New findings on the prototypical Of?p stars
Y. Naze, A. ud-Doula, M. Spano, G. Rauw, M. De Becker, N.R. Walborn
aa r X i v : . [ a s t r o - ph . S R ] J un Astronomy&Astrophysicsmanuscript no. newsofp c (cid:13)
ESO 2018November 10, 2018
New findings on the prototypical Of?p stars ⋆ Ya¨el Naz´e ,⋆⋆ , Asif ud-Doula , Maxime Spano , Gregor Rauw ,⋆⋆ , Micha¨el De Becker , , and Nolan R. Walborn GAPHE, D´epartement AGO, Universit´e de Li`ege, All´ee du 6 Aoˆut 17, Bat. B5C, B4000-Li`ege, Belgiume-mail: [email protected] Penn State Worthington Scranton, 120 Ridge View Drive, Dunmore, PA 18512, USA Observatoire de Gen`eve, Universit´e de Gen`eve, 51 Chemin des Maillettes, CH1290-Sauverny, Switzerland Observatoire de Haute-Provence, F04870-St Michel l’Observatoire, France Space Telescope Science Institute ⋆⋆⋆ , 3700 San Martin Drive, Baltimore, MD 21218, USAPreprint online version: November 10, 2018
ABSTRACT
Aims.
In recent years several in-depth investigations of the three Galactic Of?p stars were undertaken. These multiwavelength studiesrevealed the peculiar properties of these objects (in the X-rays as well as in the optical): magnetic fields, periodic line profile variations,recurrent photometric changes. However, many questions remain unsolved.
Methods.
To clarify some of the properties of the Of?p stars, we have continued their monitoring. A new XMM -Newton observationand two new optical datasets were obtained.
Results.
Additional information for the prototypical Of?p trio has been found.HD 108 has now reached its quiescent, minimum-emission state, for the first time in 50–60yrs.The ´echelle spectra of HD 148937 confirm the presence of the 7d variations in the Balmer lines and reveal similar periodic variations(though of lower amplitudes) in the He i λ ii λ -Newton observation of HD 191612 was taken at the same phase in the line modulation cycle but at a di ff erent orbitalphase as previous data. It clearly shows that the X-ray emission of HD 191612 is modulated by the 538d period and not the orbitalperiod of 1542d - it is thus not of colliding-wind origin and the phenomenon responsible for the optical changes appears also at workin the high-energy domain. There are however problems: our MHD simulations of the wind magnetic confinement predict both aharder X-ray flux of a much larger strength than what is observed (the modeled DEM peaks at 30-40 MK, whereas the observed onepeaks at 2 MK) and narrow lines (hot gas moving with velocities of 100–200km s − , whereras the observed FWHM is ∼ − ). Key words.
X-rays: stars – Stars: early-type – Stars: individuals: HD 108,HD 148937,HD 191612 – Stars: emission-line
1. Introduction
The presence of C iii λ iii lines is not a usual feature of O-star spectra. In the past, stars displaying this feature were clas-sified by Walborn (1972) in a separate category, dubbed Of?p.In recent years, this category has attracted a lot of attention,as exemplified by numerous publications (Walborn et al. , 2003,2004; Donati et al. , 2006; Howarth et al. , 2007; Hubrig et al.,2008; Naz´e et al. , 2001, 2004, 2007, 2008a). These studies un-veiled the peculiar properties of these stars (for a summary seeNaz´e et al., 2008b) and led to a modification of the definitionof these stars. The Of?p phenomenon now covers stars present-ing recurrent spectral variations (in the Balmer, He i , C iii , Si iii lines), strong C iii λ i lines (at least at some phases), and UV windlines weaker than for typical Of supergiants (Naz´e et al. , 2008a; ⋆ Based on observations collected at the Haute-ProvenceObservatory, at the La Silla and San Pedro M´artir Observatories,and with XMM-Newton, an ESA Science Mission with instrumentsand contributions directly funded by ESA Member States and the USA(NASA). ⋆⋆ Research Associate FRS-FNRS ⋆⋆⋆
Operated by the Association of Universities for Research inAstronomy, Inc., under NASA contract NAS5-26555.
