Among Wolf-Rayet stars, those of subtype WN8 are the intrinsically most variable. We have explored the long-term photometric variability of the brightest known WN8 star, WR 40, through four contiguous months of time-resolved, single-passband optical photometry with the BRIght Target Explorer nanosatellite mission. The Fourier transform of the observed light curve reveals that the strong light variability exhibited by WR 40 is dominated by many randomly triggered, transient, low-frequency signals. We establish a model in which the whole wind consists of stochastic clumps following an outflow visibility promptly rising to peak brightness upon clump emergence from the optically thick pseudo-photosphere in the wind, followed by a gradual decay according to the right-half of a Gaussian. Free electrons in each clump scatter continuum light from the star. We explore a scenario where the clump size follows a power-law distribution, and another one with an ensemble of clumps of constant size. Both scenarios yield simulated light curves morphologically resembling the observed light curve remarkably well, indicating that one cannot uniquely constrain the details of clump size distribution with only a photometric light curve. Nevertheless, independent evidence favours a negative-index power law, as seen in many other astrophysical turbulent media.

We use a 3D Monte Carlo radiative transfer code to study the polarimetric and photometric variability from stationary corotating interaction regions (CIR) in the wind of massive stars. Our CIRs are approximated by Archimedean spirals of higher (or lower) density formed in a spherical wind originating from the star and we also made allowance for a bright Gaussian spot at the base of the CIR. Comparing results from our code to previous analytical calculations in the optically thin case, we find differences which we attribute mainly to a better estimation of the total unpolarized flux reaching the observer. In the optically thick case, the differences with the analytical calculations are much larger, as multiple scattering introduces additional complexities including occultation effects. The addition of a Gaussian spot does not alter the shape of the polarization curve significantly but does create a small excess in polarization. On the other hand, the effect can be larger on the light curve and can become dominant over the resulting CIR, depending on the spot parameters and density of the wind.

The structured winds of single massive stars can be classified into two broad groups: stochastic structure and organized structure. While the former is typically identified with clumping, the latter is typically associated with rotational modulations, particularly the paradigm of corotating interaction regions (CIRs). While CIRs have been explored extensively in the ultraviolet band, and moderately in the X-ray and optical, here we evaluate radio variability from CIR structures assuming free-free opacity in a dense wind. Our goal is to conduct a broad parameter study to assess the observational feasibility, and to this end, we adopt a phenomenological model for a CIR that threads an otherwise spherical wind. We find that under reasonable assumptions, it is possible to obtain radio variability at the 10 per cent level. The detailed structure of the folded light curve depends not only on the curvature of the CIR, the density contrast of the CIR relative to the wind, and viewing inclination, but also on wavelength. Comparing light curves at different wavelengths, we find that the amplitude can change, that there can be phase shifts in the waveform, and the entire waveform itself can change. These characterstics could be exploited to detect the presence of CIRs in dense, hot winds.

The intriguing WN4b star WR 6 has been known to display epoch-dependent spectroscopic, photometric and polarimetric variability for several decades. In this paper, we set out to verify if a simplified analytical model in which corotating interaction regions (CIRs) threading an otherwise spherical wind is able to reproduce the many broad-band continuum light curves from the literature with a reasonable set of parameters. We modified the optically thin model developed by Ignace, St-Louis & Proulx-Giraldeau to approximately account for multiple scattering and used it to fit 13 separate data sets of this star. By including two CIRs in the wind, we obtained reasonable fits for all data sets with coherent values for the inclination of the rotation axis (i0 = 166°) and for its orientation in the plane of the sky, although in the latter case we obtained two equally acceptable values (ψ = 63° and 152°) from the polarimetry. Additional line profile variation simulations using the Sobolev approximation for the line transfer allowed us to eliminate the ψ = 152° solution. With the adopted configuration (i0 = 166° and ψ = 63°), we were able to reproduce all data sets relatively well with two CIRs located near the stellar equator and always separated by ∼90° in longitude. The epoch dependence comes from the fact that these CIRs migrate along the surface of the star. Density contrasts smaller than a factor of 2 and large opening angles for the CIR (β⪆35∘) were found to best reproduce the type of spectroscopic variability reported in the literature.

