Relative abundance pattern along the profile of high redshift Damped Lyman-alpha systems

0.43 MB·English

by  E. Rodriguez

(DOI: will be inserted by hand later)
Relative abundance pattern along the profile of high redshift
α ⋆
Damped Lyman- systems
E. Rodr´iguez1, P. Petitjean1,2, B. Aracil1,3, C. Ledoux4, and R. Srianand5
1 Institut d’Astrophysiquede Paris, UMR 7095, 98 bis Boulevard Arago, 75014 Paris, France
2 Observatoire deParis, LERMA,UMR 8112, 61 Avenuede l’Observatoire, 75014 Paris, France
3 Department of Astronomy,Universityof Massachusetts, 710 North Pleasant Street, Amherst,MA 01003-9305
4 European Southern Observatory,Alonso deCordova 3107, Casilla 19001, Vitacura, Santiago, Chile
5 IUCAA,Post Bag 4, Ganesh Khind,Pune 411 007, India
6
0
0 Abstract. We investigated abundance ratios along the profiles of six high-redshift Damped Lyman-α systems,
2 threeof them associated with H2 absorption, and derived optical depthsin each velocity pixel. Thevariations of

n the pixel abundance ratios were found to be remarkably small and usually smaller than a factor of two within
a a profile. This result holds even when considering independent sub-clumps in the same system. The depletion
J factor is significantly enhanced only in those components where H2 is detected. There is a strong correlation

9 between [Fe/S] and [Si/S] abundancesratios, showing that theabundanceratio patternsare definitely related to
2 thepresence of dust.The depletion pattern is usually close to theone seen in thewarm halo gas of ourGalaxy.

1 Key words.Cosmology: observations- Quasars:absorption lines -Quasars: individual:Q0528−250, Q0013+004,
v Q 1037−270, Q 1157+014, Q 0405−443
4
6
6
1 1. Introduction dwarfgalaxies(Centur´ion et al,2000)orgalacticbuilding
0 blobs(Haehnelt, Steinmetz, & Rauch,1998;Ledoux et al,
6 Damped Lyman-α systems (hereafter DLAs) observed in 1998). The answer is probably not unique. In any case,
0 QSO spectra are characterized by strong H iλ1215 ab-
DLAsrepresentthemajorreservoirofneutralhydrogenat
h/ sorption lines with broad damping wings. Although the any redshift (Storrie-Lombardi & Wolfe, 2000), and they
p definition has been restricted for historical reasons to ab- probe the chemical enrichment and evolution of the neu-
- sorptions with log N(H i) > 20.3 (Wolfe et al. 1986),
o tral Universe (see Pettini et al. 1994; Lu et al. 1996;
damping wings are easily detected in present-day, high
r Prochaska et al. 1999;Ledoux et al. 2002a and references
t quality data for much lower column densities (down to
s therein). Since abundances can be measured very accu-
a log N(H i) ∼ 18.5).A more appropriatedefinition should
rately in DLAs, we can both discuss the connection be-
:
v berelatedtothephysicalstateofthegas.Ifweimposethe tweenobservedabundanceratiosanddustcontentandto
i conditionthatthegasmustbeneutral,thenthedefinition
X trace the nucleosynthesis history of the dense gas in the
should be limited to systems with log N(H i) > 19.5 (e.g.
universe.
r
a Viegas 1995).
Since their discovery twenty years ago (Wolfe et al,
In this context, it is helpful to compare these results
1986), DLAs clearly have something to do with galaxy
with measurements in the ISM of our Galaxy. Refractory
formation. What kind of galaxy DLAs are associated
elements that condense easily into dust grains - namely,
to is, however, still a matter of debate. Some au-
Cr,Fe, Ni- are stronglydepleted (up to a factor hundred)
thors identify these systems with large rotating discs
in the ISM, while non-refractory elements remain in its
(Prochaska & Wolfe, 1997; Hou, Boissier, & Prantzos,
gaseous phase - S, Zn, and partially Si-. The amount of
2001),whileothersthinkthatDLAsarisemostlyeitherin
depletion depends on the physical condition of the gas.
Thus, different depletion patterns are observed depend-
Send offprint requests to: E. Rodr´iguez, [email protected]
⋆ ing on whether the gas is cold or warm and/or whether
Based on observations carried out at the European
the gas is located in the disc or the halo of the Galaxy
Southern Observatory (ESO), under progs. ID 65.O-0063,
66.A-0624, 67.A-0078, 68.A-0106 and 68.A-0600, with the (Savage & Sembach, 1996). The LMC and SMC also ex-
UVES echelle spectrograph installed at the Very Large hibit different gas-phase abundance ratios (Welty et al,
Telescope (VLT) unit KUEYEN on Mount Paranal, Chile. 1999).
