Metallicity Calibrations and the Mass-Metallicity Relation for Star-Forming Galaxies
Lisa J. Kewley, Sara L. Ellison
备注：
26 pages, 5 tables, 11 figures (reduced resolution); accepted for publication in the Astrophysical Journal. . A full resolution version can be downloaded in pdf form (1 Mb) from http://www.ifa.hawaii.edu/~kewley/Metallicity/ms.pdf
(Abridged) We investigate the effect of metallicity calibrations, AGN classification, and aperture covering fraction on the local mass-metallicity (MZ) relation using 27,730 star-forming galaxies from the Sloan Digital Sky Survey (SDSS) Data Release 4. We analyse the SDSS MZ relation with 10 metallicity calibrations, including theoretical and empirical methods. We show that the choice of metallicity calibration has a significant effect on the shape and y-intercept(12+log(O/H)) of the MZ relation. The absolute metallicity scale (y-int) varies up to 0.7 dex, depending on the calibration used, and the change in shape is substantial. These results indicate that it is critical to use the same metallicity calibration when comparing different luminosity-metallicity or mass-metallicity relations. We present new metallicity conversions that allow metallicities that have been derived using different strong-line calibrations to be converted to the same base calibration. These conversions facilitate comparisons between different samples, particularly comparisons between galaxies at different redshifts for which different suites of emission-lines are available.Our new conversions successfully remove the large 0.7 dex discrepancies between the metallicity calibrations, and we reach agreement in the MZ relation to within 0.03 dex on average. We find that different AGN classification methods have negligible effect on the SDSS MZ relation. We compare the SDSS MZ relation with nuclear and global relations from the Nearby Field Galaxy Survey (NFGS). The turn over of the MZ relation depends on aperture covering fraction. We find that a lower redshift limit of z<0.04 is insufficient for avoiding aperture effects in fiber spectra of the highest stellar mass galaxies.
We present new calibrations of the widely used H_alpha, [OII], and UV luminosity vs. star formation rate (SFR) relations. Using our evolutionary synthesis code GALEV we compute the different calibrations for 5 metallicities, from 1/50 solar up to 2.5 solar. We find significant changes in the calibrations for lower metallicities compared to the standard calibrations using solar input physics.
We evaluated new colour-transformations for the Sloan photometry by 224 standards and used them to revise both the equations for photometric parallax estimation and metallicity calibration cited by Karaali et al. (2003). This process improves the metallicity and absolute magnitude estimations by [Fe/H]<=0.3 dex and M^{H}_{g} <= 0.1 mag respectively. There is a high correlation for metallicities and absolute magnitudes derived for two systems, UBV and Sloan, by means of the revised calibrations.
A modification to the absolute magnitude calibration for W~UMa-type systems, taking into account differences in metal abundances, is derived on the basis of contact binary systems recently discovered in metal-poor clusters. A preliminary estimate of the magnitude of the metallicity-dependent term for the $(B-V)$-based calibration is $ΔM_V = -(0.3 \pm 0.1) \times [Fe/H]$. The calibration based on the $(V-I_C)$ color is expected to be less sensitive with the correction term $\approx -(0.12\pm 0.05) \times [Fe/H]$.
