Using photo-ionisation models to derive carbon and oxygen gas-phase abundances in the rest UV

0.75 MB·

by  Enrique Pérez-Montero

MNRAS000,1–7(2016) Preprint13March2017 CompiledusingMNRASLATEXstylefilev3.0
Using photo-ionisation models to derive carbon and
oxygen gas-phase abundances in the rest UV
Enrique P´erez-Montero,1(cid:63) Ricardo Amor´ın,2,3,4
1Instituto de Astrof´ısica de Andaluc´ıa. CSIC. Apartado de correos 3004. 18080, Granada, Spain.
2INAF-Osservatorio Astronomico di Roma, via di Frascati 33, I-00078, Monte Porzio Catone, Italy.
3Cavendish Laboratory, University of Cambridge, 19 JJ Thomson Avenue, Cambridge, CB3 0HE, UK.
4Kavli Institute for Cosmology, University of Cambridge, Madingley Road, Cambridge CB3 0HA, UK
7
1
0 AcceptedXXX.ReceivedYYY;inoriginalformZZZ
2

r
a ABSTRACT
M We present a new method to derive oxygen and carbon abundances using the ultra-
violet (UV) lines emitted by the gas-phase ionised by massive stars. The method is

9 based on the comparison of the nebular emission-line ratios with those predicted by
a large grid of photo-ionisation models. Given the large dispersion in the O/H - C/O
]
plane, our method firstly fixes C/O using ratios of appropriate emission lines and, in
A
a second step, calculates O/H and the ionisation parameter from carbon lines in the
G
UV. We find abundances totally consistent with those provided by the direct method
. when we apply this method to a sample of objects with an empirical determination of
h
the electron temperature using optical emission lines. The proposed methodology ap-
p
- pears as a powerful tool for systematic studies of nebular abundances in star-forming
o galaxies at high redshift.
r
st Key words: methods: data analysis – ISM: abundances – galaxies: abundances
a
[

2
v 1 INTRODUCTION numberofhighionisation,metalemissionlineswithrelative
1 large equivalent widths, such as those of carbon (e.g. Ciii]
1 Establishing the gas-phase abundances of carbon, nitrogen 1908 ˚A, Civ1549 ˚A) and oxygen (Oiii] 1664 ˚A), which are
4 and oxygen in galaxies through cosmic time is key to un-
oftenrelatedtothepresenceof highequivalentwidthemis-
4 derstand not only their chemical enrichment, but also how sionofHeii1640˚AandLyα. Whilethesespectralfeatures
0 galaxiesassembleandevolve(Edmunds&Pagel1978;Gar-
. could be quite common at very high redshifts (Stark et al.
1 nett et al. 1995; Henry et al. 2000; Chiappini, Romano, &
2017), they provide tight constraints to the ionisation, age,
0 Matteucci 2003).
and metallicity properties of galaxies, as recently shown by
7 Inthelastdecadeorso,largedeepopticalspectroscopic
emission-linediagnosticsbasedondetailedphoto-ionisation
1 surveysusingground-based8-10m-classtelescopes(e.g.Stei-
models (e.g. Feltre, Charlot, & Gutkin 2016; Gutkin, Char-
: del et al. 2003; Shapley et al. 2003; Lilly et al. 2007; Kurk
v lot, & Bruzual 2016; Jaskot & Ravindranath 2016).
i et al. 2013; Le F`evre et al. 2015) led to a significant in-
X
crease in the amount and depth of rest-frame UV spectra The wealth of information contained in the UV spec-
r of high redshift galaxies (z (cid:38)1-2). These and other follow- traofstrongemission-linegalaxiesallowustoexplorenew
a
up spectroscopy of galaxies in deep fields (e.g. Erb et al. techniques for an accurate characterisation of the nebular
2010;Karmanetal.2015;Continietal.2016),inparticular chemicalcontentofstarforminggalaxiesbasedontheanal-
gravitationally lensed systems (e.g. Christensen et al. 2012; ysis of the carbon abundance, the processes governing the
James et al. 2014; Stark et al. 2014; Vanzella et al. 2016a), C/O ratio, and its relation with the total metallicity (e.g.
