Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 353, 457-464 (2000)

Previous Section Next Section Title Page Table of Contents

4. Discussion and conclusions

4.1. The Lyman limit absorber

We have detected an object, S6, with extended line emission north and east of the quasar Q1205-30. The emission line of S6 is detected in a narrow band tuned to Ly[FORMULA] of a Lyman Limit absorber in front of Q1205-30, and the most likely interpretation is, that we see the Lyman Limit absorber in Ly[FORMULA] emission. Spectroscopic follow-up is required to confirm this tentative conclusion. The large residuals from the PSF subtraction of the quasar introduce errors in both the total line luminosity observed, and the impact parameter. We have therefore considered two extreme cases described in Sect. 3.2; the minimal Ly[FORMULA] flux case (model A) and the maximal Ly[FORMULA] flux case (model B). For the Ly[FORMULA] luminosity and the impact parameter in the two models we find L=11.6[FORMULA]1.1 [FORMULA] 1041 h-2 erg s-1(14.1[FORMULA]1.3 [FORMULA] 1041 h-2 erg s-1) and d=6.5 h-1 kpc (5.4 h-1 kpc) for model A (model B). In the following we will assume the more likely model B to be correct. Fynbo et al. (1999a, their Table 4) lists the Ly[FORMULA] line luminosities of the known galaxy counterparts of high redshifts DLAs and LLSs, namely the DLAs towards PKS0528-250 (named S1), Q0151+048A (named S4) and Q2059-360, and the LLS towards Q2233+131. The observed Ly[FORMULA] line luminosity of S6 is within the range 1.1-12 [FORMULA] 1042 h-2 erg s-1 of these other systems.

The size of the line emission region is 6[FORMULA]4 arcsec2 corresponding to 22[FORMULA]14 h-2 kpc2 at z=3.036. This size and the morphology of S6 are both near-identical to those of the emission line object S4, which has been shown in a spectroscopic study (Moller et al. 1998) to be a DLA galaxy unrelated to the underlying quasar. Before the precise systemic redshifts of both the QSO and S6 have been measured we cannot exclude that S6 is related to Q1205-30.

From neither S4, nor S6, do we detect what is obviously continuum emission from a high redshift galaxy. This, however, is not surprising as continuum emission from the high redshift DLA galaxy S1 was not detected in the deep ground based images of S1 (Moller & Warren, 1993a), but a compact galaxy was subsequently found in the HST images (Moller & Warren 1998). Since the impact parameters of S1, S4, and S6 are similar, clearly HST imaging is required to resolve the question of their continuum morphology.

Both the comparatively large luminosity and the extended morphology of S6 is in contrast to six faint and compact emission line galaxies (S7-S12) found at much larger distances from the quasar. We hence consider it likely, as in the case of Q2059-360 (Leibundgut & Robertson 1998, Fynbo et al. 1999a), that the emission from S6 is significantly influenced by the proximity of the active nucleus. It is, however, possible that gravitational lensing by the red galaxy g2, if it is a foreground elliptical galaxy, makes S6 appear more extended and increases its observed flux. Here we estimate the likely strength of this effect.

Assuming that g2 is at z=0.5 and has an isothermal mass profile with a velocity dispersion of 300 km s-1, the radius of its Einstein ring is given by


where [FORMULA] and [FORMULA] are the proper angular diameter distances between g2 (Lens) and S6 (Source) and between the earth and S6 respectively, in the assumed cosmology. Gravitational lensing in this simple model moves S6 and Q1205-30 by an angle [FORMULA] in the radial direction away from g2 causing a tangential stretching of S6 as well as an increase of the observed impact parameter of S6 relative to Q1205-30 by a factor of [FORMULA]=2.3 (where [FORMULA]=2.8 arcsec is the observed impact parameter of g2 relative to the QSO). The magnification of the flux of S6 would be in the range 2-10 depending on the unknown size and orientation of the diamond caustic of g2. A possible counter image on the eastern side of g2 would make S6 look more extended in the relatively low resolution of the narrow band image, in the direction of the observed elongation which would be of order 2[FORMULA].

4.2. The field population of Ly[FORMULA] emitting galaxies

The presence of the six emission line galaxies S7-S12 in the field of Q1205-30 is interesting for several reasons. The faintness of the broad band magnitudes of the objects makes it unlikely that they are low redshift (z [FORMULA] 0.31) galaxies with oxygen emission lines in the narrow filter. The most likely identification of the emission lines is Ly[FORMULA] at the same redshift as the Q1205-30 LLS absorber, and in what follows we shall assume this to be the case.

