samedi 24 juillet 2010

Active OB stars

In Paris took place in July 2010 a key IAU (International Astronomical Union) symposium - IAU 272 - on active OB stars which includes several types of hot stars and Be stars. I participated to this meeting and I propose here a brief summary of what has been discussed.




In very high level summary, active OB stars are O-type or B-type stars which are active somehow: emission line spectrum, magnetism, variability... The term “Active B Stars” was proposed by Myron Smith in 1994, when he was elected Chair of the IAU Working Group on Be Stars. Active is equivalent of intrinsically variable. It was interesting to note at the end of the sympsosium that it was actually not easy to find non variable ("normal") OB stars with the increased level of the instrumentation available!

The list of stars which are are active OB stars is:
* B[e]: B-type stars whose spectra show forbidden emission lines due to very low density area (dust) usually further away from the stars. B[e] are classified in 5 sub-groups:
    -sgB[e]: supergiants B[e];
    -pmsB[e]: pre-main sequence B[e], similar to Herbig Ae/Be or T Tauri for cooler stars;
    -cPNB[e]: compact Planetary Nebulae B[e]
    -symbiotic B[e] where binary system fill up their Roche lobe (and are in 'symbiose')
    -unclB[e]: unclassified B[e] for other types
    -Another classification was proposed with the creation of a homogeneous FS CMa group with about 47 members and 20 additional candidates where dust formation has stopped some time ago. They maybe all binary systems (visible with Li 6708 line) or some are single stars with unusually strong winds (unlikely?). Further observations are required for this specific group; * Be: B-type stars, non supergiants, whose spectra have shown or still show optical (Balmer) emission lines. About 20% B stars are Be starsin our galaxy with higher proportion in magellan clouds;
* Bn: B-type stars displaying 'nebulous' spectrum but not in emission, usually very fast edge-on rotators;
* He-variable: early-type stars (
* Bp: B-type (>~B7-type) stars peculiar (spectra are showing unusual metallic lines) with kiloGauss magnetic field and rare-earth spots;
* SPB: Slow Pulsating B0-B3 type stars with periods between 2 to 12 hours
* beta Cep: B2-B9 type stars with periods between 0.5 and 3 days
* Herbig Ae/Be: pre main sequence stars (same as T Tauri for cooler stars)
Supergiants: post main sequence stars including:
* LBV: Luminous Blue Variables
* WR: Wolf-Rayet

On top of this classification, we can also distinguish between single stars and binaries. About 45% of active OB stars are binaries.


This symposium overall showed the increasing interest for active OB stars as several mechanisms are involved on those stars whose impact would be more or less significant depending on the star type – knowing that any star, specially Be stars, is a special case by itself!

The phenomena involved benefited from recent progress in technology, specially:
* spectropolarimetry: great tool to look for magnetism and model magnetic field around those stars. Very famous spectropolarimeters are ESPaDOns at CFHT and NARVAL at TBL telescopes;
* asteroseismology: study of pulsations at the surface of the stars as probes for internal structures. This is supported by earth-based and satellites (MOST, CoRoT, Kepler) observations – a fruitful data acquisition;
* interferometry: direct high-resolution imaging using mutiple telescopes (30-300m base usually). Astronomers can now show direct images of resolved stars (elongation), material surrounding the stars (disk/ring, polar jets), binarity (we are still impressed by the direct animation of the beta Lyrae couple moving around their center of masses)...;
* gamma Rays: some active OB stars have compact companions (pulsars, black hole) and the interaction provoques the emission of very (very!) energetic gamma rays;
* modeling: theorists have stellar models which begin to integrate more and more physical phenomena and improve their tool by continnuous matching with actual observation – we really saw at this symposium Science at work!


