Detection of a centrifugal magnetosphere in one of the most massive stars in the ρ Oph star-forming cloud
S. Hubrig, M. Schöller, S.P. Järvinen, M. Küker, A.F. Kholtygin, P. Steinbrunner
aa r X i v : . [ a s t r o - ph . S R ] D ec Received 26 April 2016; Revised 6 June 2016; Accepted 6 June 2016DOI: xxx/xxxx
ARTICLE TYPE
Detection of a centrifugal magnetosphere in one of the mostmassive stars in the 𝜌 Oph star-forming cloud
S. Hubrig* | M. Schöller | S. P. Järvinen | M. Küker | A. F. Kholtygin | P. Steinbrunner Leibniz-Institut für Astrophysik Potsdam(AIP), An der Sternwarte 16,14482 Potsdam, Germany European Southern Observatory,Karl-Schwarzschild-Str. 2, 85748 Garching,Germany Astronomical Institute, Saint-PetersburgState University, Universitetskij pr. 28,198504 Saint-Petersburg, Russia Freie Universität Berlin, KaiserswertherStr. 16-18, 14195 Berlin, Germany
Correspondence *Swetlana Hubrig. Email: [email protected]
Recent XMM-Newton observations of the B2 type star 𝜌 Oph A indicated a periodic-ity of 1.205 d, which was ascribed to rotational modulation. Since variability of X-rayemission in massive stars is frequently the signature of a magnetic field, we investi-gated whether the presence of a magnetic field can indeed be invoked to explain theobserved X-ray peculiarity. Two FORS 2 spectropolarimetric observations in differ-ent rotation phases revealed the presence of a negative ( ⟨ 𝐵 z ⟩ all = −419 ± 101 G) andpositive ( ⟨ 𝐵 z ⟩ all = 538±69 G) longitudinal magnetic field, respectively. We estimatea lower limit for the dipole strength as 𝐵 d = 1 . . kG. Our calculations of theKepler and Alfvén radii imply the presence of a centrifugally supported, magneti-cally confined plasma around 𝜌 Oph A. The study of the spectral variability indicatesa behaviour similar to that observed in typical magnetic early-type Bp stars.
KEYWORDS: stars: early-type, stars: individual: 𝜌 Oph A, stars: magnetic field, stars: variables: general, stars: formation,stars: X-rays
The 𝜌 Ophiuchus star-forming cloud is one of the closest lowto intermediate mass star-forming regions and is known tocontain the star-forming cluster LDN 1688. Due to its youth,the relative proximity (120–160 pc; e.g. Motte et al., 1998),and its richness in young stars and protostars, the 𝜌 Oph star-forming cloud has been actively investigated in recent years,using multiwavelength observations from X-ray to radio bands.The radiation field in LDN 1688 is dominated by two B-type stars, 𝜌 Oph A (HD 147933) and 𝜌 Oph B (HD 147934),both of spectral type B2V (Abergel et al., 1996), making upthe visual binary system 𝜌 Oph AB. The apparent distancebetween the two stars is about 3 . ′′
1, and their orbital periodis around 2400 yr. 𝜌 Oph AB is surrounded by the bright andextended blue reflection nebula vdB 105. The basic structureof the cloud complex and its surroundings is described in detailby Wilking et al. (2008). Massive young stars are known to emit strong X-rays. Unlikethe X-ray emission from lower mass stars, which arises in stel-lar photospheres, the X-rays from massive stars are thoughtto result from powerful shocks. Detection of hard X-ray emis-sion in massive stars appears frequently to be a signature of thepresence of strong magnetic fields (e.g. Skinner et al., 2008).The first XMM-Newton observations of 𝜌 Oph AB over 53 kswere obtained by Pillitteri et al. (2014). Their analysis showeda smooth variability of the X-ray emission, probably causedby the emergence of an extended active region on the surfaceof 𝜌 Oph A. The derived hardness ratios were periodic, withthe hardest spectrum corresponding to the highest count rate.According to Pillitteri et al. (2014), the observations are fullycompatible with the hypothesis of the presence of a regionbrighter and hotter than the average stellar surface that gradu-ally appears on the visible side of the star during the rise of thecount rate. Follow-up XMM-Newton observations with a dura-tion of 140 ks in 2016 allowed Pillitteri et al. (2017) to detect aperiodicity of 1.205 d, which they ascribed to rotational modu-lation of the X-ray emission. Analysing the time resolved X-rayspectra, the authors speculated that either intrinsic magnetism
S. Hubrig
ET AL produces a hot spot on the surface of 𝜌 Oph A, or an unknownlow mass companion is the source of the observed X-ray vari-ability. Clearly, the remarkable behavior of 𝜌 Oph A in X-raysdeserves further investigations to find out whether the pres-ence of a magnetic field can indeed be invoked to explain theobserved X-ray peculiarity.In Section 2, we give an overview about the FORS 2 spec-tropolarimetric observations of 𝜌 Oph A as well as the datareduction, and describe the results of the magnetic field mea-surements. The magnetospheric parameters and our analysis ofthe spectral variability of 𝜌 Oph A are shown in Sections 3 and4. A discussion of our results is presented in Section 5.