Walborn et al. , 2010). Stars that appear otherwise normal butdisplay strong C iii λ i line profile variations: the star is brighter when the emissionlines are stronger (Naz´e et al. , 2001; Naz´e et al., 2008b). Thevariations appear recurrent, with a probable time scale of about55 years (Naz´e et al. , 2006). The minimum emission state wasnot yet reached in 2004 (Naz´e et al. , 2004), preventing a preciseevaluation of the recurrence time. Recently, an additional obser-vational result enriched the picture of HD 108: the detection ofa strong magnetic field (observed strength is 100–150 G, corre-sponding to a potential dipolar field of 0.5–2 kG Martins et al.,2010).The behaviour of HD 191612 appears very similar tothat of HD 108, but with a shorter, well-defined period of537.6d (Howarth et al. , 2007, and references therein). A strongmagnetic field ( − ±
38 G) was detected for this star byDonati et al. (2006), who further proposed an oblique magneticrotator model for the star. Following these authors, the longperiod would result from magnetic braking of the stellar rota-tion (for an analysis of such e ff ects see ud-Doula et al., 2009).Studies are now under way to check the exact geometry of the Ya¨el Naz´e et al.: New findings on the prototypical Of?p stars magnetic field (G. Wade, private communication). Finally, it isinteresting to note that this star has a lower-mass companion ina 1542d orbit (Howarth et al. , 2007).HD 148937 does not display the large photometric changesand line profile variations unlike the two other objects of thisstudy. Instead, changes of limited amplitude (peak’s amplitudevarying by a few percents vs several 100% for the other proto-typical Of?p stars) were observed for the H α line, with a pos-sible period of about 7d, during a long-term monitoring cam-paign (Naz´e et al. , 2008a), but the temporal sampling of the op-tical dataset was not optimized for finding such a short periodand the resultant value could therefore be subject to large er-rors. In addition, no change was detected in the low-resolutiondata for other spectral lines, which di ff ers from what is ob-served in HD 108 and HD 191612. Finally, a tentative detectionof a magnetic field was reported for this star by Hubrig et al.(2008, ; B z = − ±
88 G), but a spectro-polarimetric moni-toring throughout the 7d cycle is clearly needed to ascertain thestrength and geometry of the field.In addition to dedicated optical campaigns, several XMM -Newton observations of these three objects were obtained. Thederived properties are remarkably similar: overluminosities withrespect to the ‘canonical’ L X / L BOL = − relation, soft spectra,broad X-ray lines (Naz´e et al. , 2004, 2007, 2008a). MonitoringHD 191612 further revealed small variations of the X-ray flux.However, as all data were taken during the same orbital cycle andthe same 538d cycle, the exact cause for these changes remainedunclear.Though the published studies have led to the characteriza-tion of many aspects of these Of?p stars, several questions stillremain unanswered - usually, these questions are di ff erent forthe di ff erent objects (e.g. on the variations of the X-rays ofHD 191612, on the evolution of the optical spectra of HD 108,or on the variation timescale of HD 148937), due to the inho-mogeneous datasets reported up to now. In the past few years,we have continued our monitoring of these three peculiar ob-jects, with the objective of characterizing and better understand-ing their properties. This paper reports the results of this cam-paign. Section 2 presents the datasets and Section 3 the results;Section 4 summarizes the new information.
2. Observations and data reduction
Since 2004, when optical spectra of HD 108 were last published(Naz´e et al. , 2004), we have continued the monitoring of the starwith the Aur´elie spectrograph mounted on the 1.52m telescopeof the Haute-Provence Observatory (OHP, France). The detec-tor was a thin back-illuminated CCD composed of 2048 × . µ m . We used the grating / after the stellar spectrum, and the spectra werenormalized using low-degree polynomials.An additional ´echelle spectrum obtained in June 2009 atSan Pedro M´artir with the Espresso spectrograph was also madeavailable to us by P. Eenens and L. Mahy. This dataset has beenadded to the analysis and discussion. Twenty ´echelle spectra of HD 148937 were obtained in 2008May 8–21 with the 1.2 m Euler Swiss telescope at La Silla(Chile), equipped with the Coralie spectrograph and an EEV2k ×
2k CCD with pixel size 15 µ m . The Coralie instrument isan improved version of the ´Elodie spectrograph (Baranne et al.,1996). These observations covered the spectral range 3850–6890Å with a resolving power of 55 000. During the first week,two spectra per night were taken; during the second week, onespectrum per night was taken. The integration times were fixedto 30min (except for the first two exposures, which were each20 min long) and the typical S / N ratio at 5000Å was 150. Thedata were first reduced with the Coralie pipeline and then nor-malized using low-degree polynomials. A telluric correction wasapplied for H α and He i λ telluric within IRAF. XMM -Newton observation
HD 191612 was observed anew by XMM -Newton on 2008 April2 (Rev. 1523, ObsID 050068201, PI Naz´e), for a total exposuretime of 24 ks. The instrumental configuration, processing, andextraction regions were identical to the first four observationsof the source (Naz´e et al. , 2007). We processed the data withthe Science Analysis System (SAS) software, version 8.0; fur-ther analysis was performed using the FTOOLS tasks and theXSPEC software v 11.2.0. No high background episode (due tosoft protons) was detected at high-energies during this exposure.In addition to EPIC data, XMM -Newton provides high-resolution spectra thanks to the RGS instruments. The RGS datawere reduced with the SAS and then combined with the previ-ous observations following the procedure outlined in Naz´e et al.(2007). However, the improvement in signal-to-noise of thecombined spectrum is not important as all observations are ofsimilar duration. We therefore have no new results to report onthe high-resolution spectrum of HD 191612, a quantitative leapin quality awaiting the advent of more sensitive observatories.
