ROSAT Guest Observer Facility


S. A. Drake
USRA/Goddard Space Flight Center, drake tex2html_wrap_inline64


I discuss what I believe are the most worthwhile types of X-ray observations of stars and star clusters that the ROSAT HRI can make in the years prior to the launch of AXAF. I identify three generic types of stellar observing programs that will make good use of the HRI's unique capabilities, and that will also provide the type of information needed to maximize the effectiveness of the stellar observations that will be made by AXAF.


At first thought, stars would appear to be some of the most uninteresting targets to be observed by the ROSAT HRI because their angular sizes are 3 or more orders of magnitudes smaller than it can resolve, i.e., they are X-ray point sources! In fact, as I hope to convince the reader, the HRI has already (and will continue to) made many valuable observations of stellar targets, that make use both of its imaging capabilities, its several arc-second angular resolution, and its timing capabilities. I will discuss in the next sections three generic types of stellar observations that I believe that the HRI can do well:

(i) Observations of star clusters, confused regions, and/or multiple stars in which spatial resolution of arc seconds to arc minutes is required.

(ii) Monitoring studies of stars that exhibit variability in their X-ray emission; these can be further sub-divided into (a) short-timescale studies, where the timescales of interest range from days to minutes (or less), e.g., rotational modulation and eclipse mapping programs of rapidly rotating stars, and flare-monitoring programs; and (b) long-timescale studies, where the timescales of interest range from years to weeks, e.g., studies of activity cycles in late-type stars, rotational modulation studies of stars of moderate rotation rates, and programs that monitor the emission of long-period binary stars with eccentric orbits.

(iii) Programs conducted so as to establish the X-ray flux levels of specific objects, or classes of objects, or to study their trends as a function of the stellar parameters, particularly for distant and/or intrinsically X-ray faint stars which have not previously been well-studied.

I will limit most of my specific examples illustrating these three types of HRI observation to late-type (i.e., coronal) stars, since I am most familiar with their related science issues. Almost all of the arguments that I make can, however, also be made about the various types of hot stars (e.g., OB stars, WR stars), and some of them are also relevant to studies of the various sub-types of cataclysmic variables.

Open Clusters, T and OB Associations, and Multiple Stars

Open clusters and associations typically contain stars that are much younger, and hence more active, e.g., emit more X-rays, than stars like the Sun which belong to the general galactic disk population. The spatial densities of stars in these stellar aggregates, and hence the projected surface densities as seen from the Earth, are much higher than for the general field, and source confusion can be a problem for instruments like the ASCA SIS, the PSPC, or the IPC whose point-spread functions are of the order of arc minutes. The most extreme case is that of the very young Trapezium Cluster at the heart of the Orion OB1 Association for which the stellar density reaches tex2html_wrap_inline66 stars pc tex2html_wrap_inline68 in its innermost 0.1 pc (McCaughrean & Stauffer 1994): even an HRI image of this field suffers from source confusion! It is not just the enhanced global star density that causes problems: stars often are part of multiple bound systems. Abt (1988) has found that stars near a solar mass can occur in physical systems with separations all the way up to 2500 Astronomical Units (AU): thus, for a star cluster 250 pc away, there will be an excess of stars with physical companions that are tex2html_wrap_inline70 away. Thus, to uniquely cross-identify X-ray sources in a cluster with optical cluster members in many clusters requires observations with high spatial resolution.