FOcal Reducer and low dispersion Spectrograph (FORS 2;Appenzeller et al., 1998) spectropolarimetric observations of 𝜌 Oph A were obtained on 2017 July 17 and August 11. TheFORS 2 multi-mode instrument is equipped with polarisationanalysing optics comprising super-achromatic half-wave andquarter-wave phase retarder plates, and a Wollaston prism witha beam divergence of 22 ′′ in standard resolution mode. Weused the GRISM 600B and the narrowest available slit widthof 0 . ′′ 𝑅 ≈ 2000 .The observed spectral range from 3250 to 6215 Å includesall Balmer lines, apart from H 𝛼 , and numerous helium lines.For the observations, we used a non-standard readout modewith low gain (200kHz,1 × − ◦ + ◦ , + ◦ − ◦ , − ◦ + ◦ , etc. is usually executed during theobservations. Moreover, the reversal of the quarter wave platecompensates for errors in the relative wavelength calibrationsof the two polarised spectra. According to the FORS UserManual, the 𝑉 ∕ 𝐼 spectrum is calculated using: 𝑉𝐼 = 12 {( 𝑓 o − 𝑓 e 𝑓 o + 𝑓 e ) −45 ◦ − ( 𝑓 o − 𝑓 e 𝑓 o + 𝑓 e ) +45 ◦ } , (1)where +45 ◦ and −45 ◦ indicate the position angle of theretarder waveplate and 𝑓 o and 𝑓 e are the ordinary and theextraordinary beam, respectively. Null profiles, 𝑁 , are cal-culated as pairwise differences from all available 𝑉 profiles. From these, 3 𝜎 -outliers are identified and used to clip the 𝑉 profiles. This removes spurious signals, which mostly comefrom cosmic rays, and also reduces the noise. A full descrip-tion of the updated data reduction and analysis will be pre-sented in a separate paper (Schöller et al., in preparation, seealso Hubrig et al., 2014a). The mean longitudinal magneticfield, ⟨ 𝐵 z ⟩ , is measured on the rectified and clipped spec-tra based on the relation following the method suggested byAngel & Landstreet (1970) 𝑉𝐼 = − 𝑔 eff 𝑒 𝜆 𝜋 𝑚 e 𝑐 𝐼 d 𝐼 d 𝜆 ⟨ 𝐵 z ⟩ , (2)where 𝑉 is the Stokes parameter that measures the circularpolarization, 𝐼 is the intensity in the unpolarized spectrum, 𝑔 eff is the effective Landé factor, 𝑒 is the electron charge, 𝜆 isthe wavelength, 𝑚 e is the electron mass, 𝑐 is the speed of light, d 𝐼 ∕d 𝜆 is the wavelength derivative of Stokes 𝐼 , and ⟨ 𝐵 z ⟩ is themean longitudinal (line-of-sight) magnetic field.The longitudinal magnetic field was measured in two ways:using the entire spectrum including all available lines or usingexclusively the hydrogen lines. Furthermore, we have carriedout Monte Carlo bootstrapping tests. These are most oftenapplied with the purpose of deriving robust estimates of stan-dard errors. The measurement uncertainties obtained beforeand after the Monte Carlo bootstrapping tests were found to bein close agreement, indicating the absence of reduction flaws.The results of our magnetic field measurements, those for theentire spectrum or only for the hydrogen lines, are presentedin Table 1 . As no ephemeris is known for 𝜌 Oph A – only 𝑃 rot = 1 . d is mentioned in Pillitteri et al. (2017) – we fixedthe rotation phase 0 at the date of the first FORS 2 observationat MJD 57951.2242.The magnetic field of 𝜌 Oph A was detected on both observ-ing dates. Using the entire spectrum, we measure a field ofnegative polarity, ⟨ 𝐵 z ⟩ all = −419 ± 101 G, at a significancelevel of 4.1 𝜎 in the data obtained on 2017 July 17, while themeasurement using the hydrogen lines shows a significance of2.1 𝜎 . The highest values for the longitudinal magnetic field, ⟨ 𝐵 z ⟩ all = 538 ± 69 G at a significance level of 7.8 𝜎 using theentire spectrum and ⟨ 𝐵 z ⟩ hyd = 569 ± 94 G at a significancelevel of 6.1 𝜎 using the hydrogen lines was measured in thedata obtained on 2017 August 11. No detection was achievedin the diagnostic 𝑁 spectra, indicating the absence of spuriouspolarization signatures.In Figs. 1 and 2 , we show the linear regressions in plotsof 𝑉 ∕ 𝐼 against −4 .