However,a particularnucleosynthesishistory can give the six highest S/N ratio spectra from the list of 33
rise to peculiar metallicity patterns and mimic the pres- quasars (24 DLAs and 9 sub-DLAs) observed during the
ence ofdust. Tinsley(1979)suggestedthattype Iasuper- VLT survey for molecular hydrogen in DLAs (Ledoux et
nova are the major producers of Fe. An enhancement in al. 2003). The QSOs were observed with the Ultraviolet
[α/Fe]ratios(α-elementsaremostlyO,S,Si)couldreflect andVisibleEchelleSpectrograph(D’Odorico et al,2000),
an IMF skewed to high masses and therefore a predom- mounted on the 8.2 m Kueyen telescope operated at
inant role of type II supernova. For very low metallicity Cerro Paranal,Chile, during the observation periods P65
stars([Fe/H]<−3)intheGalaxy,largevariationsinsev- to P68. The actual spectral resolution lies in the range
eral abundance ratios have been reported (McWilliam , 42500<R<53000. For all the spectra, the S/N ratio is
1997), which suggests that peculiar nucleosynthesis pro- larger than 50 per pixel (see Ledoux et al. 2003 for de-
cesses and inhomogeneous chemical enrichment are prob- tails). For Q 0528−250, we also retrieved complemen-
ably taking place. tary datafromthe ESO archive,thereby obtaininga very
As mentioned above, DLAs trace the chemical evolu- high-qualityspectrumwithSNR∼100.Molecularhydro-
tion of galaxies at early epochs in the universe. Many de- gen was detected in three of these systems. This is the
tailed studies have been performed so far, revealing that case for Q 0013−004(Petitjean et al, 2002), Q 0528−250
their metallicities range between 1/300 Z⊙ and solar val- (Srianand & Petitjean,1998),andQ0405−443-systemat
ues. The abundance pattern is fairly uniform and com- z = 2.595- (Ledoux et al. 2003). The six systems span
abs
patible with low dust content (see Pettini et al. 1994, Lu a wide range of H i column densities from log N(H i) =
et al. 1996, Prochaska et al. 1999, Ledoux et al. 2002a). 19.7 up to 21.80.
This uniformity in the relative abundance patterns ob- In all systems we chose to analyze absorptionfeatures
served from one DLA to the other has been emphasized that are well-defined and do not suffer from major blend-
by Prochaska & Wolfe (2002) and suggests that proto- ing.
galaxies have common enrichment histories.
Few studies have adressed the question of the homo-
geneity inside each particular system. Prochaska & Wolfe 2.1. Q 0013−004
(1996) first studied chemical abundance variations in a
The presence of H at z = 1.973 was first re-
single DLA, and showed that the chemical abundances 2 abs
ported by Ge & Bechtold (1997). Spread over more than
were uniform to within statistical uncertainties. Lopez et
1000 km s−1 (see Fig. 1), this system is the DLA ab-
al. (2002) confirmed this finding from analysis of another
sorber with the highest molecular fraction known so far
DLA using Voigt profile decomposition. Petitjean et al.
(−2.7 < f(H ) < −0.64, Petitjean et al. 2002) for
(2002) and Ledoux et al. (2002b) showed that the deple- 2
log N(H i) = 20.83. Molecular hydrogen has been de-
tion patterns in subcomponents were very similar along
tected in four different components spread all over the
DLA profiles except in the components where molecu-
system, at relative velocities, ∼ −615, ∼ −480, ∼ 0
lar hydrogen is detected and where depletion is larger.
and ∼ +80 km s−1. Petitjean et al (2002) also report
More recently, Prochaska (2003), performed a study of
four additional strong metal components that probably
13 systems concluding that the majority of DLAs have
make a non-negligible contribution to the total H i col-
very uniform relative abundances. This contrasts in par-
umn density. All components have similar abundance ra-
ticular with the dispersion in nucleosynthetic enrichment
tios and depletion factors independent of the presence or
of the Milky Way as traced by stellar abundances. Here,
absence of H . Only for the special molecular component
we use the best data from our survey of DLAs (Ledoux 2
at z = 1.97296 does the dust depletion turn to be im-
et al. 2003) to investigate this issue further using an in- abs
portant. The depletion of Fe is comparable to that ob-
version method to derive the velocity profiles in different
servedinthecoldinterstellarmediumoftheGalacticdisc.
abundance ratios. In particular, we investigate the conse-
We summarize the overallabundances of the principal el-
quence of the presence of molecular hydrogen in some of
ements present in the system in Table 4. We also confirm
the DLAs.Thepaperis structuredasfollows:wedescribe
the strong depletion pattern in the ∼480 km s−1 compo-
the data in Sect. 2; in Sect. 3 we briefly introduce the
nent.
method used for the analysis; and results are presented
and discussed in Sects. 4 and 5.