are pushing now UV studies towards the low-luminosity Henryetal.2000;Erbetal.2010;Bergetal.2016).Theuse
(mass) regime out to redshift z ∼3.5 (e.g. Patr´ıcio et al. of photo-ionisation models assuming different combinations
2016; Vanzella et al. 2016b, 2017; Amor´ın et al. 2017) and of C/O and O/H in an emitting ionised gas distribution is
beyond (Stark et al. 2017, and references therein). henceanalternativetoestimatethegas-phasemetallicityin
Overall,theabovestudiesconsistentlyfindthattheUV those cases where very few UV emission lines can be mea-
spectrum of high redshift galaxies is systematically bluer sured. These methodologies, entirely or partially based on
and harder than in most local galaxies. Thus, it includes a theavailableUVemissionlineratios,arenecessarytocom-
plementthewidelyusedrecipestoderivephysicalproperties
and chemical abundances from rest-frame optical spectra,
(cid:63) E-mail:[email protected](EPM) which at z ∼2-3 rely on the (usually challenging) detection
©2016TheAuthors
2 Enrique P´erez-Montero & Ricardo Amor´ın
of faint emission lines in the near infra-red (NIR, e.g. Stei- Table 1.ReferencesandnumberofHiiregions,localstarbursts
del et al. 2014; Maseda et al. 2014; Troncoso et al. 2014; (LSB),andLyman-breakgalaxies(LBG)usedinthiswork.
Amor´ınetal.2014;Shapleyetal.2015;Onoderaetal.2016;
Sanders et al. 2016; Trainor et al. 2016). They would also Reference Numberofobjects Type
be of great interest for metallicity studies of galaxies at the deBarrosetal.(2016) 1 LBG
highest redshifts, where observing the optical rest-frame is Baylissetal.(2014) 1 LBG
Bergetal.(2016) 7 LSB
extremely challenging. In particular, the future arrival of
Christensenetal.(2012) 3 LBG
unprecedentedly deep observations using the James Webb
Erbetal.(2010) 1 LBG
Space Telescope (JWST) and the new generation of 30m-
Garnettetal.(1995) 6 LSB
classtelescopes,willincreasebyahugefactorcurrentgalaxy
Garnettetal.(1997) 2 LSB
samplesat z(cid:38)3,extendingdetailedUVspectroscopicstud- Garnettetal.(1999) 6 Hii
ies towards the cosmic dawn. Izotovetal.(1999) 3 LSB
For this aim, in this paper we present an adapted ver- Jamesetal.(2014) 1 LBG
sionofthecodeHii-Chi-mistry(P´erez-Montero2014,here- Kurtetal.(1995) 2 Hii
inafter HCm), originally designed to derive O/H and N/O Kobulnickyetal.(1997) 3 LSB
fromopticalemissionlines,thatdealswithUVemissionlines Kobulnicky&Skillman(1998) 3 LSB
Steideletal.(2016) 1(stack) LBG
to derive O/H and C/O ratios. This new semi-empirical
Thuanetal.(1999) 1 LSB
method is fully consistent with direct estimations based on
Vanzellaetal.(2016a) 1 LBG
the electron temperature of star forming regions and pro-
Villar-Mart´ınetal.(2004) 1 LBG
vide a potentially powerful tool to constrain the metallic-
ity of galaxies out to very high redshifts. Even considering
the inherent limitations affecting all methods for estimat- and, in some cases, from the ratio between [Oiii]5007 and
ing chemical abundances in galaxies from their integrated 1664˚A. Ourcalibrationsampleincludesobjectswithoutob-
spectra, such as, e.g. the fact that these are based on spa- servedfluxesinthe[Oii]3727˚A line.Fortheseobjectsthe
tiallyunresolveddata(e.g.Iglesias-Pa´ramoetal.2016)and thetotaloxygenabundancewasderivedbyapplyinganem-
rely on the simplistic assumption of a single Hii region for pirical calibration between O/H and te([Oiii]) (see below).
the analysis of the observed emission lines, it is still possi- For C/O determinations using the direct method, we com-
ble to find characteristic abundance ratios (i.e. O/H, N/O, piledobjectswithmeasurementsofatleastOiii]1664˚Aand
P´erez-Monteroetal.2016)thatcorrelatewiththeintegrated Ciii] 1909 ˚A, and also Civ1549 ˚A in very high-excitation
properties of galaxies, such as stellar mass, thus providing objects. All the compiled fluxes were reddening corrected
useful constraints for their assembly histories and chemical using a Cardelli et al. (1989) extinction law based on the
evolution. extinction coefficients given in these references. The list of
Our paper is organized as follows; In Section 2 we de- references and the number of objects are listed in Table 1.
