The following text has been taken from the science section of the 1996 Senior
Review proposal for the US ROSAT Science Data Center (Dr. Rob Petre, PI).
It was compiled by members of the ROSAT User Committee and USRSDC staff.
Compact Galactic Objects
Supernovae, Supernova Remnants, and Diffuse Emission
AGN and Deep Surveys
Clusters of Galaxies
When used to observe the nearest celestial objects, those within the solar system, the ability of the HRI to perform high spatial resolution observations of low surface brightness objects has resulted in some of ROSAT's most dramatic discoveries. The HRI has detected two extraterrestrial solar system objects, and both have yielded surprising, even revolutionary, results. More importantly, ROSAT has proven the value of X-ray observations of our own back yard.
Well before the launch of ROSAT, it was known that X-rays are produced by Jupiter, predominantly from the poles (Metzger et al. 1983, JGR, 88, 7731). ROSAT HRI data have allowed a much better characterization of the spatial and temporal variations of this emission, and have shown that the variability is largely consistent with the much more completely studied UV emission from the aurora, with a strong north-south asymmetry and a brightness variation with longitude. In addition, ROSAT HRI observations have revealed X-ray emission from elsewhere in the Jovian system: they reveal a belt of equatorial X-ray emission, and provide indications of X-ray emission from the foot of the Io flux tube (IFT). The responsible mechanisms are not known (Waite et al. 1994, JGR, 99, 14788; 1995, JGR, 100; 1995, Science, 268, 1598).
The most dramatic observations of Jupiter occurred during the comet Schoemaker-Levy 9 impacts. A brightening of the X-ray emission at the north pole was associated with the impact of two comet fragments (near the south pole). Waite et al. (1995) have concluded that these represented impact-induced brightenings of the aurora. The process responsible for the brightenings is still a mystery: the brightenings did not occur at the at the conjugate points of the impact sites, and thus a more complicated process is required than precipitation of particles energized by the impact and carried along magnetic field lines.
Fundamental questions associated with the Jovian X-ray emission -- the identity and energy distribution of the particles that excite the emission, the process responsible for the equatorial emission, the possible correlation between the emission and the IFT -- can be answered largely by continued ROSAT observations of spatial and intensity variability, but require an extensive campaign of monitoring, in conjunction with observations in other wavebands.
The most exciting surprise to arise from ROSAT observations in the past year was the detection of strong, extended, and highly variable X-ray emission from Comet Hyakutake as it passed within 0.1 AU of the Earth (Lisse et al. 1996, IAUC 6373; Science, submitted). Prior to the observation, the most optimistic prediction of cometary X-ray flux, based on the published literature (e.g., Ibadov 1990, Icarus, 86, 283), translated to an expected HRI rate of ~0.01 s^-1. The detection of the comet during a real time pass, at an average count rate of ~4 s^-1, was greeted with astonishment. The count rate was observed to vary by more than a factor of five on a timescale of hours. The X-ray emission arises from a nearly hemispherical shell, symmetrical about the comet-sun direction, with a diameter of ~20,000 km. No X-ray flux is observed from the nucleus. Lisse et al. (1996) discuss a number of possible emission mechanisms, including resonant scattering of solar X-rays, collisions between cometary and interplanetary dust, and interaction between the comet and the solar wind, and show that none provides a satisfactory explanation. Follow up HRI observations are planned for late June, 1996.
Confirmation that Hyakutake is by no means unique comes from a subsequent search of the ROSAT All-Sky Survey database, which thus far has yielded detection of three comets (Dennerl et al. 1996, IAUC 6404, 6413), at least one of which has extended emission. The spectral information from the PSPC remove any ambiguity about the nature of the emission (i.e., X-rays vs. UV leakage in the HRI), and suggest a thermal origin (kT ~ 0.4+/-0.1 keV) indicative of shock (solar wind) heated gas, plus a possible oxygen fluorescence line.
Cometary X-ray observations are clearly in their infancy. How the X-ray emission scales with other cometary properties is yet to be determined, as is what X-rays can reveal about the nature of comets and the solar wind. Additional observations of comets with a range of luminosities, heliocentric distances, masses, and mass loss rates will begin to answer these questions. The key fact is that a new area of X-ray astronomical study has been opened, and will undoubtedly attract to ROSAT, ASCA, AXAF, an entirely new group of astronomers.
Stellar observations continue to represent one of the largest portions of ROSAT observing time (~25 percent). There are a number of types of stellar observations for which the ROSAT HRI offers unique capability. These include:
- Stellar clusters/associations. Star clusters and associations are the Rosetta Stone of stellar astronomy. Each provides a coeval sample of stars with identical chemical compositions. Wide field X-ray images permit us to map the X-ray emission from these clusters. Observations of a single cluster permit one to determine the dependence of coronal activity on mass and rotation; observations of a range of clusters permit the most accurate determinations of the time-decay of activity, and the dependence of activity on chemical composition and other parameters. X-ray observations also permit one to identify likely members of sparse clusters based on activity levels. The large field of view permits one to observe a significant fraction of a nearby clusters in only a few pointings. Only half of the 48 known open clusters have so far been observed with ROSAT, and most of those are fairly young. Over the next few years, observations will likely concentrate on middle-aged clusters. Such observations are necessary to determine why the Hyades and Praesepe, with similar ages, have such different X-ray luminosity functions.