3. New results on Of?p stars
The spectrum of HD 108 has been followed intensively since1986 (Naz´e et al. , 2001, 2004). While some (purely photo-spheric) lines such as He ii λ i (e.g. at λλ iii (e.g. at λ iii λλ a¨el Naz´e et al.: New findings on the prototypical Of?p stars 3 Fig. 1.
Evolution of the line profiles in recent years for HD 108.2001), it is expected that the emission will strengthen againin the future, as is seen in HD 191612. We note however that,for HD 191612, the quiescent state occurs during one-third ofthe cycle, which would correspond to ∼
18 yrs for HD 108.Assuming that the variations are due to an oblique magnetic rotator configuration, this quiescence timescale would howeverdepend on the exact geometry of the system, which may be dif-ferent between the two objects despite their obvious similarities.from literature, it is only known that H β remained in absorptionduring about 10 years, while He i λ Ya¨el Naz´e et al.: New findings on the prototypical Of?p stars
HeII4542 HeII4686
Fig. 2.
Evolution as a function of time of a few quantities relatedto HD 108. Upper panel: EW (in Å, in the range 4845–4870Å) ofH β . Middle panel: spectral type criterion log (cid:16) EW HeI EW HeII (cid:17) . Lowerpanel: RVs (in km s − ) of He ii λλ Previously, a long-term monitoring (3 yrs, with monthly obser-vations) provided a large set of low-resolution spectra. Thesedata revealed the variability of 3 lines: H α , H β , and He ii λ α , the most variable line, dis-play a dispersion which, though it was 10 times larger than thatobserved for the neighbouring narrow Di ff use Interstellar Band(DIB), correspond to a change of only 20% in EW (it also corre-sponds to a variation in the peak’s amplitude of only 7% betweenthe two extreme profiles), whereas the other prototypical Of?pstars have H α lines varying from absorption to emission, withEWs and peak’s amplitudes varying by >> i and C iii lines seemed relatively constant, a behaviour di ff er-ent from that of HD 108 and HD 191612 - though subtle changesmay have remained hidden in such low-resolution data. A periodsearch further revealed a period of 7.031 ± α . However, the datasampling, aimed at studying monthly-to-yearly variations, wasnot adequate for detecting such a short period, which thus re-quired confirmation with a more intense temporal sampling. Anew, short-term monitoring was therefore undertaken, yielding20 high-resolution spectra over 2 weeks.For consistency, the radial velocities and equivalent widths(EWs) were estimated as in Naz´e et al. (2008a); they are givenin Table 1. While the C iv and DIBs display constant EWs, theBalmer, He ii λ i λ Table 1.
Mean and dispersion of the measured RVs and EWsfor HD 148937.
