skip to content
ROSAT Guest Observer Facility


Mike Corcoran
USRA/LHEA/Goddard Space Flight Center, corcoran tex2html_wrap_inline56


We have organized an effort to address the question of how X-ray emission is related to stellar wind emission in hot young stars using a deep HRI pointing of the Carina nebula region along with coordinated ground-based observations. This effort involves a large segment of the community of hot star observers. This program will allow us to: generate high S/N X-ray lightcurves for more than 10 hot stars, correlate X-ray variability and wind variability seen at other wavelengths, spatially resolve and monitor the variable core component of tex2html_wrap_inline58 Car, monitor a selection of binary stars for X-ray emission from colliding winds, and provide the deepest sampling of the X-ray luminosity function for massive stars in the Carina nebula to date.


EINSTEIN, ROSAT and ASCA have made substantial progress in the understanding of X-ray emission from hot massive OB and WR stars, having established for example:

  • that the observed X-ray luminosity is proportional to stellar bolometric luminosity for the OB stars;
  • that Wolf-Rayet stars show unexpectedly large amounts of scatter in the tex2html_wrap_inline60 plane;
  • that certain WR+O binary systems like WR 140 and tex2html_wrap_inline62 Vel show evidence of X-ray emitting gas produced by the collision of stellar winds.

However, the most important question is still unresolved: what is the physical link between the stellar wind and the production of X-rays? Recent optical and UV spectroscopy have revealed that hot-star winds are inhomogeneous and dynamically variable (e.g., Moffat et al. 1994), with Discrete Absorption Components (DACs) appearing and disappearing on large scales and clumps or blobs on small scales. These inhomogeneities generate shocks, which should produce X-rays if the currently accepted wind-shock mechanism is correct. If so, then one expects X-ray variability to occur on a timescale corresponding to the wind flow time (a day or so) or rotation time (typically a few days). In spectroscopic binaries (SBs) one also expects phase-dependent X-ray modulation on an orbital timescale ( tex2html_wrap_inline64 days in most cases).

Currently the relation between the stellar wind and the stellar X-ray emission is poorly understood, since few X-ray observations of hot stars have the required statistical significance or timespan to look for expected X-ray variations, and in general there has been no major effort to obtain coordinated ground-based observations to relate the X-ray emission to wind variability seen in other wavebands. Previous X-ray observations, largely limited to isolated snapshots with exposures of 10 ks or less, are photon limited and revealed little definitive variability in most cases (e.g., Berghoefer and Schmitt 1995). To make the next step, to conclusively link the X-ray emission to wind inhomogeneities, we need high S/N X-ray observations along with coordinated ground-based support. The viability of such an observing campaign has already been demonstrated. A deep PSPC pointing at tex2html_wrap_inline66  Pup (O4If) combined with ground-based H- tex2html_wrap_inline68 spectroscopy showed, for the first time, that the X-ray emission from this star (previously thought to be constant) varied in a way that was correlated with changes in the state of the wind (Berghoefer et al. 1996). At present tex2html_wrap_inline66  Pup is the only normal O star for which this claim can be made.

The ROSAT HRI is the only X-ray observatory available in the forseeable future which can provide the needed monitoring of hot stars, since it is the only one with the necessary soft response (most of the stellar emission from O stars is below 1 keV) which also has sufficient imaging capability to resolve individual O stars in crowded fields in which they are typically found.

This large collaborative proposal arose out of the recent international Workshop on Colliding Winds in Binary Stars, held in La Plata, Argentina 21-24 November 1995. During this meeting, it became quite evident that, despite the enormous progress that X-ray studies have provided in revealing the dynamics of stellar winds from hot stars, none of the existing time series (normally isolated snapshots of some 10 ks) is long enough to adequately sample the dominant physical timescales in hot luminous stars. Patching together several different snapshots is simply not adequate. We note that a similar continuous (15-day) run with IUE was recently (Jan 1995) devoted to an intense variability study of 3 prototypical luminous hot stars, by a large consortium of astronomers with common interests. The results of that study are of unprecedented value (Massa et al., 1995ab, Prinja et al. 1995, Howarth et al. 1995, St. Louis et al. 1995). The present proposed X-ray study is expected to have a similar high impact on hot-star research.

