While giant extrasolar planets have been studied for more than two decades now, there are still some open questions as to their dominant formation and migration processes, as well as to their atmospheric evolution in different stellar environments. In this paper, we study a sample of giant transiting exoplanets detected by the Kepler telescope with orbital periods up to 400 days. We first defined a sample of 129 giant-planet candidates that we followed up with the SOPHIE spectrograph (OHP, France) in a 6-year radial velocity campaign. This allowed us to unveil the nature of these candidates and to measure a false-positive rate of 54.6 ± 6.5% for giant-planet candidates orbiting within 400 days of period. Based on a sample of confirmed or likely planets, we then derived the occurrence rates of giant planets in different ranges of orbital periods. The overall occurrence rate of giant planets within 400 days is 4.6 ± 0.6%. We recovered, for the first time in the Kepler data, the different populations of giant planets reported by radial velocity surveys. Comparing these rates with other yields, we find that the occurrence rate of giant planets is lower only for hot Jupiters but not for the longer-period planets. We also derive a first measurement of the occurrence rate of brown dwarfs in the brown-dwarf desert with a value of 0.29 ± 0.17%. Finally, we discuss the physical properties of the giant planets in our sample. We confirm that giant planets receiving moderate irradiation are not inflated, but we find that they are on average smaller than predicted by formation and evolution models. In this regime of low-irradiated giant planets, we find a possible correlation between their bulk density and the iron abundance of the host star, which needs more detections to be confirmed.

Compositions of rocky exoplanets: The interior compositions of small rocky exoplanets cannot be observed directly but are expected to relate to the composition of the host star. Adibekyan et al . analyzed a sample of rocky exoplanets, inferring the planets’ iron fractions by combining their masses and radii with an interior structure model. The iron fractions of the host stars were calculated from stellar elemental abundances. The two iron fractions, that of the planets and that of the stars, correlate with each other, but the slope is steeper than 1, indicating that planet formation processes modify the compositions of rocky planets. —KTS

Context. Proxima Centauri is the closest star to the Sun. This small, low-mass, mid M dwarf is known to host an Earth-mass exoplanet with an orbital period of 11.2 days within the habitable zone, as well as a long-period planet candidate with an orbital period of close to 5 yr. Aims. We report on the analysis of a large set of observations taken with the ESPRESSO spectrograph at the VLT aimed at a thorough evaluation of the presence of a third low-mass planetary companion, which started emerging during a previous campaign. Methods. Radial velocities (RVs) were calculated using both a cross-correlation function (CCF) and a template matching approach. The RV analysis includes a component to model Proxima’s activity using a Gaussian process (GP). We use the CCF’s full width at half maximum to help constrain the GP, and we study other simultaneous observables as activity indicators in order to assess the nature of any potential RV signals. Results. We detect a signal at 5.12 ± 0.04 days with a semi-amplitude of 39 ± 7 cm s⁻¹. The analysis of subsets of the ESPRESSO data, the activity indicators, and chromatic RVs suggest that this signal is not caused by stellar variability but instead by a planetary companion with a minimum mass of 0.26 ± 0.05 M_⊕ (about twice the mass of Mars) orbiting at 0.029 au from the star. The orbital eccentricity is well constrained and compatible with a circular orbit.