- Compact regions The youngest clusters/associations tend to be spatially compact, and source confusion is a serious problem, even at the ~1 arcmin resolution of the PSPC or ASCA/SIS. Recent HRI observations of the Orion nebula (Gagne et al 1995, ApJ, 445, 280; Caillault et al. 1994, ApJ, 432, 386), the Rho Oph complex (Montmerle et al., in preparation), and the R Corona Australis cloud (Walter et al., in preparation) have shown that many PSPC sources are resolved into groups of 2 or more point sources, each with stellar counterparts.
- Embedded Protostars The ROSAT HRI is surprisingly effective in penetrating high column densities to observe embedded and obscured X-ray emitters. This is most evident in studies of young stellar objects in nearby star formation regions. PSPC observations of the core of the r Ophiuchi cloud, for example, revealed several dozen faint X-ray sources from Class II and III infrared sources, which are associated with classical and weak-lined T Tauri stars (Casanova et al. 1995, ApJ, 439, 752). This is not a new result, as many unembedded CTT and WTT stars are X-ray emitters. However, they reported that up to 5 Class I infrared sources, associated with the later stages of protostellar accretion, may be been seen. However, the fields are crowded and alternative Class II/III identifications were possible. A 40 ksec follow up with the HRI in 1995 clearly shows that one Class I source, IRS 43, is coincident with an X-ray source with arc second accuracy (Boresight is aligned to visible stars). In addition, the X-ray emission showed a powerful flare on hour timescales, similar to flares seen on CTT/WTT stars. The obscuration to this star is estimated from the infrared spectrum (it is optically invisible) to be Av ~ 44, equivalent to a column density NH ~ 2x10^23 cm^-2. This observation, together with a similar discovery of flaring X-rays from a protostar made with ASCA (Koyama et al. 1996, PASJ submitted), establishes that low mass protostars are X-ray emitters.
This discovery may have great importance for our understanding of star formation and protostars. First, protostellar X-rays should photoionize the immediate environments, including the edges of the cold circumstellar disk. This may provide the magnetic coupling essential in magneto-centrifugal acceleration models of bipolar flow ejection. X-rays may thus play a central role in the flows produced by all protostars. Second, protostellar X-rays may photoionize ambient material, slowing ambipolar diffusion and thereby reducing infall. X-rays might thus be involved stopping the star formation process and thus affecting the initial mass function. Third, the X-ray flares should be similar to solar flares and be accompanied by ejection of energetic protons. There is considerable evidence from meteoritic studies that the solar nebula was bombarded by energetic particles at levels much higher than produced by the contemporary Sun. These particles might conceivably also be responsible for isotopic anomalies seen in ancient meteorites. Altogether, the HRI is providing a new window onto unexpected energetic phenomena in the star formation process.
- Visual binaries (3-30 arcsec separations), which the HRI can resolve, show that the primaries and secondaries of pre-main sequence pairs seem to have different X-ray luminosity functions, even after accounting for the mass differences (Walter & Liu, in preparation). The cause is unknown, but raises questions about the completeness of X-ray selected samples.
- Temporal variability The HRI is quite useful for tracking the variability of sources, both on short timescales (because there is no wire mesh support structure) and on long timescales (because of the instrumental stability). Ongoing programs include measurements of a stellar cycle on EK Dra (Guinan 1996, in preparation) and AR Lacertae, and daily observations of Sigma Ori (Schmitt et al. 1996, in preparation) and the Orion Nebula (Caillault et al. 1996, in preparation) over timescales of a month to study rotational modulation and flaring.
One of the more dramatic results concerns the O7 V star Theta 1 Ori C, the brightest star in the Orion Trapezium. Caillault et al. (1996) discovered a clear periodicity with an amplitude modulation of 35 percent and a period of 15-16 days. The X-ray maxima line up exactly with the maxima from Ha data, which show a period of 15.42 days. This is the first example of large, periodic X-ray variability on a single O or B-type star. A survey of archival IUE data suggests the period and phase have persisted for nearly a decade (Walborn & Nichols 1996, in preparation). The only model that might explain the observation is completely at odds with current thinking about O stars: that Theta 1 Ori C possesses a large magnetosphere whereby the X-ray variability is caused by rotational modulation of a magnetic active region.
ROSAT PSPC observations in 1992 and 1993 showed that the peculiar star Eta Carinae (probably the most massive Galactic star), a known X-ray source, was in fact a variable X-ray source as well. This discovery, as surprising as it was, was confirmed in even more dramatic fashion using HRI data. A comparison between HRI images taken in 1992 and 1994 revealed significant structural differences. In particular, the 1994 HRI image clearly showed a strong core of emission at the position of Eta Car, which was absent in the 1992 image. Subsequently it was realized that the minimum of the X-ray emission was simultaneous with a minimum seen in the He I 10830A line strength, suggesting that the X-ray emission might be periodic at the 5.52 year period of the He I line. If so, then the X-ray flux should vary continuously and reach the next minimum in 1998. More recent observations with ASCA and XTE have indeed suggested that the X-ray emission from Eta Car continues to vary. But in order to constrain models of the X-ray emission we need to have some knowledge of the size of the emitting region during the crucial interval before the predicted minimum in 1998. Until the launch of AXAF, the ROSAT HRI is the only instrument capable of providing this fundamental information about the extent of the X-ray core.