Line RV EW(km s − ) (Å)H γ ± ii λ − ± − ± β ± iv λ − ± ± i λ ± α − ± λ − ± ± sion. When plotted against time (Fig. 3), obvious modulationsare detected. H α and He ii λ β , H γ , and He i λ α , significant for H β , andbarely significant for H γ . This might explain why the variabilityof, e.g., He i λ i λ iii λ − , 0.1353d − , 0.1373d − , and 0.1397d − for He ii λ β , He i λ α , respectively. Thesefrequencies translate into a period of 7.16–7.60 ± ff erent observational timespans (2 weeks vs 2.5 yrs)while the period di ff erences from one line to the other have twoorigins: the noise, which a ff ects more the fainter lines such asHe i λ ff erent amplitudes of the variations (thesignal is indeed more di ffi cult to catch in case of low ampli-tude changes, e.g. for He i λ α line, which is bothstrong and the most variable line, yields the most reliable pe-riod, i.e. 7.16 ± , this agrees with the results reported by Naz´e et al.(2008a). The 7d period was thus not an artefact from our inad-equate time sampling. HD 148937 therefore appears quite sim-ilar to the two other Galactic Of?p stars, though with a shorterperiod, a smaller amplitude of the variations and with the soleexception of the apparent constancy in the C iii lines. A spec-tropolarimetric monitoring should be undertaken to see if thiscould be related to e.g. a low inclination of the star’s rotationaxis. Note that combining the two datasets is not possible due to thevery di ff erent spectral resolutions ( R = ± Fig. 3.
Evolution with time of the EWs (expressed in Å) for some lines of HD 148937. Left: The strongest emission lines (Balmerline of H α and He ii λ i λ α data were corrected for the contamination by telluric lines. Fig. 4.
Left: Average spectra and TVS of He ii λ β , He i λ α for HD 148937. The He i λ α data werecorrected for the contamination by telluric lines. The horizontal dotted line in the lower panels represents the 99% confidence level.Right: The 20 Coralie observations of the H α line (from May 8 to May 21, note the high noise in the data of May 19, which reflectsin the EW measurements) and the extreme states of H α , He i λ β . The black solid lines correspond to the spectra of May09 (max emission state) and May 12 (min), the dotted red lines to those of May 15 (max) and May 20 (min). The colored version ofthe figure is available on-line. The previous observations of HD 191612 (Naz´e et al. , 2007) re-vealed changes in the X-ray flux but the origin of this variabilitywas unclear. To get a better understanding of this phenomenon, we undertook a new XMM -Newton observation, whose schedul-ing was carefully chosen. While the previous observations weretaken at orbital phases φ orb = φ cyc = φ cyc = / photometric cycle of 538d and aphase φ orb = Ya¨el Naz´e et al.: New findings on the prototypical Of?p stars
Fig. 5.
Left, from top to bottom: Raw periodogram of H α as observed during the May 2008 campaign, periodogram after prewhiten-ing by the period of 7.16d, and associated spectral window in the case of HD 148937. The latter is directly related to the samplingof the time series. Right, from top to bottom: Periodograms of He ii λ β , and He i λ i λ α data werecorrected for the contamination by telluric lines.from Howarth et al. 2007). The new data thus enable us to checkwhether the flux variations are linked to the circumstellar, pos-sibly magnetic, structures surrounding the star or to collidingwinds in a binary. In the former case, the new data should benearly identical to the previous observation taken at a similarphase φ cyc . On the contrary, in the latter case, a dramatic varia-tion is expected. In a colliding wind binary, variations of the flux,correlated with orbital phase, are expected to arise from changesin the separation (in eccentric systems) and / or in the absorb-ing column (when the winds are di ff erent). Indeed, such vari-ations have been observed in several systems (for a review, seeG¨udel & Naz´e, 2009). In HD 191612, the previous observations,taken before periastron, revealed a 20% change in reddening-corrected flux (see Table 2) for a phase di ff erence ∆ φ orb of only0.12, corresponding to a change in separation of 36%. The newobservation was taken two-thirds of the orbit later, near apastron:the change in separation compared to previous data has thus in-creased (up to 60%), and so should the di ff erence in flux, if trulylinked to the colliding wind emission.The new data were processed following the same procedureas for the previous data, but using more recent XMM -Newton re-duction software. The count rates derived from a run of the task edetect chain amount to 0.122 ± − for EPIC MOS1,0.119 ± − for EPIC MOS2, and 0.392 ± − for EPIC pn. The evolution of the count rates and hardness ratiois shown in Fig. 6. The new data agree very well with previousobservations taken at the same phase of the 538 d cycle, both instrength (similar count rates) and in hardness (similar hardnessratios) ; they are in addition clearly di ff erent from the resultsof previous observations obtained at the latest phase (where thecount rates are ∼
