The Evolutionary Status of GK Subgiants

The existence of field Population I GK stars later than the Sun and 1–4 mag above the main sequence (MS) may be a problem. Has there been enough elapsed time for them to evolve from the MS? Or are those new stars contracting to the MS? If they were contracting stars, they would be in or close to star-forming nebulosities, which they are not. Snowden & Young (2005) explored the possibility that they are above the MS because they are binary stars. Actually, a pair of similar MS stars cannot be more than 0.75 mag above the MS, which is inadequate to explain subgiants, and if one star of a pair is above the MS and redder, that only transfers the problem to explaining its existence. Nevertheless, Snowden & Young obtained coudé spectra of 30 GK subgiants and found only two binaries (HD 102928, 170928), but felt that their measures were insufficiently accurate. They asked David Latham to obtain additional radial velocities, which are presented in Section 2.

The Snowden–Young sample of GK class IV stars north of −20° decl. was selected from the Yale Catalog of Bright Stars (Hoffleit 1964). Since then, particularly with the Hipparcos satellite, we learned more about those bright stars and learned that with more recent spectral types and parallaxes, 7 of their 44 stars are class V and 7 are class III. The remaining 30 class IV stars have absolute magnitudes of M = 0.1 to 3.6 and $B-V$ = 0.63 to 1.15. When the latter are converted to T_eff using the calibration by Drilling & Landoldt (2004), the range is log T_eff = 3.67–3.77. All of these stars are within the local interstellar cavity (Abt 2011), in which the interstellar density is only a few percent of that in typical regions (Sfeir et al. 1999), so no corrections are needed for interstellar reddening. The 30 Snowden–Young class IV stars are listed in Table 1 with columns (1) HD numbers, (2) spectral types preferred by SIMBAD, (3) the absolute magnitudes M derived from Hipparcos parallaxes, (4) log T_eff, and (5) log L/L_⊙, where ${L}_{\odot }$ is the luminosity of the Sun.

Only one star (HD 139641) has [Fe/H] < −0.30 from SIMBAD and is therefore is not of Population I.

In addition, Snowden & Young observed 14 additional stars as controls. They were GK III standards of radial velocities and three known spectroscopic binaries: HD 102928 (Massarotti et al. 2008), HD 150680 (Katoh et al. 2013), and HD 170829 (Katoh et al. 2013). Table 2 lists those 14 + 3 stars plus Snowden & Young stars that are not class IV. Only 2 of those 30 stars (HD 3546, 135722) have values of [Fe/H] from SIMBAD that are not characteristic of Population I stars.

A parallel problem concerned the Mt. Wilson late-type subgiants. They were calibrated by spectral line ratios, rather than Morgan-Keenan (MK) types. There were doubts about that calibration because of its failure to fit supergiants. However, Sandage et al. (2016) showed that the luminosities of the 90 G0-K3 Mt. Wilson subgiants agree with their Hipparcos parallaxes, so those are also real subgiants.

Most of the new radial velocities of these bright stars were obtained with the 1.53 m Harvard Tillinghast telescope at the Oak Ridge Station. The original digital spectrograph, which gave velocity accuracies of σ ≈ ±0.3 km s⁻¹, was replaced in 2011 with an echelle spectrograph (Tillinghast Reflector Echelle Spectrograph) giving σ = ±0.10 km s⁻¹. The improvement is shown very well in the velocity curve of HD 62345 (Figure 3). The individual velocities can be found at https://www.cfa.harvard.edu/~latham/snowden/snowden.orb. The measures are also listed in Table 3, which is available online. For each star the columns give the Julian date, radial velocity in km s⁻¹, the telescope aperature in centimeters, and the estimated error in km s⁻¹.

For the Snowden–Young program stars, the last column of Table 1 gives the mean radial velocities measured by Latham, the standard error of each, and the number of measures. The orbital elements for the spectroscopic binary HD 8375, based on new additional measures, are given in Table 4. For HD 158614 the elements were determined by Pourbaix (2000). Only 3 of the 30 Snowden–Young stars turn out to be spectroscopic binaries with derived orbital elements, so it does not look as though the location of the GK subgiants in an Hertzsprung–Russell (HR) diagram can be explained by their being binaries.

Table 2 gives the mean velocities for the 14 constant-velocity giants, the class V and class III stars in the Snowden–Young sample, and three known binaries (HD 102928, 150680, and 170829). The columns give (1) the HD number, (2) the MK type preferred by SIMBAD, (3) the absolute magnitude M derived from the Hipparcos parallaxes, (4) the mean radial velocities measured by Latham, standard errors, and the numbers of measures, and (5) information about spectroscopic binaries (SBs) with derived orbital elements (spectroscopic binaries with orbital elements; SBO).

Table 4 gives the orbital elements for the seven new SBOs discovered among the stars in Table 2 plus the Latham orbital elements for HD 8375, the previously known SBO. The radial velocity curves are shown in Figures 1–8.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

It is ironic that the 30 class IV stars, which Snowden & Young thought might be mostly binaries, has only 3 discovered binaries, but the 26 stars in Table 2 that include 14 supposed constant-velocity stars have 7 SBs.

Are the subgiants off from the evolutionary tracks? Nearly all of the candidate stars in Table 1 have approximately solar metallicities, so they are mostly Population I. Columns 4 and 5 of Table 1 show that they occur in 3.66 < Log T_eff < 3.77 and 0.0 < log L/L_⊙ < 1.5. We can compare that region with the evolutionary tracks of Girardi et al. (2000), which were computed for stars with masses of 0.15–7 M_⊙ and Z = 0.0004–0.03. That is shown in Figure 9 with the positions of the 30 subgiants superimposed on the tracks. Most of the stars fall on the evolutionary tracks. They originate on the MS from 0.9 M_⊙ for HD 190360 to more than 1.9 M_⊙, and ages of about 10¹⁰ yr. Therefore there seems to be nothing unusual about the existence of G0-K1.5 IV stars.

Zoom In Zoom Out Reset image size

This research has made use of the SIMBAD database operated at CDS Strasbourg, France. All of the new measures were provided by David W. Latham, but he was too busy with the Transiting Exoplanet Survey Satellite (TESS) plans to take part in the writing of this paper.