On the physical nature of the Wilson-Bappu effect: revising the gravity and temperature dependence

Abstract

We present a sample of 32 stars of spectral types G and K, and luminosity classes I to V, with moderate activity levels, covering four orders of magnitude of surface gravity and a representative range of effective temperature. For each star we obtained high S/N TIGRE-HEROS spectra with a spectral resolving power of $R\approx20,000$ and have measured the Ca II K line-widths of interest, $W_0$ and $W_1$. The main physical parameters are determined by means of iSpec synthesis and Gaia EDR3 parallaxes. Mass estimates are based on matching to evolution models. Using this stellar sample, that is highly uniform in terms of spectral quality and assessment, we derive the best-fit relation between emission line width and gravity $g$, including a notable dependence on effective temperature $T_{\rm eff}$, of the form $W_1 \propto g^{-0.229} T_{\rm eff}^{+2.41}$. This result confirms the physical interpretation of the Wilson-Bappu effect as a line saturation and photon redistribution effect in the chromospheric Ca II column density, under the assumption of hydrostatic equilibrium at the bottom of the chromosphere. While the column density (and so $W_1$) increases towards lower gravities, the observed temperature dependence is then understood as a simple ionization effect -- in cooler stars, Ca II densities decrease in favor of Ca I.