 24, 2014 in ApJ  • 

WASP-12b dayside emission spectrum with atmospheric models. Our comprehensive analysis of available WASP-12b data includes HST/WFC3 points (green squares) in the NIR, Spitzer/IRAC points (red circles) from 3 – 10 μm, and previously-published ground-based points (triangles). The isothermal blackbody model is rejected at 4.5 μm with a significance of 7σ. The oxygen-rich model requires 5x less H2O and 100x more CO2 than solar composition. This physically-implausible scenario is 670x less probable than the best-fitting carbon-rich model, which includes absorption due to C2H2 and HCN. We conclude that a dayside atmospheric model with a carbon-to-oxygen ratio > 1 provides the best fit to the available data.


Transmission Spectrum of the Hot Jupiter WASP-12b

Published May 14, 2014 in AJ  • 

WASP-12b transmission spectrum with best-fit enhanced scattering atmospheric models. The data include HST/STIS and Gemini-N/GMOS observations in the optical, HST/WFC3 observations in the NIR, and Spitzer/IRAC observations from 3 – 5 μm. These data rule out a cloud-free, H2 atmosphere with no additional opacity sources. The solid black line depicts an atmospheric model with a solar C/O and the dashed orange line corresponds to a planetary atmosphere with a C/O > 1, both with enhanced scattering. A linear fit representing an idealized scattering model (dotted line) achieves the best fit, suggesting the presence of clouds or hazes in the atmosphere of WASP-12b.


Clouds in the atmosphere of the super-Earth GJ 1214b

Published Dec 31, 2013 in Nature  • 

The results from our 60-orbit HST/WFC3 campaign to characterize the atmosphere of GJ 1214b have been published in Nature. The featureless spectrum is consistent with an atmosphere dominated by clouds. Top: Transmission spectrum measurements from our data (black points) and previous work (gray points), compared to theoretical models (lines). The error bars correspond to 1σ uncertainties. Each data set is plotted relative to its mean. Our measurements are consistent with past results for GJ 1214 using WFC3 . Previous data rule out a cloud-free solar composition (orange line), but are consistent with a high-mean molecular weight atmosphere or a hydrogen-rich atmosphere with high-altitude clouds. Bottom: Detail view of our measured transmission spectrum (black points) compared to high mean molecular weight models (lines). The colored points correspond to the models binned at the resolution of the observations. The data are consistent with a featureless spectrum, but inconsistent with cloud-free high-mean molecular weight scenarios. Fits to pure water, methane, carbon monoxide, and carbon dioxide models are ruled out at 16.1, 31.2, 7.6, and 5.5σ confidence, respectively.


Discovering Near-Earth Objects Using the Kepler Spacecraft

Sep 3, 2013  • 

In response to a call for white papers, we propose a new Kepler mission, called NEOKepler, that would survey near Earth's orbit to identify potentially hazardous objects (PHOs). In a simulated 12-month survey, we estimate that NEOKepler would detect ∼150 new NEOs with absolute magnitudes of less than 21.5, ∼50 of which would be new PHOs. This would increase the annual PHO discovery rate by at least 50%. Kepler would also be sensitive to objects inside Earth’s orbit, discovering more objects in its first year than are currently known to exist. As an alternative science goal, NEOKepler could employ a different observing strategy to discover suitable targets for NASA’s Asteroid Redirect Mission.


Discovery of Two Nearby Sub-Earth-Sized Exoplanet Candidates

July 18, 2012  • 

Artist's impression of UCF-1.01, an exoplanet candidate that is two-thirds the size of Earth. The planet was found by Kevin Stevenson in data he was analyzing from the Spitzer Space Telescope. The background planet is the hot Neptune GJ 436b. Because of its proximity to our solar system, the starfield shares many of our culture's cosmic landmarks. To the left, the constellation of Orion gleams, though in a distorted shape compared to our vantage point on Earth.

To take a virtual trip from Earth to UCF-1.01, watch .


Disequilibrium Chemistry in the Atmosphere of GJ 436b

April 22, 2010  • 


Left: Artist's impression of the moderately-irradiated Neptune-sized exoplanet GJ 436b. In the Nature article titled "Possible thermochemical disequilibrium in the atmosphere of the exoplanet GJ 436b" by Stevenson et al., we suggest that GJ 436b's atmosphere is abundant in CO and deficient in methane (CH4) by a factor of ~7,000. This result is unexpected because, based on current models at this temperature, the atmospheric carbon should prefer CH4 over CO.

Right: Eclipse lightcurves of GJ 436b. The six lightcurves seen here have been normalized with respect to the system's light (star + planet) and offset vertically for ease of comparison. The observed dips occur when the planet is being eclipsed by its parent star (star light only). The difference between the two levels is the contribution from the planet at that wavelength.