Have all the ``good'' clusters already been done, and, if not, what kind of typical HRI exposure times will be required to sample down to reasonable X-ray luminosity levels? The answer to the first part of the question is no: only 24 of the tex2html_wrap_inline72 open clusters within 500 pc of the Sun have been observed by ROSAT using either the PSPC or the HRI detectors, and only 12 clusters have had HRI observations devoted to them. (The unobserved clusters tend to be the sparser and/or more distant ones, admittedly). One criterion that Ted Simon and I have used in proposing to observe open clusters is that of reaching a limiting X-ray luminosity of tex2html_wrap_inline74  erg s tex2html_wrap_inline76 : this turns out to be an interesting luminosity since a significant number of cluster members will be detected above this threshold, and differences in the X-ray luminosity functions as have been found comparing the similarly aged Hyades and Praesepe clusters (cf. Randich & Schmitt 1995) would be apparent. For typical coronal temperatures and interstellar gas column densities, this threshold requires an HRI exposure time t of tex2html_wrap_inline80  ks, where D is the cluster distance, and I have used a tex2html_wrap_inline82 detection level. Thus, there are many clusters which can still be usefully observed by the HRI: e.g., M 34 at 450 pc, albeit the required exposure times will be rather longer than for most previous HRI cluster observations.

Most of the issues that I have discussed above in the context of open clusters also apply to the younger star-formation complexes, such as the OB (high- and low-mass star) associations and T (low-mass star) associations. General confusion is perhaps not so much of a problem due to their larger sizes, but multiple star systems are equally prevalent. Because of the presence of super-active ( tex2html_wrap_inline84 ) pre-main sequence stars in these complexes, the HRI could make useful observations of these stellar aggregates for distances up to 1 or 2 kpc. However, even nearby T associations have not been extensively studied with the HRI: e.g., there have only been limited number of HRI observations of the Tau-Aur T Association, such that of 82 ROSAT observations pointed within tex2html_wrap_inline86 of T Tau, only 11 of the 82 were using the HRI. Given the number and angular sizes of the various star associations within tex2html_wrap_inline88 kpc of the Sun, there are many that probably merit new and/or additional HRI observations.

Finally, multiplicity is even a potential problem when observing field stars, since the majority of stars are in fact in multiple systems, and many of these are fairly wide visual binaries that are potentially resolvable by the HRI. For example, Jurgen Schmitt and collaborators have been studying the All-Sky Survey PSPC data on nearby ( tex2html_wrap_inline90 pc) low-mass binary stars, and have (or will) re-observe tex2html_wrap_inline92 of them with the HRI in order to separate their target stars from such wide visual companions.

Monitoring Programs To Study Temporal Variability

Short-Timescale Programs

Surprisingly few stars have had long contiguous ROSAT observations, primarily, I think, because these such observations are classified as time-critical by the ROSAT Project, and only a small fraction of the total observations are permitted to be of this type. This has resulted in fierce competition for the available allocation of time-critical observing, and few stellar programs of this type have been approved. Thus, information on the temporal and spatial variability of the coronae of rapidly rotating (periods of a few days or less) single and binary stars is surprisingly sparse. Similarly, knowledge about the X-ray flaring frequencies of the various classes of active stars is surprisingly sparse, with few quantitative estimates having been published. Both types of information would be extremely useful for planning, for example, AXAF observations of such stars: if one is proposing to study the X-ray spectra of flares, for instance, it would help enormously if one knew the flare rate as a function of luminosity and whether there were preferred orbital phases and/or active longitudes when X-ray flares preferentially occur! Finally, joint HRI/RXTE observations of stellar flares should prove particularly valuable, since RXTE will probably be the first X-ray mission that will detect the hard X-ray impulsive emission of these events.

Long-Timescale Programs

Surprisingly few stars have had multi-year ROSAT observations, due to the lack of foresight of most of us stellar astronomers, with a few honorable exceptions such as Ed Guinan. The major reason for this is probably that few of the astronomers who wrote AO-1 observing proposals expected to be writing AO-7 ones! Who knew that ROSAT would last so long? In retrospect, if long-term monitoring programs had started in AO-1, we could already have gained valuable information on stellar activity cycles, since the typical periods derived from Ca II monitoring programs are in the range of 5 to 20 years. X-ray monitoring would have been particularly useful for stars in which Ca II and UV emission is very hard to observe, e.g., late A and early F stars in which the photospheric continuum swamps the emission lines, or dM stars which are often too faint in the UV due to their low luminosities. It may not be too late to start such long-term programs! If one is an optimist and can envision ROSAT operating into the 21st century, or future missions such as LOBSTER (Priedhorsky et al. 1996) coming online in the not-too-distant future, then now is the time to propose...

