Comprehensive time series analysis of the transiting extrasolar planet WASP-33b^⋆,⋆⋆

¹ Konkoly Observatory, 1121 Budapest, Hungary
e-mail: kovacs@konkoly.hu
² Department of Physics and Astrophysics, University of North Dakota, 58202-7129 Grand Forks, ND, USA
³ Department of Astrophysical Sciences, Princeton University, 08544 Princeton, NJ, USA
⁴ Harvard–Smithsonian Center for Astrophysics, 02138 Cambridge, MA, USA

Received: 22 May 2012
Accepted: 5 March 2013

Abstract

Context. HD 15082 (WASP-33) is the hottest and fastest rotating star known to harbor a transiting extrasolar planet (WASP-33b). The lack of high precision radial velocity (RV) data stresses the need for precise light curve analysis and gathering further RV data.

Aims. By using available photometric and RV data, we perform a blend analysis, compute more accurate system parameters, confine the planetary mass, and, attempt to cast light on the observed transit anomalies.

Methods. We combined the original HATNet observations and various followup data to jointly analyze the signal content and extract the transit component and used our RV data to aid the global parameter determination.

Results. The blend analysis of the combination of multicolor light curves yields the first independent confirmation of the planetary nature of WASP-33b. We clearly identify three frequency components in the 15–21 d^-1 regime with amplitudes 7–5 mmag. These frequencies correspond to the δ Scuti-type pulsation of the host star. None of these pulsation frequencies or their low-order linear combinations are in close resonance with the orbital frequency. We show that these pulsation components explain some but not all of the observed transit anomalies. The grand-averaged transit light curve shows that there is a ~1.5 mmag brightening shortly after the planet passes the mid-transit phase. Although the duration and amplitude of this brightening varies, it is visible even through the direct inspections of the individual transit events (some 40–60% of the followup light curves show this phenomenon). We suggest that the most likely explanation of this feature is the presence of a well-populated spot belt which is highly inclined to the orbital plane. This geometry is consistent with the inference from the spectroscopic anomalies. Finally, we constrain the planetary mass to M_p = 3.27 ± 0.73 M_J by using our RV data collected by the TRES spectrograph.