Photometric variability of nitrogen-rich Wolf-Rayet stars in Magellanic Clouds with OGLE

TL;DR

This study analyzes the long-term photometric variability of 47 nitrogen-rich Wolf-Rayet stars in the Magellanic Clouds using OGLE data, with cross-checks from MACHO and TESS. The authors employ generalized Lomb-Scargle periodograms and a 1-year detrending window to separate long- and short-term variability, identifying significant periodicities with periods from $2$ to $56$ days in $11$ stars, and revealing quasi-periodic outbursts in $9$ targets. They classify variability into low, moderate, and high categories, find six long-period variables, and uncover evidence for binary candidacy in several objects through cross-survey coherence and Gaia indicators. The work highlights the ubiquity and diversity of WR variability, linking it to CIRs, pulsations, wind clumping, and binarity, and underscores the need for coordinated spectroscopic follow-up and extended monitoring to fully characterize the physical processes at play.

Abstract

We present a comprehensive analysis of the photometric variability of (presumably single) nitrogen-rich Wolf-Rayet (WN) stars in the Magellanic Clouds, using long-term observations from the OGLE survey. Our sample comprises 47 stars with no nearby Gaia counterparts. We characterize both overall and short-term variabilities, examining data dispersion and identifying periodicities. To validate our findings, we also compare the OGLE light curves with data from the MACHO and TESS missions. Variability is ubiquitous in our WR sample: about one third of stars display high variability, or four fifths if we include cases with moderate variations. The observed changes are found to be periodic in 11 cases, with timescales of 2-56 d. Such periodic variations originate in corotating wind structures, binary effects, or pulsations, thereby increasing the number of systems known to show these phenomena. Surprisingly, nine targets display (quasi-periodic) outbursts, making such changes a new type of WR variability. The variability shows a transient character, in about 30$\%$ of the sample, with changing amplitudes for periodic signals or for outbursts (they even sometimes completely disappear from view). Finally, we identified six long-period variables, four of which have been confirmed by at least two independent surveys.

Figures (11)

• Figure 1: First and third rows: GLS periodograms zoomed-in to the frequencies lower than 0.1 d$^{-1}$. Second and fourth rows: Binned, phase-folded light curves for the stars BAT99-41, BAT99-47, BAT99-54, BAT99-62, BAT99-67, and BAT99-94. The period and reference epoch corresponding to the most prominent peak are noted in the title of each bin-folded light curve (see also Table \ref{['tab:ephemeris']}).
• Figure 2: Left column: OGLE (black circles) and MACHO (red and blue circles) light curves showing long-term variability trends. Right column: Binned light curves folded using the OGLE ephemerides from Table \ref{['tab:ephemeris']}. The OGLE light curve is represented in the JD-2400000 instead of HJD-2450000.
• Figure 3: Binned OGLE and TESS light curves, folded using ephemeris from OGLE (as well as the ones from TESS for BAT99-26). The number of TESS sectors used for BAT99-3 is 22; for BAT99-26 is 13; 1 sector for BAT99-31, BAT99-47, BAT99-56, and BAT99-65; and for SMC-AB4 the number of used sectors is 5.
• Figure 4: Variability level as a function of effective temperature and spectral type for the data before (left) and after (middle panel) detrending. The limits between low, moderate, and high variability levels (Sect. \ref{['subsec:variabilityCriteria']}) are indicated with dashed red lines. The right panel displays the amplitude of variability for 13 WN stars (11 showing short-term variability and 3 long-term) from our sample with significant signals as a function of effective temperature and spectral type. Stars with short periods are marked as circles, and ones with long periods as squares. Empty symbols are used to designate stars with no hydrogen in their spectra; otherwise, they are filled. BAT99-47 exhibits variability on both short and long timescales.
• Figure 5: Example of non-detrended light curves: BAT99-26