skip to content
ROSAT Guest Observer Facility


You-Hua Chu, Steven L. Snowden, & Thomas Chang
University of Illinois,
USRA/Goddard Space Flight Center,


ROSAT PSPC observations of the Large Magellanic Cloud (LMC) reveal many discrete sources and a large-scale diffuse component. To study these sources in detail, we need high-resolution images to (1) improve the point source positions so optical identifications are possible; (2) resolve small extended sources so their nature may be determined; (3) subtract point sources so the diffuse component can be accurately analyzed. Therefore an HRI survey of the LMC is needed. With an exposure time of 20 ks, a detection limit of ergs s can be reached for point sources; with a 20' spacing between adjacent fields, no positions will be observed at off-axis angles . The LMC survey starts with two contrasting regions: 30 Dor and the LMC bar. The completed HRI survey images have been mosaicked. Even the large-scale diffuse X-ray emission has been clearly detected in this high-quality HRI mosaic of the LMC.


The Magellanic Clouds (MCs) are the only two external galaxies near enough so that a substantial fraction of their X-ray sources can be adequately resolved, and their optical counterparts can be studied in detail. Compared to the Milky Way, these two galaxies have the advantages of small inclination angles that minimizes the confusion and low foreground obscuration that maximizes the visibility. The MCs offer a unique opportunity to study the detailed physical properties of individual X-ray sources as well as the global statistical properties of different types of X-ray sources of two entire galaxies. The statistical results are invaluable in helping us to explain the X-ray properties of distant unresolved galaxies or starburst regions.

Substantial ROSAT time has been used to study the LMC and this has already proven to be a wise investment. A mosaic of the PSPC pointed observations of the LMC has been presented by Snowden & Petre (1994). This mosaic shows a large number of faint sources, in addition to the bright, discrete sources of SNRs, HMXBs, and LMXBs which were detected by the Einstein survey (Long, Helfand, & Grabelsky 1981; Wang et al. 1991) and earlier missions. While some of these new sources are SNRs, HMXBs, and LMXBs, many belong to new classes of X-ray sources such as superbubbles (e.g., Chu et al. 1993; Chu & Mac Low 1996) and supersoft X-ray sources (Trümper et al. 1991). The PSPC mosaic also shows diffuse emission on scales ranging from 100 pc up to 2000 pc. The enormous amount of information contained in these data will take many years to decipher, and the knowledge gained will undoubtedly advance X-ray astronomy to new frontiers.

While the PSPC observations of the LMC have been quite thorough, the spatial resolution is not ideal. Frequently, point sources and small diffuse sources cannot be distinguished, and the optical identification of X-ray sources in crowded regions is difficult. High-resolution images are needed.

Therefore, a ROSAT HRI survey of the LMC has been proposed and is well under way. In this survey, the exposure time for each pointing is 20 ks in order to reach a detection limit of ergs s for point sources, assuming a 3 K thermal emission with an absorption column of log N of 20.5 at a distance of 50 kpc. This detection limit is similar to that achieved by the PSPC All-Sky Survey of the LMC (for an average exposure of 2 ks; Pietsch & Kahabka 1993) and that of the concurrent SMC survey.

The HRI survey will complement the PSPC observations of the LMC, and can be studied in the context of LMC databases gathered at other wavelengths, e.g., an optical emission-line CCD survey being carried out by Chris Smith using the Curtis Schmidt telescope. The proposed HRI survey of the LMC undoubtedly has tremendous archival value and will be used in a wide range of projects; therefore, the normal proprietary period has been waived.

Scientific Objectives

The proposed HRI survey of the LMC will detect sources brighter than ergs s. There are several major areas in which the proposed survey will be used to complement the existing ROSAT PSPC observations of the LMC. The following discussion is limited to three general areas of scientific interest, in order of increasing linear scale: (1) point source survey, (2) small-scale extended emission regions, and (3) large-scale diffuse emission.

Point Source Survey

The proposed HRI survey of the LMC will provide the first X-ray luminosity function complete down to ergs s for an entire galaxy. This can only be matched by the concurrent HRI survey of the SMC. The HRI survey is superior to the PSPC All-Sky Survey and pointed observations in three respects: first, the uniform coverage of the HRI survey augments the statistical value of its point source survey; second, the HRI survey is less limited by confusion; third, the high spatial resolution of the HRI eases the task of making optical identifications of the X-ray sources. Since more than 50% of the point sources are variable, this HRI survey will find totally new sources as well as variability in known sources.

A complete census of the X-ray population of a galaxy down to a limiting luminosity of 10 erg sec, with optical identifications, will be crucial to understanding the types of sources with these luminosities. Many new globular cluster X-ray sources have been found by ROSAT at these luminosities. A conclusive explanation of this population (typically too bright to be CVs, yet too faint to be LMXBs or HMXBs) has not yet been found. The LMC census of sources with these luminosities may shed valuable light on this topic.

For the sources that cannot be identified optically, one may study the underlying stellar population with optical CCD photometry, and check whether the X-ray population and the stellar population are correlated. For example, the 30 Dor region is an active star formation region and the LMC bar region is dominated by a much older stellar population, hence the X-ray populations are expected to differ as well.

Small Diffuse Sources

The HRI survey can be used to determine with greater reliability whether the faint sources detected in PSPC observations are point-like or extended. This is of great importance for the identification of SNRs and the determination of X-ray emitting mechanisms in superbubbles or HII region complexes. Since Population I SNRs and superbubbles are often found associated with active star formation regions, the high concentration of massive stars makes it statistically likely for massive X-ray binaries to be found in the same areas. For example, two point sources exist in the core of 30 Dor, one near the cluster core at R136 and the other coincident with a tight group of WR stars at R140 (Chu & Kennicutt 1994; Wang 1995). These point sources need to be recognized and subtracted before meaningful models can be constructed to explain the diffuse emission.

