Annu. Rev. Astron. Astrophys. 2000. 38: 289-335 Copyright © 2000 by Annual Reviews. All rights reserved

3.2. ROSAT Surveys of Groups

Unfortunately, the results of Mulchaey et al (1993), Ponman & Bertram (1993) came late enough in the lifetime of the ROSAT PSPC that large systematic follow-up surveys of groups were not carried out with this instrument. However, the ROSAT PSPC observed many galaxies during its lifetime, and because most galaxies occur in groups, many groups were observed serendipitously. Furthermore, the field of view of the PSPC was large enough that many groups were also observed when the primary target was a star, an active galaxy, or a QSO. In the end, over 100 nearby groups were observed by the ROSAT PSPC during its lifetime, and most of our current understanding of the X-ray properties of groups comes from this dataset.

The existence of an excellent data archive has led to many X-ray surveys of groups using ROSAT PSPC data (Pildis et al 1995, David et al 1995, Doe et al 1995, Saracco & Ciliegi 1995, Mulchaey et al 1996a, Ponman et al 1996, Trinchieri et al 1997, Mulchaey & Zabludoff 1998, Helsdon & Ponman 2000, Mulchaey et al 2000). These surveys indicate that not all poor groups contain an X-ray - emitting intragroup medium. The exact fraction of groups that contain hot intragroup gas has been difficult to quantify because of biases in the sample selection. For example, many of the samples used in archival surveys contain groups that were a priori known to be bright X-ray sources or were likely to be bright X-ray sources based on morphological selection (such as a high fraction of early-type galaxies). These samples are almost certainly not representative of poor groups in general. Furthermore, the term "X-ray detected" has a variable meaning in the literature; some authors use this term only when a diffuse, extended X-ray component (i.e. intragroup medium) is present, whereas others also include cases when emission is associated primarily with the individual galaxies.

There has been considerable interest in the Hickson Compact Groups (HCGs; Hickson 1982; for a review see Hickson 1997). The short crossing times implied for these systems has led some authors to suggest the HCGs are chance alignments of unrelated galaxies within looser systems (Mamon 1986, Walke & Mamon 1989), bound configurations within loose groups (Diaferio et al 1994, Governato et al 1996) or filaments viewed edge-on (Hernquist et al 1995). X-ray observations can potentially help distinguish between these various scenarios (Ostriker et al 1995, Diaferio et al 1995). Ebeling et al (1994) detected eleven HCGs in the ROSAT All-Sky Survey (RASS) data. For some of the detections, the X-ray emission was clearly extended and thus consistent with hot intragroup gas. However, in other cases the sensitivity of the RASS was not good enough to determine the nature of the X-ray emission. Still, Ebeling et al's sample was the first to suggest a correlation between the presence of X-ray emission and a high fraction of early-type galaxies in groups. Pildis et al (1995), Saracco & Ciliegi (1995) each analyzed ROSAT pointed-mode observations of 12 HCGs (there was considerable overlap in these two samples). Both surveys found that approximately two-thirds of the HCGs were X-ray detected, although in many cases the X-ray emission could not be unambiguously attributed to intragroup gas. (Note also that many of the X-ray detections in these two surveys overlapped with Ebeling et al's earlier RASS detections.) A much more complete study of the HCGs was presented by Ponman et al (1996). This survey combined pointed ROSAT PSPC observations with ROSAT All-Sky Survey data to search for diffuse gas in 85 HCGs. These authors detected extended X-ray emission in ~ 26% (22 of 85 groups) of the systems studied and inferred that ~ 75% of the HCGs contain a hot intragroup medium (when one corrects for the detection limits of the observations). Although this is intriguing, some caution must be expressed regarding the Ponman et al (1996) results. Given the compactness of these groups, the nature of the X-ray emission in some of the detected HCGs is far from clear. For example, although Stephan's Quintet (HCG 92) is extended in the ROSAT PSPC data (Sulentic et al 1995), a higher-resolution ROSAT HRI image suggests that most of the extended emission is associated with a shock feature and not with a smooth intragroup gas component (Pietsch et al 1997). Thus, some of the detections in the Ponman et al (1996) survey may not be related to an intragroup medium at all.

Many of the problems inherent to the study of compact groups can be avoided with loose groups. Helsdon & Ponman (2000) studied a sample of 24 loose groups from the catalog of Nolthenius (1993) and found that half of the systems contain intragroup gas. Mulchaey et al (2000) detected diffuse gas in 27 of 57 groups selected from redshift surveys (including the Nolthenius catalog). Both of these studies relied on fairly deep ROSAT pointings and therefore are sensitive to gas down to low X-ray luminosities (~ 5 × 10⁴⁰ h^-2₁₀₀ ergs s^-1). The majority of the groups in both Helsdon & Ponman (2000), Mulchaey et al (2000) were observed serendipitously with the ROSAT PSPC. Based on their velocity dispersions and morphological composition, these samples are fairly representative of groups in nearby redshift surveys. Therefore, these surveys suggest that ~ 50% of nearby optically-selected groups contain a hot X-ray-emitting intragroup medium.

ROSAT All-Sky Survey (RASS) data have also played an important role in our understanding of the X-ray properties of groups. While the RASS observations are generally not very deep, the nearly complete coverage of the sky allows for larger samples to be studied than is possible with the pointed mode data alone. Henry et al (1995) used the RASS data in the region around the north ecliptic pole to define the first X-ray selected sample of poor groups. The survey by Henry et al (1995) was sensitive to all groups more luminous than ~ 2.3 × 10⁴¹ h^-2₁₀₀ ergs s^-1. Although their sample was rather small (8 groups), Henry et al (1995) were able to show that X-ray-selected groups lie on the smooth extrapolation of the cluster X-ray luminosity and temperature functions. The X-ray selected groups also have lower spiral fractions than typical optically-selected groups, which may suggest that X-ray selection produces a more dynamically evolved sample of groups (Henry et al 1995).

The RASS data have also been used to study optically-selected group samples. Burns and collaborators have devoted considerable effort into studying the X-ray properties of the WBL poor clusters and groups (White et al 1999), which were selected by galaxy surface density. One of the more important results from these studies is the derivation of the first X-ray luminosity function for an optically selected sample of groups and poor clusters (Burns et al 1996). The luminosity function derived by Burns et al (1996) is a smooth extrapolation of the rich cluster X-ray luminosity function and is consistent with the luminosity function Henry et al (1995) derived from their X-ray-selected sample of groups. Follow-up work on some of the brighter sources in the WBL catalog indicates that many of these objects are more massive than typical groups with gas temperatures of 2-3 keV (Hwang et al 1999). These systems are important because they represent the transition objects between poor groups and rich clusters.

Mahdavi et al (1997, 2000) used the RASS database to study the X-ray properties of a large sample of groups selected from the CfA redshift survey (Ramella et al 1995). After accounting for selection effects, Mahdavi et al (2000) estimate that ~ 40% of the groups are extended X-ray sources. From these detections, the authors derive a relationship between X-ray luminosity and velocity dispersion that is much shallower than is found for rich clusters (see Section 4.2). They suggest that this result is consistent with the X-ray emission in low velocity dispersion groups being dominated by intragroup gas bound to the member galaxies as opposed to the overall group potential. Unfortunately, the RASS observations of groups typically contain very few counts, so detailed spatial studies of the emission are not possible with this dataset. A much deeper X-ray survey of an optically-selected group sample like the one used in Mahdavi et al (2000) would be very useful and should be a priority for future X-ray missions.