P CYGNI - 2000

400 Years of Progress

Armagh, 21 - 23 August, 2000

A workshop to discuss the progress towards our understanding of this peculiar hypergiant made since 1600 AD, was held in Armagh, Northern Ireland, from 21-23 August, 2000.

Scope of the workshop

P Cygni was discovered on 18th August, 1600, while in outburst. Subsequent observations of the star have now covered a time span of four centuries. Especially the more frequent and accurate observations of the 20th century have revealed much about the nature of this enigmatic object. The 400th anniversary of the discovery of P Cygni was therefore the right time to discuss the progress made towards understanding this peculiar hypergiant, to ensure that the lessons learned from its study are not lost, and that these results can be applied to the study and understanding of similar objects and of the nature and evolution of massive stars in general.

Workshop format

The workshop programme consisted of a series of invited papers about different aspects of the research concerning P Cygni, supplemented with a number of contributed papers and posters, including several about objects in one way or other similar to P Cygni - notably Eta Carinae - and ample time for discussing how the various results could best be integrated into a more complete model for the astrophysical parameters and evolution of this star.

Workshop Proceedings

The Workshop Proceedings have been published by the Astronomical Society of the Pacific in its Conference Series, Volume 233, edited by Chris Sterken, Brussels, and Mart de Groot, Armagh. A copy of the book's cover and first few pages, including the table of contents, can be obtained on request from mdg@star.arm.ac.uk. A copy of the Proceedings at a reduced price can be obtained from Mart de Groot upon payment of GBP 30.00 by UK cheque or credit card, or USD 45.00 by credit card. If you pay by credit card, please, note that we can accept VISA, MasterCard and Switch only. To order, send an e-mail to mdg@star.arm.ac.uk with the following information: "Please, send me .... (fill in number) copy/copies of P CYGNI 2000 at the sum of GBP 35.00/USD 45.00 each which includes postage and packing. My VISA/MasterCard/Switch card number is .... .... .... .... ; issue No. .. (for Switch only); name of cardholder ...............; expiry date ../.. My postal address is: ......................................................"


The following abstracts are the ones that were submitted before the beginning of the Workshop. Abstracts actually appearing in the Proceedings may vary, although, in general they give a fair account of the contents of the papers presented. For more detailed information one should contact the authors.

Photometry of P Cygni and similar objects:


M. de Groot, C. Sterken, A.M. van Genderen; Armagh, Brussels, Leiden

P Cygni's photometric behaviour since its discovery outburst in 1600 will be reviewed and conclusions about the state of the star's atmosphere drawn. The 17th-century outbursts may provide clues to the star's early mass loss. Photometric monitoring since 1985, supplemented with archived data going back to 1907, show a rather stable pulsation period of ~ 17.3 days. Other, longer, cycles of variation, up to ~ 4 years, are also found. The nature of these variations is discussed in the framework of the various S Dor phases and microvariations identified in our earlier work on other S Dor variables (LBVs).



John R. Percy and Trevor D.K. Evans, Department of Astronomy, University of Toronto, Toronto ON Canada M5S 3H8

Gregory W. Henry, Center for Excellence in Information Systems, Tennessee State University, Nashville TN 37203-3401 USA

We report two decades of photometric observations of P Cygni from a robotic telescope (VRI: JD 2446400-2451500), from the American Association of Variable Star Observers (AAVSO) photoelectric photometry program (V: 2446200-2451500), and from the AAVSO visual program (mv: JD 2443500-2451000). We have analyzed these observations using light curves, Fourier analysis, and autocorrelation analysis. Light curves of the two photoelectric datasets show variations of up to 0.1 magnitude on time scales of years, months, weeks, and as short as 10 days. Fourier analysis and autocorrelation analysis show that the variations are not periodic, but there are characteristic time scales of 50-200 days and also around 400 days. The visual data, when corrected for a small seasonal effect, are capable of revealing variations of 0.1 magnitude on time scales of years to decades. The time scales and amplitudes which we have observed are in accord with previous results. Our data provide a useful comparison with data obtained spectroscopically, and at other wavelengths.