35% smaller, corresponding to a ∼ σ di ff er- ence, and the hardness ratio HR measured on pn data is lowerby 7 σ ).The best spectral fit is further provided in Table 2,together with the previous results for comparison (repro-duced from Naz´e et al. 2007). The fitted model has the form wabs( N Hint )*[wabs( N H1 )*mekal(k T )+wabs( N H2 )*mekal(k T )] ,with wabs( N Hint ) = . × cm − (Diplas & Savage, 1994) andusing solar abundances for both absorptions and optically-thinplasma emissions - i.e. the same model as used in Naz´e et al.(2007) for homogeneity reasons . Quoted fluxes are in the0.4 − f unabsX arecorrected only for the interstellar absorbing column. For eachparameter, the lower and upper limits of the 90% confidenceinterval (derived from the error command under XSPEC) arenoted as indices and exponents, respectively. The normalisationfactors are defined as − π D R n e n H dV , where D , n e and n H are respectively the distance to the source, the electron andproton density of the emitting plasma. The phases φ orb and φ cyc correspond to the phases in the orbital cycle and the538 d line profile / photometric cycle, respectively (ephemerisof Howarth et al. 2007). Again, the spectral fits to the 2008data agree very well with the previous observations taken atthe same phase of the 538 d cycle, yielding the same X-rayluminosity, 9 × erg s − (for a distance of 2.29 kpc), and samelog[ L X / L BOL ], − ff erent or-bital phases, the new data are nearly identical to those taken 3years before at a similar phase in the 538d cycle. The modula- As in Naz´e et al. (2007), we noted the two sets of parameters givinga good fit, but we recall that a fit by a di ff erential emission measure(DEM) model clearly favors the ‘cool’ fit.a¨el Naz´e et al.: New findings on the prototypical Of?p stars 7 tion of the X-ray emission thus appears directly correlated withthe optical line-profile variations, a situation very similar to thatobserved for θ Ori C (Gagn´e et al., 2005; Naz´e et al., 2008b).
To better understand the properties of HD 191612, a fully-dynamic numerical magneto-hydrodynamic (MHD) model, tak-ing into account the impact on the wind of the magnetic fielddetection, is needed. We used the publically available ZEUS-3D code (Stone & Norman, 1992) to evolve a consistent dynam-ical solution for a line-driven stellar wind from the star witha dipolar surface field. The stellar parameters of HD 191612,adopted mostly from Naz´e et al. (2008b, and references therein),are listed in Table 3.HD 191612 is a rather slow rotator with a surface rotationspeed < − (Howarth et al. , 2007; Naz´e et al., 2008b).Such a slow rotation is unlikely to have any significant dynami-cal e ff ects on the stellar wind. Therefore, the wind was assumedto be azimuthally symmetric with the magnetic field axis alignedwith the rotation axis . This is essentially a ’2.5-D’ formulationallowing for non-zero azimuthal components of both the mag-netic field B φ and velocity v φ , while still assuming that all quan-tities are constant in the azimuthal coordinate angle φ .Much of the numerical procedures in this simulation fol-lows ud-Doula et al. (2008). The models thus include rotationale ff ects but, for the first time, we also use the full energyequation, with adiabatic index γ = / .At the initial time t =
0, the wind is represented by a relaxednon-magnetic and non-rotating model, but we then simultane-ously introduce a dipole magnetic field and a surface rotationat the lower boundary, both defined relative to a common polaraxis. The wind and magnetic field are then let free to competewith each other. To ensure that there are no numerical artefactsarising from the initial conditions, we run our model for about1.5 Msec, which is at least 50 fluid crossing times.Immediately after t =
0, the magnetic field channels windmaterial towards the tops of closed loops near the equator. There,the collision with the opposite stream leads to strong shocks, likein “magnetically confined wind shocks” (MCWS) presented byBabel & Montmerle (1997). Denser material then cools radia-tively very quickly leading to a dense disk-like structure (Fig.7). Some of the less dense post-shock material within the closedmagetic loops remain very hot (ca. 80MK) for a long period oftime. In this model, the Alfv´en radius or magnetosphere extendsabout 3–4 stellar radii above the stellar surface.Initially mass builds up in the equatorial region below theKeplerian radius estimated to be at about 4.5 R ∗ , but then thereappear repeated episodes of infall of inner disk material backonto the stellar surface. Over a somewhat longer timescale, thereappears another, somewhat di ff erent kind of disruption, one thatstarts higher up, closer to the Alfv´en radius. This is characterizedby outward ejections of the upper disk mass, leading to ‘centrifu-gally driven mass ejections’. Such ejections can lead to X-rayflares (see ud-Doula et al. 2006) but these short-term variationscannot be detected in the observations of HD 191612, since they This simplifies the computation and makes no assumption on thetrue geometry of the system, i.e. an oblique rotator model is not ex-cluded ; it can easily be simulated here by changing the viewing angleon the system. Note that models used by Gagn´e et al. (2005) made use of the fullenergy equation but did not include rotation.