So... what are those phenomena? Well, of course first the basic stellar physics with hydrogen burning then helium burning cores and CNO cycles inside the stars. In those hot and massive stars, the interior is radiative (while it is mainly convective in cool stars and a mix in stars such as our Sun). The easy way to describe it is to use a Local Thermodynamical Equilibrum (LTE) model but it doesn't work and theorists have to develop non LTE – much more complex – code. This is seen at amateur level when using Spctrum code in VisualSpec to simulae a star spectrum: it doesn't work for stars earlier (ie: hotter) than B5-type where non-LTE code has to be used.
[update 2017] For O or B stars, grid can be found on Tlusty. There are pre-calculated grids of models (OSTAR2002 and BSTAR2006) that can be downloaded from the author's website. These grids should be more than enough for our needs. Start with the file BGvispec_v2.tar available at http://nova.astro.umd.edu/Tlusty2002/tlusty-frames-BS06.html


Massive Main Sequence (MS) stars are those which have passed the Pre Main Sequence (PMS) phase and the Zero Age Main Sequence (ZAMS) time; but have not yet reached the Post Main Sequence (PMS) phase and the Terminal Age Main Sequence time (TAMS)!


Several active OB stars are fast rotators. They rotate within hours/days at a very high velocity at their equator. In the 1930's, Otto Struve explained the Be star phenomenon with this very high rotation. From the measure, their rotation is lower than the critical velocity, the breaking point. But maybe the measures are underestimated for exemple with a new "gravity darkening effect"...

About 20 years ago, Jean-Paul Zhan developped a theoritical model which combined the meridional circulation and the shear turbulence in such stars. Meridional circulation is a phenomenon linked to the fast rotation creating temperature differences which then provoque circulation (like circular wave) inside the stars. Shear turbulence are movements between star layers. Fast rotation is increasing a "mixing" effect which bring more metallic elements to the external layers of the stars which then appear more abundant (higher metalicity Z).

In a more general way, rotation is impacting a lot of the star properties (luminosity, gravity, abundance, mass-loss rates...) and high resolution spectra are required to increase the details of observations and the improvement of the models.

This rotation is sometimes slown down by a magnetic braking effect, visible by looking at the nitrogen to carbon Versus nitrogen to oxygen abundance. Magnetic field is supposed to be fossil (ie: present when the star was formed) but a dynamo effect inside the convective core was also discussed and this is a key question for the (near?) future.


 Magnetism in general was lengthly discussed during this IAUS 272 symposium. About 15% of the Massive Main Sequence stars show magnetic field. We know magnetism for some times for exemple with the magnetic field around the Earth. Our star show some magnetism activity well visible with Sun spots. But there is a difference between cool (like our Sun) and hot stars in terms of magnetism:
* cool stars with convective envelopes display fields ubiquitous, which well correlate with stellar properties. Magnetic fields are weak, local, structured, aligned, equipartition, and variable (11 years period for our Sun for exemple). They come from contemporaneous dynamos effect;
* hot stars with radiative envlopes display rare magnetic fields with weak/absent correlation with stellar properties. The field when present can be very strong (kiloGauss!), global, misaligned and static. It is still unclear if it is fossil / remnant field or not (dynamo).

Some magnetic stars have been highlighted during the symposium such as sig Ori E, a long-term steady magnetic bipole oblique rotator (P=1.19 days) with anomalous surface abundance (due to rotational mixing?). Magnetic Doppler imaging (MDI) technics have been used to model the star magnetic field in 3D. Another star of this type (HR7355) was also presented.

If an observation of a Be star has shown a magnetic field, later observation didn't show it and no other Be star has displayed a magnetic field with current instrumentation threshholds. Beta Cep, which is also classified as a Be star, seems in fact a binary system with a true beta Cep star and a Be star as companion.

MiMeS is a project to dedicate lot of spectropolarimeter observing time (640 hours at CFHT and TBL telescopes, about 3 nights per month during 6+ years). It will observe some specific/known targets and look for new one in a survey mode. Professionnal astronomers expect to learn a lot from this project about magnetism in active OB stars – a field which is still to be investigated in more details.


A key phenomenon around those hot stars are stellar winds which are structured on a variety of spatial scales. Those winds are visible in our spectra for exemple with P Cygni profiles – typical of an expanding envelop around a large star. Modern models use more complex shapes (spirals for exemple) with Narrow Absorption components (NAC), Discrete Absorption Components (DAC) and Periodic Absorption Modulations (PAM).