scribe the sample of objects with available emission lines We recalculated the physical properties and the chem-
both in the optical and in the UV whose O and C abun- ical abundances of the selected objects using expressions
dances could be derived using the direct method. In Sec- based on the pyneb software (Luridiana, Morisset, & Shaw
tion3wedescribethemodelsandthemethodusedtoderive 2015). The O/H abundances were calculated using the de-
the abundances and we compare our results with the direct rived electron temperatures and the bright optical emission
abundances.Finally,inSection4wepresentourresultsand lines [Oii] 3727 ˚A and [Oiii] 4959,5007 ˚A and the expres-
in Section 5 we summaries our conclusions. sions given in P´erez-Montero (2014). In those cases where
Throughout this paper the following convention thefluxofthe[Oii]3727˚Alinewerenotavailabletocalcu-
is adopted: O/H is 12+log(O/H), C/O is log(C/O), latetheO+abundances,weresortedtotheempiricalrelation
Oiii]λ1664˚A represents the total flux of 1661 and 1666˚A, between O/H and te([Oiii]) given by Amor´ın et al. (2015),
Ciii]λ1908˚A represents the total flux of lines at 1907 and which is valid for a wide range of metallicity and ionisation
1909˚A,andCivλ1549˚Arepresentsthetotalfluxoflinesat values.
1548 and 1551˚A. Using the pyneb code, we also defined expressions for
thetemperatureandcarbonabundancesusingUVemission
lines. Computing the emission line ratio:
2 CALIBRATION SAMPLE AND
ABUNDANCES R = I(5007 ˚A) (1)
WecompiledfromtheliteratureasampleofHiiregionsand O3 I(1664 ˚A)
starburst galaxies at different z with optical and UV lines wederivethe[Oiii]electrontemperatureusingthesameset
emitted by the gas ionised by episodes of massive star for- of atomic parameters as in P´erez-Montero (2014):
mation. Owingtothescarcenumberofobjectswithsimul-
tcaonnesoisutssomneaascuormembiennattioofnoopftiinctaelgarnatdedUsVpelcintreas,oobuserrvsaedmupsle- te([Oiii])=1.9263+6.433·10−5·RO3−0.4221·log(RO3) (2)
ingavarietyofinstrumentsandtechniques.Oursamplewas valid in the range te([Oiii]) = 0.8 to 2.5 in units of 104 K
selected in such a way that the O/H and C/O abundance and that is nearly independent on electron density (i.e. a
ratioscouldbederivedusingthedirectmethodfromthees- change from 10 to 1000 cm−3 implies a change in te of less
timation of the [Oiii] electron temperature, te([Oiii]). This than 0.1%). Nevertheless, due to the very long wavelength
was obtained from the ratio between [Oiii]5007 and 4363˚A baseline between the involved emission lines and the large
MNRAS000,1–7(2016)
Gas-phase C and O abundances in the rest UV 3
extinction correction in the UV, the used ratio has a large
uncertainty if the extinction correction is big. For instance
assuming an extinction of 1 mag at 1664 ˚A implies an elec-
trontemperatureofabout10%largerusingaCalzettietal.
(2000) than a Cardelli et al. (1989) extinction law.
ThecarbonionicratiosbasedonUVlineswerederived
using the following expressions and assuming the derived
te[Oiii]:
log(C2+/O2+)=log(cid:32)Ciii] 1908 ˚A(cid:33)−0.8361+
Oiii] 1664 ˚A
0.4801
− −0.0358·te+0.2535·log(te) (3)
te
Figure1.RelationbetweenO/HandC/Oforthestudiedsample
(cid:32) (cid:33)
log(C3+/O2+)=log Civ1549 ˚A −1.5625+ as derived from the direct method. Different symbols stand for
Oiii] 1664 ˚A the class of objects and colours for the method used to derive
abundances: direct method with 5007/4363 ratio (black), O/H
from empirical relation with te[Oiii] (red), and direct method
0.2946
+ −0.043·te+0.31·log(te) (4) usingtemperaturefrom5007/1664ratio(white).
te
and using the collisional coefficients of Berrington et al.
(1985) and Aggarwald & Keenan (2004) for C2+ and C3+, carbon lines. This is similar to the situation in the optical,
respectively. when N/O must be determined before deriving O/H using
The total C/O was then derived under the assumption [Nii]emissionlines(P´erez-Montero&Contini2009;Amor´ın
that: et al. 2010).
