The near-infrared spectral energy distribution of β Pictoris b

Context. A gas giant planet has previously been directly seen orbiting at 8-10 AU within the debris disk of the similar to 12 Myr old star beta Pictoris. The beta Pictoris system offers the rare opportunity of both studying the physical and atmospheric properties of an exoplanet placed on a wide orbit and establishing its formation scenario. Aims. We aim to build the 1-5 mu m spectral energy distribution of the planet for the first time. Our goal is to provide secure and accurate constraints on its physical and chemical properties. Methods. We obtained J (1.265 mu m), H (1.66 mu m), and M' (4.78 mu m) band angular differential imaging of the system between 2011 and 2012. We used Markov chain Monte Carlo simulations of the astrometric data to revise constraints on the orbital parameters of the planet. Photometric measurements were compared to those of ultra-cool dwarfs and young companions. They were combined with existing photometry (2.18, 3.80, and 4.05 mu m) and compared to predictions from 7 PHOENIX-based atmospheric models in order to derive the atmospheric parameters (T-eff, log g) of beta Pictoris b. Predicted properties from (hot-start, cold-start, and warm start) evolutionary models were compared to independent constraints on the mass of beta Pictoris b. We used planet-population synthesis models following the core-accretion paradigm to discuss the planet's possible origin. Results. We detect the planetary companion in our four-epoch observations. We estimate J = 14.0 +/- 0.3, H = 13.5 +/- 0.2, and M' = 11.0 +/- 0.3 mag. Our new astrometry consolidates previous semi-major axis (8-10 AU) and excentricity (e <= 0.15) estimates of the planet. The location of beta Pictoris b in color-magnitude diagrams suggests it has spectroscopic properties similar to L0-L4 dwarfs. This enables one to derive Log(10) (L = L-circle dot) = 3.87 +/- 0.08 for the companion. The analysis with atmospheric models reveals that the planet has a dusty atmosphere with T-eff = 1700 +/- 100 K and log g = 4.0 +/- 0.5. Hot-start evolutionary models give a new mass of 10(-2)(+3) M-Jup from T-eff and 9(-2)(+3) M-Jup from luminosity. Predictions of cold-start models are still inconsistent with independent constraints on the planet mass. Warm-start models constrain the mass to M >= 6 M-Jup and the initial entropies to values (S-init >= 9.3K(b)/baryon) midway between those considered for cold/hot-start models, but probably closer to those of hot-start models.