ROSAT HRI's unique combination of high spatial resolution and sensitivity is providing critical new insight into the results of stellar interactions in globular cluster cores. It has already more than tripled the number of X-ray sources known in clusters, while simultaneously providing the positional accuracy needed to identify their optical counterparts. Pointed observations of 43 Galactic globular clusters have been made or are in progress with ROSAT, 35 with the HRI. These studies have shown that:
- Low-luminosity X-ray sources are indeed far more numerous in clusters than their high-luminosity counterparts. Thirty low-luminosity sources are now known in 18 clusters (Verbunt 1996, in ASP Conf. Series 90, eds. Milone & Mermilliod, p. 163), which more than triples the number identified with Einstein.
- In nearly every globular in which one or more low-luminosity sources have been detected, sources have been detected at or near the detection limit of the observation, indicative of a rising luminosity function. Estimates of a power-law slope of the luminosity function from current data are in the range a alpha=0.5-1.0 (Grindlay et al. 1995, ApJ, 455, L47; Johnston & Verbunt 1996, A&A, preprint).
- Four of the eight clusters for which the deepest exposures have been made contain three or more faint sources (Cool et al. 1995, ApJ, 439, 695; Grindlay 1993, in ASP Conf. Series 50, eds. Djorgovski & Meylan, p.285; Hasinger et al. 1994, A&A, 136, 331). In every case, the spatial resolution of the HRI has been essential to separate the sources from one another in order to properly assess their numbers, luminosities, and spatial distribution in the cluster.
- Low-luminosity sources are present in clusters with a wide range of properties. The only cluster for which an observation sensitive to sources as faint as Lx<10^32 erg s^-1 detected no sources in the cluster is M71, which has both a low central density and a low mass.
- Some of the dim sources have highly variable X-ray luminosities (e.g., Hasinger et al. 1994), so that observations at different epochs detect a different (though generally overlapping) subset of the cluster sources. An accurate assessment of the number of sources in a cluster can thus require repeated observations.
At the same time, by providing X-ray source positions accurate to a few arc seconds, the ROSAT HRI has made it feasible to identify optical counterparts even in crowded cluster cores. Promising identifications have now been made for sources near the cores of three different clusters (Cool et al. 1995; Edmonds et al. 1996, preprint; Bailyn et al. 1996, ApJ, submitted), and follow-up spectroscopy of three stars in one cluster has revealed the emission lines characteristic of accretion disks (Grindlay et al. 1995). Studies by several groups are now underway to further characterize these once elusive binary stars through their optical properties.
The fraction of Galactic globular clusters observed to date is small, particularly observations sensitive enough to readily detect the largest class of sources. Further understanding of the nature and origin of these compact binary stars and their role in cluster dynamics will profit from systematic surveys of large, unbiased samples of clusters. A primary goal of studies of X-ray sources in globulars is to determine the conditions necessary for their formation, and to understand their evolutionary relationship to populations of other cluster stars such as millisecond pulsars, primordial binary stars, and various possible merger products in cluster cores. Intercomparisons of the relative numbers of low-luminosity X-ray sources vs. millisecond pulsars, blue stragglers and extreme blue horizontal branch stars will provide valuable constraint s on ongoing efforts to understand the consequences of stellar interactions in cluster cores. On the optical side, HST is providing greatly improved information about the distribution of unusual blue stars in a large number of clusters. Only the ROSAT HRI can provide the complementary database of low-luminosity X-ray sources in clusters. By assessing the numbers, luminosities, variability characteristics, and radial distribution of dim sources in clusters with a wide variety of properties, a more complete picture of cluster evolution can be developed.
COMPACT GALACTIC OBJECTS
Cataclysmic Variables (CVs): Our present and rapidly evolving view of Cataclysmic Variables (CVs) is based in large part upon the contributions made by ROSAT during the all-sky and pointed phases of the mission. Of particular note is the explanation of the period gap, a deficit of magnetic accreting systems typified by AM Hercules with orbital periods in the range of two to three hours. The doubling by ROSAT of the AM Her type systems has essentially shown that the gap was purely a selection effect (Beuermann & Schwope 1993; Proc. ASP Meeting, San Diego; Watson 1993, Adv. Sp. Res. 13, 125). Other discoveries of ROSAT related to magnetic CVs such as very long period systems (Shafter 1994, Proc. 1994 Padua CV meeting; Beuermann 1994, Proc. 1994 Padua CV meeting), very short period systems close to the theoretical minimum (Osborne et al. 1994, MNRAS, 270, 650), and transitional systems (Mason et al. 1992, MNRAS 258,749) provide unique laboratories to test future theoretical models.
The contribution of ROSAT to the advancement of our understanding of non-magnetic systems has also been profound. In addition to the detection of recent classical novae (Lloyd et al. 1992, Nature, 356, 222; Shore et al. 1994, ApJ, 421, 344; Ogelman et al. 1993, Nature, 361, 331) and evidence for the existence of an extended component possibly related to an accretion disk corona (Wood, J. H. 1992, PASP, 104, 780; see Beuermann, K. & Thomas, H.-C. 1993, Proc. Of the COSPAR Symp., Recent Results on X-ray and EUV Astronomy, in press), an entirely new class of X-ray emitting objects unique for their extremely soft X-ray spectra, the so-called super-soft sources, has been classified and elucidated upon by ROSAT (Greiner et al. 1991, A&A, 246, L17). The large volume of new information which ROSAT has already provided has dramatically altered the CV landscape and the present "standard model" is only standard in a transitory sense.