C C2++C3+
= (5)
O O2+ 3 MODEL AND METHOD DESCRIPTION
that can be simplified to C/O = C2+/O2+ when no Civ Both O/H, C/O and the ionisation parameter (logU) were
emission is detected. derived using a semi-empirical approach based on the com-
In Fig.1 we plot the relation between total O/H and parison between the most representative extinction cor-
C/O for our calibration sample. The objects are identified rected emission lines observed in the UV (Lyα1216˚A,
according to their nature and the method used to derive Civ1549˚A, Oiii]1664˚A, and Ciii]1908˚A) and in the opti-
O/H and C/O. We distinguish between objects with tem- cal(Hβ4861˚Aand[Oiii] 5007˚A)withalargegridofphoto-
peratures obtained from the [Oiii] ratio 5007/4363 (30), ionisation models covering a wide range of values in O/H,
objects whose O/H was calculated using the empirical re- C/O, and logU.
lation between O/H and te([Oiii]) given in Amor´ın et al. The photo-ionisation models were calculated using the
(2015)(6),andobjectswhoseO/HandC/Owerecalculated code Cloudy v.13.01 (Ferland et al. 2013). In all models
using the electron temperature from the [Oiii] 5007/1664 we adopted a spherical geometry of constant density gas
ratio(8).Forthelatter,typicalerrorsinO/Hare∼0.18dex, (100cm−3),fillingfactorof10−1,andastandardgas-to-dust
whileinthetwofirstcasestheyareslightlylower,∼0.10dex. massratioaroundanionisingpoint-sourcewiththespectral
On average, C/O tend to increase with metallicity, as energy distribution of a 1 Myr old cluster using popstar
carbon is essentially a secondary element (i.e. its relative (Molla´ et al. 2009) model atmospheres at the metallicity of
productionishigherforhigherO/H)mainlyejectedintothe thegas.Thecalculationwasstoppedineachmodelandthe
ISM by massive stars (e.g. Henry et al. 2000). This trend lines were thus retrieved when the electron density fraction
startstobeobservedinoursamplefrom12+log(O/H)>7.4. was lower than 10%.
However, the dispersion of this relation, of about 0.25 dex, Although several authors have pointed out the impor-
islargerthantheaverageC/Oerrors(∼0.15dex),owingto tanceoftheassumedSEDinthemodelsforthedetermina-
thedifferentphysicalprocessesaffectingmetallicityandthe tion of abundances (e.g. Kewley et al. 2005; Morisset et al.
ratio between a primary element as O and a secondary one 2016),asourapproachissemi-empiricalitispossibletoas-
asC.Thisisthecaseofhydrodynamicalprocesses,including sessanysystematiceffectandtheuncertaintyassociatedto
metal-enriched outflows or inflows of metal-poor gas, which the assumption of an ionizing stellar cluster of a single age
can change metallicity and keep relatively unaffected C/O in our models.
(e.g. Edmunds 1990). Other factors that could also modify Thegrid coversinputoxygenabundancesintherange
this ratio are, for instance, different star-formation efficien- [7.1,9.1]inbinsof0.1dex,logUintherange[-4.0,-1.5]inbins
cies or the presence of a non-standard IMF (e.g. Gavila´n of0.25dexandassumessolarproportionsfortherestofele-
etal.2005;Mattsson2010)Therefore,asnoclear trendcan ments,withtheexceptionofcarbon,thattakesvaluesinthe
be taken between O/H and C/O, C/O has to be previously rangeofC/O=[-1.4,0.6]inbinsof0.125dex,andnitrogen,
calculatedtoallowthederivationofO/Habundancesusing forwhichweassumeasecondaryoriginandaconstantsolar
MNRAS000,1–7(2016)
4 Enrique P´erez-Montero & Ricardo Amor´ın
ratiolog(C/N)=0.6(Asplundetal.2009). Theassumption Hβ4861˚A.AsinthecaseofC3O3,C34canbedefinedinthe
of a constant C/N ratio is justified by the expected similar absenceofCivlinesusingonlyCiii]lines.AsshowninFig.2,
behaviorbetweenprimaryandsecondaryelementsunderhy- this ratio has a very similar dependence with O/H as the
drodynamicaleffects(e.g.Edmunds1990)andsupportedby R parameter (i.e. [Oii]+[Oiii] relative to Hβ, Pagel et al.
