Wind properties of Milky Way and SMC massive stars: empirical Z dependence from CMFGEN models

TL;DR

This study empirically investigates how stellar wind properties of massive stars in the Milky Way and SMC depend on metallicity, revealing a luminosity and metallicity dependence consistent with radiatively driven wind theory, with implications for stellar evolution.

Contribution

It provides the first comprehensive empirical analysis of wind properties across different metallicities using a large sample analyzed with CMFGEN, extending previous results and highlighting a luminosity and Z dependence.

Findings

Wind-luminosity relation shows Z dependence for bright stars.

Mass-loss rate scales with Z^{0.5-0.8} at high luminosities.

Terminal velocities have a shallow Z dependence, n ~ 0.1-0.2.

Abstract

Detailed knowledge about stellar winds and evolution at different metallicities is crucial for understanding stellar populations and feedback in the Local Group of galaxies and beyond. Despite efforts in the literature, we still lack a comprehensive, empirical view of the dependence of wind properties on metallicity ($Z$). Here, we investigate the winds of O and B stars in the Milky Way (MW) and Small Magellanic Cloud (SMC). We gathered a sample of 96 stars analyzed by means of the NLTE code CMFGEN. We explored their wind strengths and terminal velocities to address the $Z$ dependence, over a large luminosity range. The empirical wind-luminosity relation (WLR) obtained updates and extends previous results in the literature. It reveals a luminosity and $Z$ dependence, in agreement with the radiatively driven wind theory. For bright objects ($\log L/L_\odot \gtrsim 5.4$), we infer that $\dot{M} \sim Z^{0.5-0.8}$. However, this dependence seems to get weaker or vanish at lower luminosities. The analysis of the terminal velocities suggests a shallow $Z^n$ dependence, with $n \sim 0.1-0.2$, but it should be confirmed with a larger sample and more accurate $V_{\infty}$ determinations. Recent results on SMC stars based on the PoWR code support our inferred WLR. On the other hand, recent bow-shocks measurements stand mostly above our derived WLR. Theoretical calculations of the WLR are not precise, specially at low $L$, where the results scatter. Deviations between our results and recent predictions are identified to be due to the weak wind problem and the extreme terminal velocities predicted by the models. The Z dependence suggested by our analysis deserves further investigations, given its astrophysical implications.