In an extended mission life the ROSAT HRI and its unique photometric capabilities can focus on longer time scale projects which are not normally carried out during the early phases of a space based mission. In particular light curve mapping of eclipsing systems can further elucidate the extent of the X-ray emitting regions in these systems and constrain the parameters of the accretion disk corona which the previous ROSAT observations uncovered. It is indeed ROSAT's unique soft X-ray response which makes this type of program possible and insures a fruitful scientific outcome.
Neutron Stars: Neutron stars, by virtue of the energetics associated with these dense compact objects, are arguably the canonical stellar system in high energy astrophysics. Isolated neutron stars convert mechanical energy of rotation into electromagnetic radiation via their large (B~10^12 G) intrinsic surface magnetic fields; in many ways a cosmic dynamo. The advent of soft X-ray detectors in orbit, initially Einstein but primarily ROSAT, opened a whole new channel for the study of neutron stars and their structure through the detection of the cooling radiation from the stellar surface, one of the original goals of X-ray astronomy.
Prior to the launch of ROSAT there were only a handful of known X-ray emitting isolated neutron stars. Presently, thanks to the unprecedented sensitivity of the suite of ROSAT detectors, that number is now more than 20 with over a dozen having pulsations either detected or confirmed during the mission. A brief list of some of the highlights would read: the solution to the mystery of Geminga (Halpern & Holt 1992, Nature, 357, 222) which was recognized to be a radio quiet middle-aged isolated neutron star and had been an enigma since its discovery in the gamma-ray domain over 20 years ago, the discovery of X-ray pulsations from the Vela pulsar (Ogelman et al. 1993, Nature, 361, 163) as well as the detection of a jet emanating from the pulsar which may be responsible for carrying away the bulk of the spin down energy of the neutron star (Markwardt & Ogelman 1995, Nature, 375, 40), the detection of the middle-aged pulsars (age ~ 10^5-10^6 years) PSRs B0656+14 and B1055-52 (Finley et al. 1992, ApJ, 394, L21; Ogelman & Finley 1993, ApJ, 413, L31) which, along with Geminga, constitute the best candidate cooling neutron stars and are the crucial experimental data necessary to begin to understand the equation of state of matter at supra-nuclear densities, and the spectral details of the old neutron stars (age ~ 10^6-10^7 years) PSRs B1929+10 and B0950+08 (Yancopoulos et al. 1994, ApJ, 429, 832; Manning & Willmore 1994, MNRAS, 266, 635) which are indicative of emission from a magnetospherically heated polar cap. In short the ROSAT mission has provided detailed data spanning some six orders of magnitude in energetics which address very fundamental physics questions related to neutron star dynamics and structure. This invaluable database covers the entire X-ray life of a neutron star from youth through old age and will continue to be a rich canvas upon which theorists can paint their impressions.
The unique soft X-ray response and spatial capabilities of the ROSAT HRI will allow the aforementioned database to be continually enriched during an extended mission life. The excellent spatial response and photometric capabilities of the HRI insure that with deep pointings more isolated neutron stars will be detected. In fact, this assertion is supported by the recent detection of a newly discovered middle-aged neutron star PSR J0538+2817 which is a twin of the cooling candidate PSR B1055-52 (Finley 1996, private communication). This discovery was possible as a result of the deeper and more sensitive radio surveys which are being conducted and the neutron star turns out to be one of the brighter X-ray emitting neutron stars. Thus, the HRI will serve as a valuable tool as the pulsar catalog continues to rapidly grow. The excellent spatial response of the HRI will also be useful in revealing the morphology of neutron stars and the interactions of these high velocity objects with their interstellar environs. A good recent example of these type of studies are the detection of compact nebulae, similar to that in which the Vela pulsar is embedded, associated with the young neutron stars PSRs B1823-13 and B1706-44 (Finley 1996, Head meeting 1996, San Diego, CA).
Another important role of future HRI observations is discovering young neutron stars in supernova remnants. These detections are important because it facilitates a more complete understanding of the supernova phenomenon. An example of this is what has been learned from a single such discovery, in Puppis A (Petre et al. 1996, ApJL, in press). The oxygen abundances determined from spectroscopy of the filaments requires the progenitor star mass to be >325Mo; making this the most massive star known to leave a neutron star remnant. In addition, the high proper motion of the neutron star relative to the kinematic center (~1,000 km s^-1) and the fact that it is moving diametrically opposite from the fast moving optical filaments whose proper motion were used to determine the center, offers some of the strongest evidence for an asymmetric supernova explosion. There are many bright Galactic SNR for which HRI searches for neutron stars remain to be done.