23
recentobservationsofgalaxiesatlow(Bergetal.2016)and 1979) as it is double-valued (i.e. it increases with O/H for
high redshifts (Steidel et al. 2016), which show a constant lowO/HanditdecreasesforhighO/H).Besides,itpresents
trend in C/N with metallicity. Under the above conditions, additional dependence on logU and C/O that contribute to
the grid consists on a total of 3,927 models that can be ob- enlargethedispersion.ThedependenceonC/Oislimitedin
tained from the 3MDB database (Morisset et al. 2015) 1. ourmethodbecauseC/Oisconstrainedinthefirstiteration
A Python-based routine called Hii-Chi-mistry-UV 2 of our calculations. As shown in Fig. 2, theC34 parameter
(hereafter HCm-UV was developed in order to calculate also presents a large dependence on log U. The latter is
O/H, C/O and logU comparing the observed relative UV obtainedinthisiterationastheaveragevalueoftheresulting
andopticalemissionlineswiththepredictionsfromthemod- distribution.ThedependenceonlogU isalleviatedusing an
els. The methodology is the same as that described for the additional parameter in the total χ2 that depends on the
optical in P´erez-Montero (2014). In a first step, the code C3C4index,definedastheratiobetweentheCiii]andCiv
calculates C/O in order to constrain the space of models in lines:
which O/H abundances can be calculated using UV carbon
emission lines. (cid:32)I(Ciii] 1908 ˚A)(cid:33)
The C/O abundance ratio is derived as the average of C3C4=log (9)
the χ2-weighteddistributionoftheC/Ovaluesinthemod- I(Civ1549 ˚A
els. The χ2 values are calculated as the relative quadratic
differencesbetweenobservedandmodel-predictedemission-
line ratios sensitive to C/O. The ratio R is used when
O3
Oiii] 1664 and 5007 ˚A are observed. In all cases, we also
4 RESULTS AND DISCUSSION
use theC3O3 parameter, defined as:
In Figure 3 we show the comparison between the chemical
(cid:32)I(Ciii] 1908 ˚A) + I(Civ1549 ˚A)(cid:33) abundances derived from both the direct method and from
C3O3=log (6) HCm-UV. In the two upper panels, the latter uses all the
I(Oiii] 1664 ˚A) available optical (Hβ and [Oiii]5007) and UV (Civ, Ciii],
whichcanbealsodefinedinabsenceoftheCivlineasthera- Oiii])emissionlines.Overall,wefindagoodagreementbe-
tiooftheCiii]and[Oiii]lines.AsshowninFig.2,according tweenthesetwodatasets, bothforC/OandO/H>7.5,with
tomodelsC3O3haslittledependenceonO/H, whilethede- anmeanoffset thatislowerthantheuncertaintyofthedi-
pendenceonlogU isonlysignificantforverylow-excitation. rect method (i.e. <0.10 dex) both for O/H and C/O. The
Figure 2 also shows the calibration sample, which follow a overall standard deviation of the residuals is 0.15 dex for
linear trend that can be used to derive C/O empirically: O/H and 0.16 dex for C/O.
Interestingly,usingHCm-UVitcouldbepossibletoob-
tainabundancesfromUVlinesaddingonlyfollow-upobser-
log(C/O)=−1.069+0.796·C3O3 (7) vations of the blue part of the rest-frame optical spectrum
with a dispersion of 0.20 dex, calculated as the standard correspondingto[Oiii]5007˚AandHβ,withoutanylowex-
deviation of the residuals. citationemissionlinesuchasOii]3727˚A. Thoughforvery
Once C/O is fixed in the grid of models, both O/H lowvaluesofO/H((cid:46)7.5)theabundancesderivedbyHCm-
and logU are calculated as the mean of the model input UVdonotfollowanycorrelationwiththosefromthedirect
values in the χ2-weighted distribution, where again the χ2 method,thecodeessentiallyidentifiesalloftheseobjectsas
valuesforeachmodelarederivedfromthequadraticrelative extremely metal-poor galaxies.