67 10 −13 𝜆 (1∕ 𝐼 )(d 𝐼 ∕d 𝜆 ) together with theresults of the Monte Carlo bootstrapping tests. In Fig. 3 , wepresent Stokes 𝐼 , 𝑉 , and diagnostic 𝑁 spectra of 𝜌 Oph Aobtained on 2017 August 11 in the spectral region around theH 𝛽 line and several He I lines. . Hubrig ET AL TABLE 1
Logbook of the FORS 2 spectropolarimetric observations of 𝜌 Oph A, including the modified Julian date of mid-exposure, followed by the achieved signal-to-noise ratio in the Stokes 𝐼 spectra around 5200 Å, and the measurements of themean longitudinal magnetic field using the Monte Carlo bootstrapping test, for all lines and for the hydrogen lines. In the lastcolumns, we present the results of our measurements using the null spectra for the set of all lines and the orbital phases. Therotation phases of 𝜌 Oph A are calculated relative to a zero phase corresponding to the date of the first FORS 2 observation atMJD 57951.2242 assuming 𝑃 rot = 1 . d. All quoted errors are 1 𝜎 uncertainties.MJD SNR ⟨ 𝐵 z ⟩ all ⟨ 𝐵 z ⟩ hyd ⟨ 𝐵 z ⟩ N 𝜑 orb 𝜆 − ± − ± − ±
92 057976.0708 3005 538 ±
69 569 ±
94 29 ±
68 0.620 -0.5 0.0 0.5-4.67 10 -13 λ (1/ I ) (d I /d λ ) [10 -6 G -1 ]-0.4-0.20.00.20.4 V / I [ % ] -600 -500 -400 -300 [G]0.00.20.40.60.81.0 N o r m a li ze d o cc u rr e n ce s FIGURE 1
Left panel : Linear fit to Stokes 𝑉 obtained forthe FORS 2 observation of 𝜌 Oph A on MJD 57951.2242.
Rightpanel : Distribution of the longitudinal magnetic field values 𝑃 ( ⟨ 𝐵 z ⟩ ) , which were obtained via bootstrapping. From thisdistribution follows the most likely value for the longitudinalmagnetic field ⟨ 𝐵 z ⟩ all = −419 ± 101 G.We can estimate the dipole strength of 𝜌 Oph A followingthe model by Stibbs (1950) as formulated by Preston (1967): 𝐵 d ≥ ⟨ 𝐵 z ⟩ max ( 𝑢 𝑢 ) ) −1 . (3)Assuming a limb-darkening coefficient 𝑢 of 0.3, typical forthe spectral type B2V (Claret & Bloemen, 2011), we can givea lower limit for the dipole strength of 𝐵 d ≥ . . kG. Similar to the small number of previously studied early-Btype stars, the X-ray emission in 𝜌 Oph A detected in XMM-Newton observations can be generated via magnetically con-fined shocks. Babel & Montmerle (1997) suggested that instars with large-scale magnetic fields wind streams from oppo-site hemispheres are channeled toward the magnetic equator,where they collide, leading to strong shocks and associated X-rays. To investigate whether the wind plasma is locked to themagnetic field, we need to know several physical parameters -0.5 0.0 0.5-4.67 10 -13 λ (1/ I ) (d I /d λ ) [10 -6 G -1 ]-0.4-0.20.00.20.4 V / I [ % ]
400 500 600 700 [G]0.00.20.40.60.81.0 N o r m a li ze d o cc u rr e n ce s FIGURE 2
Left panel : Linear fit to Stokes 𝑉 obtained forthe FORS 2 observation of 𝜌 Oph A on MJD 57976.0708.