2.2. Q 0405−443
2. Observations and sample
Lopez et al. (2001) first discovered three DLA sys-
As emphasized by Prochaska (2003), a high S/N ratio tems along the line of sight to this quasar at redshifts
is needed to investigate variations along the profile. In z = 2.550, 2.595 and 2.621. Later, Ledoux et al (2003)
abs
addition, we want to investigate the difference between confirmed the damped nature of these three absorptions.
systems where molecular hydrogen is and where is not We adopted their H i column density values. From the
detected because it has been shown (e.g. Petitjean et three systems observed in this line of sight, we included
al. 2002) that depletion is larger in components where the systems at z = 2.549, log N(H i) = 21.0 and
abs
H is detected. We therefore restricted the sample to z =2.595,logN(Hi)=20.9inoursample.Inthethird
2 abs
Quasar zem zabs logN(H i)a logN(H2) [Fe/H] Metallicity
0013−004 2.09 1.973 20.83 17.72 −1.46±0.01 −0.68±0.03
0405−443 3.00 2.595 20.90 18.16 −1.30±0.10 −0.93±0.02
0405−443 3.00 2.549 21.00 <13.9 −1.49±0.01 −0.96±0.04
0528−250 2.80 2.811 21.35 18.22 −1.44±0.01 −1.07±0.01
1037−270 2.31 2.139 19.70 <13.7 −0.51±0.03 −0.19±0.02
1157+014 1.99 1.944 21.80 <14.5 −1.84±0.03 −1.41±0.02
Table 1. Description of the sample. The neutral and molecular hydrogen column densities were taken from Ledoux
etal.(2003).Both[Fe/H] andmetallicity values werefromthis workandcorrespondto the values integratedoverthe
systems. Metallicity was derived in all cases from [S/H], except for Q 1157+014 for which [S/H] determination was
uncertain, so we used [Zn/H] instead.
ZnII_2026 SiII_1808
1.0 1.0
0.5
0.5
0.0
0.0 1.0 SII_1250
SiII_1808
1.0
0.5
0.5
0.0
NiII_1709
1.0
0.0
SII_1250
1.0 0.5
0.0
0.5 H2_L1P3
1.0
0.0 0.5
H2_L1P3
1.0
0.0
FeII_2260
1.0
0.5
0.5
0.0
FeII_1608 0.0
1.0
CrII_2056
1.0
0.5
0.5
0.0 0.0
−600 −400 −200 0 −60 −40 −20 0 20
Fig.1. Absorption spectra plotted on a velocity scale Fig.2. Absorption spectra plotted on a velocity scale for
for a few transitions -including H2L1P3- observed in the a few transitions -including H2L1P3 transition- observed
z = 1.973 DLA system towardQ 0013−004.Molecular in the z = 2.595 DLA system toward Q 0405−443.
abs abs
transitionshavebeendetectedinfourcomponentslocated Numerousmoleculartransitionshavebeendetectedintwo
at v ∼ −615 km s−1, v ∼ −480 km s−1, v ∼ 0 km s−1, componentslocatedatv ∼−24kms−1and∼−11kms−1
and ∼85 km s−1 (Petitjean et al, 2002). (Ledoux et al, 2003).
2.3. PKS 0528−250
system,metallicityismuchlower,[Fe/H]=−2.15,which We collected complementary data from the ESO archive
renders the analysis much more uncertain. Ledoux et al available on this QSO and added them together with
(2003) detect the presence of H at z = 2.595 from our own data. The result is a very high SNR spectrum
2 abs
J = 0, 1, 2, 3 rotational transitions. They measure that extends over the wavelength range 3000−10000 ˚A ,
log N(H ) = 18.16, one of the largest H column den- apart from a few gaps. The absorption redshift is
2 2
sities ever seen in DLAs, although the molecular frac- slightly higher than that of the emitting source, sug-
tion in the corresponding cloud is not all that large gesting that the absorbing system could be associated
(log f(H ) = −2.44) due to the high column density of with the quasar. A useful consequence of this is that a
2
neutral hydrogen. large number of metallic transitions are redshifted out-
SiII_1808 SiII_1808
1.0
1.0
0.5
0.5
0.0
SII_1253
1.0
0.0
NiII_1741 0.5
1.0
0.0
NiII_1370
1.0
0.5
0.5
0.0 0.0
FeII_1608 H2_L0R1
1.0
1.0
0.5
0.5
0.0
FeII_1608
1.0
0.0
CrII_2056 0.5
1.0
0.0
CrII_2056
1.0
0.5
0.5
0.0 0.0
−50 0 50 100 150 −100 0 100 200 300
Fig.3. Absorption spectra plotted on a velocity scale for Fig.4. Absorption spectra plotted on a velocity scale for
a few transitionsobservedin the z = 2.55DLA system afewlow-ionizationspecies-includingH2L0R1transition-
abs
toward Q 0405−443.(Ledoux et al, 2003) observed at z = 2.811 along the line of sight to
abs
PKS 0528−250. The system extends over more than
400 km s−1.
