A Search for Exoplanets around Northern Circumpolar Stars X. The origin of radial velocity variations in the evolved star HD 216595

Technical Summary: The Origin of Radial Velocity Variations in the Evolved Star HD 216595

Problem Statement
Detecting exoplanets around evolved stars, particularly those on the Asymptotic Giant Branch (AGB), is hindered by intrinsic stellar variability. Processes such as surface convection, pulsations, and mass loss generate radial velocity (RV) signals that can mimic Keplerian motion caused by substellar companions. While over 8,000 exoplanets have been discovered, confirmed planets around AGB stars remain elusive, with most reported companions in such systems falling into the substellar mass regime. This study investigates the origin of long-period, low-amplitude RV variations observed in the AGB star HD 216595 to determine whether they arise from a genuine substellar companion or intrinsic stellar processes.

Methodology
The authors analyzed high-resolution spectroscopic data spanning approximately 16 years (2010–2025) obtained from two instruments:

1. BOAO/BOES: The Bohyunsan Optical Astronomy Observatory Echelle Spectrograph (R = 90,000) using an iodine cell for precise RV measurements.
2. LCO/NRES: The Las Cumbres Observatory Network of Robotic Echelle Spectrographs (R = 53,000) from McDonald Observatory (USA) and WISE Observatory (Israel).

Stellar parameters were derived using the isoclassify package, integrating Gaia DR2/DR3 data, 2MASS photometry, and MESA Isochrones & Stellar Tracks (MIST). To isolate the source of RV variations, the authors employed:

• Generalized Lomb-Scargle (GLS) periodograms to identify periodic signals in the RV time series.
• Activity Diagnostics: Analysis of line bisectors (BVS, BVC), chromospheric indicators (H$\alpha$, H$\beta$, Na D1/D2, Ca II H&K), and TESS photometric light curves to correlate RV variations with stellar activity.
• Orbital Modeling: Fitting Keplerian models to the RV data and performing N-body numerical integrations (using the Exo-Striker toolbox) to test the long-term dynamical stability of multi-companion scenarios.

Key Results

• Stellar Properties: HD 216595 is confirmed as an AGB star with a mass of $\sim 4.0 M_\odot$, a radius of $\sim 112 R_\odot$, and a luminosity of $\sim 3236 L_\odot$.
• RV Periodicity: The RV data reveal a statistically significant periodic signal at 567 days. A secondary signal at 1623 days was detected after pre-whitening the primary signal.
• Activity Correlation: No strong correlations were found between the 567-day RV signal and activity indicators. While weak signals near 567 days appeared in H$\alpha$ and H$\beta$ equivalent widths, they were only marginally significant (False Alarm Probability $\sim 5\%$). Bisector indicators showed no significant periodicity at the RV period.
• Companion Hypothesis: The 567-day signal fits a Keplerian model consistent with a substellar companion with a minimum mass of $39.3^{+5.2}_{-5.0} M_J$ and a semi-major axis of $2.14^{+0.14}_{-0.14}$ AU. This is broadly consistent with previous astrometric constraints.
• Stability Analysis: Numerical simulations of a two-companion system (including the 1623-day signal) demonstrated that such a configuration is dynamically unstable, leading to orbital disruption within $\sim 300$ years. This suggests the 1623-day signal is likely not a genuine Keplerian companion.
• Pulsation vs. Companion: The observed RV period (567 days) is significantly longer than the fundamental radial pulsation periods predicted by Period-Mass-Radius (PMR) and Period-Luminosity Relations (PLR) for this star ($\sim 18–57$ days). However, it falls within the range of Long Secondary Periods (LSPs) observed in AGB stars, though the RV amplitude is smaller than typical LSP variations.

Significance and Conclusions
The paper concludes that the origin of the RV variations in HD 216595 remains ambiguous. While the 567-day signal is statistically significant and can be described by a Keplerian model indicative of a $\sim 39 M_J$ companion, the authors cannot definitively rule out intrinsic stellar processes.

The study highlights the "intrinsic degeneracy" in evolved stars, where convection, granulation, and pulsations can produce RV signals with amplitudes and timescales comparable to orbital motions. The authors identify three viable interpretations:

1. Intrinsic variability driven by pulsations and convection.
2. A companion-induced Keplerian signal.
3. Chromospheric activity contributing to the variability.

Given the limitations of current diagnostics for AGB stars, the paper asserts that the observed signal lies in a regime where all three mechanisms are plausible. The authors emphasize that distinguishing genuine orbital signals from stellar jitter in such evolved systems requires improved diagnostics, more sophisticated modeling of stellar variability, and coordinated high-precision observations. The study does not claim a definitive detection of a planet but rather underscores the complexity of interpreting RV data around AGB stars.