Rightpanel : Distribution of the longitudinal magnetic field values 𝑃 ( ⟨ 𝐵 z ⟩ ) , which were obtained via bootstrapping. From thisdistribution follows the most likely value for the longitudinalmagnetic field ⟨ 𝐵 z ⟩ all = 538 ± 69 G.of 𝜌 Oph A. Currently, the only information can be found in thework of Pillitteri et al. (2017), who assumed a stellar radius of ∼ 8 𝑅 ⊙ , without mentioning the method of estimation, usedthe rotational modulation of the X-ray emission to determine arotation period of 1.205 d, and applied a Fourier transform tothe He I 𝑣 sin 𝑖 = 239 . km s −1 .Based on revised trigonometric parallaxes from the Hip-parcos data (van Leeuwen, 2008), Mamajek (2008) concludedthat a distance of
131 ± 3 pc to the Ophiuchus star-formingregion is the best available derived from Hipparcos data. Usingthis distance, the extinction 𝐴 v = 3 from Pillitteri et al.(2016), assuming 𝑇 eff = 21 000 K for the spectral type B2(Böhm-Vitense, 1981), and the corresponding bolometric cor-rection 𝐵𝐶 = −2 . . (Flower, 1996), we estimate log ( 𝐿 ∕ 𝐿 ⊙ ) = 4 . . , taking into account estimationinaccuracies of the distance determination and the bolometriccorrection. From the position of 𝜌 Oph A in the H-R dia-gram, using evolutionary tracks from Ekström et al. (2012) andassuming that 𝜌 Oph A is still in the hydrogen fusing stage,we find a mass of about
10 ± 0 . 𝑀 ⊙ . From the Stefan–Boltzmann law we find log( 𝑅 ∕ 𝑅 ⊙ ) ≈ 0 . , i.e. a value of S. Hubrig
ET AL
I/I C −0.10.00.1 x V /I C −0.10.00.1 x N /I C FIGURE 3
Stokes 𝐼 , 𝑉 , and diagnostic 𝑁 spectra (from bot-tom to top) of 𝜌 Oph A in the vicinity of the H 𝛽 line. Note thatthe Stokes 𝑉 and the diagnostic 𝑁 spectra were magnified bya factor of 80. . 𝑅 ⊙ for the stellar radius. Using this radius, 𝑣 sin 𝑖 =240 ± 10 km s −1 , and the rotation period 𝑃 rot = 1 . d, weobtain 𝑣 eq = 360 ± 40 km s −1 and an inclination angle of thestellar rotation axis to the line of sight 𝑖 = 42 ± 6 ◦ . Fur-ther, the rotation period of 1.205 d corresponds to an angularvelocity Ω = 6 .
035 × 10 −5 s −1 . To determine the propertiesof the stellar wind and the magnetosphere, we first computethe escape velocity, 𝑣 esc = √ 𝐺𝑀 ∕ 𝑅 = 671 km s −1 . Assum-ing that 𝑣 ∞ ∕ 𝑣 esc = 1 . (Vink et al., 2001), we find 𝑣 ∞ =873 km s −1 for the terminal velocity. Using equation (25) ofVink et al. (2001), we calculate a mass loss rate ̇𝑀 = 8 .
69 ×10 −9 𝑀 ⊙ ∕yr for solar metallicity. With a polar magnetic fieldstrength 𝐵 𝑝 ≥ . kG, we obtain a confinement parameter 𝜂 ∗ = ( 𝐵 𝑅 )∕( ̇𝑀𝑣 ∞ ) ≥ .