Acknowledgements: NASA, NSF, NSERC (Canada), the AAVSO visual and photoelectric photometry programs, and Lou Boyd (Fairborn Observatory).



N. Markova, National Observatory, Bulgaria

An extensive study of quasi-simultaneous spectroscopic and photometric data of P Cygni revealed the presence of continuous wind and photospheric activity in this star. Wind variability is diagnosed through variations in the profiles of a large number of optical lines while photospheric variability is detected through changes in the stellar brigthness and colour indices.
Four different kinds of line-profile variability (lpv) were established: i) Very Long Term (VLT) variability traced by variations in the Halpha emission-line strength; ii) Long-Term (LT) variability diagnosed through variations in velocity of the absorption cores and intensity of the emission peaks of HI and HeI lines of relatively large optical depth; iii) systematic variability due to propagating Discrete Absorption Components (DACs), and iv) ``swaying'' variability consisting of continuous modulations in velocity and strength of lines of intermediate and high excitation.
The ``swaying'' variability makes up the fastest pattern of P Cygni's wind variability (time-scale of four to five monts) and affects all layers of the supersonic wind starting at its base up to layers where the Halpha line forms. The DAC-induced variability repeats on a time-scale of 200 days and operates in a relatively outer part of the wind. The LT and the VLT variability are both incident to only the outermost part of the wind with the former being considerably faster - time-scale of about 2 years - than the latter (time-scale of more than 7 years). No indication for any clear relation between different kinds of lpv was found. Even when the variations operate in one and the same region of the wind (in velocity space) it is not obvious whether and how they interact.
The analysis showed that these lpv are not just due to the redistribution of a fixed amount of line absorption, as would occur if, e.g., the variability was caused by a macroscopic velocity field alone, but reflects variations in the density and temperature structure of the wind as well. Stellar rotation does not seem to play a fundamental role in setting the time-scale of the VLT-, the LT- or the DAC-induced variability either. The relationship between the ``swaying'' variability and the rotation is still not clear but it may well be that this variability is rotationally modulated.
Photometric analysis reveals the presence of a Very Long-Term (VLT) brightness variation with an amplitude of about 0.2 mag and a duration of about 7.3 years. During this variation, i.e. when the star brightens, the effective temperature decreases and the radius increases, thus suggesting that P Cygni has more likely experienced a normal S Dor-type variation with a maximum phase in 1992 and a minimum in 1996. This suggestion is supported by the behaviour of the Halpha luminosity, which showed systematically higher values (resp. mass-loss rate) in 1992 than in 1996. Superimposed on the VLT variation we also observed a Short-Term (ST) brigthness variation with an amplitude between 0.1 and 0.2 mag which seems to recur on a time scale of three to four months. In its colour behaviour (redder in B-V and bluer in U-B, bluer when the star brightens in V) this variation resembles the so-called "100 d-type micro-variations", recognized in other LBVs by van Genderen et al. (1997a,b). Non-radial pulsations (NRPs), either of g-mode or of s-mode oscillations are a possible cause for this variability.
Evidence for a close relationship between line-profile variations and changes in the stellar brightness and temperature was found, implying that P Cygni's wind variability is likely coupled to processes in the stellar photosphere.



C. Sterken, L.M. Freyhammer, T. Arentoft, Brussels
A.M. van Genderen, Leiden

We outline the major problems which photometrists face when performing long-term monitoring of a composite and extended object like Eta Car, viz.:

* the interruptions by seasonal gaps

* annually recurrent, high air masses which induce very significant colour errors, especially in broad-bandpass measurements

* non-availability of a suitable photometric measuring instrument (a modest-size telescope with suitable photometric instrumentation)

* incompatible photometric filter systems

* presence of variable and strong emission lines in the spectrum of Eta Car

* stellar photometers with diaphragms of different sizes, and aperture photometry of CCD frames



C. Sterken, L.M. Freyhammer, T. Arentoft, A. Jones, W. Liller, A.M. van Genderen, E. Brogt, M. de Wildt, R. Dijkstra, M. Janson, A. van der Meer, C. Ponsioen, T. Voskes, F. Marang, G. Roberts, F. van Wyk, D. Kilkenny

We present the actual state of the light- and colour variations of Eta Car as derived from our ground-based photometric monitoring program.