X-Ray Binaries (XRBs): Due to their presence predominantly in the Galactic plane the XRBs are typically heavily absorbed sources and as such have not been the target of a large number of pointed observations. This is not to say that ROSAT has not had its say in the matter. Timing studies of eclipsing systems have determined the secular behavior of the orbital period in at least one LMXB (Hertz et al. 1995, ApJ, 438, 385), a parameter crucial in sorting out the possible evolutionary scenarios leading to the their formation, and confirmed a persistent periodicity of ~ 3 hours which is not correlated to the orbital period in a HMXB (Finley et al. 1992, A&A, 262, L25). The excellent spatial resolution has also been useful in narrowing the search space for the counterparts of XRBs. An excellent example was the recent HRI observation of the bursting XRB GRO J1744-28 (Kouveliotou et al. 1996, IAUC 6286), which suggested a source position different from that of all optical counterpart candidates considered up to that time.
In an extended mission the HRI will prove a useful tool in the XRB area owing to its excellent spatial resolution and photometric capabilities. Long time scale projects such as orbital monitoring and eclipse mapping will be useful in discriminating between various evolutionary scenarios for the formation of XRBs. Follow-up observations of the transient sources which are now routinely discovered by CGRO and RXTE will enable a cataloging of these objects with sub-arcminute uncertainty, and thus a more constrained optical counterpart searcy. Finally, deep HRI pointed observations of nearby galaxies will be very important in understanding the distribution and evolution of XRBs.
Supernova (SN) explosions are the most energetic events in the interstellar medium (ISM). They play a major role in the chemical and dynamical evolution of the ISM in a galaxy. ROSAT HRI observations have provided essential information on three evolutionary stages of SNe and their remnants in the ISM: (1) the transition from SNe to supernova remnants (SNRs), (2) the structure and evolution of SNRs, and (3) the large-scale diffuse X-ray emission powered by multiple SNRs.
Supernovae and Their Young Remnants: Soon after a SN explosion, the expanding ejecta encounter the circumstellar material that was shed by the SN prior to the explosion. The interaction produces a fast outward moving shock as well as a relatively slow backward moving shock into the stellar envelope. Both shocks generate X-ray emission that carries information about the distribution of circumstellar material and the development of a young SNR. Several <20 yr old Type II SNe (with massive progenitors) have been detected in X-rays. Massive stars are usually located in Population I environments that are likely to contain SNRs and massive X-ray binaries. Therefore, it is crucial that X-ray observations have adequate spatial resolution to resolve the SNe from their background X-ray sources. The ROSAT HRI is up to the task. For example, SN1986J in NGC 891 is resolved from a fainter source 28" away (Houck & Bregman 1996, c). SN1988Z, recently detected by the HRI (Fabian & Terlevich 1996, MNRAS, 280, L5), is especially noteworthy as the most distant and most luminous SN detected in X-rays, with an X-ray luminosity exceeding 10^41 erg s^-1.
The ROSAT HRI has been used to monitor the evolution of young SNe. The most exciting object is SN1987A in the Large Magellanic Cloud (LMC). Its SN ejecta are expected to encounter the progenitor's circumstellar shell within the next few years. Because of its proximity, SN1987A offers a unique opportunity for us to watch the spatially resolved interaction between SN ejecta and the circumstellar material. A few young SNe in more distant galaxies have also been monitored. For example, SN1978K in NGC 1313 shows a surprisingly constant X-ray light curve (Schlegel et al. 1996, ApJ, 456, 187). The currently available X-ray observations of SNe are insufficient for a full understanding of SNe evolution. Deep HRI observations are needed to detect and to monitor a larger number of young SNe in nearby galaxies.
Supernova Remnants: In a fully developed SNR, X-ray emission is expected to peak behind the shock fronts. However, if the ISM is clumpy and if a SNR has engulfed many cloudlets, the SNR may show centrally peaked X-ray emission (White & Long 1991, ApJ, 373, 543). The high spatial resolution of the HRI provides an excellent means to study the physical condition and the shock structure of SNRs in the Galaxy as well as in the Magellanic Clouds.
The temporal baseline of high resolution imaging is now long enough (six years since the launch of ROSAT, eighteen since the launch of Einstein) to facilitate measurements of the X-ray expansion of nearby, young remnants. This has now been done for both Tycho (Hughes 1996 presented in the ASCA conference, Tokyo, March 1996) and Cas A (Gotthelf & Keohane 1996, in preparation). Further measurements of the expansion of these remnants, plus measurements of other young remnants like Kepler and SN1006, can still be carried out using the HRI. Deep images of the shells of nearby, older remnants such as Puppis A may serve as a baseline for expansion velocity measurements using AXAF.
The Cygnus Loop is the nearest SNR, hence can be studied at the highest spatial resolution. The HRI mapping of the Cygnus Loop (a large observing program requiring 10^6 s) has indeed produced a spectacular image allowing detailed comparison between the X-ray morphology and optical images (Graham & Aschenbach 1996, ROSAT Workshop at the HEAD meeting, San Diego). It is possible to trace the shock wave as a continuous surface and to closely examine its interaction with interstellar clouds (Graham et al. 1995, ApJ, 444, 787). The analysis of HRI mapping of the entire Cygnus Loop will yield the most complete information about the structure and evolution of a SNR. Several apparently centrally filled SNRs are being mapped by the HRI. A comparison between the X-ray and Ha morphologies of their interiors will allow a critical comparison with White & Long's (1991) model predictions.
