Pith. Sign in

REVIEW

The Mass-Metallicity Relation at z~2

PaperPDF from arXiv

data files

ReviewThe objections and verdict, as JSON Paper breakdownThe authors' claims, conditions and measurements, as JSON Signed receiptProof this review has not been altered
X Bluesky LinkedIn Reddit Hacker News
Dawn K. Erb (CfA), Alice E. Shapley (Princeton), Max Pettini (IoA), Charles C. Steidel (Caltech), Naveen A. Reddy (Caltech), Kurt L. Adelberger (McKinsey)

Astrophysics · Submitted 21 Feb 2006 · 103 references

Pith's verdict: The paper's real contribution is the direct z~2 mass-metallicity relation; the outflow-rate claim in the abstract is stronger than the degenerate (y, f) model actually allows.

T0 · AI review

How this review was made: An AI referee (deepseek-v4-flash) read the paper on August 28, 2026 and answered a fixed list of questions about it. T0 is the starting level every paper gets. The levels describe how deeply Pith has checked the paper, never the authors.

  1. T0 AI review An AI read the paper and answered a fixed list of questions.
  2. T1 Audit Automatic checks that the references and files the paper relies on exist and match.
  3. T2 Compute The paper's code or calculations re-run with the same results.
  4. T3 Formal statement The claims written out in a formal math language a computer can read.
  5. T4 Proof Those claims proved, and the proof checked by computer.

How Pith checks papers →

· 2 major objections · 3 minor · Ask about this paper · The compiled record · Read the full review →
Challenge this review Comment (0) Read on arXiv ↗
Cite ▾
arXiv IDastro-ph/0602473 (version 1)
Pith IDODT3LRRI, Pith's permanent record of this paper and its review
CitationAlice E. Shapley (Princeton), Charles C. Steidel (Caltech), Dawn K. Erb (CfA), Kurt L. Adelberger (McKinsey), Max Pettini (IoA), Naveen A. Reddy (Caltech). "The Mass-Metallicity Relation at z~2." arXiv:astro-ph/0602473 (2006). https://pith.science/paper/ODT3LRRI
Cite this reviewBibTeX for Pith's review of this paper
Claim paper

Why Pith says this

Highlight any text to comment on it.

Referee report

How Pith checks papers → · Full audit of this paper →

The paper

The paper claims

At redshift 2, star-forming galaxies show a monotonic mass–metallicity relation, offset about 0.3 dex below local galaxies of the same stellar mass.

The reading

This paper establishes that star-forming galaxies at redshift ~2 follow a mass–metallicity relation: the oxygen abundance rises monotonically with stellar mass, from 12+log(O/H) < 8.2 at ~3×$10^{9}$ M_sun to 8.6 at ~$10^{11}$ M_sun. The relation sits about 0.3 dex below the local one, meaning a galaxy of fixed stellar mass was more metal-poor ten billion years ago. Using Schmidt-law gas fractions, the authors show this offset and the relation's slope trace higher gas fractions in the past, and they fit the metallicity–gas-fraction trend with a model of strong outflows (mass-loss rate ~4 times the star-formation rate) and a supersolar yield. The result matters because it ties the mass–metallicity relation to gas consumption and galactic winds, and it argues against the preferential loss of metals from low-mass galaxies at high redshift.

What carries the argument

The argument rests on composite spectroscopy and the closed-box/outflow chemical-evolution model. Metallicities come from the N2 index, log([NII]/Hα), calibrated by Pettini & Pagel (2004) to oxygen abundance; stacking spectra in six stellar-mass bins raises the signal-to-noise needed to detect weak [NII]. Gas fractions are derived from the Kennicutt (1998) Schmidt law, converting Hα surface density to gas surface density. These two inputs feed the effective yield, $y_{\rm eff} = Z/\ln(1/\mu)$, and the outflow-modified relation $Z = y(1+f)^{-1} \ln[1+(1+f)(\mu^{-1}-1)]$, where $f$ is the outflow rate in units of the star-formation rate; comparison of the observed $Z$–$\mu$ trend with these models yields the best-fit supersolar yield and $f \sim 4$.

Extended reading notes

Core claim

The central discovery is a monotonic relation between stellar mass and gas-phase oxygen abundance at z~2, measured from composite Hα+[NII] spectra of 87 UV-selected galaxies divided into six mass bins. The relation is offset from the local (SDSS) relation by ~0.3 dex at fixed stellar mass, and the effective yield—the metallicity per unit log gas fraction—does not fall with decreasing baryonic mass as it does locally; if anything it rises slightly. The paper interprets this as evidence that the z~2 mass–metallicity relation is driven by the conversion of gas to stars (gas fractions fall from ~85% to ~20% across the mass range), modulated by metal loss from outflows in galaxies of all masses, with no preferential metal loss from low-mass systems.

Load-bearing premise

The derivation of gas fractions assumes that the local relation between star-formation rate per area and gas density applies unchanged at z~2; if that relation does not hold there, the inferred gas fractions, effective yields, and outflow rate are all compromised, while the observed mass–metallicity trend would remain.