Co-Investigators in this project (referred to as the ``XMEGA'' group) are Dominique Ballereau (Meudon), Rodolfo Barba, Virpi Niemela, and Nidia Morrell (LaPlata), Thomas Berghoefer and Ulf Wessolowski (MPE), David Cohen (Wisconsin), Mike Corcoran (USRA), Augusto Damineli (IAGUSP), Mart de Groot (Armagh), Philippe Eenens (INAOE), Thomas Eversberg, Sergey Marchenko, Tony Moffat, and Nicole St.-Louis (Montreal), Ken Gayley (Bartol), Doug Gies and Michelle Thaller (Georgia State), Kenji Kawashima, Kenzo Kinugasa, Shunji Kitamoto, Takashi Toneri, and Ken-ichi Torii (Osaka), Gloria Koenigsberger (UNAM), Yoji Kondo (GSFC), Siegfried Luhrs (Munster), Felix Mirabel (Saclay), Nancy Morrison (Toledo) , Ralf Palsa (Munich), Andy Pollock (C&S Ltd), Gregor Rauw, Jean-Marie Vreux (Liege), Ian Stevens (Birmingham), Olga Tsiopa (Pulkova), Roberto Viotti (IAS), Peredur Williams (ROE), Wayne Waldron (ARC), Stephen White (UMd), and Juan Zorec (IAP).

Proposed: An HRI Deep Survey of the Carina Nebula Field

We will use the HRI to obtain a deep pointing (500 ks) at the Carina nebula, one of the most important Galactic massive-star nurseries. The Carina nebula, a ``starburst-like'' region of massive-star formation at a distance of 2.7 kpc, contains some of the hottest, most massive stars in the Galaxy, including an assortment of some six O3 stars, three WR stars and the extreme Luminous Blue Variable tex2html_wrap_inline58 Carinae. Previous short (50 ks) PSPC and HRI images of this field (Corcoran 1994) show a large number of detected X-ray sources ( tex2html_wrap_inline74 ) most of which are identified massive stars. These new HRI observations, along with coordinated ground-based observations, will allow us to obtain high S/N lightcurves of the 10 or more brightest sources over an interval of many days, and combined with the ground-based work will let us examine in unprecedented detail the physical link between the X-ray emission from these stars and the state of the stellar winds.

Science Objectives

This deep HRI survey will for the first time test in detail the relation between X-ray emission and the stellar wind from these massive stars. We have 4 main objectives.

We will obtain high S/N X-ray lightcurves of the 10 brightest sources to allow us to look for variability on timescales (one to several days) that are truly relevant to the flow times in hot star winds. Coordinated ground-based (optical spectra and photometry, and radio flux) observations will be used to look for clear correlations between the state of the wind and the X-ray emission. A particularly important star in this field is the apparently single WR star HD 93162. This WN star has the highest X-ray flux known of any single WR star. This is a mystery, since in the optical and UV its flux levels are not unusually high. Clearly, intense monitoring on hour-day timescales of this star (and others) will prove to be extremely interesting.
We will spatially resolve and monitor the X-ray variable core of tex2html_wrap_inline58  Carinae. tex2html_wrap_inline58  Car is presently undergoing an outburst visible in all energy bands from the radio (Duncan et al. 1995) through optical (Damineli 1996) and X-ray range (Corcoran et al. 1995). There is evidence that the X-ray variations are correlated with changes in the He I 1083nm line (which is known to be periodic with a period of 5.52 years, Damineli 1996). Confirming this correlation provides hard constraints on the star's small and large-scale instabilities and may entirely alter our understanding of the physical nature of this unique object. We point out that tex2html_wrap_inline58  Car is currently being monitored by RXTE and ASCA; the spatially-resolved HRI observation proposed here will provide a perfect complement to the temporally-resolved RXTE and spectrally-resolved ASCA data.
We will obtain at the same time the deepest, most precise X-ray luminosity function ever, for a complete sample of luminous, hot stars of the same age and at a unique location in the galaxy. This deep observation should enable us to detect at tex2html_wrap_inline82 mid-B main sequence stars, of which about 60 can be found in the field. This will allow us to significantly extend our current knowledge based on the previous ROSAT pointings to look in a statistically reliable way, never before achievable, at the dependence of X-ray emission on wind strength.
We will obtain high S/N X-ray lightcurves of a selection of the most massive binary systems known in order to look for orbital-dependent X-ray emission resulting from wind collisions. Colliding wind emission provides a direct, localized probe of the dependence of X-ray emission on wind mass-loss rate and the physics of wind acceleration by the line-driving mechanism. One particularly important system is HD 93205 (O3V+O8V), a highly eccentric (e=0.5) 6-day binary containing the most massive main sequence star ever directly weighed. Evidence of colliding-wind X-ray emission for HD93205 has already been found in a 50 ks PSPC observation (Corcoran 1996). However, this orbital dependence could not be confirmed in a 50 ks HRI observation already in hand, due to the factor of 4 difference in sensitivity of the HRI which washed out the variations. Another important system is HD 93206, a quadruple system, containing an O9II + B0Ib eclipsing pair in a 6-day orbit, along with a second non-eclipsing O9.7Ib (+B2V?) pair in a 21-day orbit. Both pairs are likely revolving around each other in less than 25 years. We expect any X-ray variability to be dominated by the shorter binary.

