A hot mini-Neptune in the radius valley orbiting solar analogue HD 110113

Abstract

We report the discovery of HD 110113 b (TOI-755.01), a transiting mini-Neptune exoplanet on a 2.5-day orbit around the solar-analogue HD 110113 (Teff= 5730K). Using TESS photometry and HARPS radial velocities gathered by the NCORES program, we find HD 110113 b has a radius of 2.05 ± 0.12 R⊕ and a mass of 4.55 ± 0.62 M⊕. The resulting density of $2.90^{+0.75}_{-0.59}$ g cm−3 is significantly lower than would be expected from a pure-rock world; therefore, HD 110113 b must be a mini-Neptune with a significant volatile atmosphere. The high incident flux places it within the so-called radius valley; however, HD 110113 b was able to hold on to a substantial (0.1-1 per cent) H-He atmosphere over its ∼4Gyr lifetime. Through a novel simultaneous Gaussian process fit to multiple activity indicators, we were also able to fit for the strong stellar rotation signal with period 20.8 ± 1.2 d from the RVs and confirm an additional non-transiting planet, HD 110113 c, which has a mass of 10.5 ± 1.2 M⊕ and a period of $6.744^{+0.008}_{-0.009}$ d.

Investigating the architecture and internal structure of the TOI-561 system planets with CHEOPS, HARPS-N, and TESS

“…Considering this, we performed an additional dynesty fit assuming a 4-planet model and including a Gaussian Process (GP) regression with a quasi-periodic kernel, as formulated in Grunblatt et al (2015), to account for the long-term signal. We modelled simultaneously the RVs and the S-index time series in order to better inform the GP (Langellier et al 2021;Osborn et al 2021), using two independent covariance matrices for each dataset with common GP hyper-parameters except for the amplitude of the covariance matrix, assuming uniform, non-informative priors on all of them. The fit suggests a periodicity longer than in the case of the 5-Keplerian fit, the HARPS-N jitter is significantly improved ( HARPS−N ∼ 1.30 m s −1 ) when including the GP model.…”

Investigating the architecture and internal structure of the TOI-561 system planets with CHEOPS, HARPS-N and TESS

“…Considering this, we performed an additional dynesty fit assuming a 4-planet model and including a Gaussian Process (GP) regression with a quasi-periodic kernel, as formulated in Grunblatt et al (2015), to account for the long-term signal. We modelled simultaneously the RVs and the S-index time series in order to better inform the GP (Langellier et al 2021;Osborn et al 2021), using two independent covariance matrices for each dataset with common GP hyper-parameters except for the amplitude of the covariance matrix, assuming uniform, non-informative priors on all of them. The fit suggests a periodicity longer than ∼ 570 d, but the GP model is too flexible to derive a precise period value, considering also that the global RV baseline (∼ 768 d) is comparable with the periodicity of the long-term signal.…”

pyaneti II: A multidimensional Gaussian process approach to analysing spectroscopic time-series

“…A slight variation on the training-GP approach involves modelling the ancillary time-series and the RVs simultaneously, using independent GPs sharing the same covariance function (see e.g., Osborn et al 2021;Suárez Mascareño et al 2020). In this case, both the RVs and the ancillary time-series are used to constrain the GP hyper-parameters, and the functions describing them share exactly the same covariance properties, but they are still independent of each other and have unrelated shapes.…”