Only one of the six emission line galaxies (S9) would have been selected as an LBG in current ground based programmes. The remaining five are too faint in the broad bands (see Table 4). The number of LBGs brighter than R(AB)=25.5 with redshifts between 3.0 and 3.5 selected from ground based surveys is 0.4[FORMULA]0.07 galaxies arcmin-2 (Steidel et al. 1996), and we hence expect 11 in our 27.6 arcmin2 field. Assuming that LBGs have redshifts uniformly distributed between 3.0 and 3.5 we would expect on average 0.4 LBGs within the redshift range [FORMULA] corresponding to the width of the narrow filter. In the Hubble Deep Field North the LF of LBGs has been extended to R(AB)=27 (Steidel et al. 1999). Integrating this LF leads to an expected number of LBGs of about 0.3 galaxies arcmin-2 within the redshift range [FORMULA] or on average 9 in our 27.6 arcmin2 field (see Sect. 4.2 in Fynbo et al. 1999a for a discussion on the relation between the redshift density [FORMULA] of DLAs and the LF of LBGs). Hu et al. (1998) have reported on detections of Ly[FORMULA] emitting galaxies by means of narrow band imaging and long slit spectroscopy in random fields with densities about 4 arcmin-2 per unit redshift down to similar depths in observed flux as in our sample, but at somewhat higher redshifts.

Hence, narrow/broad band Ly[FORMULA] imaging indicate the existence of a significant population of high redshift galaxies that is not included in current ground based LBG samples and which is possibly identical to the faint LBGs detected by HST. Those galaxies, which make up the faint end of the high redshift galaxy LF, have large Ly[FORMULA] equivalent widths and are about a factor of 10 more numerous than the bright LBGs of current ground based samples. This result is in good agreement with our earlier conclusion based on searches for DLA galaxies (Moller & Warren 1998; Fynbo et al., 1999a) that there is a significant population of faint high redshift galaxies which, despite their very small HI absorption cross-section, make up most of the high redshift DLA absorbers.

Total Ly[FORMULA] luminosities for S7-S12 are in the range 3.3-9.5 [FORMULA] 1041 h-2 erg s-1 for [FORMULA]=1.0 and 12-34 [FORMULA] 1041 h-2 erg s-1 for [FORMULA]=0. Using the Kennicutt (1983) prescription SFR = L(H[FORMULA])/1.12 [FORMULA] 1041 erg s-1 and assuming L(Ly[FORMULA])/L(H[FORMULA])=10 and negligible dust extinction, we find star formation rates in the range 0.2 - 0.8 h-2 [FORMULA] yr-1 for [FORMULA]=1.0 and 1.1-3.0 h-2 [FORMULA] yr-1 for [FORMULA]=0. The large equivalent widths of the emission lines make strong dust obscuration unlikely. The star formation rates for R(AB)[FORMULA]25.5 LBGs as estimated from Balmer and OII emission line strengths by Pettini et al., (1998) fall in the range 20 - 270 [FORMULA] yr-1 for [FORMULA]=0.2 and h=0.7. For [FORMULA]=1.0 this corresponds to 4-55h-2 [FORMULA] yr-1. Hence, the Ly[FORMULA] emitting galaxies have star formation rates about an order of magnitude lower than what is measured for the bright LBG population. Nevertheless, since the population of Ly[FORMULA] emission line galaxies is much more common than the brighter LBG population, it is still an open question if the integrated SFR is dominated by the faint or the bright end of the high redshift galaxy LF.

As seen in Fig. 1 the faint galaxies are all located in a region north west of the QSO. Their projected impact parameters from Q1205-30 range from 156 to 444 h-1 kpc, and four of the six galaxies are found in a small region of projected size 100 [FORMULA] 100 h-2 kpc2. As for the compact group S1-S2-S3 associated with a z=2.81 DLA absorber, this compact group of faint galaxies associated with a z=3.036 LLS absorber is destined to merge on a short timescale (Navarro et al. 1995; Haehnelt et al. 1998). This, then, supports the suggestion (Warren & Moller 1996; Moller & Warren 1998) that high redshift, high column density absorbers predominantly are proto-galactic sub-clumps in the process of early galaxy assembly, rather than fully formed large rotating disks.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: December 17, 1999