differences between the observed and predicted emission- Whenthe RO3 ratio–usedtoderivete([Oiii])inthedi-
line ratios sensitive to O/H and log U. Also in this case, rectmethod–cannotbederivedduetotheabsenceofoptical
R is used when Oiii]1664 and [Oiii]5007 are observed. lines,thecodefollowsthesamestrategyasinP´erez-Montero
O3
Similarly to the case of C/O, the χ2 values for R are (2014):itconstrainsthegridofmodelsbyassuminganem-
O3
added in quadrature to other emission-line ratios from the piricallawbetweenO/HandlogU insuchawaythatmetal-
UV, such as theC34 parameter, defined as: poor objects have higher logU and metal-rich objects have
lowerlogU.Thisempiricalrelationwasalreadyobservedby
Dopita&Evans(1986)andcouldberelatedwithanevolu-
C34=log(cid:32)I(Ciii] 1908 ˚A) + I(Civ1549 ˚A)(cid:33) (8) tionarysequence.Sinceoursamplesmallcomparedwiththe
I(Hi) samplestudiedinP´erez-Montero(2014)weadoptthesame
constrainforourgrid. InFig.4weshowthecomparisonbe-
where I(Hi) is the intensity of a Hydrogen recombination
tweentheempiricalO/H-logUrelationempiricallyderived
line.Whenopticalemissionlinesareavailable,theHIlineis
for Sa´nchez Almeida et al. (2016) and P´erez-Montero et al.
(2016)andthevaluesderivedforoursampleusingalllines.
1 3MdB project and the models can be found in As can be seen the trend is well followed by the points,
https://sites.google.com/site/mexicanmillionmodels/ Byusingthisconstrainedgrid,thecodecalculatesC/O
2 The routine HCm-UV is publicly available at using only the C3O3 ratio and the same procedure as de-
http://www.iaa.es/∼epm/HII-CHI-mistry-UV.html. scribed above and then derives O/H and logU using the
MNRAS000,1–7(2016)
Gas-phase C and O abundances in the rest UV 5
Figure 2.Relationsbetweentherelevantemissionlineratios,models,andderivedabundancesforthecompiledsample.Upperpanels:
RelationbetweenC3O3andC/OatfixedO/H(Left)andfixedlogU (Right).Thedashedblacklinerepresentsthelinearfittingtothe
observations.Lowerpanels:RelationbetweenC34andO/HatfixedC/O(Left)andatfixedlogU (Right).
C3C4 and C34 parameters, but taking Lyα as the hydro- caseofN/O(Amor´ınetal.2010;P´erez-Monteroetal.2013,
gen recombination line. This approach, however, should be 2016).
used with caution. The Lyα line is resonant and photons When no previous determination of C/O can be made
scatter in the neutral Hydrogen, thus having a complex ra- (i.e. the C3O3 ratio cannot be measured), we follow again
diation transfer. Especially for faint Lyα emission (i.e. low theproceduredescribedinP´erez-Montero(2014)forthecase
equivalent width) this may produce an additional source of of N/O: the code adopts a new constrain for the space of
systematic uncertainty. models assuming a given empirical relation between O/H
InthetwolowerpanelsofFig.3weshowthecomparison and C/O to calculate O/H using the C34 parameter. The
betweentheabundancesderivedfromthedirectmethodand codethenusestherelationbetweenO/HandN/Oshownin
fromHCm-UVinabsenceoftheopticalemissionlines.The Fig. 3 of P´erez-Montero (2014) and assumes the C/N solar
number of points with a reliable measurement of Lyα in ratio for all the sample.
our sample is very low, but all the objects are identified as
metal-poorinthissample.RegardingC/O,asthisratiohas
a very little dependence on the electron temperature, the
5 CONCLUSION
agreement is still very good, i.e. the mean offset is lower
thantheuncertaintyofthedirectmethodandthestandard In this work we present a semi-empirical method to derive
deviation of the residuals is 0.20 dex, fairly similar to that oxygen (O/H) and carbon abundances (C/O) from rest-
obtained using optical lines. frame UV emission lines, which are consistent with those
ThisresultunderlinestheimportanceoftheC/Oabun- obtained using the direct method.
danceratioasanindicatorofthechemicalcontentofgalax- We highlight the potential use of this method for (Te-
ies at different epochs. Apart from the reliability of the re- consistent) metallicity inferences at high redshifts (z > 2),
sults only using Oiii] and Ciii] UV emission lines, C/O is e.g. through deep optical spectroscopy with large ground-
notaffectedbyhydrodynamicalprocesses,suchasinflowsor based telescopes (Amor´ın et al. 2017) or using future space
outflows, and can be more tightly related with other inte- telescopes,suchastheJamesWebbSpaceTelescope.Forex-
grated properties of galaxies such as stellar mass, as in the ample, the NIRSpec spectrograph onboard the JWST, will
MNRAS000,1–7(2016)
6 Enrique P´erez-Montero & Ricardo Amor´ın
Figure 3. Comparison between C/O (Right) and O/H (Left) abundances derived from the direct method and from HCm-UV under
different assumptions. The HCm-UV-based abundances in lower panels were derived using only UV lines, while for upper panels they
alsoincludeopticalinformation,i.e.Hβ and[Oiii]5007˚A.Theredsolidlinesrepresentinallcasesthe1:1relation.