We point out that the absolute magnitude of the Galactic Cepheids, which have been used for local calibrations for Cepheid period-luminosity relation, shows an unexpectedly large metal abundance effect. This indicates that either Cepheids period luminosity relation has a large metallicity dependence, much larger than expected in stellar pulsation models, or the unreddened distances to clusters contain a large systematic error arising from the neglect of the metallicity effect. This effect may introduce an uncertainty of --0.4 to +0.1 mag into the distance to the LMC estimated using the Cepheid…
We present an improved uvby-metallicity relation calibrated for F, G, and early K dwarfs, and an analogous uvby-T(eff) relation, both derived using a Levenberg-Marquardt minimization scheme. Our calibrations are based on 1533 stars which appear in both the Cayrel de Strobel (2001) metallicity compilation, and in the Hauck-Mermilliod (1998) catalog of uvby photometry. We also examine the speculative possibility of using uvby photometry to produce a uvby-planeticity calibration. We conclude that while there is likely no strong photometric indicator of the presence or absence of short-period plan…
We adopted the procedure of Carney to obtain a metallicity calibration for dwarfs for the RGU-photometry. For this purpose we selected 76 dwarfs of different metallicities from Carney, and Strobel et al., and evaluated their δ(U-G) ultra-violet excess relative to Hyades by transforming their UBV magnitudes to RGU via metallicity dependent equations of Ak-Gungor. The δ_{0.6}/\deltaM normalized factors of Sandage transform δ(U-G) excess at any G-R to δ=δ_{1.08}, i.e.: the ultra-violet excess at G-R=1.08 mag, corresponding to B-V = 0.60 mag in the UBV-system. Finally, the (δ, [Fe/H]) couples were…
We performed a new calibration of the Stroemgren metallicity index m1 based on the b-y color of cluster red giant stars. The current Metallicity-Index-Color relation is not linear in the color range 0.40 < b-y < 1.0, but provides iron abundances of cluster and field red giants with an accuracy of ~ 0.25 dex.
We consider the effect of stellar activity, as measured by X-ray luminosity, on metallicities of Solar-neighborhood F and G dwarfs derived from Stromgren photometry. Rocha-Pinto & Maciel found evidence that Stromgren colors systematically underpredict [Fe/H] for stars with extremely high Ca II H & K emission. We investigate whether a recent photometric metallicity calibration derived by Martell & Laughlin might be subject to this effect, and whether the amount of underprediction could reliably be expressed as a function of log (L_X/L_bol). Among those calibration stars used by Martell & Laughl…
We obtained high resolution ELODIE and CORALIE spectra for both components of 20 wide visual binaries composed of an F-, G- or K-dwarf primary and an M-dwarf secondary. We analyse the well understood spectra of the primaries to determine metallicities ([Fe/H]) for these 20 systems, and hence for their M dwarf components. We pool these metallicities with determinations from the literature to obtain a precise (+- 0.2 dex) photometric calibration of M dwarf metallicities. This calibration represents a breakthrough in a field where discussions have had to remain largely qualitative, and it helps u…
The pulsational method to estimate the absolute magnitudes of RR Lyrae stars is updated with new data on field-star metallicities by Layden, a new calibration of the (B-V)_0-temperature correlation from recent atmospheric models by Bell and Tripicco, and new mass estimates by Bono et al. New linear and non-linear calibrations of M_V(RR)=f([Fe/H]) are derived that depend on the nature of the adopted envelope in a diagram of log period versus metallicity fitted to the shortest period field variables in each metallicity range, together with the stated assumptions on the colour of the stars at tha…
88 dwarfs, covering the colour index interval 0.37 <(B-V)o <= 1.07 mag, with metallicities -2.70 <= [Fe/H] <= +0.26 dex have been taken from three different sources for new metallicity calibration. The catalogue of Cayrel de Stroble et al. (2001) which includes 65% of the stars in our sample supplies detailed information on abundances for stars with determination based on high resolution spectroscopy. 77 stars which supplies at least one of the following conditions have been used as "corner stones" for constructing the new calibration: (i) the parallax is larger than 10 mas (distance relative …