The Magellanic Clouds, having little foreground absorption and being at small distances, provide an excellent sample of SNRs that can be studied in detail at multiple wavelengths. ROSAT HRI observations are needed to resolve these SNRs; existing HRI observations have produced interesting results. The SNR 0101-7226 in the Small Magellanic Cloud (SMC) does not seem to emit any X-rays (Ye et al.1995, MNRAS, 275, 1218). The double shells of DEM L 316 are thought to be a pair of colliding SNRs (Williams et al. 1996, BAAS, 28, 924). The SNR N63A has an X-ray size much larger than its optical size; its progenitor must have been a massive star whose wind blew a bubble within which it exploded (Chu et al. 1996, submitted to AJ).
Diffuse X-Ray Emission: SNRs are responsible for heating the ~10^6 K X-ray-emitting gas in a galaxy. This hot ionized phase of the ISM has been the least well-known component because of the difficulties in making observations of large-scale diffuse X-ray emission. The HRI survey of the Magellanic Clouds (another large observing program, requiring 2.8x10^6 s) has been extremely successful in not only detecting but also resolving the diffuse X-ray emission (Snowden & Chu 1996; Chu et al. 1996, ROSAT Workshop at the HEAD meeting, San Diego). Numerous SNRs, superbubbles, supergiant shells, and large-scale diffuse emission are detected in the HRI survey. Both optical emission line images and H I 21 cm line images of the Magellanic Clouds are available (Smith et al. 1996, BAAS, 28, 900; Staveley-Smith & Kim 1996, BAAS, 28, 900). These multi-wavelength surveys of the Magellanic Clouds will allow us to examine the global structure of an ISM for the first time. The HRI survey is essential in providing information on the hot ionized medium.
One of ROSAT's observational strengths has always been the study of individual galaxies. While the surface brightness and spectral capabilities of the PSPC, which facilitated searches for diffuse components and determination of the nature of some discrete sources, are missed, the HRI is a sensitive instrument for detecting and precisely locating discrete sources and tracking their variability, and mapping the morphology of diffuse emission. An example of these capabilities as well as some fascinating results is the nearby galaxy NGC 1313, for which ~120 ks of HRI exposure has been accumulated over the past several years (Petre & Schlegel 1996, in preparation). A total of 11 discrete sources have been detected, all of those observed by the PSPC, plus some additional ones in the confused center of the galaxy. The light curves of the five brightest of these have been traced since early in the mission; all have luminosities above 1x10^39 erg s^-1 (5 times brighter than any Galactic binary), and all vary, except (strangely) the unusual supernova SN1978k. The most variable, and fifth brightest, source is coincident with the nucleus. In addition, the central region of the galaxy contains diffuse emission tracing the spiral arms.
One of the very first HRI pointings provides another demonstration of the power of the HRI. A 50 ks HRI pointing of the central region of M31 detected 86 sources above a threshold of ~1.4x10^36 erg s^-1 (Primini et al. 1993, ApJ, 410, 615). Extrapolation of a flattening luminosity function below 2x10^37 erg s^-1 indicates that only 15-26 percent of the residual diffuse emission observed within 1 kpc of the center can be attributed to unresolved sources from this distribution. Comparison of the ROSAT data with data from Einstein indicates that ~42 percent of the detected sources are variable. Deep monitoring of the central region of M31 continues, and a large observing program has been proposed to widen the spatial coverage within the galaxy.
In M32, a companion to M31, and the nearest elliptical galaxy, emission from the central region of the galaxy is seen with a luminosity of a few x10^38 erg s^-1 (Eskridge et al. 1996, ApJL, 463, L59). A brief HRI pointing suggests the presence of at least three discrete sources; a proposed deep HRI pointing can determine whether more are present.
Several key results involving HRI observations of diffuse emission have recently been reported:
- The HRI observation of NGC 1399 reveals an anticorrelation between the radio lobes and X-ray emission (Kim et al. 1995, ApJ, submitted). This suggests than the hot, X-ray emitting gas provides a thermal pressure confinement for the radio jets.
- The deep HRI observation of the colliding elliptical galaxies NGC 4782/4783 (3C278) shows the complexity of the X-ray emitting plasma in the system (Colina & Borne 1995, ApJL, 454, L101). The image clearly details emission from the galaxies, a bridge between the galaxies, tidal-like tails, and a sheet of emission at their interaction boundary. In addition, models of the radio jet indicate that it is likely being deflected by ram pressure from the hot gas.
- The edge-on starburst galaxy NGC 2146 displays an extensive extended emission resolved in both PSPC and HRI observations with a total luminosity is ~3x10^40 erg s^-1 (Armus et al. 1995, ApJ, 445, 666). In the HRI observation, the emission is further divided into several bright knots on top of the general diffuse emission. The extent of the X-ray emission is much greater than the starburst activity observed at longer wavelengths. The X-ray emission along the galactic minor axis is associated with Ha emission and dust filaments.
- The HRI image of the SABbc galaxy NGC 4258 (M106) shows strong features consistent with the twisted nuclear jets previously observed in radio continuum and optical emission-line studies (Cecil et al. 1995, ApJ, 440, 181). The southeast jet is unresolved over its 2.5' (5 kpc) length which contrasts to the more diffuse northwest jet. The data are consistent with temperature of 0.3 keV with a luminosity of~1.6x10^40 erg s^-1. The strong outflow of NGC 4258 provides a tight correlation between thermal X-rays and the radio jet. The inferred temperature is consistent with that expected from shocks with velocities observed in optical data.
