Long‐Term Spectroscopic and Precise Radial Velocity Monitoring of Arcturus

This dissertation presents an analysis of data obtained from long‐term spectroscopic monitoring of the bright, nearby K giant star Arcturus (α Bootis, K1.5 III, HR 5340, HD 124897), conducted using the University of Western Ontario's Elginfield Observatory. Over 1300 high‐resolution spectra (λ/Δλ ≈ 10⁵) were recorded in the 6250 Å spectral region during the 1984–2006 observing seasons, with 311 simultaneous, precise radial velocity measurements (σ ∼ 25 m s⁻¹) obtained during the 2002–2006 observing seasons. Temperature variations were measured using the well‐established line‐depth ratio V i λ6251.83 to Fe i λ6252.56. Velocity spans were obtained from the line bisector of the Fe i λ6252.56 line. Precise radial velocities were measured using the telluric absorption technique outlined by Gray & Brown (2006, PASP, 118, 399).

Spectroscopic observations and precise radial velocity measurements reveal variability in Arcturus from hours to years. From velocity span observations, three distinct periods of variability are observed: (1) a semiregular 5 ± 1 yr period interpreted as evidence for a long‐term stellar magnetic cycle, (2) a 2.0 ± 0.2 yr modulation identified as the rotation period (see also Gray & Brown 2006, PASP, 118, 1112), and (3) a semicoherent 200 ± 5 day period interpreted as one‐fourth of the rotation period, indicating the presence of four magnetically active longitudes on the surface of Arcturus. One of these four active longitudes is postulated to dominate the other three, inducing the observed 2 yr rotational modulation in the velocity span measurements. Line‐depth ratio measurements indicate no evidence for variations in temperature in excess of 10 K over the observed magnetic cycle, rotation period, and short‐term variability of Arcturus.

Precise radial velocity measurements exhibit a long‐term period of 222 ± 3 days, with an amplitude of 150 ± 20 m s⁻¹. This long‐term period and the corresponding amplitude are consistent with the radial velocity observations of Hatzes & Cochrane (1993, ApJ, 413, 339). The 222 day period is coherent over the 14 yr time span between the Hatzes & Cochrane (1993, ApJ, 413, 339) data and the observations of this work, and does not correlate with observed line bisector or line‐depth ratio variations, suggesting that this observed period is not associated with stellar variability. Hatzes & Cochrane (1993, ApJ, 413, 339) suggested that the observed long‐term radial velocity period was due either to rotational modulation or a low‐mass companion, but were unable to differentiate between the two explanations, due to a lack of ancillary spectroscopic data. A detection of a low‐mass companion is postulated by this work. Assuming a mass of 1 M_⊙ for Arcturus, a best‐fit Keplerian orbit to the 222 day period is found to have approximately zero eccentricity, a projected semimajor axis (asin i) of 0.72 AU, and a companion mass (M₂ sin i) of ≥4.4 M_J.

Precise radial velocity observations also demonstrate a scatter (σ) of ∼145 m s⁻¹ superimposed on the observed 222 day period, resulting from short‐term variability on the timescale of individual nights to days. This short‐term variability has been interpreted as being the result of solar‐type p‐mode oscillations by previous studies, such as Hatzes & Cochrane (1994, ApJ, 422, 366), Merline (1995, Ph.D. thesis, Univ. Arizona), and Retter et al. (2003, ApJ, 591, L151). Observed scatter during individual nights ranges from ∼50 to 180 m s⁻¹, while the scatter over consecutive nights ranges from ∼70 to 160 m s⁻¹. These observations are consistent with previous studies of the short‐term radial velocity variability of Arcturus. Ultimately, this scatter of ∼145 m s⁻¹ limits the accuracy of studies of the long‐term radial velocity variability of Arcturus. Evidence for eruptive events is also suggested from the observed radial velocities.