With a deep Chandra/HETGS exposure of WR 6, we have resolved emission lines whose profiles show that the X-rays originate from a uniformly expanding spherical wind of high X-ray-continuum optical depth. The presence of strong helium-like forbidden lines places the source of X-ray emission at tens to hundreds of stellar radii from the photosphere. Variability was present in X-rays and simultaneous optical photometry, but neither were correlated with the known period of the system or with each other. An enhanced abundance of sodium revealed nuclear-processed material, a quantity related to the evolutionary state of the star. The characterization of the extent and nature of the hot plasma in WR 6 will help to pave the way to a more fundamental theoretical understanding of the winds and evolution of massive stars.

Oxygen sequence Wolf-Rayet stars (WO) are thought to be the final evolution phase of some high-mass stars, as such they may be the progenitors of Type Ic SNe as well as potential progenitors of broad-lined Ic and long gamma-ray bursts. We present the first spectropolarimetric observations of the Galactic WO stars WR93b and WR102 obtained with FORS1 on the Very Large Telescope. We find no sign of a line effect, which could be expected if these stars were rapid rotators. We also place constraints on the amplitude of a potentially undetected line effect. This allows us to derive upper limits on the possible intrinsic continuum polarization and find Pcont < 0.077 per cent and Pcont < 0.057 per cent for WR93b and WR102, respectively. Furthermore, we derive upper limits on the rotation of our WO stars by considering our results in the context of the wind compression effect. We estimate that for an edge-on case the rotational velocity of WR93b is vrot < 324 km s−1 while for WR102 vrot < 234 km s−1. These correspond to values of vrot/vcrit < 19 per cent and j) < 18.0 cm2 s−1 for WR93b and 2 s−1 for WR102. The upper limits found on vrot/vcrit and log(j) for our WO stars are therefore similar to the estimates calculated for Galactic Wolf-Rayet (WR) stars that do show a line effect. Therefore, although the presence of a line effect in a single WR star is indicative of fast rotation, the absence of a line effect does not rule out significant rotation, even when considering the edge-on scenario.

Malgré leur rareté et leur vie éphémère, les étoiles massives, par l’entremise de leur forte émission de rayonnements ultraviolet et de leurs vents stellaires importants, ont un fort impact sur l’enrichissement du milieu interstellaire en éléments lourds, ainsi que sur son ionisation, sa structure et sa cinématique. L’étude des étoiles massives et de leur perte de masse est essentielle pour mieux comprendre l’évolution stellaire. En particulier en fin de vie, lors de la phase Wolf-Rayet, on constate parfois la présence d’une nébuleuse sous forme de bulle autour de ces étoiles. Dans le cadre de cette étude, nous analysons la présence et les caractéristiques de ces bulles autour d’étoiles Wolf-Rayet dans la galaxie M33, avec les données de SITELLE, un spectromètre imageur installé au télescope Canada-France-Hawaii. Sur les 211 étoiles Wolf-Rayet connues dans cette galaxie, nous avons pu en observer 178 et détecter la présence de bulles autour de 43 d’entre elles grâce à des critères que nous avons déterminés à partir de nos données. Nous avons analysé chacune d’entre elles et nous présentons notre analyse et les résultats obtenus. Nous remarquons notamment l’absence de biais favorisant la présence d’une bulle autour d’un type spectral par rapport aux autres, et la présence de bulles intriquées autour de certaines étoiles WR pouvant attester de l’évolution des vents stellaires durant les différentes phases de la vie de l’étoile. / Despite their scarcity and their short lifetime, massive stars have a strong impact on the enrichment of the interstellar medium via their heavy element production, ultraviolet radiation and stellar wind emission. Hence, the study of massive stars and their mass-loss history is essential for the understanding of stellar evolution. Especially at the end of their lives, during the Wolf-Rayet stage, we often observe bubble-like nebulae surrounding massive stars. In this study, we analyze the presence and characteristics of these Wolf-Rayet bubbles in M33, with data from SITELLE, an instrument mounted on the Canada-France-Hawaii telescope. Of the 211 known Wolf-Rayet stars in M33, we were able to observe 178 of them and we detected the presence of a bubble around 43, according to the criteria we adopted based on our data. We present here the analysis and the results of our study. We find notably that the presence of bubbles doesn’t seem to be correlated with the spectral type of the progenitor star, and we observe a structure of nested bubbles around some of the WR stars, enlightening us on the evolution of stellar winds during the different phases of a massive star’s life.

