Bayesian approach to cyclic activity of CF Oct

TL;DR

This study applies Bayesian statistical methods to a 45-year dataset of CF Oct to analyze its activity cycles, confirming known rotational periods and revealing long-term cycles indicative of stellar surface differential rotation.

Contribution

It introduces Bayesian approaches to study stellar activity cycles, providing new evidence for long-term cycles and differential rotation in CF Oct.

Findings

Confirmed rotational period of 20.16 days with variations from 19.90 to 20.45 days.

Identified a long-term cycle of approximately 7.1 to 9.8 years.

Detected evidence of stellar surface differential rotation.

Abstract

Bayesian statistical methods of Gregory-Loredo and the Bretthorst generalization of the Lomb-Scargle periodogram have been applied for studying activity cycles of the early K-type subgiant star CF Oct. We have used a ~45 year long dataset derived from archival photographic observations, published photoelectric photometry, Hipparcos data series and All Sky Automated Survey archive. We have confirmed the already known rotational period for the star of 20.16 d and have shown evidences that it has exhibited changes from 19.90 d to 20.45 d. This is an indication for stellar surface differential rotation.The Bayesian magnitude and time--residual analysis reveals clearly at least one long-term cycle. The cycle lenght's posterior distributions appear to be multimodal with a pronounced peak at a period of 7.1 y with FWHM of 54 d for time-residuals and at a period of 9.8 y with FWHM of 184 d for magitude data. These results are consistent with the previously postulated cycle of 9+/-3 years.