Physical Insights Derived From 20th Century Photometric Monitoring

A.M. van Genderen (Leiden), C. Sterken (Brussels), M. de Groot (Armagh)

This is a review of the results obtained from studies into the physical nature of the central star(s). It is based on photometric observations beginning with the nice fit to all 1935-2000 photometry of the predicted secular increase in brightness due to the thinning of the Homunculus' expansion and the consequent decrease in self-extinction of the Homunculus and circumstellar matter (van Genderen, de Groot, The 1994), and ending with a model featuring a luminous accretion disk.
Most of the brightness variations are undoubtedly due to an S Dor-type variable (LBV) inside Eta Carinae. Additional variations of up to 0.15 mag in the near-UV seem to be of non-stellar origin. The most likely explanation for these variations is a luminous accretion disk around a nearby companion star. Boundary limits for a model of the binary and the disk, based on energy-balance arguments, will be presented.



Roberta M. Humphreys and Nathan Smith, University of Minnesota

We present recent optical and near-infrared photometry of Variable 12 (SN 1954j) in NGC 2403, a possible Eta Car-type variable. Prior to its eruption in 1954, V12 exhibited remarkable and erratic oscillations. It rapidly faded in 1955 and there are apparently no observations of it since the late 1950s.



M. Corcoran, K. Ishibashi, J. Swank, R. Petre, Goddard SFC

Using the ground-based orbital parameters as a starting point, we use the observed X-ray lightcurve of the extreme object Eta Carinae to refine the orbit. We show that the X-ray data require that the time of periastron passage occurs on 1997.9, not 1998.13 as derived from the ground-based radial velocities. This helps resolve a discrepancy between the ground-based radial velocities and spatially-resolved velocity measures obtained by STIS. The X-ray data also suggest that the mass function f(M) ~ 1.5, so that the masses of the two stars are about 80 and 30 Msun. The extended width of the X-ray minimum cannot be fitted with a standard colliding-wind model in which the wind is spherically symmetric and the mass-loss rate is constant. However, we show that we can get a very satisfactory fit to the observed X-ray minimum if we assume that there is an increase of a factor of ~50 in the mass-loss rate from the primary for a short interval following periastron. If real, this excess in mass-loss rate could be due to tidally-enhanced mass flow off the primary when the two stars are close. If this model is correct, then it is likely that the X-ray spectra measured by ASCA and RXTE near the X-ray minimum are contaminated by unresolved hard emission (E >= 2 keV) from some other nearby source, perhaps associated with fast shocks near the Homunculus. Based on the observed X-ray fluxes, the distance to Eta Car is 2300 pc.


Spectroscopy of P Cygni (and similar objects):


I. Kolka, Tartu Observatory, Estonia

The observed variability of P Cygni at different wavelength regions will be reviewed. Starting with optical spectroscopy, we point to a sample of diverse characteristic time-scales (from 15 to 600 days) extracted by different authors from this type of data.
We show the consistency of IUE spectroscopic data of P Cygni with the results in the optical region. The time-scales found in photometric, polarimetric and radio flux time series of P Cygni will be compared with spectroscopic ones.
It is stressed that the obvious time delay between variability cycles of many measurable quantities of P Cygni makes it difficult to search for reliable correlations among them. The most plausible explanation of the variability should be the changes in the stellar wind structure of P Cygni - both at the base of the wind and in the outer wind regions. Structured (non-smooth) wind models for P Cygni are not yet available. There is an obvious need for such models to explain both the persistent spectroscopic features and the different kinds of variability patterns.



C. Rossi, G. Muratorio, R. Viotti, the SARG-TNG team

We describe thhe high-resolution (R = 50 000) spectrum of P Cyg near 1 micron, observed with AURELIE at OHP in July 1998. The main spectral features are analysed, with special emphasis on the fluorescence Fe II 999.7 nm emission line. The appearance of this line in early-type stars is discussed. We also present the preliminary results of the very-high-resolution (R = 86 000) optical spectrum of P Cyg obtained as a first-light observation of the new SARG spectrograph attached to the Italian Galileo National Telescope



F. Najarro, Madrid

We present a full UV to IR quatitative spectroscopic analysis of P Cygni using a new generation of line-blanketed models for extended atmospheres with stellar winds. The observational constraints are given by averaged high-resolution IUE SWS-LWP spectra, high-resolution optical and near-IR spectra (4600 - 8900 A), and ISO-SWS IR spectra (2.38 - 45.2 micron). Our models show excellent agreement with the observations at all wavelength regions. We confirm results from previous He-abundance studies and present first results on abundance determinations for C, N, O, Mg, Al, Si and Fe.



