next previous
Up: High matter density peaks abundances


Subsections

3 Properties of the main absorption systems

In the following, we briefly discuss the relative abundances of some chemical elements for two DLASs and a sub-DLAS detected in the present spectra. All the abundances are given relative to the solar values of Grevesse & Anders (1989) and Grevesse & Noels (1993), in the notation $[X/Y]=\log~(X/Y)_{\rm
obs}-\log ~(X/Y)_{\odot}$ (see Table 2). A summary of the obtained chemical abundances is reported in Table 3.

 

 
Table 2: Adopted solar abundances for the relevant chemical elements.
H Hydrogen 0.00
C Carbon -3.45
N Nitrogen -4.03
O Oxygen -3.13
Si Silicon -4.45
S Sulphur -4.79
Fe Iron -4.49



 

 
Table 3: Measured relative chemical abundances. If not otherwise stated the error on the logarithmic column densities and on the relative abundances is 0.1.
Object Redshift $\Delta v^{a}$ log N(H I)b [Fe/H] [Si/H] [N/H] [S/H] [N/S] [C/Fe] [0/Fe]  
UM681 1.78745 -150.5 18.6     $-0.6\pm0.2$ $-0.3\pm0.2$        
Q2343+1232 averagec   20.35 $-1.2\pm 0.2$   -1.1 -0.7        
  2.43125 0.0           -0.3      
Q2344+1228 averagec   20.4 $-1.8\pm 0.2$ -1.85 -2.75          
  2.53746 -35.6             <0.06 <0.2  
a Relative velocities as reported in the corresponding figures.
b Since it is not possibile to disentangle the velocity structure of the H I Lyman-$\alpha $ absorption lines, the profiles have been fitted with
one or two components at the redshift of the stronger components observed in the neutral and singly ionised absorption lines.
c Average value obtained considering the sum of the column densities of all the components of the iron absorption profile.


3.1 The sub-DLAS at ${z_{a}\sim 1.788}$ in UM681


  \begin{figure}
\par\includegraphics[width=8.8cm,height=8cm,clip]{MS2352f1.eps}\end{figure} Figure 1: $z \sim 1.788$ - coincidence between the sub-damped Lyman-$\alpha $ absorption line at $z_{\rm a}\simeq 1.788$ in the spectrum of UM681 (bottom panel) and a weak Lyman-$\alpha $ absorption line ($\log
N($H I $) \simeq 13.81 \pm 0.05$) in the spectrum of UM680 (top panel). The spatial separation between the two LOSs at this redshift is $\sim $ $870\ h^{-1}$ kpc. The dotted lines mark the position of the 2 components fitting the sub-DLAS at redshifts $z_{\rm a}=1.78745$ and 1.78885 (origin of the velocity axes).


  \begin{figure}
\par\includegraphics[width=8.8cm,height=11cm,clip]{MS2352f2.eps}\end{figure} Figure 2: UM681: ionic transitions of the system at $z_{\rm a}\simeq 1.788$. The dotted vertical lines mark the position of the components discussed in the text. The two thick, dashed lines corresponds to the redshifts of the two components used to fit the H I Lyman-$\alpha $ absorption.

This is a newly detected metal absorption system. The H I Lyman-$\alpha $ absorption at this redshift shows a clear Lorentzian wing on the blue side of the velocity profile (see Fig. 1). In order to obtain an estimate of the H I column density, we fit the profile with two components at the redshifts of the two strongest groups of lines seen in the singly ionised element profiles (see Fig. 2). At $z_{\rm a} \simeq 1.78745$ ( $v \simeq -150$ km s-1 in the figures), we measure a column density $\log
N$(H I) $\sim 18.6\pm0.1$ and at $z_{\rm a} \simeq
1.78885$ ( $v \simeq 0$ km s-1), $\log
N$(H I) $\sim 19.0
\pm 0.1$. The associated heavy element transitions have a complex structure spread over about 250 km s-1 with at least 8 narrow components in Fe II, Si II and Al III and 4 components in N I and S II (see Fig. 2). O I, C II and Si III are either saturated or blended, Si IV, C IV and Al II are outside our wavelength range.

The observed H I column density is not high enough to assure that no ionisation corrections have to be applied to get the relative chemical abundances (see e.g. Viegas 1995). We derive an indicative measure of the nitrogen and sulphur abundances by using the Cloudy software package (Ferland 1997) to build a photoionisation model and estimate the ionisation corrections. We consider the UV background flux due to quasars and galaxies (Madau et al. 1999) and try to recover the observed column densities of the components at $z_{\rm a} \simeq 1.78745$ and 1.78765 ( $v \simeq -150$ and -130 km s-1 in Fig. 2). The results do not change significantly if we adopt an ionizing spectrum due to a single stellar population of solar metallicity and 0.1 Gyr. The resulting ionisation corrections are $\sim $2.3, 0.2 and 14 to be multiplied for the ratios N I/H I, S II/H I and N I/S II respectively. Since the corrections are relatively small in the first two cases, we compute the corresponding abundance ratios. N I $\lambda~1199$ is partially blended (see Fig. 2), thus we estimate the nitrogen abundance from the two components at lower redshift ( $v \simeq -150$ and -130 km s-1 in Fig. 2), associated with the $z_{\rm a} \simeq 1.78745$ H I component with $\log
N($H I $) \simeq
18.6\pm 0.1$. From this we derive [N/H]  $\simeq -0.6 \pm
0.2$ corrected for the ionisation. This value is about one order of magnitude larger than the higher value measured in DLASs published in the literature (Centurión et al. 1998; Lu et al. 1998). The corrected sulphur abundance ratio for the same components is [S/H]  $\sim -0.3 \pm 0.2$ which again is about one order of magnitude larger than what is observed in DLASs.