A new calibration of the Stroemgren (b-y),m_1 diagram in terms of iron abundance of red giants is presented. This calibration is based on a homogeneous sample of giants in the globular clusters omega Centauri, M22, and M55 as well as field giants from the list of Anthony-Twarog & Twarog (1998). Towards high metallicities, the new calibration is connected to a previous calibration by Grebel & Richtler (1992), which was unsatisfactory for iron abudances lower than -1.0 dex. The revised calibration is valid for CN-weak/normal red giants in the abundance range of -2.0 < [Fe/H]r $< 0.0 dex, and a c…
We present an improved calibration of photometric metallicity indicators, derived from the new metallicity scale for Globular Clusters presented by Carretta & Gratton (1997) and based on direct high resolution spectroscopy of 160 stars in 24 globular clusters. We have carefully recalibrated the traditional abundance indices based upon the red giant branch (RGB) morphology, both in the V,B-V and V,V-I planes, namely the dereddened colour at the luminosity level of the horizontal branch (HB), and the magnitude difference between the HB and the RBG at a given dereddened colour. Finally, we give n…
We construct three extreme different scenarios of the star formation histories applicable to a sample of dwarf galaxies, based either on their present metallicity or their luminosity. The three possible scenarios imply different mechanical energy input rates and these we compare with the theoretical lower limits established for the ejection of processed matter out of dwarf galaxies. The comparison strongly points at the existence of extended gaseous haloes in these galaxies, acting as the barrier that allows galaxies to retain their metals and enhance their abundance. At the same time our find…
We present a spectrophotometric study of circumnuclear star forming regions (CNSFR) in the early type spiral galaxies: NGC 2903, NGC 3351 and NGC 3504, all of them of over solar metallicity according to standard empirical calibrations. A detailed determination of their abundances is made after careful subtraction of the very prominent underlying stellar absorption. It is found that most regions show the highest abundances in HII region-like objects. The relative N/O and S/O abundances are discussed. In is also shown that CNSFR, as a class, segregate from the disk HII region family, clustering …
We show the dependence of the Bayesian and the N2-derived oxygen abundances on the N-enrichment history of the star-forming galaxies. We select 531 metal-poor and ~20,000 metal-rich galaxies from the SDSS database for this study. Their "accurate" O/H abundances are obtained from Te and P-method, respectively. The discrepancies of the Bayesian and N2-derived abundances from these two "accurate" abundances show obvious correlations with log(N/O) abundance ratios.
We extend the colour calibration of the widely used BaSeL standard stellar library (Lejeune, Cuisinier, & Buser 1997, 1998) to non-solar metallicities, down to [Fe/H] ~ -2.0 dex. Surprisingly, we find that at the present epoch it is virtually impossible to establish a unique calibration of UBVRIJHKL colours in terms of stellar metallicity [Fe/H] which is consistent simultaneously with both colour-temperature relations and colour-absolute magnitude diagrams (CMDs) based on observed globular cluster photometry data and on published, currently popular standard stellar evolutionary tracks and isoc…
We present 4-meter Kitt Peak telescope and 6.5-meter MMT spectrophotometry of the extremely low-metallicity galaxy SBS 0335-052W, the western companion of the blue compact dwarf galaxy SBS 0335-052E. These observations have been combined with published 10-meter Keck data to derive for the brightest region of SBS 0335-052W an oxygen abundance 12+logO/H=7.12+/-0.03. This makes SBS 0335-052W the lowest metallicity star-forming galaxy known in the local universe. Using a Monte Carlo technique, we fit the spectral energy distribution of SBS 0335-052W to derive the age of the oldest stars contributi…
The metallicity dependence of the primary indices of the uvby photometric system for cooler dwarfs (Teff ~ 6500 K to 5000K) is investigated. The data base for the analysis is composed of the overlap between a composite catalog of selected, high-dispersion spectroscopic abundances for 1801 stars on the metallicity scale of Valenti and Fischer (2005) and a merged catalog of high-precision uvbyHbeta photometry for over 35,000 stars. While [Fe/H] for F dwarfs is best estimated from m_1, with a modest dependence on c_1 as expected, for hotter G dwarfs the pattern reverses and c_1 becomes the domina…
