SpringerLink
Forum Springer Astron. Astrophys.
Forum Whats New Search Orders


Astron. Astrophys. 346, 359-368 (1999)

Previous Section Next Section Title Page Table of Contents

5. Discussion and conclusions

The two sources H3 and H5 are at the same redshift, but their stellar populations are noticeably different. Both of them are well fitted by burst or continuous star-formation models. Metallicities much higher than [FORMULA] seem to be excluded by the photometric data in the case of H3, and they have a lower probability in H5. Both galaxies are roughly compatible with solar metallicities within the [FORMULA] confidence level: [FORMULA] for H3 and [FORMULA] for H5. The stellar population seen in H5 is younger than in H3, on the basis of a burst model with [FORMULA]. The main source of uncertainty when deriving the restframe properties of these objects is the value of [FORMULA], and this is clearly shown in the case of H3 where the [FORMULA] uncertainties are [FORMULA] magnitudes. The two sources are intrinsically bright: roughly [FORMULA] for H5 and [FORMULA] ([FORMULA]) magnitudes for H3 (here [FORMULA] is the local value, from Loveday et al. 1992). This, combined with the high gravitational amplification and the presence of relatively strong emission lines, has allowed us to obtain a spectroscopic redshift for these sources using a 4m telescope. Among the high redshift candidates in our sample behind A2390, 4 additional ones are at the same photometric redshift. A subsequent spectroscopic survey using a 8m class telescope is needed to go further on this study, especially to estimate more precisely the metallicities from UV absorption lines.

When comparing our results on H3 to those by Frye & Broadhurst (1998) and Bunker et al. (1998), we find them in fairly good agreement, after correction for the difference in the amplification factor, which is 3 times higher in their lens model. The age of the system and the reddening value given by Bunker et al. for H3 are included within our [FORMULA] best solution region, which is a remarkable result taking into account that both the photometric data (HST excepted) and the analysis are completely independent. The high amplification factor obtained by Frye & Broadhurst (1998) corresponds to the maximum value attained in our model for H3, on the region neighbouring the critical line, but the surface-averaged value in our case is 3 times lower.

Compared to other known [FORMULA] galaxies, H3 and H5 belong to the bright end of the field population (Lowenthal et al. 1997; Steidel et al. 1996a and 1996b). They are slightly brighter than the [FORMULA] objects found by Trager et al. (1997) in the cluster lens Cl0939+4713, although the amplification factor is poorly known in this case. They are also intrinsically brighter than the [FORMULA] galaxy found by Dey et al. (1998). These sources are not resolved in their width, where the lens inversion is limited by the resolution of the WFPC2 images (about [FORMULA], [FORMULA] kpc with [FORMULA]), but they are resolved on their length. H3 and H5 are splitted into several small, compact and bright subclumps, all aligned towards the same direction. The total length of these emitting regions is similar to that of the compact cores of the [FORMULA] field sample by Steidel et al. (1996a). H3 is more elongated than H5, and probably more dusty. The linear separation between H3 and H5 ([FORMULA] with [FORMULA]) and their peculiar morphologies strongly point towards a hierarchical merging process as a likely scenario for the formation of the brightest spheroids. Their photometric SEDs are compatible with a wide range of metallicities at a [FORMULA] level, all with [FORMULA]. Nevertheless, as shown in Sect. 4, broad band photometry does not allow a precise estimate of the metallicity. A spectroscopic study using UV restframe absorption lines, or near-IR spectroscopy, is urgently needed for these purposes. A relatively high metallicity for the bright subclumps would imply that we are actually seeing an advanced step in this merging process, where stars are forming from a metal enriched gas (see also Trager et al. 1997; Lowenthal et al. 1997, Baugh et al. 1997; Moscardini et al. 1997). It is worth noting that we are dealing with local conditions in these star forming systems. The light detected is mainly emitted in very small and compact regions ([FORMULA] kpc), which could be highly enriched compared to the remainder of the source. This is fully compatible with the best fit metallicities being [FORMULA] in these systems, because in this case the star formation becomes a local and efficient process, where the cooling rate is enhanced by metallic atoms allowing the formation of molecules and increasing the dust opacity. Such a process is discussed in details in a recent paper by Spaans & Carollo (1998).

The uncertainty in the amplification factor for H3 and H5 is 0.3 magnitudes. This means that the intrinsic luminosities and SFRs are known with an accuracy of [FORMULA]. This source of error has the same importance than the model uncertainties for a relatively well constrained SED (i.e., the uncertainty of [FORMULA] magnitudes in [FORMULA] for H3). It is worth noting that H3 and H5 are multiple images, well modelled compared to other images in this or other cluster lenses. This gives an idea of the limitation arising from lens modelling when using clusters as gravitational telescopes to access the background sources. Only the well constrained clusters are actually useful for this programme.

H3 and H5 are the first spectroscopically confirmed images of sources at [FORMULA] in this cluster, a redshift domain which is well constrained by the set of filters used here. The selection of high-redshift candidates using a photometric redshift approach, including the near-IR bands, is strongly supported by the present results. For most statistical purposes, photometric redshifts should be accurate enough to discuss the properties of these extremely distant galaxies. Conversely, the spectroscopic confirmation of the redshifts of such gravitationally amplified sources could help on the calibration and improvement of the photometric redshifts techniques up to much fainter limits in magnitude compared to field surveys.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: May 21, 1999
helpdesk.link@springer.de