In recent years, it has been discovered that massive stars commonly exhibit a non-coherent form of variability in their light curves referred to as stochastic low frequency (SLF) variability. Various physical mechanisms can produce SLF variability in such stars, including stochastic gravity waves excited at the interface of convective and radiative regions, dynamic turbulence generated in the near-surface layers, and clumpy winds. Gravity waves in particular are a promising candidate for explaining SLF variability as they can be ubiquitously generated in main sequence stars owing to the presence of a convective core, and because they provide the large-scale predominantly tangential velocity field required to explain macroturbulence in spectral line fitting. Here, I provide an overview of the methods and results of studying SLF variability in massive stars from time series photometry and spectroscopy.

We used interferometric observations made with the CHARA Array of 25 B-type stars and 6 O-type stars to obtain precise measurements of angular size, radius, and effective temperature to test stellar atmospheric models for massive stars. Our measured angular diameters range from 1.09 milli-arcseconds (mas) for β Tau down to 0.11 mas for 10 Lac, the smallest star yet resolved with the CHARA Array. The rotational oblateness of the rapidly rotating star ζ Oph is directly measured for the first time. We collected ultraviolet to infrared spectrophotometry for all sample stars and derived temperatures, angular diameters, and reddening estimates that best fit the spectra. There is generally good agreement between the observed and spectral fit angular diameters for the O and B stars, indicating that the fluxes predicted from model atmospheres are reliable. The derived and model temperatures for the O stars are also in fair agreement, however the sample size is small and several of the O stars results we consider to be preliminary. On the other hand, the temperatures derived from angular diameters and fluxes tend to be larger (by ≈ 4%) for the B stars than those from published results based on analysis of the line spectrum (Gordon et al. 2018, 2019).

In this contribution we present the results from a1 homogeneous quantitative spectroscopic analysis of ∼400 Galactic O-type stars targeted by the IACOB and OWN surveys. The ultimate objective is to perform a modern reassessment of one of the long-standing problems in the field of massive stars: the elusive detection of mid O-type stars close to the “canonical” theoretical ZAMS. We first provide statistically significant evidence of the existence of a clear lack of stars in our sample populating the region of the spectroscopic HR diagram approximately delimited by the theoretical ZAMS, the ∼ 40 and ∼ 70 M⊙ single evolutionary tracks and the 2 Myr isochrone. We then evaluate if this empirical result could be a result of possible limitations of our analysis strategy and/or the existence of potential observational biases affecting the compiled sample. Once both explanations are investigated, we evaluate the possibility that a modification of the efficiency of mass accretion during the star formation process could lead to a new (corrected) theoretical ZAMS in better accordance with our empirical results.

We present in-progress resolution test and parameter space studies for very massive stars using MESA, showcasing current MESA version convergence studies.

This contribution presents new results on two members of the class of post-Red Supergiants, IRAS 17163-3907, the central star of the Fried Egg nebula and IRC +10420. New optical spectra in the blue spectral range confirm their spectral type to be of A-supergiant class. Our VLTI/GRAVITY K-band interferometry reveals that the neutral Na i 2.2 μm line emitting region is smaller than that of the hydrogen Brγ emission. This can be explained with the hydrogen emission the result of collisional excitation populating the higher levels in a neutral region instead them being populated through recombination in an ionised environment as mostly inferred in stellar winds. Finally, the central star of the Fried Egg nebula, has undergone 3 distinct mass loss episodes over the last hundreds of years. As it is likely that at least the last mass loss event occurred when the star was already a Yellow Hypergiant and not a Red Supergiant, we put forward the bi-stability mechanism as explanation for the mass loss.

Despite the important role mass loss in the red supergiant phase plays in controlling stellar evolution and massive stars’ final supernova fates, a theoretical explanation of the mechanism driving this mass loss has been elusive. In this contribution we present a recent breakthrough [Kee et∼al., 2021] showing that turbulent pressure alone is sufficient to markedly extend the atmospheres of red supergiants and allow a wind to be launched. The resulting theory provides a fully analytic prescription for red supergiant mass-loss rates. Moreover, the theoretical mass-loss rates computed from observationally inferred turbulent velocities are in overall good agreement with observationally inferred red supergiant mass loss. A particularly interesting aspect of this theory is that it is not sensitive to metallicity, providing important implications for stellar evolution and the so-called “red-supergiant problem” for supernova progenitors in various environments.