The mean metallicity of the Milky Way thin disc in the solar neighbourhood is still a matter of debate, and has recently been subject to upward revision (Haywood, 2001). Our star sample was drawn from a set of solar neighbourhood dwarfs with photometric metallicities. In a recent study, Reid (2002) suggests that our metallicity calibration, based on Geneva photometry, is biased. We show here that the effect detected by Reid is not a consequence of our adopted metallicity scale, and we confirm that our findings are robust. On the contrary, the application to Stromgren photometry of the Schuster…
Evolution of galaxies through cosmic time has been widely studied at high redshift, but there are a few studies in this field at lower redshifts. However, low-redshifts studies will provide important clues to the evolution of galaxies, furnishing the required link between local and high-redshift universe. In this work we focus on the metallicity of the gas in spiral galaxies at low redshift looking for signs of chemical evolution. We analyze the metallicity contents of star forming galaxies of similar luminosities at different redshifts, we studied the metallicity of star forming galaxies from…
Infrared JKs-band photometry of the Galactic globular clusters M15 and, for the first time, M56 is presented. We estimated the reddening (E(B-V)=0.18+/-0.08 mag) and distance modulus ((m-M)_V=15.43+/-0.30 mag) towards the poorly studied globular cluster M56. We combined our data with observations of other clusters from the literature (12 in total) to extend the [Fe/H] vs. Red Giant Branch (RGB) slope relation towards metal-poor clusters. Our best fit yields to [Fe/H]=-3.40(+/-0.22)-27.74(+/-2.35)x(RGB Slope), with an r.m.s.=0.20. The broader metallicity baseline greatly reduced the uncertainti…
A brief review concludes that there is now good overall agreement between theoretical estimates of the energy associated with the production of the observed metal content of the Universe and the observed extragalactic background light. In addition the overall form of the star-formation history over 0 < z < 5 is reasonably well constrained. The study of emission line gas in galaxies as a function of redshift provides a complementary view of chemical evolution to that obtained from studies of absorption line systems in quasar spectra. Emission line gas is more relevant for some questions, includ…
We present an empirical calibration of the Mixing-Length free parameter $α$ based on a homogeneous Infrared database of 28 Galactic globular clusters spanning a wide metallicity range (-2.15$<$[Fe/H]$<$-0.2). Empirical estimates of the red giant effective temperatures have been obtained from infrared colors. Suitable relations linking these temperatures to the cluster metallicity have been obtained and compared to theoretical predictions. An appropriate set of models for the Sun and Population II giants have been computed by using both the standard solar metallicity $[Z/X]_{\odot}=0.0275$ and …
Ages, metallicities, space velocities, and Galactic orbits of stars in the Solar neighbourhood are fundamental observational constraints on models of galactic disk evolution. We aim to consolidate the calibrations of uvby photometry into Te, [Fe/H], distance, and age for F and G stars and rediscuss the results of the Geneva-Copenhagen Survey (Nordstrom et al. 2004; GCS) in terms of the evolution of the disk. We substantially improve the Te and [Fe/H] calibrations for early F stars, where spectroscopic temperatures have large systematic errors. Our recomputed ages are in excellent agreement wit…
New empirical calibrations of the Red Giant Branch Tip in the I,J,H and K bands based on two fundamental pillars, namely omega Centauri and 47 Tucanae, have been obtained by using a large optical and near infrared photometric database. Our best estimates give M_I(TRGB)= -4.05 pm 0.12, M_J(TRGB)= -5.20 pm 0.16, M_H(TRGB)= -5.94 pm 0.18 and M_K(TRGB) = -6.04 pm 0.16 at [M/H]=-1.5 (omega Cen) and M_I(TRGB)= -3.91 pm 0.13, M_J(TRGB)= -5.47 pm 0.25, M_H(TRGB)= -6.35 pm 0.30 and M_K(TRGB)= -6.56 pm 0.20 at [M/H]=-0.6 (47 Tuc). With these new empirical calibrations we also provide robust relations of…
Cloud evolution for various metallicities is investigated by three-dimensional nested grid simulations, in which the initial ratio of rotational to gravitational energy of the host cloud β_0 (=10^-1 - 10^-6) and cloud metallicity Z (=0 - Z_\odot) are parameters. Starting from a central number density of n = 10^4 cm^-3, cloud evolution for 48 models is calculated until the protostar is formed (n \simeq 10^23 cm^-3) or fragmentation occurs. The fragmentation condition depends both on the initial rotational energy and cloud metallicity. Cloud rotation promotes fragmentation, while fragmentation t…