Nathan Smith et al., Minnesota

We present a collection of recent ground-based infrared observations of P Cygni and its nebula, including near-infrared spectroscopy, near-infrared imaging, and 1 to 20 micron photometry.



T. Nugis, Tartu Observatory, Estonia

Radio emission of P Cygni is known to be variable and this is usually explained as the variation of the ionization fraction of hydrogen far in the wind. We modelled the formation of radio fluxes in the clumped wind of P Cygni and looked for the mechanisms which can lead to the variations of radio fluxes. The mass-loss rate of P Cygni as determined from radio-emission power was found to lie in the range 1.5-2 x 10^{-5} Msun/yr.



Th. Rivinius, ESO, and the Heidelberg Hot Star Group

The spectral variability of OBA super- and hypergiants will be reviewed, based on the monitorings conducted by the Landesternwarte Heidelberg Hot Star Group. These data sets of a sample of nearly two dozen stars cover typically a couple of months in the spectral range from 4000 to 6700 Angstrom with a resolving power of 20,000. Data sets taken after 1994 extend to the Balmer and Paschen discontinuities. Using these data sets, a scheme of how the variability changes as a function of the position in the Hertzsprung-Russell-Diagram (HRD) will be proposed. For some well-studied cases of early B hypergiants correlations between the simultaneously observed photospheric and wind variability will be reviewed, leading to an acceleration curve of the wind base. This might be connected to the evolution of discrete wind components seen at higher velocities. Part of the variability of the photosphere might also be attributed to stellar pulsation. For other stars of later spectral type a co-rotation hypothesis will be shown to be a more promising explanation for the observed variability than the acceleration of wind features. The spectral variability of P Cygni will be compared to these results from stars in neughbouring regions of the HRD.



A. Damineli and C. Barbosa, Univ. of Sao Paulo

We present high-resolution spectra of the star P Cygni and several other related stars. H-band spectra (R~10000) display the HeI line at 1.700 mu, Bracket-11 and [FeII]. The spectra around the HeI 1.083 mu line (R~15000) are of particular interest because this line samples structures in the stellar wind. Temporal variability in the HeI 1.083 line of eta Car and AG Car are shown, for high and low excitation phases.



S. Tubbesing, A. Kaufer, O. Stahl, B. Wolf, Heidelberg

R81 (B2.5Iab:e) of the LMC is the only known P Cygni type eclipsing binary. During our guaranteed time with the new Fiber fed Extended Range Optical Spectrograph (FEROS: spectral range 3700 -- 9200 A; resolution = 48000) at the ESO 1.5m telescope we obtained a spectroscopic time series covering the whole period of 74.55d with a S/N ~ 100 and a time resolution of typically one spectrum per night. Simultaneously, we obtained the lightcurve of R81 at the Danish 0.5m telescope with a total phase coverage of typically one measurement per night. Due to the large spectral coverage of FEROS a huge amount of strategic lines could be indentified which allow to trace the phase-locked photospheric, wind and envelope behaviour of R81. We present a first overview of the photometric and spectroscopic data. Furthermore, we present preliminary orbital parameters from our first analyses of the radial-velocity curve.



C. de Jager, A. Lobel, G. Israelian, H. Nieuwenhuijzen, Utrecht, Harvard, Tenerife

As long as the yellow hypergiant HR8752 has been observed spectroscopically it has shown erratic and conspicuous fluctuations in its effective temperature. But an impressive and hitherto never observed rise in its temperature started around 1988. Since that time T_eff rose from 4600 K to 9700 K. Regular further observations are needed to see if and when this rise will stop, and what will happen thereafter. The instability is related to the atmospheric compressibility. The atmospheric average value of Gamma_1 decreased in the mid-eighties to below 4/3 and it has decreased further since that time. The atmosphere apparently became dynamically unstable. Since 1998 the effective acceleration in the upper atmosphere became negative, which stimulates outward motions. We hypothesise that HR8752 will traverse the Yellow Evolutionary Void and thereafter, entering the second region of dynamic instability, may become a star like P Cygni.