side the Lyman-α forest and therefore can be used for
our analysis. This system has been known for many
years to be the only system at high redshift where
molecules were detected (Levshakov & Varshalovich,
1985; Songaila & Cowie, 1996; Srianand & Petitjean,
from Srianand & Petitjean (2001). This is the lowest col-
1998).NewdatahavebeenobtainedwithVLTbyLedoux
umn density in our sample. However,this system has the
etal.(2003).TheseauthorsderivelogN(H )=17.93and
2 highest metallicity known for DLAs at such a redshift
18.0intwocomponentsatz =2.81100and2.81112,re-
abs (z =2.139), [Zn/H] = −0.26. As a consequence of this
abs
spectively. Given the large neutral hydrogencolumn den-
high metallicity, many low-ionization lines were detected,
sity (log N(H i) = 21.35), the molecular fraction is only including C i, though no molecules were detected. Some
f(H )=9×10−4.TheexcitationtemperaturefortheJ=1
2 of the corresponding species are shown in Fig. 5. The de-
rotationallevelisbetween150and200K,andthedensity
pletion pattern derivedby Srianand& Petitjean(2001)is
is probably quite large (Srianand & Petitjean 1998). As
compatible with very low dust content, if any at all. The
mentionedbyLu et al. (1996),themetalabsorptionlines
gas seems to be warm and halo-like.
are unusually wide and complex, spreading over about
400 km s−1. They appear to be structured in two main
sub-clumps above and below +140 km s−1 (see Fig. 4). 2.5. Q 1157+014
This system has the highest H i column density in our
2.4. Q 1037−270
sample (log N(H i) = 21.8 cm−2) and is close to the
The continuumof the QSO is difficult to constraindue to emission redshift of the quasar. Absorption in 21 cm has
the presence of a complex system of BAL troughs (see beendetectedbyBriggs et al(1984)andthespintemper-
Srianand & Petitjean 2001). A first limit on the neu- ature is constrained by Kanekar & Chengalur (2003) to
tral hydrogen column density was derived by Lespine & be 865±190K. Neither H nor C i were detected and the
2
Petitjean (1997) from the absence of damped wings in a metallicityis the smallestinthe sample([Zn/H]=−1.41,
low-resolution spectrum. We adopted log N(H i) = 19.7 see Table 1).
SiII_1808 SiII_1808
1.0 1.0
0.5 0.5
0.0 0.0
SII_1259 NiII_1751
1.0 1.0
0.5 0.5
0.0 0.0
NiII_1741 FeII_2260
1.0 1.0
0.5 0.5
0.0 0.0
FeII_1608 CrII_2056
1.0 1.0
0.5 0.5
0.0 0.0
−60 −40 −20 0 20 40 60 80 −100 −50 0 50
Fig.5. Absorption spectra plotted on a velocity scale for Fig.6. Absorption spectra plotted on a velocity scale for
selectedtransitionlinesintheDLAsystematz =2.139 a few transitions observed in the DLA system at z =
abs abs
towards Q 1037−270.The system shows a “smooth” pro- 1.944 towards Q 1157+014.
file for all the elements spread over nearly 100 km s−1
will choose profiles that are apparently not blended with
interlopers).Asthedifferenttransitionsarefromthesame
3. Analysis method
species, the quantity log Nk (or simply τXk/fkλk) is the
a i 0
We want to estimate the column density per unit veloc- same for all transitions k.
ity along the absorption profile of species X using several The fitted optical depth before any instrumental con-
transitions. Following Savage & Sembach (1991), we can volution and without adjusting for any overlapping be-
write that the apparent optical depth per unit velocity of tween regions, is then:
species X in the QSO spectrum at wavelength λ for any
τk =τXk + τBk ∀i∈{1,...,n}, ∀k ∈{1,...,m}. (1)
transition of oscillator strength f and rest wavelength λ0 i i i
is τ (λ) = Ln(1/F(λ)), and the column density per unit
a Tosimplifythenotation,weusethedatavectorOofm∗n
velocity is log N (λ) = log τ (λ)−log fλ −14.976. Given
a a 0 elements, obtained by stacking the observed flux of each
a set of m transitions of the same species X, we can use region (i.e. Fik = Oi+kn−n), the parameter vector P of
the duplication of the informationoverthe different tran- n+m∗n elements, obtained by stacking τXk/fkλk and
i 0
sitionlinestoderivetheopticaldepthprofileofthespecies τBk (i.e.τXk/fkλk =P andτBk =P ),andthefitted
watchingoutinaddition,forpossibleblending.Letλk0 and viector F ciorrespo0ndingito thei fittedkflnu+xi. The χ2 of the
fk be,respectively,the laboratorywavelengthandthe os-
fit is then:
cillator strength of transition k. First, the regions corre-
sponding to these transitions arerebinned to the smallest χ2 =tr(cid:16)[W·(O−F)]T·[W·(O−F)](cid:17) (2)
pixelsizesothattheyspanthesameredshiftrangeoveran with F =C·exp∗(−M·P). (3)
identicalnumbernofpixels.Letλk,Fk bethewavelength
i i
andthenormalizedfluxatpixelioftheregioncorrespond- ThematrixMappliesthelinearrelation(1)totheparam-
ing to transitionk.The value λk/λk−1is independent of eters and takes care of possible overlap between different
i 0
k, thanks to the rebin, and is the redshift at pixel i. The regions.Thefunction exp∗ replaceseachelementofa ma-
observedoptical depth, −ln Fik, is consideredas the sum trix by its exponential value (i.e. exp∗(X)ij =eXij). The
of the optical depth, τXk, of transition k and the optical matrix C is used to convolve the calculated flux with the
i
depth, τBk, of a possible intervening absorption blended instrumental profile and, finally, W is the inverse of the
i
with the absorption of interest (of course in practice we variance-covariancematrix of the data.
