Probing Polaris’ puzzling radial velocity signals

We investigate temporally changing variability amplitudes and the multi-periodicity of the type-I Cepheid Polaris using 161 high-precision radial velocity (RV) and bisector inverse span (BIS) measurements based on optical spectra recorded using Hermes at the 1.2 m Flemish Mercator telescope on La Palma, Canary Islands, Spain. Using an empirical template fitting method, we show that Polaris’ RV amplitude has been stable to within ∼30 m s−1 between September 2011 and November 2018. We apply the template fitting method to publicly accessible, homogeneous RV data sets from the literature and provide an updated solution of Polaris’ eccentric 29.3 yr orbit. While the inferred pulsation-induced RV amplitudes differ among individual data sets, we find no evidence for time-variable RV amplitudes in any of the separately considered, homogeneous data sets. Additionally, we find that increasing photometric amplitudes determined using SMEI photometry are likely spurious detections due to as yet ill-understood systematic effects of instrumental origin. Given this confusing situation, further analysis of high-quality homogeneous data sets with well-understood systematics is required to confidently establish whether Polaris’ variability amplitude is subject to change over time. We confirm periodic bisector variability periods of 3.97 d and 40.22 d using Hermes BIS measurements and identify a third signal at a period of 60.17 d. Although the 60.17 d signal dominates the BIS periodogram, we caution that this signal may not be independent of the 40.22 d signal. Finally, we show that the 40.22 d signal cannot be explained by stellar rotation. Further long-term, high-quality spectroscopic monitoring is required to unravel the complete set of Polaris’ periodic signals, which has the potential to provide unprecedented insights into the evolution of Cepheid variables