V.F. Polcaro, R. Viotti et al., Rome

It is well known that the Sumeric civilization was the first to start a systematic study of the sky in the third millennium BC. Most of the Western "classical" astronomy, as well as possible Chinese astronomy, derives from these early studies that were transmitted, first to subsequent Mesopotamic civilizations, and then to the Greeks through the Hellenistic Kingdoms. It is thus not surprising that most of the Greek constellations are strictly linked with the Mesopotamic ones. This rule fails in just a few cases, the most significant one being the Ursa Major constellation. In the Sumeric sky, this asterism is lacking and it is substituted by two other ones, the first made by the brightest stars of Ursa Major, and the second centered around (and named from) the star now named 80 UMa. This is quite strange, since this star is not too bright (V = 4.0) and is not now the brightest one of the sky field covered by the Sumeric constellation. A possible explanation is that, in the Sumer era, 80 UMa was much brighter than it is now. Actually, 80 UMa is classified as a variable star, but its V variability is reported to be only of 0.1 mag. On the other hand, it is an IR source and it is surrounded by a significant amount of circumstellar material. We thus explored the possibility that this star underwent a huge outburst arond 5000 years ago, similar to that experienced by P Cyg in the XVII century and by Eta Car in the past century, by using, on the one hand, the available spectroscopic and photometric data of 80 UMa and, on the other hand, the analysis of the Mesopotamic myths linked to this asterism. We argue that, despite 80 UMa's classification as a main-sequence (A5V) spectroscopic binary, we cannot rule out the possibility that it is actually a more exotic object.



Th. Gaeng, NASA/GSFC

We present and discuss spectra of P Cygni and similar objects (i.e. LBVs and LBV-candidates) with high resolution (resolving power = 50,000 - 100,000) and high signal-to-noise ratios (S/N = 100 - 1000, depending on wavelength region and object). The spectra cover the optical ranges of 4320-6602, 3600-6800 and 3920-7030 Angstroem.
Like other authors who created spectral catalogues, we averaged multiple spectra to obtain the best possible S/N ratio. In contrast to most recent spectral atlases of LBVs (see e.g. for P Cygni: Stahl et al. 1993, Markova 1994), the individual spectra in our observations were taken over a relatively short time frame (i.e. within a few hours) so that no smearing of spectral features contaminated the final averaged spectra; LBV variablity time scales can be as short as a few days but are not on the order of minutes to hours so that our observations are not influenced by changing stellar atmospheric conditions. (poster in absentia)


Observations of P Cygni's (and similar) nebula :


J. Meaburn et al., Manchester

A series of optical observations have been made over several years with the Manchester Occulting Mask Imager and Manchester Echelle Spectrometer of the environs of P Cygni. Inner and outer shells and a 7-arcmin long giant lobe were discovered.
Longslit spectra and an image in the light of the fluorescently excited N I line have revealed the knottiness of the inner shell. Well-defined `velocity ellipses' for the position/velocity arrays across the outer shell indicate that its expansion velocity is 160 km/s which is somewhat lower than the previously reported value.
The kinematical behaviour of the giant lobe, both in the vicinity of P Cygni, and further away, emphasizes its strange nature, though close association with P Cygni is again suggested by these new observations.
A variety of possible explanations of the observed phenomena are explored e.g. radio knots in the outer stellar wind either coalesce or expand to become the N II emitting knots in the inner shell; that a mass-loaded wind overunning slower moving clumps explains the kinematics of the inner shell; that the one-sided giant lobe could be a `trail' of ejected material from P Cygni funnelled through the outer shell.



