2 The EMSS arc survey cluster sample

The Einstein Observatory Medium Sensitivity Survey (EMSS; Gioia et al.1990a; Stocke et al.1991; Maccacaro et al.1994) has been for many years the only large and sensitive X-ray catalog from which an X-ray selected sample of distant clusters can be drawn. Consequently, the EMSS cluster sample has been under study by a number of groups for a variety of cosmological investigations (Annis 1994; Gioia et al.1990b; Henry et al.1992; Vikhlinin et al.1998b; Henry 1997; Fahlman et al.1994; Luppino & Kaiser 1997; Clowe et al.1998; Luppino & Gioia 1995; Carlberg et al.1997a, 1997b and 1998; Donahue et al.1998). These investigations range from the study of galaxy evolution in $z\simeq0.3$ X-ray selected clusters, in evolution of the X-ray luminosity and of the temperature function of clusters, in detection of weak gravitational lensing and consequent mass estimates, to determination of cluster virial masses and of the cosmological density parameter $\Omega_{0}$. Some of these studies have made X-ray followed up observations of individual clusters using the ROSAT or ASCA satellites (Donahue & Stocke 1995; Donahue 1996; Donahue et al.1998; Gioia et al.1998a).

Until the ROSAT North Ecliptic Pole (NEP) region (Henry et al.1995; Mullis et al.1998) of the all-sky survey is completely identified, the EMSS is the only large and sensitive X-ray catalog from which an X-ray selected sample of distant clusters can be drawn. Even if there are several EMSS-style cluster surveys working from the ROSAT data archive of pointed observations (Rosati et al.1995 and 1998; Scharf et al.1997; Collins et al.1997; Vikhlinin et al.1998a), most of these are still works in progress and do not cover a large area of sky (> 700 square degrees) as the EMSS, with the exception of the Vikhlinin et al.sample which covers about 160 square degrees, a fourth of the EMSS area.

During the course of this survey, the total number of clusters in the sample has fluctuated slightly. We added new distant EMSS clusters that previously had no measured redshift, but were obviously at distances z$\,\gt\,$0.5 and, from their detected X-ray flux, clearly met our selection criteria. We also removed sources that, after follow-up observations, turned out not to be clusters. For example, three of the sources listed in Gioia & Luppino (1994), MS1209+3917, MS1333+1725 and MS1610+6616, were removed from the cluster list after ROSAT observations revealed that they are unresolved (MS1209+39 has been identified with a Bl Lac object by Rector et al.1998, MS1333+1725 is a star and MS1610+66 is still unidentified).

There are presently 100 EMSS sources classified as clusters. The cluster subsample we chose for the arc survey was subjected to the following criteria. First, the sources had to lie North of declination $\delta$$\,\geq\,$$-40 \hbox{$^\circ$}$ to be observable from Mauna Kea. Second, the fluxes of the sources in the $2.4\thinspace \hbox{$^\prime$}\times 2.4 \thinspace \hbox{$^\prime$}$ IPC detection cell had to exceed 1.33 10^-13 ergcm^-2s^-1 after converting from IPC counting rates with a thermal spectrum of 6 keV temperature and correcting for the galactic absorption in the direction of each source, but with no IPC point response function correction applied. Third, we restricted our sample to clusters with redshifts z$\,\geq\,$0.15, and fourth, we required that the X-ray luminosity be greater than $L_{\scriptsize 0.3-3.5\,{\rm keV}}$$\,\geq\,$$2.0\ 10^{44}$ ergs^-1 to select for deep potential wells which are most likely to exhibit gravitational lensing. These criteria resulted in 38 clusters spanning a large redshift range ($0.15\leq z\leq 0.823$) and an order of magnitude in X-ray luminosity ($2\ 10^{44}\; {\rm erg~ s}^{-1} \leq L_{\rm x} \leq 2\ 10^{45}\; {\rm erg~ s}^{-1}$). For each of the 38 clusters optical, radio and X-ray data can be found in Gioia & Luppino (1994), including wide field CCD images (>1Mpc$\times$1Mpc in the cluster frames).

Although we used X-ray selection to avoid obvious optical selection effects, it would not be fair to claim that the EMSS cluster sample is completely free from selection biases. As pointed out by Donahue et al. (1992; hereafter DSG), the EMSS is not, strictly speaking, a flux limited sample. Instead, the detection of EMSS sources is limited by central surface brightness. Consequently, there may be a bias toward clusters with centrally peaked X-ray surface brightness since the EMSS detection algorithm was optimized for detecting point sources in the IPC $2\hbox{$.\mkern-4mu^\prime$}4 \times 2\hbox{$.\mkern-4mu^\prime$}4$ detection cell. Clusters with more extended emission at lower surface brightness would have been resolved by the detector and may have been missed by failing to meet the minimum detection criterion in the central cell (note that this problem is more severe at lower redshifts). One might argue that the EMSS might preferentially select cooling flow clusters (Pesce et al.1990; Edge et al.1992) given the possible bias toward centrally peaked objects. DSG, however, make compelling arguments for why EMSS clusters cannot all be cooling flow clusters; namely that there do exist non-cooling flow clusters with large $L_{\rm x}$ and small core radii. Moreover, clusters with large core radii tend to have lower $L_{\rm x}$ and thus would be excluded from the EMSS sample for that reason. Additional evidence that the EMSS does not miss clusters which are not cooling flow clusters is given by the agreement between the X-ray luminosity function of the EMSS clusters with z$\,=\,$0.14$\,\rightarrow\,$0.2, and the X-ray luminosity function derived by earlier studies (Piccinotti et al.1982) which use large-beam, non-imaging detector fluxes or with the X-ray luminosity function of the ROSAT Brightest Cluster Sample (Ebeling et al.1997). The agreement indicates that this bias, if present in our sample, is small, or at least it is at work in the same manner in the non-imaging and in the ROSAT data.