Observations to Measure X-ray Flux Levels of Stars or Classes of Stars

There are a couple of important issues that come to mind here. The first is the question of whether ROSAT has ``missed'' any classes of X-ray emitting stars? The ROSAT All-Sky Survey (RASS) had a typical detection threshold of tex2html_wrap_inline94  erg s tex2html_wrap_inline76  cm tex2html_wrap_inline98 , that corresponds to tex2html_wrap_inline100 at 10 pc, 29.4 at 100 pc, and 31.4 at 1 kpc. Thus, X-ray dim (e.g., tex2html_wrap_inline106 ) objects would only have been detected within 3 pc. Also, rare (a few per cubic kpc) and X-ray luminous objects may also have been missed by RASS, and, of course, RASS data is still not publicly available, anyway. ROSAT PSPC pointed observations are on the average 10 times deeper than RASS exposures but only cover a small fraction ( tex2html_wrap_inline108 ) of the sky. Thus, if one believes that, for example, R CrB variables may be emitting X-rays, before one has a chance of observing any of these rather rare stars with AXAF, its HRI detection, assuming there is no previous PSPC data of relevance, would seem to be essential!

The second issue relates to studies of well-defined classes of stars, such as G V stars, W U Ma binaries, etc. For any given class one is interested in, the question of whether or not enough X-ray detections of members of this class have been made so that a well-defined X-ray luminosity function can be constructed needs to be addressed. I suspect that, in many cases, the answer is in the negative. For example, in a recent paper on Algol binaries which I co-authored (Singh et al. 1996), we found only 23 detections of such stars in the WGACAT Catalog (White et al. 1994) of X-ray sources. This was sufficient for us to make a bulk comparison of Algols versus RS CVn binaries with similar periods, but clearly insufficient to do analyses of the X-ray luminosity function of subgroups of Algols. The HRI could easily perform volume-limited surveys of interesting classes of objects, such as Algols, or symbiotics, or supergiant binaries, in a relatively modest amount of observing time, provided that the survey parameters are well-chosen. For example, if one believed it was important to do, one could obtain a complete X-ray observed and volume-limited (to 100 pc) sample of all known late-type binaries with orbital periods between 2 and 6 days in tex2html_wrap_inline110 ksec of HRI time.


There are still, in my opinion, many stars and groups of stars for which ROSAT HRI observations would be extremely useful. Because of the wide range of beasts in the stellar menagerie, I have mentioned only a selection of the potential targets, namely mostly those of which I have thought about writing proposals myself! In AO-7 and future AO cycles, I hope that a reasonable amount of observing time will be devoted to such studies. Despite the impressions of some X-ray astronomers who specialize in observing much more distant objects, our understanding of the X-ray emission properties of stars is still rather limited, and I believe that most stellar X-ray astronomers would agree that HRI observations can help to fill the gaps in our present knowledge.


Abt, H. A. 1988, ApJ, 331, 922

McCaughrean, M. J., & Stauffer, J. R. 1994, AJ, 108, 1382

Priedhorsky, W. C., Peele, A. G., & Nugent, K. A. 1996, MNRAS, 279, 733

Randich, S., & Schmitt, J. H. M. M. 1995, A&A, 298, 115

Singh, K. P., Drake, S. A., & White, N. E., AJ, in press

White, N. E., Giommi, P., & Angelini, L. 1994, IAUC, 6100

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 drake.tex.

The translation was initiated by Michael Arida on Mon Jun 3 11:51:08 EDT 1996

Michael Arida
Mon Jun 3 11:51:08 EDT 1996