Table 3.
Stellar parameters used for the MHD simulation ofHD 191612, adopted mostly from Naz´e et al. (2008b, and ref-erences therein).
Parameter Value R × cm = ⊙ M × g =
35 M ⊙ L × erg s − = × L ⊙ α Q δ M × − M ⊙ yr − v rot − B polar η ∗ ∼ √ lack the required sensitivity to analyze the lightcurve on veryshort timescales.Fig. 8 shows the di ff erential emission measure (DEM, or therelative amount of material in a given temperature bin) calcu-lated in the MHD simulation. The DEM peaks at very high tem-peratures (about 10 MK and larger), suggesting that the very hotgas dominates the X-ray emission of the system. Of course, thematerial in the unshocked regions remain at the e ff ective temper-ature of the star due to the intense UV radiation from the parentstar that keeps the wind nearly isothermal. Applying a magnetically-confined wind model to HD 191612yields four clues. First, the X-ray emission should clearly bedominated by very hot plasma, as was the case for θ Ori C (seeFig. 6 of Gagn´e et al., 2005). Second, on average, the totalEM calculated for HD191612 in the 1-100MK range amountsto 5.3 × cm − . This agrees well with the MHD simulationof θ Ori C, considering the di ff erence in the stellar parametersand modelling. Indeed, the simulated EM for this star, in thesame temperature range, reached 8.2 × cm − (Gagn´e et al.,2005), i.e. a factor 6.5 lower than for HD 191612 which is closeto the factor of 4 expected from the EMs scaling with the thirdpower of the radius. Considering an average cooling factor Λ ( T )of ∼ × − erg cm s − (Raymond et al., 1976), this yields a totally unabsorbed X-ray luminosity of ∼ × erg s − in the0.5–10. keV range. Third, in the MHD models, the hot gas gen-erally moves with velocities of about 100–200km s − and onlyvery few hot plasma reach 400km s − , whatever the viewingangle on the equatorially-confined region. This is unsurprisingsince the hot gas always remains close to the star: the associatedX-ray lines are thus expected to be narrow, again a similar resultas that found for θ Ori C. Finally, the X-ray emission should bemodulated throughout the cycle of 538d, as the viewing angle onthe disk-like structure changes.The observations (both old and new) do not fully agree withthis picture. On the one hand, the emission is modulated withthe 538d cycle, as expected for MCWS. The observed X-rayemission is indeed bright ( L X = × erg s − when correctedfor the interstellar absorption), in fact on the luminous edgecompared to “normal” O-type stars (for O stars observed withXMM -Newton , log[ L X / L BOL ] = − .
45 with a dispersion of 0.51dex, see Naz´e 2009). Considering the uncertainty in the spec-tral models (see Table 2), the observations yield a totally un-absorbed
X-ray luminosity of ∼ × erg s − , for the ‘hot’ Ya¨el Naz´e et al.: New findings on the prototypical Of?p stars
Table 2.
Best-fitting models and X-ray fluxes at Earth for each XMM -Newton observation of HD 191612.
Date Rev. φ cyc φ orb N H1 k T norm N H2 k T norm χ ν (dof) f absX f unabsX cm − keV 10 − cm − cm − keV 10 − cm − (10 − erg cm − s − )‘cool’ model05 / /
05 975 0.09 0.84 0 . . . . . . . . . . . . . . . . . . / /
05 981 0.12 0.84 0 . . . . . . . . . . . . . . . . . . / /
05 1004 0.20 0.87 0 . . . . . . . . . . . . . . . . . . / /
05 1068 0.44 0.96 0 . . . . . . . . . . . . . . . . . . / /
08 1523 0.13 0.55 0 . . . . . . . . . . . . . . . . . . / /
05 975 0.09 0.84 0 . . . . . . . . . . . . . . . . / /
05 981 0.12 0.84 0 . . . . . . . . . . . . . . . . . . / /
05 1004 0.20 0.87 0 . . . . . . . . . . . . . . . . . . / /
05 1068 0.44 0.96 0 . . . . . . . . . . . . . . . . . . / /
08 1523 0.13 0.55 0 . . . . . . . . . . . . . . . . . . Fig. 6.