55 × 10 , where 𝐵 eq = 0 . 𝐵 𝑝 (ud-Doula & Owocki, 2002). The impact of rotation is mea-sured by the parameter 𝑊 = Ω 𝑅 ∕ 𝑣 K = 0 . , where 𝑣 𝐾 = √ 𝐺𝑀 ∕ 𝑅 . Using equation (9) from ud-Doula et al. (2008), wethus arrive at 𝑅 𝐴 = (0 . 𝜂 ∗ + 0 . ) 𝑅 ∗ ≥ . 𝑅 ∗ forthe Alfvén radius and 𝑅 𝐾 = ( 𝐺𝑀 ∕Ω ) = 1 . 𝑅 ∗ for theKepler (corotation) radius.Petit et al. (2013) divided magnetic massive stars into twogroups, those with dynamical magnetospheres (DMs) with 𝑅 𝐴 < 𝑅 𝐾 and those possessing centrifugal magnetospheres(CMs) with 𝑅 𝐴 > 𝑅 𝐾 . Since for 𝜌 Oph A the radial extentof the magnetic confinement of the wind given by the Alfvénradius is much larger than the Kepler radius, material caughtin the region between 𝑅 𝐴 and 𝑅 𝐾 is centrifugally supportedagainst infall, and so builds up to a much denser CM.The rotational modulation of the X-ray emission with 𝑃 rot =1 . d detected by Pillitteri et al. (2017) possibly indicatesthat the wind plasma is predominantly locked to the magnetic field and that the magnetosphere of 𝜌 Oph A can be inter-preted in the context of the rigidly rotating magnetospheremodel (RRM; Townsend & Owocki, 2005). A few rapidlyrotating early-B type stars with RRM magnetospheres show-ing comparable rotation periods and magnetic field strengthswere discovered in the last years (e.g., Rivinius et al., 2013,Eikenberry et al., 2014). As is shown in Fig. 4 , the spectralappearance of 𝜌 Oph A in our FORS 2 observations is very sim-ilar to the spectral appearance of two other rapidly rotatingearly-B type stars, HD 23478 ( 𝑃 rot = 1 . d) and HD 345439( 𝑃 rot = 0 . d), for which the presence of RRMs was recentlydetected (Eikenberry et al., 2014). In all stars with RRMs, thecooler and denser postshock material trapped in the stellarmagnetospheres is typically detected in the H 𝛼 line or in thenear infrared hydrogen recombination lines. Since our FORS 2polarimetric spectra of 𝜌 Oph A do not cover the spectralregion containing the H 𝛼 line, it would be valuable to monitorthe variability of the H 𝛼 line profile in future observations toconfirm that circumstellar gas is locked to the magnetosphereand is in corotation with the stellar surface. In early-B type magnetic Bp stars, the global magnetic dipole-like field is tilted to the rotation axis by the angle 𝛽 and thesurface distribution of certain chemical elements, such as sili-con or carbon, displays a spotted structure, which is, as a rule,disjunct from the helium distribution. As the star rotates, weshould detect variations in the strength of the longitudinal mag-netic field and the intensities of spectral line profiles of variouselements with the rotation period of the star. Since our obser-vations correspond to two different rotational phases, whereopposite magnetic field polarities are detected, we have com-pared the line profile shapes on these two different epochs. Dueto the rather low FORS 2 spectral resolution, we checked thevariability only for hydrogen, helium, and silicon lines. Ourcomparison of the line profiles belonging to these elementspresented in Fig. 5 reveals distinct changes in the line inten-sities of the helium lines, supporting the assumption of thepresence of an inhomogeneous helium distribution on the stel-lar surface. We also detect that the H 𝛽 line intensity is lower inthe phase where the helium line intensities are stronger. On theother hand, silicon lines are very weak and noisy and do notpresent any obvious variability. To check variability on a timescale of a few minutes, we compared the stability of the lineprofiles belonging to these elements over the full sequencesof sub-exposures obtained on that time-scale. No short-termvariability was detected in both FORS 2 observations. . Hubrig ET AL N o r m . F l u x Wavelength (Å)
HD 23478HD 345439 ρ Oph A
BalmerSeries H − H − H ε − H e I − H δ − H e I − C II − H γ H e I − H e I − H e I − H β − H e I − H e I − H e I − FIGURE 4
The normalized FORS 2 Stokes 𝐼 spectrum of 𝜌 Oph A is displayed together with the normalized FORS 2 spectraof two other rapidly rotating early-B type stars, HD 23478 and HD 345439, for which the presence of a rigidly rotating mag-netosphere was recently detected (Eikenberry et al., 2014). Well known spectral lines are indicated. The spectra of HD 345439and 𝜌 Oph A were vertically offset for clarity. D i ff e r en c e I/I C D i ff e r en c e I/I C FIGURE 5
Left panel : Variability of the H 𝛽 and He I lines in the spectra of 𝜌 Oph A. On the top one can see the normalizedFORS 2 spectra obtained on two different nights. On the bottom, the difference between the two spectra is shown.