One indication of the power of the HRI for the study of nearby galaxies is that nearly half the large observing programs proposed in AO7 are to study galaxies. Topics include the detailed mapping of the LMC, a complete census of Local Group galaxies, detailed study of M31, and a survey of previously unobserved nearby spirals. All these programs complement the science that will be performed by AXAF as well as identify objects meriting deep imaging and spectroscopy by AXAF.
AGN AND DEEP SURVEYS
Current HRI observations of active galactic nuclei (Seyfert galaxies, quasars, radio galaxies, blazars) focus on three aspects: temporal variability on scales from minutes to months, morphological studies, and detection of faint objects.
The HRI is an excellent photometer, and the X-ray variability of AGN has not been well characterized on very short (less than one day) or very long (longer than one week) timescales, nor has a comprehensive understanding of the relationship between X-ray variability and that in other bands been established. A number of ambitious observations of AGN, many in coordination with one or more other orbiting observatories, seek to provide this information. The most impressive such study was the monitoring of 3C390.3 every three days over a nine month interval (Leighly et al. 1996, in Roentgenstrahlung from the Universe, eds. Zimmermann et al. p. 467). The X-ray flux varies with much higher amplitude, and over much shorter time scales, than does the optical emission. Monitoring of 4 low-redshift quasars with the HRI over a period of about 1 month showed significant variability on timescales of days -- both gradual dimming/brightening, and outburst. Quite unexpected, if such variability proves to be common, this will place a crucial constraint on models for X-ray emission in AGN. (Elvis et al. 1996, in preparation).
HRI observations of nearby Seyfert galaxies show that kpc-scale extended emission with typical luminosity 10^40-10^42 erg s^-1 is common. The extended X-ray emission aligns with radio and/or narrow line region axes and has spatial extent similar to that of the extended emission line gas. The X-rays most likely arise as thermal emission from hot gas in pressure equilibrium with optical line emitting gas and the synchrotron emitting plasma responsible for the kpc scale radio jets and lobes (Wilson et al. 1996, in Roentgenstrahlung from the Universe, eds. Zimmermann et al. p. 529).
HRI observations are also revealing important new information about diffuse X-ray emission from radio galaxies. Both for the low luminosity radio galaxy NGC 1275 in the Perseus Cluster (Hasinger et al. MNRAS...) and for the high luminosity (FR II) radio galaxy Cygnus A (Carilli et al. 1994, MNRAS, 270, 173), the expansion of the radio lobes evacuates a cavity in the hot (cluster) gas surrounding the central galaxy. This allows the use of pressure balance arguments to constrain models for the lobes. In Cygnus A, Harris et al. (1994, Nature 367, 713) found for the first time a spatially resolved structure for which the synchrotron self-Compton model provides the best explanation of the data. This allows an estimate of the magnetic field strength which does not rely on equipartition arguments, and implies that the jets of Cygnus A (and presumably other radio sources) may be comprised of e+/e- rather than p+/e- (relativistic) plasma.
A number of deep surveys have been conducted using the HRI, many of which are follow up observations of PSPC deep survey pointings. The best known of these is the Lockman Hole pointing, with a total exposure approaching 5x10^5 s, and a point source flux sensitivity of 1x10^15 erg cm^-2 s^-1. The precise HRI positions ease the search for optical counterparts, and split confused sources, even at the detectability threshold. The HRI fluxes for the resolved sources allow more precise estimates of the unresolved flux in the PSPC images. The deep surveys are revealing that at low X-ray fluxes different source populations contribute more strongly, especially the narrow line X-ray galaxies (Hasinger 1996, in Roentgenstrahlung from the Universe, eds. Zimmermann et al. p. 291).
An extraordinary number of projects remain to be performed on AGN using the HRI, as chronicled in the contribution by M. Elvis to the San Diego workshop. Among these are surveys of various classes of objects, from nearby galaxies potentially housing micro-AGN, to the radio galaxies in the 3C catalog. Several proposals for new deep survey fields were proposed in AO7, including more than one for the Hubble deep field. An observation of this region would dovetail with the high level of observational activity at all wavebands in this field. The HRI is uniquely positioned to resolve the large number of sources expected. The X-ray observation will help characterize cluster evolution, and the contribution of starbursting galaxies and AGN to the X-ray background.
The long exposures available during the HRI-only phase of the ROSAT mission are enabling qualitatively new studies of clusters of galaxies to be made. The higher background of the HRI compared to the PSPC (or the Einstein IPC before it) typically meant that previously most observers sacrificed the improved angular resolution for increased sensitivity to the low surface brightnesses characteristic of clusters. Now, there are several situations where the HRI is being used to advantage for cluster observations.
The first is to resolve fine details in low redshift objects. Huang and Sarazin (1996, ApJ, 461, 622) have presented a high angular resolution study of the nearby Hercules cluster. This cluster exhibits substructure on a rich variety of scales, the smallest of which can only be studied with the HRI. Two papers have presented observations of the interaction between X-ray emitting cluster gas and the radio lobes of the brightest cluster galaxy, which also happens to be a radio galaxy (Boehringer et al. 1993, MNRAS, 264, L25 - NCG 1275; Carilli et al. 1994, MNRAS, 270, 173 - Cygnus A). Both papers find that the X-ray emitting gas is displaced by the radio emitting material. These observations provide strong support for the idea that jets of material emanating from the nucleus power the radio galaxies. The effects of the jets are clearly seen in the displaced X-ray gas.