Table 2. Atomic data more parameters (n∗(m+1)) than data points (n∗m),
the problem is underestimated. Nevertheless,the regular-
Transition λvac f Ref. ization introduces a constraint that diminishes the real
(˚A) number offree parameters.Moreover,inpractice,mostof
Siiiλ1020 1020.6989 0.01680 a
the profiles we use here are free of blending and, in that
Siiiλ1190 1190.4158 0.25000 a
case, we impose a zero optical depth for the intervening
Siiiλ1304 1304.3702 0.08630 a
absorption.
Siiiλ1526 1526.7070 0.12700 a
Siiiλ1808 1808.0129 0.00208 a Once the different profiles have been fitted, we com-
Siiλ1250 1250.5840 0.00545 a pute the ratios between the most important elements in
Siiλ1253 1253.8110 0.01090 a ordertoanalyzepossiblevariationsoftheratiosalongthe
Siiλ1259 1259.5190 0.01620 a profile.Themethodiswell-suitedtostudyingthecomplex
Criiλ2056 2056.2690 0.1050 a absorptionprofiles ofDLAs. Indeed, itis well-knownthat
Criiλ2062 2062.2361 0.0780 a
the decomposition in discrete components is not unique
Criiλ2066 2066.1640 0.0515 a
due to the fact that the observed absorption profile is a
Feiiλ1143 1143.2260 0.01770 a
Feiiλ1144 1144.9379 0.10600 a convolution of absorptions from density fluctuations in a
Feiiλ1608 1608.4508 0.05800 a continuous medium. In addition, absorptions from differ-
Feiiλ1611 1611.2003 0.00136 a ent overdensities can be superimposed by peculiar kine-
Feiiλ2249 2249.8768 0.00182 a matics, making any decomposition in components an ad
Feiiλ2260 2260.7805 0.00244 a hoc representation of reality. The pixel-by-pixel method
Feiiλ2344 2344.2130 0.11400 a instead does not assume any model and emphasizes devi-
Feiiλ2374 2374.4603 0.0313 a ations from the mean value when a fit tends to smooth
Niiiλ1370 1370.1320 0.0765 b such deviations out. Although the mean depletion mea-
Niiiλ1454 1454.8420 0.0300 b suredbybothmethods,thepixel-by-pixelanalysisandthe
Niiiλ1709 1709.6042 0.0324 b
sub-componentfitting, isapproximativelythe same,devi-
Niiiλ1741 1741.5531 0.0427 b
ations from the mean and in particular locations within
Niiiλ1751 1751.9157 0.0277 b
Zniiλ2026 2026.1371 0.489 a the gas, can only be revealed by a pixel-by-pixel analy-
Zniiλ2062 2062.6604 0.256 a sis. In particular, the scatter in the data points is a good
estimate of the (in)homogeneity of the gas.
References:Vacuum wavelengths from Morton (2003).
Oscillator strengths: (a) Morton (2003); (b) Fedchak et al.
(2000).
4. Results for individual objects
A regularizationconstraintisaddedto selectthe solu-
Several of the systems in the sample are spread over
tions P that are correlated over a specific length l0. The more than 200 km s−1. The absorption profiles are of-
selection is done by a minimization of the high frequency
ten clearly structured in several well-detached clumps
coefficientsofthediscreteFouriertransformoftheparam-
with no or very little absorption inbetween. To investi-
eters P, i.e. by minimizing ζ2 (Eq.4 below) at the same
gate whether the properties differ from one clump to the
time as χ2 (Eq.2).
other, we performed the analysis on each of the clumps.
Indeed,itisreasonabletobelievethatdifferentpartofthe
ζ2 =tr [S ·T·P]T·[S ·T·P] , (4)
(cid:16) l0 l0 (cid:17) DLAs could have different characteristics (e.g. Haehnelt
et al. 1998). Column densities of several species, inte-
where T is the matrix of the discrete Fourier transform
grated over the system, are given for each system in
and S the high frequency filter that cancels the coeffi-
l0 Table 4. Abundance ratios relative to solar are defined as
cients of the Fourier transform corresponding to a length
[X/Y] ≡ log[N(X)/N(Y)] - log[N(X)/N(Y)]⊙, adopting
(in the parameter space) greater than l .