Current models of stellar evolution predict that stars more massive than ∼6 M⊙ should have completely depleted all lithium (Li) in their atmospheres by the time when they reach the He core burning phase. Against this, a non-negligible number of red giants with masses ≳6 M⊙ presenting strong Li lines have recently been reported. Motivated by this finding, we have carried out a spectroscopic survey of red supergiants (RSGs) in the Perseus Arm and a selection of young open clusters in the Magellanic Clouds to assess the presence of the Li <sc>i</sc> 6708Å doublet line. Based on a sample of> 70 objects, close to one third of RSGs in the Perseus Arm display noticeable Li lines, with perhaps a trend towards a lower fraction among more luminous stars. The samples in the Magellanic Clouds are not as large, but hint at a metallicity dependence. Twenty one RSGs in 5 LMC clusters show a very high fraction of Li detection, around 40%. Conversely, 17 RSGs in 5 SMC clusters give only one secure detection. The interpretation of these observational results is not straightforward, but a mechanism for Li production seems most likely. Further characterisation work is ongoing, while theoretical studies into this matter are urgently needed.

The temperature independent part of the Humphreys-Davidson (HD) limit sets the boundary for evolutionary channels of massive stars that either end their lives as red supergiants (RSGs) or as the hotter blue supergiants (BSGs) and Wolf-Rayet stars. Recent downward revision of most luminous RSGs the Galaxy below log(L / L⊙) ≈ 5.5, more in line with the Magellanic Clouds, might hint towards a metallicity (Z)-independent HD limit. We present MESA single star models in the 15-40 M⊙ range and study the different Z-dependent processes that could potentially affect the location of the upper luminosity limit of RSGs.

The empirical upper limit to Red Supergiant (RSG) luminosity, known as the Humphreys-Davidson (HD) limit, has been commonly explained as being caused by the stripping of stellar envelopes by metallicity-dependent, line-driven winds. As such, the theoretical expectation is that the HD limit should be higher at lower metallicity, where weaker mass-loss rates mean that higher initial masses are required for an envelope to be stripped. In this work, we test this prediction by measuring the luminosity function of RSGs in M31 and comparing to those in the LMC and SMC. We find that $\[\log ({L_{{\rm{m}}ax}}/{L_ \odot }) = 5.53 \pm 0.03\]$ in M31 (Z ≳ Z⊙), consistent with the limit found for both the LMC (Z ∼ 0.5 Z⊙) and SMC (Z ∼ 0.25 Z⊙), while the RSG luminosity distributions in these 3 galaxies are consistent to within 1σ. We therefore find no evidence for a metallicity dependence on both the HD limit and the RSG luminosity function, and conclude that line-driven winds on the main sequence are not the cause of the HD limit.

There is evidence that some red supergiants (RSGs) experience phases of episodic mass-loss. These episodes yield more extreme mass-loss rates, further stripping the envelope of the RSG, significantly affecting the further evolution towards the final collapse of the star. Mass lost through RSG outbursts/superwinds will flow outwards and form dust further out from the surface and this dust may be detected and modelled. Here, we aim to derive the surface properties and estimate the global properties of Mid-IR bright RSGs in the Magellanic Clouds. These properties will then be compared to evolutionary predictions and used for future spectral energy distribution fitting studies to measure the mass-loss rates from present circumstellar dust.

The role of mass loss from massive stars, especially episodic mass loss, is one of the outstanding open questions facing stellar evolution theory. Multiple lines of evidence are pointing to violent, episodic mass-loss events being responsible for removing a large part of the massive stellar envelope, especially in low-metallicity galaxies. The ERC ASSESS project aims to determine whether episodic mass loss is a dominant process in the evolution of the most massive stars by conducting the first extensive, multi-wavelength survey of evolved massive stars in the nearby Universe. The project hinges on the fact that mass-losing stars form dust and are bright in the mid-infrared. We aim to investigate the properties of evolved targets in nearby galaxies and estimate the amount of ejected mass, which will constrain evolutionary models. In this work we present some of our first observational results from the galaxies NGC 6822 and IC 10 obtained with OSIRIS (GTC).