Although stellar SED models are constantly improving and seem to show similar differential tendencies as measured spectra, their absolute properties (eg. absolute flux in different wavelength regions) can still differ significantly from the measured ones. Therefore, it is necessary to "color-calibrate" the theoretical spectra to obtain realistic model-based stellar spectra, which can then be used for many purposes, such as population synthesis models. In order to calibrate a hybrid library (consisting of the Kurucz -, the Allard and Hauschild -, and the Scholz-models) of theoretical SEDs, UBVR…
We present synthetic ultraviolet spectra of metal-poor star-forming galaxies which were calculated with the Starburst99 package. A new spectral library was generated from HST observations of O stars in the Large and Small Magellanic Clouds. The corresponding mean metallicity of the synthetic spectra is approximately 1/4\Zs. The spectra have a resolution of 1Åand cover the spectral range 1200-1600Å. A set of model spectra was calculated for a standard initial mass function and star-formation history and is compared to synthetic spectra at solar metallicity. We find that the spectral lines are g…
Generally the gas metallicity in distant galaxies can only be inferred by using a few prominent emission lines. Various theoretical models have been used to predict the relationship between emission line fluxes and metallicity, suggesting that some line ratios can be used as diagnostics of the gas metallicity in galaxies. However, accurate empirical calibrations of these emission line flux ratios from real galaxy spectra spanning a wide metallicity range are still lacking. In this paper we provide such empirical calibrations by using the combination of two sets of spectroscopic data: one consi…
We adopted uvby Stroemgren photometry to investigate the metallicity distribution of Omega Cen Red Giant (RG) stars. We provided a new empirical calibration of the Stroemgren m1 = (v-b)-(b-y) metallicity index based on cluster stars. The new calibration has been applied to a sample of Omega Cen RGs. The shape of the estimated metallicity distribution is clearly asymmetric, with a sharp cut-off at low metallicities ([Fe/H] < -2.0) and a metal-rich tail up to [Fe/H] ~ 0.0. Two main metallicity peaks have been identified, around [Fe/H] ~ -1.9 and -1.3 dex, and a metal-rich shoulder at ~ 0.2 dex.
We utilize Sloan Digital Sky Survey imaging and spectroscopy of ~53,000 star-forming galaxies at z~0.1 to study the relation between stellar mass and gas-phase metallicity. We derive gas-phase oxygen abundances and stellar masses using new techniques which make use of the latest stellar evolutionary synthesis and photoionization models. We find a tight (+/-0.1 dex) correlation between stellar mass and metallicity spanning over 3 orders of magnitude in stellar mass and a factor of 10 in metallicity. The relation is relatively steep from 10^{8.5} - 10^{10.5} M_sun, in good accord with known tren…
The chemical composition of high redshift galaxies is an important property which gives clues to their past history and future evolution and yet is difficult to measure with current techniques. In this paper we investigate new metallicity indicators, based upon the strengths of stellar photospheric features at rest-frame ultraviolet wavelengths. By combining the evolutionary spectral synthesis code Starburst99 with the output from the non-LTE model atmosphere code WM-basic, we have developed a code that can model the integrated ultraviolet stellar spectra of star-forming regions at metalliciti…
Based on detailed 2D numerical radiation hydrodynamics (RHD) calculations of time-dependent compressible convection, we have studied the dynamics and thermal structure of the convective surface layers of solar-type stars. The RHD models provide information about the convective efficiency in the superadiabatic region at the top of convective envelopes and predict the asymptotic value of the entropy of the deep, adiabatically stratified layers. This information is translated into an effective mixing-length parameter alpha suitable to construct standard stellar structure models. We validate the a…
We present new empirical and theoretical calibrations of two photometric metallicity indices based on Walraven photometry. The empirical calibration relies on a sample of 48 Cepheids for which iron abundances based on high resolution spectra are available in the literature. They cover a broad range in metal abundance (-0.5 < [Fe/H] < +0.5) and the intrinsic accuracy of the Metallicity Index Color (MIC) relations is better than 0.2 dex. The theoretical calibration relies on a homogeneous set of scaled-solar evolutionary tracks for intermediate-mass stars and on pulsation predictions concerning …