0 solar abundances from Anders & Grevesse (1993). These
Actually, the overall fit with the regularization con-
solarabundancesandtypicalgalacticdepletionvaluesare
straint is done by using a Lagrange parameter f and
0 summarizedin Table 3. We computed the quantity [X/Y]
minimizing the following quantity,
ineachpixelanditsaverage([X/Y])overeachsubclump,
Q=χ2 + f ζ2 (5) andresultsaregiveninTable5.Eachsubclumpisreferred
0
to by the velocity range over which it is spread relative
Theparameterf controlstherelativestrengthofthereg- to the main redshift of the system, with a reference z
0 abs
ularisation over the χ2 fit of the data. If f is too small, taken from (Ledoux et al, 2003). Then [X/Y], averaged
0
thenonlyχ2 isminimizedandthedatawillbeover-fitted. over this velocity range, is given for each ratio. In each
In contrast, if f is too large, the minimization is done row, the number next to [X/Y] is the error on the mean
0
only for ζ2 to give a very smooth solution that does not ratio(σ = n σ2/n),andthe secondnumber(inital-
pPi=1 i
fit the data. Thus, f is chosen to obtain a χ2 value ap- ics) is the scatter of [X/Y] around the mean calculated
0
proximately equal to 1. In the worst case, since there are over the subclump, [ ([X/Y]−[X/Y] )2/n]1/2.
i
P
1.8
10−3
SSSSiiiiIIIIIIII
1.6
1] SSIIII
− 1.4
)
s
/ FFeeIIII
m
1.2
k
(

2
− 1.0
m
c
[ 0.8

)
v
,
X
0.6
(
N
0.4
0.2
0.0
−700 −600 −500 −400 −300 −200 −100 0 100
Fig.7. This panel shows the optical depth profiles in the z = 1.973 DLA toward Q 0013−004. The apparent
abs
columndensity per velocity bin -N(X,dv) (y axis values are scaledto ×10−3)- is representedona velocity scale,with
v =0 km s−1 centered at z =1.973.
abs
SSSSSIIIIIIIIII
−1m/s)] FFFFFeeeeeIIIIIIIIII
−2m (k0.001
X,v) [c
N(
H2 H2 H2 H2
0.00 00
S]
Fe/ −1
[
−650 −600 −550 −490−480−470 −200 −150 −100 0 50 60 80 100
Relative Velocity [km/s]
Fig.8. In this figure, top panels show the same Fe and S profiles than Fig. 7, divided in five regions. Lower pannels
show the ratio [Fe/S] computed for each pixel. Points with errors that are too big are rejected to compute the mean,
represented here as empty squares, while valid points are represented as filled squares. The mean value is plotted as
a solid line. Dashed lines represent the ±σ level, and the dotted line the typical scatter of the points. H is detected
2
in four components at ∼−615,−480, 0, and 85 km s−1.
4.1. Q 0013−004 thesystemexhibitsfourmolecularcomponentsat∼−615,
−480, 0 and 85 km s−1 relative to the central redshift.
Fig. 7 illustrates this complexity. The apparent column
AsmentionedinSect.2.1,theabsorptionsystemfoundin
the line of sight of Q 0013−004 at z = 1.973 shows a densitypervelocitybin,referredtosolarvalues,isplotted
abs for Fe ii, S ii, and Si ii. All have complex but remarkably
complex multicomponent structure. Besides the fact that
similarprofiles.Thebottompanelsrepresenttheratioper
the system is composed of a DLA at z = 1.96753 and
abs
velocitybin[Fe/S]fordifferentvelocityranges.Themean
a sub-DLA at z = 1.9733 (Petitjean et al. 2002), the
abs
low-ionization absorptions span about 1000 km s−1, and values(see Table5)aresmallerthaninothersystemsbut
Table 3. Solarabundances and typicalgalactic depletion
values of elements investigated in this paper
Ion log(X/H)+12a [X/S]b,c 0.004
Colddisc Warmdisc Halo
HSiiii 7.5152±.000.02 −.1...3 −.0...4 −.0...3 −1s)]
Sii 7.27±0.05 +0.0 +0.0 +0.0 m/
MCrniiii 55..5638±±00..0043 −−12..51 −−11..02 −−00..76 −2m (k SiII
ZNFeniiiiiii 467...625551±±±000...000221 −−−022...422 −−−011...244 −−−000...166 N(X,v) [c0.002 FeII
a Solar & Meteorite abundancesfrom Anders(1993)
b [X/H]= log[Z(X)]-log[Z⊙(X)]
c Galactic values are from Welty et al. (1999) H2 H2
0.000.00
are again quite similar from one sub-clump to the other.
