2023A&A...677L..15S

The nearby mid-K dwarf HIP 66074 was recently identified as host to a candidate super-Jupiter companion on a ∼300 day, almost edge-on, orbit, based on Gaia Data Release 3 (DR3) astrometry. Initial attempts at confirming the planetary nature of the signal based on publicly available radial-velocity (RV) observations uncovered an intriguing conundrum: the inferred RV semi-amplitude appears to be a factor of 15 smaller than the one predicted based on the Gaia solution (corresponding to a 7-M_Jup companion on a close to edge-on orbit). We present the results of intensive RV monitoring of HIP 66074 with the HARPS-N spectrograph. We detected the companion at the Gaia period, but with an extremely eccentric orbit (e = 0.948 ± 0.004), a semi-amplitude $ K = 93.9^{+9.4}_{-7.0} $ m s^–1, and a minimum mass m_b sin i_b = 0.79 ± 0.05 M_Jup. We used detailed simulations of Gaia astrometry with the DR3 time-span to show that the conundrum can be fully resolved by taking into account the combination of the initially sub-optimal RV sampling and systematic biases in the Gaia astrometric solution, which include an underestimation of the eccentricity and incorrect identification of orbital inclination, which has turned out to correspond to a close to face-on configuration (i ≲ 13°). With an estimated mass in the approximate range of 3 - 7 M_Jup, we find that HIP 66074b (≡Gaia-3b) is the first exoplanet candidate astrometrically detected by Gaia to be successfully confirmed based on RV follow-up observations.