Luminous blue variables (LBVs) and B[e] supergiants (B[e]SGs) are some of the most massive stars that display extreme and puzzling behavior. Their rarity indicates that they belong to short evolutionary phases or short-lived phenomena in the post-main sequence evolution of massive stars. However, their strong mass loss and episodic mass eruptions may be crucially impacting massive star evolution. LBVs are a group of evolved massive stars that exhibit irregular variability and eruptive mass loss. Various subtypes, including S Doradus variables, giant eruptions, and pre-supernova outbursts, exist. The physical cause of the LBV phenomenon remains heavily debated. B[e]SGs have strong forbidden line emission and infrared excess from dust that are thought to arise in a circumstellar disk or torus. The formation mechanism of their disk-like structures is yet to be settled. The evolutionary phases of LBVs and B[e]SGs and their connection to other evolved massive stars are important unanswered questions in massive star evolution.

Episodic mass loss is not understood theoretically, neither accounted for in state-of-the-art models of stellar evolution, which has far-reaching consequences for many areas of astronomy. We introduce the ERC-funded ASSESS project (2018-2024), which aims to determine whether episodic mass loss is a dominant process in the evolution of the most massive stars, by conducting the first extensive, multi-wavelength survey of evolved massive stars in the nearby Universe. It hinges on the fact that mass-losing stars form dust and are bright in the mid-infrared. We aim to derive physical parameters of ∼1000 dusty, evolved massive stars in ∼25 nearby galaxies and estimate the amount of ejected mass, which will constrain evolutionary models, and quantify the duration and frequency of episodic mass loss as a function of metallicity. The approach involves applying machine-learning algorithms to select dusty, luminous targets from existing multi-band photometry of nearby galaxies. We present the first results of the project, including the machine-learning methodology for target selection and results from our spectroscopic observations so far. The emerging trend for the ubiquity of episodic mass loss, if confirmed, will be key to understanding the explosive early Universe and will have profound consequences for low-metallicity stars, reionization, and the chemical evolution of galaxies.

Mass loss is a key property to understand stellar evolution and in particular for low-metallicity environments. Our knowledge has improved dramatically over the last decades both for single and binary evolutionary models. However, episodic mass loss although definitely present observationally, is not included in the models, while its role is currently undetermined. A major hindrance is the lack of large enough samples of classified stars. We attempted to address this by applying an ensemble machine-learning approach using color indices (from IR/Spitzer and optical/Pan-STARRS photometry) as features and combining the probabilities from three different algorithms. We trained on M31 and M33 sources with known spectral classification, which we grouped into Blue/Yellow/Red/B[e] Supergiants, Luminous Blue Variables, classical Wolf-Rayet and background galaxies/AGNs. We then applied the classifier to about one million Spitzer point sources from 25 nearby galaxies, spanning a range of metallicites (). Equipped with spectral classifications we investigated the occurrence of these populations with metallicity.

The evolutionary link between Red Supergiants and Luminous Blue variables is interesting, but still poorly understood. We present the results of a study of the Galactic candidate luminous blue variable Wray 15-906, revealed via the detection of its infrared circumstellar shell (of ≍2 pc in diameter) with the Wide-field Infrared Survey Explorer (WISE) and the Herschel Space Observatory. Using the stellar atmosphere code CMFGEN and the Gaia parallax, we found that Wray 15-906 is a relatively low-luminosity, log(L/Lȯ) ≍ 5.4, star with a temperature of 25±2 kK. In the framework of single star evolution, the obtained results suggest that Wray 15-906 is a post-red supergiant star with an initial mass of ≍ 25Mȯ and that before exploding as a supernova it could transform for a short time into a WN11h star. The presence of a shell with a mass 2.9±0.5Me indicates that Wray 15-906 has suffered substantial mass loss in the recent past.

Wolf-Rayet (WR) stars comprise a class of stars whose spectra are dominated by strong, broad emission lines that are associated with copious mass loss. In the massive-star regime, roughly 90% of the known WR stars are thought to have evolved off the main sequence. Dubbed classical WR (cWR) stars, these hydrogen-depleted objects represent a crucial evolutionary phase preceding core collapse into black holes, and offer a unique window into hot-star wind physics. Their formation is thought to be rooted in either intrinsic mass-loss or binary interactions. Results obtained from analyses using contemporary model atmospheres still fail to reconcile the derived properties of WR stars with predictions from stellar evolution. Importantly, stellar evolution models cannot reproduce the the bulk of cWR stars, a problem that becomes especially severe at subsolar metallicity. Next-generation model atmospheres and upcoming observational campaigns to hunt for undetected companions promise a venue for progress.