mic reionisation (at z >6). In these presumably metal-poor
objects, rest UV emission lines with relatively high EWs
(such as Ciii], Civ, and Oiii]) are apparently ubiquitous
(e.g. Stark et al. 2015; Stark et al. 2017), but the access to
faintkeyopticallines(suchas[Oiii]4363˚Aand[Oii]3727˚A
in the case of high ionisation objects) will be very limited,
thus making Te-consistent metallicity inferences extremely
challenging.
ACKNOWLEDGEMENTS
Wethanktheanonymousrefereeforconstructingandhelp-
ful comments. EPM acknowledges support from the Span-
ish MICINN through grants AYA2010-21887-C04-01 and
AYA2013-47742-C4-1-P and the Junta de Andaluc´ıa for
Figure 4. Relation between O/H as derived from the direct
grant EXC/2011 FQM-7058. RA acknowledges the support
methodandlogUderivedusingHCm-UVusingallavailablelines
fromtheERCAdvancedGrant695671“QUENCH”andthe
andthecompletegridofmodelsfortheselectedsampleofobjects.
FP7SPACEproject“ASTRODEEP”(Ref.No:312725),sup-
Theredsolilineencompassesthegridofmodelsempiricallycon-
strainedinP´erez-Montero(2014). ported by the European Commission.
beabletodetectthefullsetofrestUVandopticallinesre-
quiredtousethismethodforgalaxiesspanningfrom z∼3.5 REFERENCES
to z ∼9 in relatively short exposure times. Particularly in- AggarwaldK.M.,KeenanF.P.,2004,Phys.Scr.,69,385
terestingcouldbetheuseofthistechniquetoprovidemetal- Amor´ınR.,etal.,2017,NatAs,1,0052
licity constraints for primeval galaxies in the epoch of cos- Amor´ınR.etal.,2015,A&A,578,A105
MNRAS000,1–7(2016)
Gas-phase C and O abundances in the rest UV 7
Amor´ınR.,etal.,2014,ApJ,788,L4 P´erez-MonteroE.etal.,2013,A&A,549,A25
Amor´ınR.O.,P´erez-MonteroE.,V´ılchezJ.M.,2010,ApJ,715, P´erez-MonteroE.etal.,2016,A&A,595,A62
L128 Sa´nchez Almeida, J., P´erez-Montero, E., Morales-Luis, A. B., et
AsplundM.,GrevesseN.,SauvalA.J.,ScottP.,2009,ARA&A, al.2016,ApJ,819,110
47,481 SandersR.L.,etal.,2016,ApJ,825,L23
Bayliss M. B., Rigby J. R., Sharon K., Wuyts E., Florian M., ShapleyA.E.,etal.,2015,ApJ,801,88
GladdersM.D.,JohnsonT.,OguriM.,2014,ApJ,790,144 ShapleyA.E.,SteidelC.C.,PettiniM.,AdelbergerK.L.,2003,
Berg D. A., Skillman E. D., Henry R. B. C., Erb D. K., Carigi ApJ,588,65
L.,2016,ApJ,827,126 StarkD.P.,etal.,2017,MNRAS,464,469
Berrington K. A., Burke P. G., Dufton P. L., Kingston A. E., StarkD.P.etal.,2015,MNRAS,450,1846
1985,AtomicDataandNuclearDataTables,33,195 StarkD.P.etal.,2014,MNRAS,445,3200
CalzettiD.,ArmusL.,BohlinR.C.,KinneyA.L.,KoornneefJ., SteidelC.C.,etal.,2014,ApJ,795,165
Storchi-BergmannT.,2000,ApJ,533,682 SteidelC.C.,AdelbergerK.L.,ShapleyA.E.,PettiniM.,Dick-
CardelliJ.A.,ClaytonG.C.,MathisJ.S.,1989,ApJ,345,245 insonM.,GiavaliscoM.,2003,ApJ,592,728