Left: Position of the new XMM -Newton observation in the cycles of HD 191612. The orbital cycle, calculated using orbitalparameters of Howarth et al. (2007) and with the primary (resp. secondary) curve drawn in solid (resp. dotted) line, is shown on topwhile the bottom panel shows the line profile variation cycle of 538d, from the analytical approximation of Howarth et al. (2007).The phases of the previous observations are shown by the thin short-dashed lines and the phase of the new observation by the thicklong-dashed line. Right: Evolution of the count rate in the 0.4–10.0 keV band (top) and of two hardness ratios (middle, bottom) withphase from the 538d cycle. Open symbols refer to the old observations, filled red symbols to the new dataset; triangles and squareswere used for the EPIC MOS data (note that the MOS count rates are multiplied by a factor of 3 in this figure) and circles for theEPIC pn data. The hardness ratio HR and HR are defined as ( M − S ) / ( M + S ) and ( H − M ) / ( H + M ), respectively, where S is thecount rate in the 0.4–1.0 keV band, M in 1–2 keV, and H in 2–10 keV. The colored version of the figure is available on-line.and ‘cool’ model respectively. Total emission measures, derivedfrom the normalization factors of Table 2 or as in Gagn´e et al.(2005) for a spectral fitting considering a single absorbing col-umn for both thermal components, amount to 3–7 × cm − foran equivalent ‘cool’ model and ten times lower for an equivalent‘hot’ model. Theoretical predictions thus agree rather well withthe ‘cool’ model, which is also the one favored by the DEMmodelling(Naz´e et al. , 2007). On the other hand, however, the X-ray emission ofHD 191612 displays broad lines ( FWHM obs ∼ − ,Naz´e et al. , 2007) and is far from being as hard as predictedby the model. In fact, a fit of the EPIC data by a DEM modelyields a strong emission peak at about 2 MK, not beyond 10 MK,as shown for example in Fig. 3 of Naz´e et al. (2007). In ad- a¨el Naz´e et al.: New findings on the prototypical Of?p stars 9 Fig. 7.
Snapshots of density (left) and temperature (right), in cgs units on a logarithmic color-scale, shown with field lines (solidlines) at arbitrarily chosen time t = τ =
10 for κ =
30 cm g − , seeCohen 2009, a density of 10 − g cm − and a disk height of 0.03 R ∗ ) but it is totally negligible in the dense equatorial region outsidemagnetosphere. The observations show only a small absorption in addition to the interstellar extinction and it is not significantlyvarying with phase (i.e. there is no clear evidence of excess absorption in the magnetic equatorial plane, or the e ff ect is much smallerthan would be expected from a dense equatorial cooling disk, as found in Gagn´e et al. 2005). Fig. 8.
Volume emission-measure distribution (DEM) per log T = derived in the 2XMM surveyfor HD 191612 amounts to ∼ < kT > below 1 keV,Naz´e 2009) - it does not resemble the extreme case of θ Ori Cfor which < kT > amounts to 2.5 keV in the same 2XMM sur-vey and for which the MCWS model works well. The relativelysoft character of that emission despite the large magnetic con-finement parameter might be reminiscent of the characteristicsobserved in some B-type stars (G¨udel & Naz´e, 2009): furtherinvestigation of the MCWS model and magnetic field geome-try is thus required to understand the high-energy properties ofthe Of?p stars.
4. Conclusions
Our continued monitoring in the X-ray and optical domains ofthe prototypical Of?p stars has now clarified some of their ob-servational properties.For HD 108, the Balmer, C iii λ i , and Si iii lines dis-play minimum emissions since 2005: the star has thus finallyreached its quiescent state, for the first time in 50–60yrs. Long-term monitoring of this quiescence will further constrain the ge-ometry of the line emission region.For HD 148937, the presence of the 7d variations in theBalmer lines is confirmed by a short-term monitoring. Duringthis cycle, the amplitude of the emission peak in H α , the mostvariable line, changes by only 7%. Thanks to the high spectralresolution of the new data, similar periodic variations, thoughof even smaller amplitude, are detected in the He i λ ii λ -Newton observation clearlyshows that its X-ray emission is modulated by the 538d periodand not the orbital period - it is thus not of colliding-wind ori-gin: the phenomenon responsible for the optical changes appearsalso at work in the high-energy domain. Contrary to observa-tions, our MHD simulations predict a hard spectrum, dominatedby very hot plasma ( >
10 MK) and narrow lines. Further refine-ments to the modelling will be needed in order to fully explainthis discrepancy. P kT i × norm i / P norm i for models using a single absorption infront of the thermal emission components Acknowledgements.