To investigate if magnetic fields are present in Wolf–Rayet stars, we selected a few stars in the Galaxy and one in the Large Magellanic Cloud (LMC). We acquired low-resolution spectropolarimetric observations with the European Southern Observatory FORS 2 (FOcal Reducer low dispersion Spectrograph) instrument during two different observing runs. During the first run in visitor mode, we observed the LMC Wolf–Rayet star BAT99 7 and the stars WR 6, WR 7, WR 18, and WR 23 in our Galaxy. The second run in service mode was focused on monitoring the star WR 6. Linear polarization was recorded immediately after the observations of circular polarization. During our visitor observing run, the magnetic field for the cyclically variable star WR 6 was measured at a significance level of 3.3σ (〈Bz〉 = 258 ± 78 G). Among the other targets, the highest value for the longitudinal magnetic field, 〈Bz〉 = 327 ± 141 G, was measured in the LMC star BAT99 7. Spectropolarimetric monitoring of the star WR 6 revealed a sinusoidal nature of the 〈Bz〉 variations with the known rotation period of 3.77 d, significantly adding to the confidence in the detection. The presence of the rotation-modulated magnetic variability is also indicated in our frequency periodogram. The reported field magnitude suffers from significant systematic uncertainties at the factor of 2 level, in addition to the quoted statistical uncertainties, owing to the theoretical approach used to characterize it. Linear polarization measurements showed no line effect in the stars, apart from WR 6. BAT99 7, WR 7, and WR 23 do not show variability of the linear polarization over two nights.

Ce mémoire présente une analyse de données photométriques d’étoiles de type Wolf-Rayet ayant pour objectif de caractériser les causes de la variabilité stochastique à court terme omniprésente chez ces étoiles, ainsi que la variabilité périodique due aux éclipses atmosphériques dans deux systèmes particuliers. À l’aide de données de la mission spatiale BRITE, les éclipses initialement supposées atmosphériques des systèmes WR22 et WR48 ont été investiguées avec un modèle analytique, suivi d’une modélisation numérique de l’éclipse pour WR22. Le taux de perte de masse de l’étoile Wolf-Rayet, la profondeur optique du vent et l’angle d’inclinaison du système ont été déterminés avec précision pour WR22, tandis que des valeurs-plafonds pour ces paramètres ont été obtenues avec un angle i tiré de la littérature pour WR48. L’étude de la variabilité stochastique dans le vent de 55 étoiles Wolf-Rayet galactiques a été conduite en utilisant 70 séries temporelles obtenues avec les satellites MOST, BRITE-Constellation et TESS. L’application d’un modèle semi-Lorentzien aux périodogrammes des observations a permis d’obtenir des paramètres caractéristiques : l’amplitude typique, la fréquence caractéristique et le taux d’augmentation de l’amplitude de variabilité. Des comparaisons entre les paramètres obtenus et ceux de la littérature nous ont permis de contextualiser le rôle de divers processus physiques dans l’origine de la variabilité, tel l’instabilité du processus d’entraînement du vent, d’une zone de convection subsurfacique engendrée par l’ionisation partielle des éléments du groupe du fer, ainsi qu’une possible contribution d’ondes de gravité générées à l’interface du coeur convectif et de la zone radiative. / We present an analysis of photometric observations of Wolf-Rayet stars in an effort to characterize their ubiquitous short-term stochastic variability and the origin of the periodic variability due to atmospheric eclipses in two specific systems. With data from BRITE-Constellation, we investigated what we initially thought to be atmospheric eclipses in the WR22 and WR48 systems. Both stars’ datasets were first analyzed using an analytical atmospheric eclipse model, and then a fully numerical core and wind eclipse model was used to better characterize the ambiguous case of WR22. The mass-loss rate of the Wolf-Rayet component, the optical depth of the wind and the inclination angle of the system were accurately determined for WR22, whereas for WR48 an upper limit was found for those parameters using an inclination angle from the litterature. The study of the short-term stochastic variability in the winds of 55 Wolf-Rayet stars was obtained with 70 time-series produced by MOST, BRITE and TESS. The amplitude spectra of these stars were analyzed using a semi-Lorentzian function that allowed us to extract characteristic parameters. Comparisons between the values we obtained and those from the litterature were carried out and we discuss the possible contribution to the variability of various physical processes, namely the line de-shadowing instability, the presence of a subsurface iron-group driven convection zone and possibly of internal gravitational waves generated at the boundary between the convective core and the radiative layer to the observed variability.

