5 Discussion

In this paper we presented the first detailed X-ray observation of the distant Abell cluster A33, obtained with the Beppo-SAX satellite. We have closely examined and clarified the complex X-ray emission in the direction of A33. The analysis of the X-ray data revealed the presence of three different X-ray sources in the field of A33. The X-ray counterpart of the cluster is 1SAXJ0027.2-1930. We present a spectroscopic redshift for A33, applying a $\sim 20\%$ correction to the previous photometric estimate. From optical spectra of six cluster galaxies we measure a redshift $z=0.2409\pm0.0009$ and a velocity dispersion along the line of sight $\sigma_{{\rm los}}= 472^{+295}_{-148}$ km s^-1. The dominant X-ray component (incorrectly linked with A33 in the past) is associated with a blend of an AGN and M star, while the X-ray emission from A33 is $\sim 4$ times fainter. Using the proper X-ray flux and measured redshift, we determine a more realistic cluster luminosity of $L_{2-10~{\rm keV}} = (7.7\pm0.93)~10^{43} ~h^{-2}_{50}$ erg s^-1, one to two orders of magnitude lower than previous attempts. The MECS spectral resolution also allows us to determine that the intracluster gas temperature is $T\ =\ 2.91^{+1.25}_{-0.54}$ keV. No useful information on the cluster abundance is given due to the low count rate of the source in the MECS detector.

In the following we will focus on measured quantities such as the low temperature and low velocity dispersion. We are dealing here with a moderately rich (R = 1) and distant (D = 3) Abell cluster but with X-ray luminosity and temperature more typical of nearby (z < 0.1) poor clusters. The temperature of A33 is commensurate with the predictions from its X-ray luminosity from the $L_{\rm X}-T$ relation by David et al. (1993) and Arnaud & Evrard (1999). There is an extensive literature on the correlation between these two basic and measurable quantities (Edge & Stewart 1991; Ebeling 1993; David et al. 1993; Fabian et al. 1994; Mushotzky & Scharf 1997; Markevitch 1998; Arnaud & Evrard 1999). Comparing the bolometric luminosity of A33 with the best fit relation, log( $L_{{\rm X}}$)=(2.88 $\pm$ 0.15) log(T/6 keV) +(45.06 $\pm$ 0.03) obtained by Arnaud & Evrard (1999), analyzing a sample of 24 low-z clusters with accurate temperature measurements and absence of strong cooling flows, we would expect for the A33 a temperature of 3.4 keV, as compared with our deduced value $\ 2.9\ ^{+1.25}_{-0.54}$. The $L_{\rm X}-T$ relation does not seem to evolve much with redshift since z=0.4 (Mushotzky & Scharf 1997). Note however that the ASCA data that they use show a strong bias at the low-luminosity end of the distribution due to the absence of objects in the lower luminosity range in the ASCA database. The present data on a cluster at about 0.2 are thus important to fill in the gap in the $L_{\rm X}-T$ relationship found among rich clusters and groups (see Mushotzky & Scharf 1997).

The measured velocity dispersion of A33 is also commensurate with the predictions from the $\sigma-T_{{\rm X}}$ relationship. A large number of authors (see Table 5 in Girardi et al. 1996, or Table 2 in Wu et al. 1998, for an exhaustive list of papers on the subject) have attempted to determine the $\sigma-T$ using different cluster samples in order to test the dynamical properties of clusters. Girardi et al. (1996) have derived a best fit relation between the velocity dispersion and the X-ray temperature, with more than 30$\%$ reduced scatter with respect to previous work (Edge & Stewart 1991; Lubin & Bahcall 1993; Bird et al. 1995; Wu et al. 1998, among others). If we substitute the temperature of 1SAXJ0027.2-1930 in the best fit relation log($\sigma$) = (2.53 $\pm$ 0.04)+(0.61 $\pm$ 0.05)log(T), derived by Girardi et al. (1996) a value of 650 km s^-1would be expected for the 1-D velocity dispersion, somewhat higher but within the uncertainties of the measured value from six cluster members of A33. If we assume energy equipartition between the galaxies and the gas in the cluster ($\beta$ = 1) and we use the measured temperature of 2.9 keV from the SAX data in the equation $\beta = \mu {\rm m_p} \sigma_v^{2} / k T_{{\rm gas}}$ (where $\mu {\rm m_p} =0.62$, for solar abundance), we obtain a velocity dispersion of 665 km s^-1.

The data for A33 are also consistent with the relation $\sigma_{{\rm los}} \propto (T/{\rm keV})^{0.6\pm0.1}$ found by Lubin & Bahcall (1993) and increase its statistical significance in the low temperature ( $T \ \raise -2.truept\hbox{\rlap{\hbox{$\sim$ }}\raise5.truept \hbox{$<$ }\ }3$ keV) range and at intermediate redshifts ( $z \sim 0.2$) where only a few clusters have measured values of $\beta$. This issue will be discussed in a forthcoming paper.

We have also found that the bright source 1SAXJ0027.1-1926 has an extended appearance which is due to the blending of two different sources: an AGN at z = 0.227 and approximate B magnitude $M_B \approx -23.9$ (derived from the apparent B magnitude as given in the APM scans) and an M-type star. The X-ray spectrum does not show any line features, and it is contaminated by the emission of the M star. Given the low statistics we did not try to disentangle the two contributions but we consider an upper limit to the AGN emission using the $F_{\rm X}/F_{V}$for the M star. The ROSAT BSC source found at a position consistent with the coordinates of 1SAXJ0027.1-1926 is most probably associated with the AGN. The distance between the foreground AGN and the cluster is $\Delta ~d_{\rm L} \approx 89.2~h^{-1}_{50}~{\rm Mpc}$. At the redshift of the AGN, the observed total flux corresponds to a luminosity $L_{\rm X} \ \raise -2.truept\hbox{\rlap{\hbox{$\sim$ }}\raise5.truept \hbox{$<$ }\ }4.5~10^{43}$ erg/s, which can be considered as an upper limit to the AGN luminosity.

We also detected a point-like faint source, 1SAXJ0027.0-1928, for which no X-ray spectroscopic identification was possible. The 2-10 keV spectrum of this source can be fitted by both thermal and non-thermal models (see Table 5) but we do not elaborate further given the poor statistics.

Acknowledgements

S.C. acknowledges useful discussions with G. Hasinger and C. Sarazin. Partial financial support from ASI, NASA (NAG5-1880 and NAG5-2523) and NSF (AST95-00515) grants is gratefully acknowledged. We appreciate the generosity of B. Tully who allowed us to take some images and spectra during his observing runs.