Another situation is the observation of distant clusters where the PSPC does not resolve the object. The availability of substantial amounts of HRI time has vastly improved our knowledge of the X-ray emission properties of these objects. Studies of distant clusters provide information on the rare objects whose space densities are so small that nearby examples are unlikely to be found. Edge et al. (1994, MNRAS, 270, L1) report HRI observations of Zw3146, a cluster at z=0.291 which contains the largest cooling flow known, 1000 Mo y^-1. Schindler et al. (1996, A&A, submitted) describe the morphology of the most luminous cluster known, RXJ1347.5-1145, which is at a redshift of 0.451 and has a bolometeric luminosity greater than 10^46 ergs s^-1. This cluster may contain an even larger cooling flow, except its ASCA temperature is a hot 9.3 keV. Donahue (1996, ApJ, in preparation) is analyzing the morphology of the most luminous cluster in the Einstein Medium Survey, MS0451.6-0305 at z=0.539 with an ASCA temperature of 10.4 keV. This cluster, although elongated, appears to be relaxed because there are no clumps containing more than 1 percent of the total luminosity.
Another rare class of objects are clusters containing gravitationally lensed arcs. Fortunately X-ray luminosity is correlated with a high likelihood of the cluster being massive enough to image background galaxies. However, even the high masses of very luminous objects only provide a weak lens so that their focal lengths correspond to redshifts > 0.2 in most cases. A symbiotic relationship has arisen between mass determinations of clusters by X-ray methods and by gravitational lensing, both weak and strong. Each method has its own systematic errors, so the combination of both yields more reliable results. For example, one of the first applications of the weak lensing technique was to MS 1224.7+2007 which yielded a mass corresponding to a density parameter ~ 2 if this cluster were typical. HRI data (Henry et al. 1996, in preparation) show that there are two clusters superimposed on each other but whose centers are displaced by 7'. The weak lensing result assumed that the mass was zero at the edge of its field of view, 7' in this case, and so is undoubtedly biased. After other apparent disagreements, more careful analyses are showing that the two methods give the same results for the mass, at least to better than factor of two (Allen et al. 1996, MNRAS, 279, 615 - PKS 0745-191; Squires et al. 1996, ApJ, 461, 572 - A 2218). Much work remains to be done, because the detailed morphology of the mass deduced by the two methods is not always in agreement, e.g., A2218 (Squires et al., 1996; Kneib et al. 1995, A&A, 303, 27). Here even the morphology deduced from the weak lensing disagrees with that from the strong. However, all is not lost, sometimes the mass distributions coming from the two methods are in excellent agreement, such as for A370.
ROSAT has provided a unique window on gravitational lensing systems. There are a number of such systems in which the lensed object is observed in other bands but the object serving as the lens is invisible. If the so-called dark lens is a cluster of galaxies, its massive halo can in principle be detected in X-rays. Such a detection has been made in the lens system MG2016+112 in a 20 ks observation (Hattori et al. 1996, in X-Ray Imaging and Spectroscopy of Cosmic Hot Plasmas). The lens is now identified as a cluster of galaxies at z=1. If the mass of the lensing cluster can be estimated based on the X-ray luminosity and extent, it can be used to estimate Ho (Grogin and Narayan 1996, ApJ submitted). In the system 0957+561, X-ray counterparts of the two images of the lensed quasar are detected, and while variability of both components have been seen in both X-ray and other bands, the X-ray variability does not track that seen at longer wavelengths (Chartas et al. 1995, ApJ, 445, 140). Chartas et al. suggest that at least one image is being microlensed by stars or clumps of dark matter. In 0957+561, only an upper limit can be placed on the luminosity of the lensing cluster. The success of these observations demonstrates that an HRI campaign to observe more gravitational lenses could yield valuable cosmological information; a number of observations of lens systems were proposed in AO7.
The classical reason for observing distant objects is to study the evolution of their properties. Henry et al. (1996, in preparation) have obtained HRI observations of a complete sample of 10 Einstein Medium Survey clusters in the range 0.3 - 0.4 (they also have ASCA observations of these same clusters). Given the high percentage of substructure in low redshift clusters, ~40 percent, nearly every object in the distant sample should have such structure if the density parameter were ~ 1. The morphologies of objects in the distant sample contain the usual suspects: some with central peaks indicating cooling flows with no substructure and some with evidence for mergers. These two types are in about the same proportions as at low redshift (within the limited statistics). There are two intriguing objects with peaked surface brightnesses not centered on the overall cluster emission. Such morphology is rare if nonexistent at low redshifts.
As the above examples show, studies of distant clusters with the ROSAT HRI have reached a level comparable to that of early work on their nearby counterparts. If ROSAT would continue to operate, X-ray images could be obtained for a substantial fraction of all clusters known at redshifts beyond 0.3. Combined with ASCA temperatures, these data will permit investigation into a range of problems of cosmological interest.
HEASARC Guest Observer Facility
Please use the Feedback link if you have questions on ROSAT.