4. Metal absorption systems in HS 1216+5032 A and B
We now give a description of the most remarkable metal absorption systems observed in HS 1216+5032 A and B: a MgII system at , and a strong CIV system at .
4.1. The CIV systems at in HS 1216+5032 A and B
Three strong CIV systems at , 0.721 and 0.726 are identified through the doublets at Å in both spectra of HS 1216+5032. Although some of the CIV lines are under the significance level, the well-determined redshifts of the doublet components make their identification unambiguous; consequently, all six CIV lines appear in Tables 1 and 2.
The rest frame equivalent widths of the lines range between 0.36 and 1.03 Å in the A spectrum, and between 0.72 and 1.19 Å in B. The large column densities implied by these line strengths, make this system a firm candidate for a Lyman-limit system (Sargent et al. 1989). Also, the large line widths in A and B, FWHM Å, suggest that these CIV systems will reveal several narrower components at higher resolution.
No further metal lines are found at these redshifts. Line A13 could be identified with SiIV at , but two strong Ly lines at Å make the detection of the second doublet line impossible. No transitions by low-ionization species (e.g., CII ) are detected. In the low-resolution optical spectrum of HS 1216+5032 B (see Fig.  in Hagen et al. 1996), a significant absorption feature at Å could tentatively be identified with a MgII doublet associated to these CIV systems, but no absorption feature is seen at this wavelength in the optical spectrum of A.
The small velocity differences of km s-1, km s-1, and km s-1 at , 0.721 and 0.726, respectively, between the CIV lines in A and the corresponding ones in B suggest strongly that these CIV systems are physically associated. Therefore, the large velocity span of roughly 1500 km s-1 along each LOS can only be explained if the gas is virialized by a cluster of galaxies. The virial mass within the radius given by the half separation between the LOSs is derived to be . Consequently, if CIV systems arise in the extended halos of galaxies (Bergeron & Boisse 1991), the present data gives evidence that the CIV absorbers at in HS 1216+5032 A and B arise either in the highly ionized intra-group gas of a galaxy cluster, or in small structures associated with the cluster.
The whole absorption complex shows an asymmetry in the sense that the weakest system in spectrum A has the same redshift as the strongest in B: the ratio of the line's equivalent width in A to that in B varies between 1.4 and 3.3. If CIV systems in A and B with similar redshifts are physically associated, then the different line strengths have direct implications for the size of these absorbers. This is because such gas inhomogeneities imply that the LOSs sample the clouds on spatial scales similar to the transverse cloud sizes. The interpretation remains valid if the lines in the present FOS spectra result from many unresolved velocity components, because in that case stronger lines result from more narrow components than weaker lines do, a picture that is still compatible with gas inhomogeneities. All these three CIV absorption systems should therefore arise in clouds with characteristic transverse lengths of kpc, larger than the statistical lower limits of kpc derived by Smette et al. (1995) for CIV absorbers at .
Alternatively, it is still possible that the CIV absorption is correlated in redshift, but occurs in distinct, separated structures. At higher resolution, Rauch (1997) has shown that in gas associated with metal systems density gradients on sub-kpc scales are not uncommon. The logical conclusion would be that C IV absorbers are composed of a large number of small cloudlets. However, if that is the case of the absorbers, the correlation in redshift between lines in A and B is difficult to explain without invoking cloudlets aligned along a filamentary or sheet-like structure.
4.2. A MgII system at in HS 1216+5032 B?
There is a strong and blended feature at Å in the B spectrum for which we do not find any plausible identification with other observed metal systems. One possible identification is absorption by two MgII doublets at and . Although the Gaussian profile fit is not able to resolve the single lines in the red trough of the absorption feature, the identification is supported by a good match between the line positions and profiles of the doublet. The red wing of the whole absorption feature could be blended with a Ly line because it coincides quite well with a strong Ly line at in A (A47). However, no FeII lines are detected at this redshift, nor further transitions by low-ionization species. The MgI line is possibly present, but the match in wavelength with line B50, Å, is not very good. Under the MgII -hypothesis, the total rest-frame equivalent width of the stronger MgII doublet component would be Å, which is somewhat larger than typical values found in gas associated with damped Ly (DLA) systems at high redshift (Lu et al. 1996) or in the Milky Way (Savage et al. 1993). This could be explained if the lines are indeed made up of several narrower components, as usually seen in DLA systems. The redshift difference between both MgII systems implies a velocity span of roughly 300 km s-1, also typical of DLA systems. Therefore, there is some evidence that these MgII lines might be associated with a DLA system at .
Regardless of the DLA-system interpretation, however, even the absence of FeII and - possibly - also MgI lines associated with MgII of such a strength is difficult to explain considering the incidence of the former ions in MgII -selected samples (e.g., Bergeron & Stasinska 1986; Steidel & Sargent 1992). Consequently, an alternative identification of the Å feature as HI BAL at km s-1 (see Sect. 5 and Fig. 2) must be considered as well. The absence of a corresponding HI BAL profile, however, is also in this case remarkable, but could be explained if the absorber does not completely cover the continuum source (see below).
© European Southern Observatory (ESO) 2000
Online publication: May 3, 2000