We point out that while the previous 50 ks HRI exposure already in hand is too short to achieve these goals by itself, it will be fully utilized by us to examine issues of stability of emission and look for long-term changes in objects like tex2html_wrap_inline58 Car, HD 93205, HD 93162 and other important sources.

Supporting Observations

Supporting observations will include Coude spectra of important line regions in the tex2html_wrap_inline58  Car optical & near-IR spectrum along with UBVRI photometry of important target stars using the 1.6m telescope at the LNA observatory (Brazil). In addition, we plan to obtain optical spectra of all X-ray sources visible in the ROSAT HRI field of view at San Juan, Argentina, with the CASLEO 2.15m telescope and REOSC echelle spectrograph. Also optical direct CCD images will be obtained in order to identify fainter X-ray sources. Spectrograms of several stars in the region have already been obtained, as part of a search for spectroscopic binaries among the previously detected X-ray sources at San Juan. We also plan to obtain one-channel precision optical photometry of the brighter stars in the Carina field at the U. of Toronto 0.6m telescope at Las Campanas, Chile. We will also apply for observing time at the 1.5m ESO telescope equipped with a B&C spectrograph providing medium resolution spectroscopy (reciprocal dispersion of 33A/mm) over a range of 1000A, allowing us to investigate line profile variability in the brighter stars. At the same time, we will also apply for high resolution spectroscopy at the Coude Auxiliary Telescope (CAT) at ESO for targets brighter than 9th magnitude to provide very accurate radial velocities, e.g., for those stars that show RV variations but are not known to be binaries.

Radio observations of tex2html_wrap_inline58 Car at the ATNF which are part of a long term monitoring campaign will continue. X-ray emission from tex2html_wrap_inline58  Car will also be monitored by ASCA and RXTE. The ground based observations will be scheduled as close in time as possible to the HRI observations. Because we have a fairly large number of observing sites participating, we feel confident that a fairly large degree of simultaneity will be achieved.


Berghoefer, et al., 1996, A & A, 306, 899

Berghoefer, T., and Schmitt, J., 1994, Ap&SS, 221, 309

Corcoran, M., 1996, RMAA Conf Ser, in press

Corcoran, M., et al., 1994, AIP Conference Proceedings 313, 159

Corcoran, M., et al., 1995, ApJL, 445, 121

Damineli, A., 1996, ApJL, 460, 49

Duncan, R., et al., 1995, ApJL, 441, 73

Howarth, I., Prinja, D., Massa, D., 1995, ApJL, 452, 65

Massa D, et al., 1995a, ApJL ,452, 53

Massa, D., Prinja, R., Fullerton, A., 1995b, ApJL, 452, 842

Moffat, A., Owocki, S., Fullerton, A., and St-Louis, N., 1994, ApSS, 221

Prinja, R., Massa, D., Fullerton, A., 1995, ApJL, 452, 61

St.-Louis, N., Dalton, M. J., Marchenko, S. V., Moffat, A., and Willis, A., 1995, ApJL, 452, 57

Return to the Title Page
About this document ...

This document was generated using the LaTeX2HTML translator Version 96.1 (Feb 5, 1996) Copyright © 1993, 1994, 1995, 1996, Nikos Drakos, Computer Based Learning Unit, University of Leeds.

The command line arguments were:
latex2html -split 0 corcor.tex.

The translation was initiated by Michael Arida on Mon Jun 3 11:50:05 EDT 1996

Michael Arida
Mon Jun 3 11:50:05 EDT 1996