The IACOB project XV. Updated calibrations of fundamental parameters of Galactic O-type stars

TL;DR

The paper provides updated, empirically derived calibrations for fundamental parameters of Galactic O-type stars by integrating homogeneous high-resolution spectroscopy with Gaia DR3 distances for 234 robust calibrators out of 358 surveyed stars. It delivers spectral-type and luminosity-class based relations for $T_{ m eff}$, $\log g_{\rm true}$, $M_V$, $R$, $\log L$, and $M_{sp}$, and introduces empirical bolometric corrections and FW3414-based diagnostics for $M_V$ when spectral classification is incomplete. Compared with Martins 2005, the study finds systematic shifts in $T_{ m eff}$ for dwarfs and in $M_V$ across all classes, while radii and luminosities are largely consistent for LC V and show class-dependent adjustments for LC I/III; the scatter in $M_V$ remains substantial even with precise distances, highlighting the impact of binarity and classification uncertainties. The FW3414 parameter emerges as a practical, fast diagnostic for large surveys lacking full spectral classifications, enabling reliable $M_V$ estimates and enabling propagation to radii and luminosities. These calibrations provide a robust reference for Galactic and extragalactic massive-star population studies in the era of large spectroscopic surveys, and the authors make the data publicly accessible to support community use.

Abstract

Modern spectroscopic surveys combined with Gaia distances are enabling reliable estimates of fundamental parameters for hundreds of Galactic O-type stars and the full range of spectral types and luminosity classes. Here we provide updated, statistically robust empirical calibrations of the fundamental parameters of Galactic O-type stars, as well as of their absolute visual magnitudes (Mv) and bolometric corrections (BC), based on high-quality observational data. We perform a homogeneous analysis of a sample of 358 Galactic O-type stars, combining high-resolution spectroscopy and Gaia distances. A subset of 234 stars meeting strict quality criteria involving parallax, extinction, and multi-band photometry was used to derive empirical calibrations of fundamental parameters. For those same stars, calibrated parameters were estimated from their measured Mv using the derived relations, allowing us to assess the internal consistency and predictive power of the calibrations. We present updated spectral-type-based calibrations of fundamental parameters for luminosity classes V, III, and I. Compared to previous works, we find systematic shifts, particularly in effective temperature for dwarfs and in Mv across all classes, which propagate into derived quantities. Applying the Mv calibrations to the full sample yields consistent estimates of radius and luminosity, while spectroscopic mass (Msp) shows significant scatter. We also evaluate the FW3414 parameter (from the Hbeta line) as a calibrator for Mv, useful in large surveys lacking reliable spectral classification. Excluding SB1 systems has a noticeable impact only on the Msp calibration for LC V. These updated empirical calibrations offer a robust reference for Galactic O-type stars and will support studies of massive star populations in both Galactic and extragalactic contexts, particularly in the era of large spectroscopic surveys.

Figures (16)

• Figure 1: Cumulative histograms of the nominal relative error in parallax of increasing large samples of Galactic O-type stars with available parallaxes following the Tycho-2 Hog2000 astrometric catalogs, and the various Gaia data releases, including the Gaia-TGAS solution Michalik2015. For each catalog we indicate the total number of known stars of this type with available entries, along with those in which a relative error in parallax $<$10% was achieved (see also vertical line).
• Figure 2: Distribution of the various samples described in Sect. \ref{['section2']} as a function of $G_{\rm corr}$, the Gaia G magnitude MaizApellaniz2018b. Vertical dashed lines indicate the 99% and 85% completeness limits of the full IACOB sample, using GOSC v4.1 as a reference. Stars dimmer than $G_{\rm corr}$ = 10.5 mag are highlighted with a dash pattern, providing a reference for Figs. \ref{['Hist_SpT_Gaia']} and \ref{['Hist_d_Gaia']}.
• Figure 3: Similar to Fig. \ref{['Hist_G_Gaia']}, but showing the distribution of SpTs. Percentages indicate the completeness of stars marked in pink and pink+black with respect to GOSC v.4.1, excluding stars dimmer than $G_{\rm corr}$ = 10.5 mag (marked with dash pattern).
• Figure 4: Similar to Fig .\ref{['Hist_G_Gaia']}, but showing the distribution of distances (from ALS, see Sect. \ref{['section32']}). A vertical dashed line indicates the 3000 pc limit used as a quality cut (see Appendix \ref{['AppCuts']} for explanation). The large concentration of stars missing in the IACOB spectroscopic database at $\sim$1600 pc corresponds to relatively highly extinguished O-type stars in the Cygnus OB2 region Berlanas2020.
• Figure 5: SpT -- $T_{\rm eff}$ (middle panels) and SpT -- log $g_{\rm true}$ (bottom panels) calibrations and the 358 O-type stars within the working sample, i.e., with available estimations of $T_{\rm eff}$ and log $g$. 234 stars comprising the calibrator sample (solid points) and 124 eliminated (semitransparent) due to quality-cut criteria (Appendix \ref{['AppCuts']}). Columns divided by LC groups. Dashed lines represent the "observational scales" proposed by M05, while solid lines correspond to the calibrations derived in this work (see definitions in Table \ref{['Cab1']}). Stars identified as SB1 are highlighted with open symbols. The top panels display the number of stars per SpT bin, where number in brackets (and hatched bars) indicates stars excluded from the calibrator sample (see Sect. \ref{['section2']}).