Architecture of planetary systems with and without outer giant planets I. Inner planet detections around HD 23079, HD 196067, and HD 86226

TL;DR

The paper investigates whether outer giant planets influence the occurrence of inner light planets by constructing two homogeneous RV samples and conducting long-term, high-precision observations with CORALIE, HARPS, and ESPRESSO. Using both periodogram analysis and Bayesian nested sampling (kima), the authors detect three ILPs (two new detections and one confirmation) in systems hosting outer giants, with all ILP signals achieving high significance ($\text{FAP}<10^{-7}$ and $\text{TIP}>99\%$). Joint RV and HGCA astrometry for at least one system yields refined outer-planet masses and inclinations, illustrating the system architectures. While the current small-number results do not decisively confirm or refute a correlation between ILP and OGP populations, the study establishes a rigorous, homogeneous framework and a path toward a robust statistical assessment in forthcoming articles.

Abstract

Understanding the link between outer giant planets (OGPs) and inner light planets (ILPs) is key to understanding planetary system formation and architecture. The correlation between these two populations of planets is debated both theoretically -- different formation models predict either a correlation or an anticorrelation -- and observationally. Several recent attempts to constrain this correlation have yielded contradictory results, due to small-number statistics and heterogeneous samples. We present an ongoing long-term observational effort with CORALIE, HARPS, and ESPRESSO to probe the ILP occurrence in systems with and without OGP. In this first article of a series, we discuss how, from the design to the observations, we ensured the homogeneity of the samples, both in terms of stellar properties and observing strategy. We also present the first three detections of ILPs in our OGP host sample. We find a 8.3 mE planet at 5.75 d around HD 23079, a 10.4 mE planet at 4.6 d around HD 196067, and we confirm the 7.5 mE planet at 3.98 d around HD 86226. While a rigorous statistical analysis of our samples will be performed in subsequent studies, the relatively low number of detections in our sample seems to contradict previous studies that found a strong OGP-ILP correlation.

Figures (11)

• Figure 1: Masses (top panel) and metallicities (bottom panel) of the stars in our Giant and Flat samples (see Sect. \ref{['sec:sample']}).
• Figure 2: Periodograms of the raw RV time series (top panel) as well as of the residuals after subtracting the OGP (middle panel) and the ILP (bottom panel), for HD 23079 (left), HD 196067 (center), and HD 86226 (right).
• Figure 3: Posterior distributions for the orbital periods from the analysis of the three targets (left: HD 23079, center: HD 196067, right: HD 86226). The plots show the true inclusion probability (TIP), highlighting the peaks corresponding to the detected inner planets. Because the Keplerian signal of the outer planet is always included in the model, the TIP value for the outer planet period is, by definition, unity.
• Figure 4: Time series (top panel) and residuals (bottom panel) of the maximum-likelihood fit for each target (HD 23079, HD 196067, and HD 86226).
• Figure 5: Top panel: Residuals of the maximum-likelihood fit phase-folded at the periods of the two detected planets, showing the outer giant (left) and lower-mass inner planet (right) for each target (HD 23079, HD 196067, and HD 86226). Bottom panel: Same residuals, but showing only the HARPS data.