Stellar winds seem also to be clumpy, maybe due to meridional circulation. This clumpiness has been discussed several times during the symposium and seems to be an improvement of both observations and theories/models. Are those winds spherical, with small scale structure or with large scale structure? It seems they are all of this together!

A big question, specially for theorists as this is a key parameter for their model, is the mass loss rate. Different methods used to determine this parameter give different results!

Mass loss record is hold by eta Car: a LBV. Luminous Blue Variables seem to be the latest stage after the Wolf-Rayet stage and before the SuperNova stage. A massive star (above about 22 solar mass), near the end of its life, will spend about half million years in Wolf Rayet state and about 25000 years (but it could be more, still in debate) in LBV state. It will eject material in violent outburst – P Cygni latest one was 400 years ago; eta Car every 75 years approximately. Expansion winds are about 32000 km/sec for eta Car. LBV mass loss rates are between 10-3 and 10-6 solar mass per year! It's not a wind anymore... but a storm! Result gives very nice bipolar, hourglass type nebulae.


During last few years, tremendous progress have been made in studying stellar pulsations. Asteroseismology consists in observing the waves at the photosphere of the stars due to internal wave propagation. The single and easiest wave is a radial pulsation such as for beta Cep or Slow Pulsating B-type (SPB) stars. But astronomers have found more complex pulsations on the stars (our Sun is showing hundred of pulsating modes). SPB stars (B2-B9 main sequence stars) have gravity pulsation mode (wave inside the star, called g-mode) with 0.3-5 days period and less than 20mmag variations and Radial Velocities below 10km/s usualy; they display kappa-mechanism with Z-bumps. Very close to those are beta Cep stars (B0-B3 main sequence stars) with g-mode and pulsations due to pressure near stellar surface  (p-mode); they vary in 2-12 hours with light variation below 40mmag and RV below 20km/s; they are also due to kappa-mechanism (Z-bumps).

In spectroscopy, for exemple with MuSiCoS multi-site campaigns, the detailed line profile changes due to Non Radial Pulsations (NRP) have shown some pulsating mode in different types of active OB stars. Higher resolution spectrographs such as ESPADONS (CFHT) and NARVAL (TBL, 2 meter telescope at Pic du Midi observatory) have recently improved the power of resolution. In addition, astronomers can observe in spectropolarimetry mode the magnetic fields too.

Those NRP are also visible in photometry by looking at frequencies of brightness variations and by looking at the power spectrum of the light curve. Few years ago, astronomers were able to predict an outburst on µ Cen star by looking at the NRP of the star. But it has not been reproduced for other stars.

Satellites such as MOST (first canadian space telescope), CoRoT and Kepler has provided tons of data still beeing analyzed. With a real jump in photometric accuracy, astronomers can now 'see' more pulsating modes and are making some nice discovery. This includes HD49330 observed with CoRoT and a combination of g-mode (gravity, waves around the core not reaching the surface) and p-mode (pressure, waves below but reaching the surface). A 0.03mag outburst was also detected during the observing run and some spectroscopic data were taken from the ground (including few amateur observations). The open question is to understand if the pulsations are creating the outburst or if, provoqued by another mechanism, outburst is disturbing the star and creating pulsations... Extension of CoRoT observing time and continuous spectroscopic survey where amateur can play an important role will hopefully provide more information on Be stars in the near future.


What best than actually 'look' directly at a star. This is what interferometry is now bringing with higher and higher resolution direct images of stars and surrounding environnement. Achernar star direct imaging with interferometry have shown unusually ellongated star due to its rotation with a equator/pole ratio above 1.5 which was supposed to be the limit – but nature is more complex than our models! More recent images show polar jets but no equatorial disk; a comanion has been recently identified from VLT/VISIR observations. A gravity darkening effect (beta coefficient... a greek letter used for lot of other parameters and we could easily get lost by all those 'beta' in equations!) was introduced to explain some physical phenomenon observed.