Right panel :No variability is detected for the Si
III lines at wavelengths 4553, 4567, and 4574 Å.
The analysis of FORS 2 spectropolarimetric observations of 𝜌 Oph A on two different rotation phases reveals the presenceof a rather strong magnetic field with a dipole strength of 𝐵 d ≥ . . kG. Using physical parameters of this B2 type star, wecalculated Kepler and Alfvén radii and concluded that a cen-trifugally supported, magnetically confined plasma is presentaround 𝜌 Oph A. Since the magnetic field of 𝜌 Oph A was mea-sured only on two occasions, it should be a prime candidate fora follow-up spectropolarimetric study that would lead to moreaccurate magnetospheric parameters.A comparison of line profiles on two different rotationphases shows a clear variability of helium lines similar to that observed in typical magnetic early-type Bp stars. The variabil-ity of the H 𝛽 line can probably be explained by the presenceof an extended magnetosphere around 𝜌 Oph A. Based on ourdetection of the presence of a magnetic field in 𝜌 Oph A, weconclude that the most likely reason for the variations ofthe X-ray emission observed by Pillitteri et al. (2017) is theoccultation of parts of the magnetosphere by the stellar body.The origin of magnetic fields in massive stars is still a majorunresolved problem in astrophysics. Only a small fraction ofstars (5–7%, e.g. Schöller et al., 2017) with radiative envelopespossess strong large-scale organized magnetic fields. Suchfields can probably be generated during the star formation pro-cess, by dynamo action taking place in the rotating stellarcores, or they could be products of a merger process. While
S. Hubrig
ET AL the first two scenarios are unable to explain a number of obser-vational phenomena (e.g. Ferrario et al., 2015), the magneticfields might form when two protostellar objects merge late dur-ing their evolution towards the main sequence and when atleast one of them has already acquired a radiative envelope(Ferrario et al., 2009). 𝜌 Oph A is not the only magnetic star detected in a com-plex star forming region. In 2013, using the High Accu-racy Radial velocity Planet Searcher polarimeter (HARPSpol;Snik et al., 2008) attached to ESO’s 3.6 m telescope (LaSilla, Chile) and FORS 2 observations, Hubrig et al. (2014b)searched for the presence of a magnetic field in the threemost massive central stars in the Trifid nebula, HD 164492A,HD 164492C, and HD 164492D. These observations indicatedthe presence of a strong longitudinal magnetic field of about500–600 G in the poorly studied component HD 164492C.Later, González et al. (2017) showed that HD 164492C is aspectroscopic triple system consisting of an eccentric closespectroscopic binary with a period of 12.5 d, and a massivefast rotating He-rich tertiary possessing a variable kG ordermagnetic field. Similar to 𝜌 Oph A, also in HD 164492C theX-ray emission was firmly detected using Chandra observa-tions (Rho et al., 2004). The detection of magnetic massivestars in the youngest star-forming regions implies that thesetargets may play a pivotal role in our understanding of theorigin of magnetic fields in massive stars. It is striking thatboth magnetic massive stars located in young star-formingregions show similar characteristics such as fast rotation andthe presence of X-ray emission. Also their stellar surfacesshow helium abundance variations typical for He-rich Bp starswith large-scale organized magnetic fields. Especially intrigu-ing is the presence of an extended blue reflection nebula aroundthe system 𝜌 Oph AB: A few years ago, Hubrig (2013a) dis-cussed the variability of the longitudinal magnetic field inthe O6.5f?p star HD 148937 (Hubrig et al., 2008, 2013b), sug-gesting that this target may provide a smoking gun, as itis surrounded by the 3000 yr old, nitrogen-rich bipolar neb-ula NGC 6164/5 (Leitherer & Chavarria-K., 1987), which waslikely created through strong binary interaction. In parallelwith HD 148937, we can speculate that 𝜌 Oph A can similarlybe a merger product and that the surrounding nebula is cre-ated by the ejected material. Obviously, it would be importantto investigate the chemical composition of the material of thenebula around 𝜌 Oph A to determine its origin. Furthermore,since star formation in molecular clouds is assumed to be trig-gered by the dynamical action of winds from massive stars,we need to understand how magnetized winds from magneticmassive stars formed during the first episodes of star forma-tion influence their environments, including nearby sites ofstar formation and protoplanetary disks surrounding low-masspre-main-sequence stars.
