A Hubble Space Telescope transit light curve for GJ 436b^*

¹ Institut für Astrophysik, Georg-August-Universität Göttingen, Friedrich-Hund-Platz 1, 37077 Göttingen, Germany e-mail: bean@astro.physik.uni-goettingen.de
² McDonald Observatory, University of Texas, 1 University Station, C1402, Austin, TX 78712, USA
³ Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138, USA
⁴ Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218, USA

Received: 21 April 2008
Accepted: 4 June 2008

Abstract

We present time series photometry for six partial transits of GJ 436b obtained with the Fine Guidance Sensor instrument on the Hubble Space Telescope (HST). Our analysis of these data yields independent estimates of the host star's radius , and the planet's orbital period  d, orbital inclination , mean central transit time  HJD, and radius . The radius we determine for the planet is larger than the previous findings from analyses of an infrared light curve obtained with the Spitzer Space Telescope. Although this discrepancy has a 92% formal significance (1.7σ), it might be indicative of systematic errors that still influence the analyses of even the highest-precision transit light curves. Comparisons of all the measured radii to theoretical models suggest that GJ 436b has a H/He envelope of ~10% by mass. We point out the similarities in structure between this planet and Uranus and Neptune and discuss possible parallels between these planets' formation environments and dynamical evolution. We also find that the transit times for GJ 436b are constant to within 10 s over the 11 planetary orbits that the HST data span. However, the ensemble of published values exhibits a long-term drift and our mean transit time is 128 s later than that expected from the Spitzer ephemeris. The sparseness of the currently available data hinders distinguishing between an error in the orbital period or perturbations arising from an additional object in the system as the cause of the apparent trend. Assuming the drift is due to an error in the orbital period we obtain an improved estimate for it of  d. This value and our measured transit times will serve as important benchmarks in future studies of the GJ 436 system.