Some stars show very nice equatorial disk on the interferometric images: gamma Cas, phi Per, zeta Tau, kappa Dra, chi Oph, 48 Per, psi Per, kappa CMa (non keplerian disk while most Be stars disk are keplerian), kappa Dra, beta Psc, nu Cyg, alpha Ara, HD62623 (supergiant B[e]), beta Cmi... Interferometry associated with spectroscopy provide information on inclination then stellar masses. The future will provide more astonishing images such as an animation shown of beta Lyrae with the two stars clearly visible during their rotation.


It was also interesting to see during the symposium to have a portion dedicated to (very) high energy observations. Some systems provoque gamma rays above 100 GeV! One object (HESS J0632°057) seem to be a long term binary compact object associated with a Be. PSR B1259-63 is a Be and a pulsar couple. LSI+613003 also or with a black hole. LS5039 is a O star with a neutron star or a black hole. Cyg X-3 is a Wolf-Rayet with a neutron star or a black hole. Cyg X-1 is a O star with a black hole. Those are extreme systems that can be used as extreme conditions test beds.

In general, this symposium was a great moment for theorists and observers to combine their work to improve their understanding of stellar evolution and mechanisms involved for exemple during mass loss events.


Be stars are really at the cross-road of many things – they have been mentionned many times during the workshop. Their critical rotation play a role for their evolution, mass-loss mechanisms are key to understand, they are sometimes at the frontier or limit (maximal rotational velocities, Omega/Gamma limit). Circum stellar disk could have a veiling effect on our observations. Be phenomenon is intermittent (30%-50% stars are still missing as they have not shown emission to us... yet). Line are sometimes saturated. Pulsations seems a property depending on metallicity. They have weak magnetic field, consistent with star formation. Their mass loss seems to be mainly through outburst and not a continuous process; but equatorial versus polar ejections were still debated. In short, the overall Be phenomenon remains a mystery!


Amateur contribution to active OB stars were shown in multiple talks and posters during the symposium. BeSS (Be Star Spectra) database was mentionned several times and about 16% of the posters could have had an amateur contribution. Several had some contribution direct (Rigel/Deneb, P Cygni, WR140, MONS campaign...) or through BeSS database.

More observations from amateurs are requested. Several professional were asking how to contact amateurs and Spectro-L (http://groups.yahoo.com/group/spectro-l/) was the best answer. Be stars community benefits also from a structured spectra collection, validation, and archiving process (ARASBeAm and BeSS). Not only Be stars but also B[e], Bp, beta Cep, SPB, supergiants B (Deneb, Rigel...), LBV post main sequence stars (ex: eta Car, P Cyg...), Herbig Ae/Be pre-main sequence stars (AB Aur...), etc...

A key target was specially mentionned: delta Sco. This binary system will reach periastron during spring-summer 2011; at this time, the companion will hit the disk. A date of July 4th was mentionned but prediction are still uncertain. Periastron passage will be very short – couple of weeks during which the Radial Velocity of the Be star will switch from 20km/s to -60km/s. This is well in reach of current spectrographs. But the RV shift will start very quickly; continuous and very acurate observations are required starting in march 2011 to alert the community when the RV starts to fall down (alert within few km/s change!). With echelle spectrographs, this can be measured on He II 4550 line for exemple (He II 4686 line may be in emission). H-alpha can be used but will be less accurate; still, it will be very interesting to follow up on H-alpha line during the full periastron passage (starting at least one month before the periastron). This is a key project for amateurs to contribute to!


The symposium was overall a great success and as several people mentionned the best that ever occured! It was well organized thanks to the Local Organizing Committee and scientifically extremely useful thanks to the Scientific Organizing Committee led by Coralie Neiner. She was highly efficient in leading this group effort and everyone left IAUS272 with the feeling of a great, successful and fruitful meeting...


Olivier Thizy
24-july-2010

Note: this article summarizes my amateur understanding of what was discussed during this 5 days meeting – of course all mistake done will be mine!


References:
-IAUS 272 website: http://iaus272.obspm.fr/
-IAUS 272 proceedings which should be distributed before end of 2010
-Spectro-L: http://groups.yahoo.com/group/spectro-l/
-ARASBeAm: http://arasbeam.free.fr/
-BeSS: http://basebe.obspm.fr/basebe/Accueil.php