We present a new calibration of the Stroemgren metallicity index m1 using red giant (RG) stars in 4 globular clusters (GCs:M92,M13,NGC1851,47Tuc) with metallicity ranging from [Fe/H]=-2.2 to -0.7, marginally affected by reddening (E(B-V)<0.04) and with accurate u,v,b,y photometry.The main difference between the new metallicity-index-color (MIC) relations and similar relations available in the literature is that we adopted the u-y/v-y colors instead of the b-y.These colors present a stronger sensitivity to effective temperature, and the MIC relations show a linear slope. The difference between …
Archival Spitzer observations of 41 starburst galaxies that span a wide range in metallicity reveal for the first time a correlation between the [FeII]/[NeII] 26.0/12.8 micron ratio and the electron gas density as traced by the 18.7/33.4 micron [SIII] ratio, with the [FeII]/[NeII] ratio decreasing with increasing gas density. The correlations of the [FeII]/[NeII] ratio, the PAH peak to continuum strength & metallicity found in an earlier paper were confirmed for a larger sample. We also find a strong correlation between the gas density and the PAH peak to continuum strength. Using shock and ph…
Archival Spitzer observations of 41 starburst galaxies that span a wide range in metallicity reveal for the first time a correlation between the [FeII]/[NeII] 26.0/12.8 micron ratio and the electron gas density as traced by the 18.7/33.4 micron [SIII] ratio, with the [FeII]/[NeII] ratio decreasing with increasing gas density. We also find a strong correlation between the gas density and the PAH peak to continuum strength. Using shock and photoionization models, we see the driver of the observed [FeII]/[NeII] ratios is metallicity. The majority of [FeII] emission in low metallicity galaxies may…
I present deep spectroscopy of four HII regions in the inner, metal-rich zone of the spiral galaxy M101 obtained with the LRIS spectrograph at the Keck telescope. From the analysis of the collisionally excited lines in two of the target HII regions, H1013 and H493, I have obtained oxygen abundances 12+log(O/H)=8.52 and 12+log(O/H)=8.74, respectively. These measurements extend the determination of the oxygen abundance gradient of M101 via the direct method to only 3 kpc from the center. The intensity of the CII 4267 line in H1013 leads to a carbon abundance 12+log(C/H)=8.66, corresponding to ne…
The average near-infrared (K-band) luminosity of 238 Hipparcos red clump giants is derived and then used to measure the distance to the Galactic center. These Hipparcos red clump giants have been previously employed as I-band standard candles. The advantage of the K-band is a decreased sensitivity to reddening, and perhaps also a reduced systematic dependence on metallicity. In order to investigate the latter, and also to refer our calibration to a known metallicity zero-point, we restrict our sample of red clump calibrators to those with abundances derived from high-resolution spectroscopic d…
Estimates of the [O/H] metallicity of the star-forming gas in a sample of 66 CFRS galaxies between 0.47 < z < 0.92 have been made based on the flux ratios of bright emission lines. Most galaxies (> 75%) have the [O/H] ~ 8.9 metallicities that are seen locally in galaxies of similar luminosities. However, a minority (< 25%) appear to have significantly lower metallicities [O/H] < 8.6 as indicated by high values of the Pagel R23 parameter. The high metallicities of the majority of the galaxies suggest that they do not fade to be low metallicity dwarf galaxies today. Only one of the 66 galaxies h…
We test the hypothesis that the host galaxies of long-duration gamma-ray bursts (GRBs) as well as quasar-selected damped Lyman-alpha (DLA) systems are drawn from the population of UV-selected star-forming, high-z galaxies (generally referred to as Lyman-break galaxies). Specifically, we compare the metallicity distributions of the GRB and DLA populations to simple models where these galaxies are drawn randomly from the distribution of star-forming galaxies according to their star-formation rate and HI cross-section respectively. We find that it is possible to match both observational distribut…
We present mid-infrared Spitzer Space Telescope observations of a complete sample of star-forming dwarf galaxies selected from the KPNO International Spectroscopic Survey. The galaxies span a wide range in mid-infrared properties. Contrary to expectations, some of the galaxies emit strongly at 8 micron indicating the presence of hot dust and/or PAHs. The ratio of this mid-infrared dust emission to the stellar emission is compared with the galaxies' luminosity, star-formation rate, metallicity, and optical reddening. We find that the strength of the 8.0 micron dust emission to the stellar emiss…