These measurements, although slightly uncertain, suggest that sub-DLAS could have higher metal abundances than DLASs, as already observed in LLS (e.g. D'Odorico & Petitjean 2001), and probe a more evolved chemical stage of high redshift galaxies when gas has been partly consumed by star formation.

  \begin{figure}
\par\includegraphics[width=8.8cm,height=11cm,clip]{MS2352f3.eps}\end{figure} Figure 3: Q2343+1232: ionic transitions associated to the DLAS at $z_{\rm a} \simeq 2.43125$. The dotted line at the extreme left marks the position of the satellite sub-system at $z_{\rm a} = 2.42536$ (see text). The dotted lines on the right are drawn at the redshifts $z_{\rm a} =2.42834$ and 2.43125 (origin of the velocity axes) of the components discussed in the text.

3.2 The DLAS at ${z_{a}\simeq 2.43125}$ in Q2343+1232

This DLAS, as well as the one seen along the LOS of Q2344+1228 (see next section), has first been detected by Sargent (1987). The two QSOs were observed recently with HIRES+Keck and the relative chemical abundances of the two systems were used in statistical samples (Rauch et al. 1997; Lu et al. 1998; Prochaska & Wolfe 1998, 2001; Prochaska et al. 2001). The metal absorption complex corresponding to this damped system is made of two groups of lines. The major one counts at least 8 components, with the strongest one at $z_{\rm a} \simeq 2.43125$ (v = 0 km s-1 in Fig. 3). This component is heavily saturated in C II, O I, Mg II, and Si II. It shows absorption due to the two triplets of N I, $\lambda~1134$ Å and $\lambda~1200$ Å, and to the S II triplet, $\lambda\lambda\lambda~1250,1253,1259$ Å. The Si IV doublet is clearly identified in this complex, the C IV doublet is outside our spectral range but was detected by Sargent et al. (1988). A satellite sub-system is observed at more than 500 km s-1 from the centre of the main one, at $z_{\rm a} \simeq
2.42536$. It is very weak and shows transitions due to C II, Mg II, Si II and Si III.

Since we cannot disentangle the velocity structure of the hydrogen absorption, we assume a single component at the redshift of the strongest component observed in singly ionised lines at $z_{\rm a} \simeq 2.43125$. The total H I column density is $\log
N($H I $)
\simeq 20.35 \pm 0.05$ and the error is mainly due to the uncertainty in the position of the continuum. The average iron abundance is [Fe/H]  $\simeq -1.2\pm 0.2$. While, we obtain [S/H]  $\simeq -0.7 \pm 0.1$ and [N/H]  $\simeq -1.1 \pm 0.1$. Those estimates are $\sim $0.2 and 0.5 dex higher respectively, than those reported by Lu et al. (1998). The difference for sulphur is within the uncertainties (at the $3~\sigma$ level), while the larger one for nitrogen could be due to the fact that Lu et al. (1998) used the saturated N I triplet at $\lambda~1200$ Å.

As previously stated, the main component is badly saturated for all the observed transitions due to C, O and Si, so relative abundances for these elements cannot be determined. On the other hand, we can study the abundance ratios of S, N and Fe at this redshift, assuming that ionisation corrections are negligible. This hypothesis is supported by the large column density characterising this component and by the absence of Si IV absorption. We measure column densities: $\log
N($S II $)\simeq 14.75\pm0.05$, $\log
N($N I $)\simeq 15.16\pm0.05$ (where we do not consider the N I triplet at $\lambda~1200$ Å because it is saturated and affected by the wing of the damped H I Lyman-$\alpha $ line) and $\log
N($Fe II $)\simeq
14.49\pm 0.08$. From which we derive the abundance ratios: [S/Fe]  $\simeq 0.6\pm 0.1$, and [N/S]  $\simeq
-0.3\pm 0.1$. The latter abundance ratio, which is not affected by dust, can be compared with the ratios [N/O] measured for metal poor Galactic stars making the assumption, [S/O] $\equiv 0$ as reported by Centurión et al. (1998). Our measurement is larger than any other for DLASs present in the literature (see also Lu et al. 1998) and it is consistent with values obtained for H II regions in dwarf galaxies.

