We analyse a recently published array of 288 refined and new radial velocity (RV) measurements of υ And. This star is currently believed to host 3 giant planets having orbital periods about 4.6 days, 240 days и 1280 days [1]. These data cover about 15 years and have intrinsic (instrumental) precision near 5-10 m/s (starting approximately from 1997). Periodogram analysis of these data reveals two extra, previously unknown, RV variations for this star, both have about 10 m/s semi-amplitude and high statistical significance.

The first oscillation has annual character, and it can be treated as having period of either 360 days or 180 days. This ambiguity is a consequence of the aliasing effect, caused by periodic seasonic gaps in the observation sequence. Due to the dynamical stability argument, this oscillation cannot be treated as a hint of an extra planet with the corresponding orbital period. Therefore, this seasonic periodicity is caused, most probably, by systematic errors. The source of these errors can lie in some instrumental drifts or in imperfections of the algorithm of the spectrum analysis and deriving of the actual barycentric radial velocity 'measurement'. Such seasonic errors are frequently met in the exoplanet search practice [2].

The second variation has period of roughly 12 years, if we consider it is sinusoidal. Under assumption of the planetary nature of this RV variation, more long orbital periods become admissible as well (from the viewpoint of pure data), up to parabolic orbits (with pericentric distance of 3-6 AU). Taking into account the requirement of the dynamical stability of the system allows us to mark a few regions in the parameter space, where the orbit of the putative fourth planet (e) can be located. These may be moderately eccentric orbits in the 3:1, 4:1, 5:1, 6:1 mean-motion resonances with the known planet d, or highly eccentric orbits with pericentric distance of 5.5 AU (but such orbits are less probable from the viewpoint of exoplanetary statistics).

Nonetheless, given the small amplitude of the variation with respect to the errors background, other (non-planetary) interpretations of the long-term RV variation of υ And are possible. These include: instrumental systematic errors, hints of the stellar magnetic activity cycle (an 11-years Solar cycle analog). In recent years several different authors have already conducted long-term observation series, which can clarify this question in future (e.g., monitoring of the activity of this star based on the emission level in H and K Ca II lines, paper [3]). However, currently available data can neither confirm nor reject the planetary interpretation of the long-period RV oscillation of υ Andromedae. It is necessary to continue observations of this star by means of heterogeneous methods (high-precision radial velocity measurements at different observatories, magnetic activity monitoring, high-precision astrometric observation) for a few extra years.

 

1. Wright J.T., Upadhyay S., Marcy G.W., Fischer D.A., Ford E.B., Johnson J.A. 'Ten new and updated multi-planet systems, and a survey of exoplanetary systems', ApJ, 2009, 693, 1084-1099

2. Baluev R.V. 'Accounting for velocity jitter in planet search surveys', MNRAS, 2009, 393, 969-978

3. Shkolnik E., Bohlender D.A., Walker G.A.H., Cameron A.C. 'The on/off nature of star-planet interactions', ApJ, 2008, 676, 628-638