One of the great contributions of ROSAT in the study of faint diffuse X-ray sources is the detection of a large number of superbubbles in the LMC. Most of these superbubbles have surface brightnesses below Einstein's detection threshold. The wide range of observed X-ray surface brightness indicates different physical conditions and heating mechanisms. Models incorporating SNRs and/or a clumpy interstellar medium have been suggested to explain the X-ray-bright superbubbles (Chu & Mac Low 1990, 1996; Wang & Helfand 1991). In order to distinguish among these models, one must know how the X-ray emission is distributed relative to the shell morphology and whether there are accidentally coincident point sources as described above. The HRI images would be ideal for this morphological comparison.

Large-Scale Diffuse Emission

The 30 Dor region and the LMC bar region constitute the most conspicuous large-scale diffuse emission in the LMC. The underlying stellar and interstellar environments are quite different in these regions. The bar has an older stellar population and a lower gas content; numerous star clusters are clearly visible along the bar (see chart 47V of Hodge & Wright 1967). It is not well known whether the emission in this region originates from stars or hot gas. If the emission is due to a stellar population, two possible signatures may be found in the proposed observations. First, with the higher resolution capabilities of the HRI, the diffuse emission in the bar may be resolved into individual sources. Second, to investigate the presence of unresolved stellar sources, small-scale variations in surface brightness can be correlated with the variations in the stellar density, i.e. with the clusters. Furthermore, optical photometry (e.g., CCD images in the UBV bands) can be used to determine the properties of the associated stellar populations, which may offer clues to the origin of the X-ray population.

For all bright diffuse emission regions, a higher sensitivity search for point sources in these regions is needed to better exclude them from the truly diffuse flux. In this scientific objective, the HRI observations are perfectly complementary to those of the PSPC. After removal of additional sources identified by the HRI, the PSPC data can be more accurately analyzed to provide the best possible spectral information for studying the diffuse emission. This in turn will make it more likely to correctly identify the source and emission mechanisms.

Progress Report

Clearly, it will take a lot of observation time to survey the entire LMC; therefore, the LMC has been divided into several contiguous regions according to their diffuse X-ray emission properties. These regions have been prioritized and grouped into manageable parts. The survey starts with two contrasting regions: a area centered on 30 Dor and a area along the LMC bar. Both regions are rich in discrete X-ray sources, but the 30 Dor region has a high surface density of young/massive stars and the LMC bar has a high surface density of older/less-massive stars.

The HRI survey of the LMC commenced in AO-5. A mixture of B- and C-time for 65 pointing were awarded in AO-5 and AO-6. However, circumpolar C-targets were preferentially excluded in the scheduling, hence none of the priority C targets were completed. Only 27 priority B pointings will be actually completed in AO-5 and AO-6. In AO-7, both the US time and German time have been requested for the HRI survey of the LMC. The US proposal (Chu, Cowley, Ogelman, Petre, Smith, Snowden, & Sanders) requests 45 pointings to complete the 30 Dor and the LMC bar regions, while the German proposal (Haberl, Pietsch, & Dennerl) requests 12 pointings to extend the coverage to the north of 30 Dor.

The HRI survey of the LMC provides a high-quality dataset. The survey has detected not only point sources and small diffuse sources as expected, but also the faint, large-scale diffuse emission. The completed HRI survey observations have been mosaicked. The HRI mosaic is presented in Fig. 1, and the PSPC mosaic of the same region is displayed in Fig. 2 for comparison. Most of the diffuse emissions detected in the PSPC mosaic are also detected in the HRI mosaic; furthermore, the HRI data have resolved several bright diffuse emission regions into fine structures with which more accurate estimates of the emitting volume, density, and thermal energy are finally possible.

gif Figure 1: HRI mosaic of the LMC survey; binned into 20'' pixel and smoothed with a gaussian of =1.5 pixels.

gif Figure 2: PSPC mosaic of pointed observations of the LMC.


Chu Y.-H. & Kennicutt, R.C. 1994, ApJ, 425, 720

Chu Y.-H. & Mac Low M.-M. 1990, ApJ, 365, 510

Chu Y.-H. & Mac Low M.-M. 1996, in ``Röntgenstrahlung from the Universe," p.241

Chu Y.-H., Mac Low M.-M., Garcia-Segura G., Wakker B., & Kennicutt R.C. 1993, ApJ, 414, 213

Hodge, P.W. & Wright, F.W. 1967, ``The Large Magellanic Cloud", Smithsonian Publication 4699

Long, K. S., Helfand, D.J., & Grabelsky, D.A. 1981, ApJ, 248, 925

Pietsch, W., & Kahabka, P. 1993, Lecture Notes in Physics 416, New Aspects of Magellanic Cloud Research, p.59

Snowden, S.L. & Petre, R. 1994, ApJL, 436, L123

Trümper et al. 1991, Nature, 349, 579

Wang, Q.D. 1995, ApJ, 453, 783

Wang, Q., Hamilton, T., Helfand, D.J., & Wu, X. 1991, ApJ, 374, 475

Wang, Q., & Helfand, D. J. 1991, ApJ, 373, 497

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 chu.tex.

The translation was initiated by Michael Arida on Mon Jun 3 13:57:46 EDT 1996

Michael Arida
Mon Jun 3 13:57:46 EDT 1996