The strong Si IV absorption imply that ionisation corrections could be necessary to determine the abundance ratios in the other components of the system. Nothing can be said on the H I column density corresponding to the single components. This makes hard the realization of a photoionisation model to determine the ionisation corrections. We report the measured column densities of sulphur and nitrogen (a faint absorption is observed for the transition N I $\lambda~1200$) relative to the component at $z_{\rm a} \simeq 2.42834$ ( $v \simeq -250$ km s-1 in Fig. 3), $\log
N($S II $)\simeq
14.22\pm0.05$ and $\log
N($N I $)\simeq 13.5\pm0.05$.

  \begin{figure}
\par\includegraphics[width=8.8cm,height=11cm,clip]{MS2352f4.eps} \end{figure} Figure 4: Q2344+1228: ionic transitions associated to the DLAS at $z_{\rm a} = 2.53788$ (origin of the velocity axes). The left dotted line marks the position of the component at $z_{\rm a} = 2.53746$ discussed in the text.

3.3 The DLAS at ${z_{a}\simeq 2.53788}$ in Q2344+1228

The damped H I Lyman-$\alpha $ absorption line of this system has been fitted with a single component at the redshift of the strongest component observed in the neutral and singly ionised lines of associated heavy elements ( $z_{\rm a} \simeq 2.53788$). The H I column density is $\log
N($H I $) \simeq 20.4
\pm 0.1$, where the error on the column density is due mainly to the positioning of the continuum level.

In Fig. 4, we show the ionic transitions observed for the DLAS. The C IV doublet is outside our wavelength range and it was not observed in the low resolution spectrum by Sargent et al. (1988). The iron and silicon column densities can be derived from non saturated lines to obtain the average values, [Fe/H]  $\simeq -1.8 \pm 0.2$, and [Si/H]  $\raisebox{-5pt}{$\;\stackrel{\textstyle >}{\sim}\;$ }-1.85 \pm
0.1$. While, [N/H]  $\simeq -2.75 \pm 0.11$, which is in good agreement with the value found by Lu et al. (1998).

The absence of high ionisation lines and the simple velocity profile of the system allow the assumption that ionisation corrections are negligible in this case. We can thus obtain reliable abundance ratios from the column densities of the transitions observed in the single components. N I is observed in the central component, we compute the abundance ratio [N/Fe]  $\simeq -0.8 \pm 0.1$, which is consistent with previous measurements for DLAS. The central component is unusable to derive reliable abundance measures for other chemical elements because all the observed lines are heavily saturated.

In the component at lower redshift ( $z_{\rm a} \simeq
2.53746$, $v\simeq -35.6$ km s-1 in Fig. 4), the ratio [Si III/Si II $\simeq -0.66\pm 0.07$ implies that ionisation corrections are small at this velocity. We observe transitions due to Si II and Fe II that are not saturated, from which the relative abundance [Si/Fe]  $\simeq 0.2 \pm 0.1$ is obtained. On the other hand, C II and O I are slightly saturated but not going to zero, this results in the upper limits: $\rm -0.02< [C/Fe]<0.06$ and $\rm -0.06< [O/Fe]<0.2$. Reliable measures of O and C abundances are quite rare. We discuss the implications of our result in the following section.

3.4 Comments on the measures of [O/Fe] and [C/Fe]

$\alpha $-capture elements are mainly produced by type II SNe which should dominate in the early stages of the chemical evolution of galaxies, while type I SNe contribute iron peak elements later on. Therefore, the [$\alpha $/Fe] abundance ratio can be used to trace the chemical evolution history and, to a certain extent, the nature of galaxies. Oxygen and sulphur are more reliable estimators of $\alpha $-element abundances than silicon which is subject to dust depletion. Abundance studies of carbon (e.g. Tomkin et al. 1995) indicate that in the disk of our Galaxy [C/Fe] and [$\alpha $/Fe] show similar trends with [Fe/H].

Measures of C and O in DLASs are generally complicated by the fact that often the only available lines are C II $\lambda~1334$ and O I $\lambda~1302$ which most of the times are heavily saturated. In the DLAS described in Sect. 3.3, we constrain the values of the O/Fe and C/Fe ratios, considering a single component which is only mildly saturated and not going to zero. We derive that the C/Fe ratio is consistent with solar while the O/Fe and Si/Fe ratios are consistent among them and show a very small enhancement. The average iron abundance of the system is about 1/100 solar. Our result, together with the recent measures by Molaro et al. (2000) for the DLAS in the spectrum of Q0000-26, indicates that there is no evidence for the [O/Fe] ratio to be over-solar in DLAS. This is at a variance with what is observed in the atmosphere of Galactic stars at the same metallicity (but see also Dessauges-Zavadsky et al. 2001).

The abundance pattern which is closest to the above data is that of an old starburst, as is observed at the boundaries of our galactic disk, although, in general, for larger iron abundances (Chiappini et al. 1999).


next previous
Up: High matter density peaks abundances

Copyright ESO 2002