YN, MDB and GR acknowledge support from the FondsNational de la Recherche Scientifique (Belgium), the Communaut´e Franc¸aisede Belgique, the PRODEX XMM and Integral contracts, and the ‘Action deRecherche Concert´ee’ (CFWB-Acad´emie Wallonie Europe). The authors thankL. Mahy and P. Eenens for providing the 2009 ´echelle spectrum. ADS and CDSwere used for preparing this document.
References
Babel, J., & Montmerle, T. 1997, A&A, 323, 121Baranne, A., et al. 1996, A&AS, 119, 373Cohen, D. H. 2009, American Institute of Physics Conference Series, 1161, 132Diplas, A., & Savage, B. D. 1994, ApJS, 93, 211Donati, J.-F., Howarth, I.D., Bouret, J.-C., Petit, P., Catala, C., & Landstreet, J.2006, MNRAS, 365, L6Fullerton, A.W., Gies, D.R., & Bolton, C.T. 1996, ApJS, 103, 475Gagn´e M., Oksala M.E., Cohen D.H., et al. 2005, ApJ, 628, 986Gosset, E., Royer, P., Rauw, G., Manfroid, J., & Vreux, J.-M. 2001, MNRAS,327, 435G¨udel, M., & Naz´e, Y. 2009, A&A Rev., 17, 309Heck, A., Manfroid, J., Mersch, G. 1985, A&AS, 59, 63Hubrig, S., et al. 2008, A&A, 490, 793Hinkle, K., Wallace, L., Valenti, J., & Harmer, D. 2000, Visible and Near InfraredAtlas of the Arcturus Spectrum 3727-9300 A ed. Kenneth Hinkle, LloydWallace, Je ff Valenti, and Dianne Harmer. (San Francisco: ASP) ISBN: 1-58381-037-4, 2000.,Howarth, I.D., Walborn, N.R., Lennon, D.J., et al. 2007, MNRAS, 381, 433MacDonald, J., & Bailey, M. E. 1981, MNRAS, 197, 995Martins, F., Donati, J.-F., Marcolino, W.L.F., Bouret, J.-C., Wade, G.A.,Escolano, C., Howarth, I.D., and the MiMeS Collaboration 2010, MNRAS,in press (arXiv:1005.1854)Naz´e, Y. 2009, A&A, 506, 1055Naz´e, Y., Vreux, J.-M., Rauw, G. 2001, A&A 372, 195Naz´e, Y., Rauw, G., Vreux, J.-M., & De Becker, M. 2004, A&A, 417, 667Naz´e, Y., Barbieri, C.,Segafredo, A., Rauw, G., & De Becker, M. 2006, IBVS,5693Naz´e, Y., Rauw, G., Pollock, A.M.T., Walborn, N.R., & Howarth, I.D. 2007,MNRAS, 375, 145Naz´e, Y., Walborn, N.R., Rauw, G., Martins, F., Pollock, A.M.T., & Bond, H.E.2008a, AJ, 135, 1946Naz´e, Y., Walborn, N. R., & Martins, F. 2008b, Revista Mexicana de Astronomiay Astrofisica, 44, 331Rauw, G., Morrison, N. D., Vreux, J.-M., Gosset, E., & Mulliss, C. L. 2001,A&A, 366, 585Raymond, J. C., Cox, D. P., & Smith, B. W. 1976, ApJ, 204, 290Stone, J. M., & Norman, M. L. 1992, ApJS, 80, 753ud-Doula, A., & Owocki, S. P. 2002, ApJ, 576, 413ud-Doula, A., Townsend, R. H. D., & Owocki, S. P. 2006, ApJ, 640, L191ud-Doula, A., Owocki, S. P., & Townsend, R. H. D. 2008, MNRAS, 385, 97ud-Doula, A., Owocki, S. P., & Townsend, R. H. D. 2009, MNRAS, 392, 1022Walborn, N.R. 1972, AJ, 77, 312Walborn, N.R., Howarth, I.D., Herrero, A., & Lennon, D.J. 2003, ApJ, 588, 1025