Nous présentons les résultats de trois campagnes d'observation d'un mois chacune dans le cadre de l'étude de la collision des vents dans les systèmes binaires Wolf-Rayet + OB. Ce travail se concentre sur l'étude des objets de l'hémisphère sud n'ayant jamais encore fait l'objet d'études poussées dans ce contexte. À cela, nous avons ajouté l'objet archétype pour ce type de systèmes : WR 140 (WC7pd + O5.5fc) qui a effectué son dernier passage périastre en janvier 2009. Les deux premières campagnes (spectroscopiques), ont permis une mise à jour des éléments orbitaux ainsi qu'une estimation de la géométrie de la zone de collision des vents et d'autres paramètres fondamentaux des étoiles pour 6 systèmes binaires : WR 12 (WN8h), 21 (WN5o+O7V), 30 (WC6+O7.5V), 31 (WN4o+O8), 47 (WN6o+O5) et 140. Une période non-orbitale courte (probablement reliée à la rotation) a également été mesurée pour un des objets : WR 69 (WC9d+OB), avec une période orbitale bien plus grande. La troisième campagne (photométrique) a révélé une variabilité étonnamment faible dans un échantillon de 20 étoiles WC8/9. Cela supporte l'idée que les pulsations ne sont pas courantes dans ce type d'étoiles et qu'il est peu probable que celles-ci soient le mécanisme dominant de formation de poussière, suggérant, par défaut, le rôle prédominant de la collision des vents. / We present the results from three month-long observational campaigns related to WR+O colliding-wind binaries. This work focuses on southern-hemisphere systems that have not yet been previously studied in this context. To this we add the northern archetype colliding-wind system WR 140 (WC7pd + O5.5fc), which recently suffered another active periastron passage in 2009 January. The first two (spectroscopic) campaigns allowed us to update the orbital elements and to constrain the geometry of the colliding wind region and other fundamental parameters, for 6 binary systems: WR 12 (WN8h), WR 21 (WN5o+O7V), WR 30 (WC6+O7.5V), WR 31 (WN4o+O8), WR 47 (WN6o+O5) and WR 140. A short non-orbital (probably rotational) period was also detected in the WR component of one of our objects, WR 69 (WC9d+OB), with a much longer orbital period. The third (photometric) campaign revealed a surprisingly low level of variability in our sample of 20 WC8/9 stars. This supports the idea that pulsations are not very frequent in this type of star and are unlikely to be the main mechanism for dust formation, suggesting, by default, a predominant role of wind-wind collision.