O. Chesneau, F. Vakili, L. Abe, M. Roche, C. Aime, H. Lanteri, Cote d'Azur Obs./Univ. of Montreal

In contrast to other LBVs, the P Cygni nebula is rather faint and precludes any observation of the closest regions by classical imaging. We report the first observation of the inner wind of P Cygni in H-alpha using experimental adaptive optics at the 1.52 m telescope of Observatoire de Haute Provence in 1997. We emphasize that the observations were done without using any coronagraphic mask by means of short exposures provided by a photon-counting camera. A deconvolution with a nearby reference star reveals a few 50-100 mas strongly-emissive clumps whose structure qualitatively corresponds to images obtained by large baseline radio interferometry, around an unresolved core. The closest clumps could have been observed by means of the GI2T stellar interferometer 3 years before, if we suppose a velocity of the structures near the terminal wind speed.
As usual, this first result represents only an instantaneous state of P Cygni's highly variable environment, especially at these spatial scales. To precise the clumps mass and nature, there is a strong need (and now the possibility) of monitoring the clumps formation and evolution from the base of the wind to larger scales. These results enlighten the capabilities of new high-resolution techniques in visible and IR to monitor the spatial activity from the innermost regions to the outer nebula. I present the possibilities opened by techniques such as adaptive optics, long-baseline interferometry and interferopolarimetry which can cover a large range of spatial and temporal resolution particularly well suited to the study of LBV ejecta.



K. Exter, Univ. College London

I will present a history of the radio observations of P Cygni, concentrating on a series of relatively recent VLA and MERLIN observations that have yielded interesting, new information about P Cyg's circumstellar environment (Skinner etal, 1997, 1998). I will discuss some of the concerns of interpreting radio observations, and especially when comparing those taken at different resolutions, by different people, and with different arrays. An overview of the published radio work done on P Cyg will be followed by the details of our MERLIN/VLA work, including a recent series of quite fine-timescale images which show rapid variation in the apparent structure of the inner environment (300 mas). This is work still in progress, so our conclusions on these data have yet to be finalized.



K. Nordsieck et al., Wisconsin

We report on new data, including four vacuum ultraviolet spectropolarimetric observations by the Wisconsin Ultraviolet Photo-Polarimeter Experiment ("WUPPE") on the Astro-1 (one observation) and Astro-2 (three observations) shuttle missions, and 15 new visible-wavelength observations obtained by the HPOL CCD spectropolarimeter at the Pine Bluff Observatory of the University of Wisconsin. This includes three HPOL observations made within 12 hours of each of the three Astro-2 WUPPE observations, giving essentially simultaneous observations extending from 1500 to 10500 Angstroms. Among the results to be discussed are 1) a polarimetric timescale of ~3 days, 2) the lack of a Balmer or Paschen Jump in the polarized flux, 3) an enhancement in the polarized flux of the FeIII wind driving lines at 1800 Ang, and of the absorption components of the visible P Cygni lines, and 4) the lack of polarization of the HeI emission lines. We will suggest constraints these data imply for the nature of the strong inhomogeneity of the P Cygni wind.



K. Davidson, T. Gull, K. Ishibashi, R. Humphreys et al.

I will discuss the unique andd extreme stellar wind of Eta Carinae and its changes since 1997 based on high spatial and spectral resolution spectra of the central star.