Line-driven stellar winds are ubiquitous among hot massive stars. In some cases they can become so strong, that the whole star is cloaked by an optically thick wind. The strong outflow gives rise to large emission lines, defining the class of so-called Wolf-Rayet (WR) stars. While being major players in the evolution of massive stars, the formation of heavy black holes, and the distribution of elements, the occurrence and nature of WR winds is still quite enigmatic.

A promising instrument towards a better theoretical understanding are stellar atmospheres allowing for a consistent inclusion of the hydrodynamics. By coupling stellar and wind parameters and the inclusion of a detailed non-LTE radiative transfer, they allow us to go beneath the observable layers and study the onset of WR-type winds. Establishing larger sets of models, we were able to make ground-breaking progress by identifying trends with mass and metallicity that deviate significantly from present empirical descriptions. Our modelling efforts reveal a complex picture for WR-type winds with strong, non-linear dependencies. Besides covering metallicity and mass, we further identify surface hydrogen as an important ingredient to retain WR-type mass loss at lower metallicity. Here, we present a summary of recent insights on the nature and onset of WR-type winds in massive stars including the consequences for stellar evolution, remaining open questions, and current efforts to overcome them.

Massive stars emit X-rays. Despite modest X-ray luminosities of single hot massive stars, the ongoing large observing campaigns combining X-ray and UV spectroscopy provide a tomographic view of stellar winds. It is now established that X-ray radiation is modulated with stellar rotation and shows the same period as discrete absorption components (DACs) in the UV resonance lines. The latter are associated with corotating interaction regions (CIRs) in stellar winds, therefore the mechanisms responsible for generation of X-rays and CIRs appear to be physically linked. Binary massive stars with accreting compact companions – high-mass X-ray binaries (HMXBs) – are routinely observed by modern X-ray observatories at Mpc distances. Joint observations in X-ray and UV allow to determine realistic properties of these systems. The brightest sources among HMXBs are ultraluminous X-ray sources (ULXs). Their powerful radiation is an important source of stellar feedback. HMXBs are the products of massive binary evolution and are typically found in the vicinity of young massive star clusters. The superstar clusters blow hot superbubbles which fill large areas in star-forming dwarf galaxies. Recent models show that X-ray emission from superbubbles is likely the dominant source of He ii ionization in metal-poor star-forming dwarf galaxies. To conclude, X-ray observations provide an important window for studying massive stars and their feedback near and far.

We report on a study of 9 nearby primitive galaxies observed by Hubble’s COS far-UV spectrograph that can serve as templates of high-z galaxies to be observed by JWST. By “primitive galaxies,” we mean galaxies having a low stellar mass, and low gas metallicity, , whether they are local or at high redshift. We find that far-UV spectra of these galaxies show evidence of hard radiation, including X-rays. Following Thuan et al. (2004), we identify these galaxies as massive X-ray binaries containing a massive accreting stellar black hole. We further find that the lower the metallicity, the higher the probability of extremely strong X-radiation. Following Heger et al. (2003), we suggest that the accreting black hold is produced by direct collapse of stars having initial masses greater than 50 . The X-radiation produced by black hole disk directly affects the surrounding interstellar medium, and many of these effects are observable in far-UV spectra.

Wolf-Rayet stars are regarded as candidates for progenitors of core-collapse supernovae, and they are expected to be progenitors of long gamma-ray bursts. These types of stars are considered to be fast rotators. Their high rotation speed breaks the sphericity of the star and leads to an axisymmetric wind density structure. In such a case, the electron scattering takes place in a nonspherical environment, and as a result, we might expect an intrinsic polarization. We present a 2.5D radiation hydrodynamic stellar wind model of these stars. The model simulations account for the deformation of the stellar surface due to rotation, gravity darkening, and nonradial forces. We computed the polarization from the density variable of the hydrodynamic model, derived the upper limit of rotational velocities, and found no conflict with the previous studies of Wolf-Rayet stars.