Withinapeculiarsub-clump,scatterissmall,ontheorder
of 0.2 dex. In the components where H is detected, the Si]
depletion factor is of the order on [Fe/S2] = −1, therefore [Fe/−0.5
slightly smaller but not much less than in the overallsys-
tem, except in the component at −480 km s−1, which is
theonlyregioninDLAs knowntodatewheredepletionis
−60 −50 −40 −30 −20 −10 0 10
similar to the cold gas of the Galactic disc.
Relative Velocity [km/s]
Fig.10. Top panel: optical depth profiles in the z =
abs
4.2. Q 0405−443 2.595 DLA toward Q 0405−443. The apparent column
densitypervelocitybin-N(X,dv)-isrepresentedonave-
We analyzed the systems at z = 2.549 and z =
abs abs locity scale, with v = 0 km s−1 centered at z = 2.595.
2.595 toward Q 0405−443. In the latter system, molec- abs
Thelowerpanelsshowtheratio[Fe/Si]computedforeach
ular hydrogen has been detected with a column density,
pixel. The mean value is plotted as a solid line. Dashed
log N(H ) = 18.16 (Ledoux et al, 2003). Relatively nar-
2
row,low-ionizationtransitionsspanabout70kms−1.Itis lines represent the ±σ level and the dotted line the typ-
ical scatter of the points. This DLA has two molecular
apparentfromFig.10thattherearesomeinhomogeneities
components at v ∼−24 and ∼−11 km s−1.
inside the system and there is more depletion (by about
a factor of two) in the two components where H is de-
2
tectedatv∼−8and−24kms−1.Notethatthedepletion plex with many components. Note that the profiles of Fe
is larger in these components, but not by a large amount and S follow each other remarkably well over most of the
anddefinitivelynotasincoldgasofthe Galaxydisc.The profile. The analysis alongthe profile gives similar results
scatter of the pixel values is much larger than the errors for eachof the subclumps, suggestingthat once againthe
for all element ratios (see Table 5). The overall depletion mixing of heavy element must have been very efficient in
pattern is again similar to that of warm gas of the halo, thissystem(seeTable5).Thecentralpartat[0,20]hasa
with a slight enhancement of Cr. largerdepletioncoefficient,downto[Fe/S]=−0.55,when
The system at zabs = 2.549, presented in Fig. 9, therestofthesystemhas[Fe/S]∼−0.3.Thisistheplace
shows absorptions spread over ∼ 120 km s−1, and is where, again, H is detected. Even though the depletion
2
well-structuredinthree subclumps. FrombothFig.9 and factoris largerinthis component,it is muchsmallerthan
Table 5, it can be seen that the three subclumps at ve- in the coldgas of the Galactic ISM. The overalldepletion
locities [-30,20], [40, 60], and [60, 90] are quite homoge- pattern in the system is very close to what is observed in
neous but have slightly different mean depletion factors, the Galactic halo (see Table 5).
[Fe/S] = −0.26, −0.08, and −0.33, respectively. However
again,thedifferencesaresmallerthan0.2dex.Inaddition,
4.4. Q 1037−270
the depletion values are close to those of the z =2.595
abs
system, which is located ∼−3000 km s−1 away. Figure 12 shows the column density per pixel along the
profile for different species, including Fe ii, Si ii, S ii, and
Ni ii. All species followthe same pattern, andtheir ratios
4.3. Q 0528−250
remain fairly constant across the profile.
This systemis spreadovermore than350km s−1 andex- We considered two subclumps for −10 < v < 30 and
hibits two molecular components located at ∼ 0 km s−1. 40 < v < 60 km s−1 (see Table 5). There is essentially
We observe in Fig. 11 that the internal structure is com- no difference between the two subclumps. Note that the
SSSiiiIIIIII
NNNiiiIIIIII
SSSIIIIII
−1] FFFeeeIIIIII
s)
m/
k0.001
−2 (
m
c
v) [
X,
N(
0.000
0.0
S]
e/
F
[
−0.5
−30 −20 −10 0 10 20 40 45 50 55 60 65 70 75 80 85
Relative Velocity [km/s]
Fig.9. Top panel: optical depth profiles in the z =2.549 DLA toward Q 0405−443.The apparent column density
abs
per velocity bin -N(X,dv)- is represented on a velocity scale for Si, Ni, S, and Fe; with v = 0 km s−1 centered at
z = 2.549. The lower panels show the ratio [Fe/S] computed for each pixel. The mean value is plotted as a solid
abs
line. Dashed lines represent the ±σ level and the dotted line the typical scatter of the points.
1.0
SSSIIIIII
−1s)]10−3 FFFeeeIIIIII
m/
k
−2m ( 0.5
c
v) [
X,
N(
0.0 H2
0.0
S]
Fe/ −0.5
[
−30−20−10 0 10 20 30 40 50 60 70 80 90 100 110 180 200 220 240 260 280
Relative Velocity [km/s]
Fig.11.Toppanel:opticaldepth profiles inthe z =2.811DLA towardQ 0528−250.The apparentcolumndensity
abs
per velocity bin -N(X,dv)- is represented on a velocity scale, with v =0 km s−1 centered at z =2.811. The lower
abs
panels showthe ratio[Fe/S]computedfor eachpixel.The meanvalue is plottedas asolidline.Dashedlines represent
the ±σ level and the dotted line the typical scatter of the points. H is detected in two components at ∼ 0 km s−1.