T. Gull, K. Ishibashi, K. Davidson, and the HST/STIS Gang, NASA/GSFC

We have been studying the structure of the ejecta surrounding Eta Carinae taking advantage of the high angular resolution of HST, coupled with good spatial PSF properties, and moderate spectral resolution of STIS with the CCD detectors. In these studies we primarily have used a 52"x0.1" slit and medium dispersion gratings yielding ~5000 resolving power and 0.1"x0.1" angular resolution.
We have monitored for three epochs one very specific orientation extending through the star and Weigelt blobs B and D. We find that indeed the Homunculus is a reflection nebulosity with very little emission structure. We can trace the expansion of the dust-scattering component by the stellar broad Balmer emission.
A very strong P-Cygni absorption component is visible and gives us a measure of line profiles at greatly different viewing angles from line of sight. The edge of the Homunculus is seen in emission of Balmer alpha and [N II] lines, likely due to a shock front of the Homunculus expanding into the previous stellar wind. We see external blue-shifted and red-shifted structures connecting to the strings seen by Kristin Weis. Exterior to the biconical lobes is a large disk, easily detected by narrow Balmer absorption lines against the dust-scattered broadened stellar Balmer emission. In the northern lobe, we see both the front and back portions of the lobe. We do not see the back portion of the southern lobe. Within the Homunculus is another biconical structure, much like the Russian doll within a doll within a doll. The middle structure, originally called the Integral Nebula, is an emission nebula, seen in hundreds of emission lines throughout the visible and near-ultraviolet portion of the spectrum. Most of the emission lines are singly-ionized iron. Next are the Weigelt blobs, which are extraordinarily bright emission blobs, complete with thousands of emission lines. The Lund University atomic physics group, especially Torgil Zethson and Sveneric Johanson, are busily identifying the various lines and studying the complexes of lines that appear to be pumped. Katya Verner and Fred Hamann are in the beginning stages of applying CLOUDY models for comparison to the observations. Deep within the structure is a very small structure <0.5" diameter that is a filled hydrogen emission region, with [Fe II] shell, as seen in March 1998. John Hillier, Kazunori Ishibashi and I have noticed that this may be the resolved stellar wind structure as the broadened stellar lines change in intensity relative to the stellar continuum. The higher ionization lines drop with distance while the lower ionization lines remain constant, or increase. Then of course is the central star(s?) which drives the complete system. Line profiles have changed drammatically between March 1998 and February 1999. The STIS acquisition images demonstrate an increase of nearly 2.5-fold flux between December 1997 and March 2000 for the star, but 4.5-fold of the nebulosity within 5" of the star.
These observations were done through support by NASA by the STIS GTO and the HST GO programs.


Theory, pulsations, evolution:


C. de Jager, Utrecht

Successive investigations based on various spectral data have yielded a fairly consistent set of atmospheric parameters. I use T_eff = 19100; log (L/Lo) = 5.8; v_turb = 20 km/sec; log(Mass loss) = -4.5. The resulting atmospheric model is characterized by negative values of g_eff in the higher parts of the atmosphere (optical depth below 0.7). In addition, the sonic point is reached already at a Rosseland optical depth of 0.18. These properties explain the outward motions in the atmosphere and the observed microturbulent velocity. While the body of the star has a Gamma_1 value below 4/3 the average atmospheric Gamma_1 is just above this limit: Gamma_1 = 1.338. Hence the atmosphere is marginally dynamically stable and the source for the long-term dynamic instability must be seated in the body of the star. The luminosity variations of relative short period (< a year) are pressure waves with wavelengths of tens of millions of km. The large observed envelope(s) are probably due to occasional very large-scale gravity waves. As far as its evolution is concerned I suggest the sequence Blue supergiant - Red supergiant - Yellow hypergiant - LBV - WR star or supernova.



M. Friedjung, G. Muratorio, N. Markova, G. Israelian, Paris, Marseille, Bulgaria, Tenerife

We are using the SAC (self-absorption-curve method) to analyze emission lines of P Cyg. This method gives information on the self-absorption and optical thickness of emission lines; when applied to the Fe II emission lines in the spectra of various objects; it has also given information about the column densities and the radii of their line-emitting regions. In the case of P Cyg we have studied lines of NII and FeIII. More highly-excited lines, without blue-shifted absprption components, appear to be optically thick. We shall present and discuss in detail what is seen. This surprising counter-intuitive result can be understood if lines of different excitation potential are preferentially formed in different regions of the wind. We shall discuss the nature of these regions and whether we need to suppose the presence of clumps.



W. Glatzel, S. Chernigovski, Goettingen

We report on recent numerical simulations of the evolution of strange-mode instabilities into the nonlinear regime. Multiple shocks are formed and velocity amplitudes above 100 km/sec are reached which can imply direct mass loss of the objects. The appearance of regular or chaotic variability of stellar parameters found in the final state of the strange-mode unstable objects considered will be discussed. Finally, we comment on the numerical treatment of the problem and its extension to multi-dimensional simulations.

Last Revised: 29 November, 2001
Workshop contact: mdg@star.arm.ac.uk
Go to HOME Page