ACKNOWLEDGMENTS
Based on observations made with ESO Telescopes at the LaSilla Paranal Observatory under programme 099.D-0067(A).AK acknowledges financial support from RFBR grant 16-02-00604A.
REFERENCES
Abergel, A., Bernard, J. P., Boulanger, F., et al. 1996, A&A, 315, L329Angel, J. R. P., Landstreet, J. D. 1970, ApJ, 160, L147Appenzeller, I., Fricke, K., Fürtig, W., et al. 1998, The ESO Messen-ger, 94, 1Babel, J., Montmerle, T. 1997, A&A, 323, 121Böhm-Vitense, E. 1981, Ann. Rev. Astron. Astrophys., 19, 295Claret, A., Bloemen, S. 2011, A&A, 529, A75Eikenberry, S. S., Chojnowski, S. D., Wisniewski, J., et al. 2014,ApJL, 784, L30Ekström, S., Georgy, C., Eggenberger, P., et al. 2012, A&A, 537,A146Ferrario, L., Pringle, J. E., Tout, C. A., Wickramasinghe, D. T. 2009,MNRAS, 400, L71Ferrario, L., Melatos, A., Zrake, J. 2015, Space Sci. Rev., 191, 77Flower, P. J. 1996, ApJ, 469, 355González, J. F., Hubrig, S., Przybilla, N., et al. 2017, MNRAS, 467,437Howarth, I. D., Stevens, I. R. 2014, MNRAS, 445, 2878Hubrig, S., Kurtz, D. W., Bagnulo, S., et al. 2004a, A&A, 415, 661Hubrig, S., Szeifert, T., Schöller, M., et al. 2004b, A&A, 415, 685Hubrig, S., Schöller, M., Schnerr, R. S., et al. 2008, A&A, 490, 793Hubrig, S. 2013a, in: “Massive Stars: From 𝛼 to Ω ”, held 10–14 June2013 in Rhodes, Greece; online at http://a2omega-conference.net,id. 39Hubrig, S., Schöller, M., Ilyin, I., et al. 2013b, A&A, 551, A33Hubrig, S., Schöller, M., Kholtygin, A. F. 2014a, MNRAS, 440, 1779Hubrig, S., Fossati, L., Carroll, T. A., et al. 2014b, A&A, 564, L1Leitherer, C., Chavarria-K., C. 1987, A&A, 175, 208Maheswaran, M., Cassinelli, J. P. 2009, MNRAS, 394, 415Mamajek, E. E. 2008, AN, 329,10Motte, F., Andre, P., Neri, R. 1998, A&A, 336, 150Petit, V., Owocki, S. P., Wade, G. A., et al. 2013, MNRAS, 429, 398Pillitteri, I., Wolk, S. J., Goodman, A., Sciortino, S. 2014, A&A, 567,L4Pillitteri, I., Wolk, S. J., Chen, H. H., Goodman, A. 2016, A&A, 592,A88Pillitteri, I., Wolk, S. J., Reale, F., Oskinova, L. 2017, A&A, 602, A92Preston, G. W. 1967, ApJ, 150, 547Rho, J., Corcoran, M. F., Hamaguchi, K., Lefloch, B. 2004, ApJ, 607,904Rivinius, T., Townsend, R. H. D., Kochukhov, O., et al. 2013,MNRAS, 429, 177Schöller, M., Hubrig, S., Fossati, L., et al. 2017, A&A, 599, A66Skinner, S. L., Sokal, K. R., Cohen, D. H., et al. 2008, ApJ, 683, 796Snik, F., Jeffers, S., Keller, Ch., et al. 2008, Proc. SPIE, 7014, E22Stibbs, D. W. N. 1950, MNRAS, 110, 395Townsend, R. H. D., Owocki, S. P. 2005, MNRAS, 357, 251ud-Doula, A., Owocki, S. P. 2002, ApJ, 576, 413ud-Doula, A., Owocki, S. P., Townsend, R. H. D. 2008, MNRAS, 385,97van Leeuwen, F. 2008, VizieR Online Data Catalog: Hipparcos, theNew Reduction, Cat. 1311 . Hubrig ET AL7