2
scatter is comparable to or smaller than the errors (for 4.5. Q 1157+014
the first [-10,30] subclump, σ = 0.04, σ = 0.07 and for
e s This system is spread over about 100 km s−1 and no as-
the second, [40,60], σ [Fe/S] = 0.07 and σ = 0.02). This
e s
sociatedmolecules weredetected.We consideredonlyone
systemisthereforefairlyhomogeneouswithin0.1dex,and
clump. Pixel analysis is shown in Fig. 13. Unfortunately,
theabundancepatternsimilartowhatisobservedinwarm
S ii lines lie at the bottom of a BAL trough, so we used
gas of the Galactic halo.
zinc as the non-depleted species reference instead.
The system has a low mean metallicity,
[Zn/H] = −1.40 (see Table 1), and shows a depletion
patternsimilartothehaloofthegalaxyforeveryelement
analyzed.PixelanalysisisshowninFig.13.Itcanbeseen
SSiiIIII
0.02 NNiiIIII
SSIIII 1.0
−1s)] FFeeIIII −1s)]
m/ m/
−2m (k −2m (k10−3 SZniIIII
X,v) [c0.01 X,v) [c 0.5 NiII
N( N( FeII
0.00
0.0
0.0
−0.2
[Fe/S]−0.4 [Fe/Zn]−0.5
−0.6
−10 0 10 20 30 40 45 50 55 60 −60 −40 −20 0
Relative Velocity [km/s] Relative Velocity [km/s]
Fig.12. Top panel: optical depth profiles in the z = Fig.13. Top panel: optical depth profiles in the z =
abs abs
2.139 DLA towards Q 1037−270. The apparent column 1.944 DLA towards Q 1157+014. The apparent column
density per velocity bin -N(X,dv)- is represented on a density per velocity bin -N(X,dv)- is represented on a
velocityscale,withv =0kms−1 centeredatz =2.139. velocity scale for Zn, Si, Ni, and Fe species, with v =
abs
The lowerpanelshowsthe ratio[Fe/S]computedfor each 0 kms−1 centeredatz =1.944.The lowerpanelshows
abs
pixel. The mean value is plotted as a solid line. Dashed the ratios [Fe/Zn] computed for each pixel. The mean
linesrepresentthe±σlevelandthedottedlinethetypical value is plotted as a solid line. Dashed lines represent the
scatter of the points. ±σ level, and the dotted line the typical scatter of the
points.
thatdepletionis larger,[Fe/Zn]∼ −0.65,in the strongest
small if we take the observational and fitting uncertain-
absorption component. It should be noticed that 21 cm
ties into account .
absorption has been reported by Kanekar & Chengalur
Moreover,whenwecomparedepletionvaluesfromone
(2003) at this velocity (∼ −28 km s−1), revealing
sub-clump to another in the same system, differences are
dense gas. The difference in depletion between the
small. A distinct depletion pattern is observed only for
strongest subcomponent (at −28 km s−1) and the gas
some molecular components .This is remarkably summa-
at v ∼ −40 km s−1 is significant (about a factor of two
rizedbyFig.15,where we plotthe [Fe/S]ratioversusthe
larger than 3σ). This is clearly a sign of more depletion
[Si/S]ratioinallsubclumps.First,asalreadyemphasized
in the dense gas producing the 21 cm absorption. The
by Petitjean et al. (2002) and Ledoux et al. (2002b), the
overallscatter is only about 0.12 dex (see Table 5 ). This
sequence seen in this figure is a dust-depletion sequence.
is fairly small compared to what is observed through the
Indeed, there is a correlation between the two quantities
ISM of our Galaxy (Welty et al, 1999). In addition, Ni
which is expected if the depletion is due to the presence
and Fe follow each other very well.
ofdust.Secondly,the valuesmeasuredindifferentclumps
of the same system are gathered at the same place in
the figure. The only exception is the H component at
5. Discussion 2
−480 km s−1 in Q 0013−004 (see above). Thirdly, most
5.1. Relative abundance ratios of the depletion pattern is similar to that of the gas ob-
served in the Galactic halo. Finally, it seems that silicon
Relative abundance ratios for each sub-clump considered is overabundant by about 0.2 dex even relative to sulfur.
in the six systems analyzed in this work are given in In all this, however, it must be recalled that we do not
Table 5. In the same way, results are summarized in have access to the absolute metallicity in the subclumps,
Fig. 14. It is apparent that, within each subsystem, large becausewearenotabletodisentangletheHiabsorptions
departures from the mean ratio are rare: the scatter is of the different subclumps.