ChiappiniC.,RomanoD.,MatteucciF.,2003,MNRAS,339,63 SteidelC.C.,StromA.L.,PettiniM.,RudieG.C.,ReddyN.A.,
ChristensenL.etal.,2012,MNRAS,427,1973 TrainorR.F.,2016,ApJ,826,159
ContiniT.,etal.,2016,A&A,591,A49 ThuanT.X.,IzotovY.I.,FoltzC.B.,1999,ApJ,525,105
deBarrosS.etal.,2016,A&A,585,A51 Trainor R. F., Strom A. L., Steidel C. C., Rudie G. C., 2016,
DopitaM.A.,EvansI.N.,1986,ApJ,307,431 arXiv,arXiv:1608.07280
EdmundsM.G.,1990,MNRAS,246,678 TroncosoP.,etal.,2014,A&A,563,A58
EdmundsM.G.,PagelB.E.J.,1978,MNRAS,185,77P VanzellaE.etal.,2016,ApJ,821,L27
Erb D. K., Pettini M., Shapley A. E., Steidel C. C., Law D. R., VanzellaE.,etal.,2016b,ApJ,825,41
ReddyN.A.,2010,ApJ,719,1168 VanzellaE.,etal.,2017,MNRAS,465,3803
FeltreA.,CharlotS.,GutkinJ.,2016,MNRAS,456,3354 Villar-Mart´ın M., Cervin˜o M., Gonza´lez Delgado R. M., 2004,
FerlandG.J.etal.,2013,RMxAA,49,137 MNRAS,355,1132
Garnett D. R., Shields G. A., Peimbert M., Torres-Peimbert S.,
Skillman E. D., Dufour R. J., Terlevich E., Terlevich R. J.,
ThispaperhasbeentypesetfromaTEX/LATEXfilepreparedby
1999,ApJ,513,168
theauthor.
GarnettD.R.,SkillmanE.D.,DufourR.J.,PeimbertM.,Torres-
Peimbert S., Terlevich R., Terlevich E., Shields G. A., 1995,
ApJ,443,64
GarnettD.R.,SkillmanE.D.,DufourR.J.,ShieldsG.A.,1997,
ApJ,481,174
Gavil´anM.,BuellJ.F.,Moll´aM.,2005,A&A,432,861
GutkinJ.,CharlotS.,BruzualG.,2016,MNRAS,462,1757
HenryR.B.C.,EdmundsM.G.,K¨oppenJ.,2000,ApJ,541,660
Iglesias-P´aramoJ.etal.,2016,ApJ,826,71
Izotov Y. I., Chaffee F. H., Foltz C. B., Green R. F., Guseva
N.G.,ThuanT.X.,1999,ApJ,527,757
JamesB.L.etal.,2014,MNRAS,440,1794
Jaskot,A.E.,&Ravindranath,S.2016,ApJ,833,136
KarmanW.,etal.,2015,A&A,574,A11
KewleyL.J.,JansenR.A.,GellerM.J.,2005,PASP,117,227
KobulnickyH.A.,SkillmanE.D.,1998,ApJ,497,601
KobulnickyH.A.,SkillmanE.D.,RoyJ.-R.,WalshJ.R.,Rosa
M.R.,1997,ApJ,477,679
KurtC.M.,DufourR.J.,GarnettD.R.,SkillmanE.D.,Mathis
J. S., Peimbert M., Torres-Peimbert S., Walter D. K., 1995,
inRevistaMexicanadeAstronomiayAstrofisicaConference
Series,Vol.3,RevistaMexicanadeAstronomiayAstrofisica
ConferenceSeries,PenaM.,KurtzS.,eds.,p.223
KurkJ.,etal.,2013,A&A,549,A63
LeF`evreO.etal.,2015,A&A,576,A79
LillyS.J.etal.,2007,ApJS,172,70
LuridianaV.,MorissetC.,ShawR.A.,2015,A&A,573,A42
MasedaM.V.,etal.,2014,ApJ,791,17
MattssonL.,2010,A&A,515,A68
Molla´M.,Garc´ıa-VargasM.L.,BressanA.,2009,MNRAS,398,
451
Morisset C., Delgado-Inglada G., Flores-Fajardo N., 2015,
RMxAA,51,103
MorissetC.etal.,2016,A&A,594,A37
OnoderaM.,etal.,2016,ApJ,822,42
Pagel B. E. J., Edmunds M. G., Blackwell D. E., Chun M. S.,
SmithG.,1979,MNRAS,189,95
Patr´ıcioV.,etal.,2016,MNRAS,456,4191
P´erez-MonteroE.,2014,MNRAS,441,2663
P´erez-MonteroE.,ContiniT.,2009,MNRAS,398,949
MNRAS000,1–7(2016)