Editorial extensions

If this is right

  • If the mass–metallicity relation at z~2 is universal, the local relation is not primordial: metallicity tracks the cosmic decline in gas fraction as galaxies consume their gas.
  • The ~0.3 dex offset provides a direct benchmark for galaxy formation simulations, which must reproduce the factor-of-two under-enrichment at fixed stellar mass ten billion years ago.
  • The absence of a significant metallicity–luminosity correlation at z~2, combined with a strong mass–metallicity correlation, implies stellar mass is the fundamental driver of chemical enrichment.
  • An outflow rate of order four times the star-formation rate means high-redshift galaxies expel most of their metals into the surrounding medium, contributing to early metal enrichment of the intergalactic and intragroup gas.
  • The supersolar yield preferred by the model, if coupled to a top-heavy initial mass function, would also raise the expected stellar masses and star-formation rates derived from spectral energy distributions.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The gas fractions are derived by assuming the local star-formation-rate–gas-density relation holds at z~2; a direct CO-based measurement of gas masses at this redshift would either confirm or overturn the outflow-rate inference, while leaving the observed mass–metallicity trend intact.
  • The supersolar yield is an indirect inference; if a top-heavy IMF is responsible, it would simultaneously shift the SED-derived stellar masses, potentially altering the zero-point of the mass–metallicity relation and the magnitude of the metallicity offset from local galaxies.
  • The paper's scenario implies that today's early-type galaxies, which show stellar metallicities near solar, are plausible descendants of the massive z~2 galaxies; comparing stellar metallicities of quiescent galaxies at z~0 with these gas-phase abundances could test that continuity.
  • A testable extension would measure electron-temperature metallicities (via [OIII]λ4363) or use the O3N2 index on a subset of z~2 galaxies; a calibration shift would change the absolute metallicity scale and the size of the offset, but the monotonic mass–metallicity trend itself would remain.
Original abstract

We use a sample of 87 rest-frame UV-selected star-forming galaxies with mean spectroscopic redshift z=2.26 to study the correlation between metallicity and stellar mass at high redshift. Using stellar masses determined from SED fitting to 0.3-8 micron photometry, we divide the sample into six bins in stellar mass, and construct six composite H-alpha+[NII] spectra from all of the objects in each bin. We estimate the mean oxygen abundance in each bin from the [NII]/H-alpha ratio, and find a monotonic increase in metallicity with increasing stellar mass, from 12+log(O/H) < 8.2 for galaxies with <M_star> = 2.7e9 Msun to 12+log(O/H) = 8.6 for galaxies with <M_star> = 1e11 Msun. We use the empirical relation between star formation rate density and gas density to estimate the gas fractions of the galaxies, finding an increase in gas fraction with decreasing stellar mass. These gas fractions combined with the observed metallicities allow the estimation of the effective yield y_eff as a function of stellar mass; in constrast to observations in the local universe which show a decrease in y_eff with decreasing baryonic mass, we find a slight increase. Such a variation of metallicity with gas fraction is best fit by a model with supersolar yield and an outflow rate ~4 times higher than the star formation rate. We conclude that the mass-metallicity relation at high redshift is driven by the increase in metallicity as the gas fraction decreases through star formation, and is likely modulated by metal loss from strong outflows in galaxies of all masses. There is no evidence for preferential loss of metals from low mass galaxies as has been suggested in the local universe. [Abridged]

The compiled record

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-09-23: 1 request, 6,700 input tokens.

What the abstract itself states, as probabilities: claims first 0.03 · states limitation 0.18 · human participants 0.01 · controlled comparison 0.24 · randomized assignment 0.01 · states quantitative result 0.98 · compares with prior results 0.96 · mentions code or data availability 0.02.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; .

We use a sample of 87 rest-frame UV-selected star-forming galaxies with mean spectroscopic redshift z=2.26 to study the correlation between metallicity and stellar mass at high redshift. Using stellar masses determined from SED fitting to 0.3-8 micron photometry, we divide the sample into six bins in stellar mass, and construct six composite H-alpha+[NII] spectra from all of the objects in each bin. We estimate the mean oxygen abundance in each bin from the [NII]/H-alpha ratio, and find a monotonic increase in metallicity with increasing stellar mass, from 12+log(O/H) < 8.2 for galaxies with <M_star> = 2.7e9 Msun to 12+log(O/H) = 8.6 for galaxies with <M_star> = 1e11 Msun. We use the empirical relation between star formation rate density and gas density to estimate the gas fractions of the galaxies, finding an increase in gas fraction with decreasing stellar mass. These gas fractions combined with the observed metallicities allow the estimation of the effective yield y_eff as a function of stellar mass; in constrast to observations in the local universe which show a decrease in y_eff with decreasing baryonic mass, we find a slight increase. Such a variation of metallicity with gas fraction is best fit by a model with supersolar yield and an outflow rate ~4 times higher than the star formation rate. We conclude that the mass-metallicity relation at high redshift is driven by the increase in metallicity as the gas fraction decreases through star formation, and is likely modulated by metal loss from strong outflows in galaxies of all masses. There is no evidence for preferential loss of metals from low mass galaxies as has been suggested in the local universe. [Abridged]

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#sentencerolethe authors' own workstates the main conclusionstrength
s0 We use a sample of 87 rest-frame UV-selected star-forming galaxies with mean spectroscopic redshift z=2.26 to study the correlation between metallicity and stellar mass at high redshift. contribution 0.72 0.96 0.02 asserted 0.71
s1 Using stellar masses determined from SED fitting to 0.3-8 micron photometry, we divide the sample into six bins in stellar mass, and construct six composite H-alpha+[NII] spectra from all of the objects in each bin. method 1.00 0.97 0.01 asserted 0.95
s2 We estimate the mean oxygen abundance in each bin from the [NII]/H-alpha ratio, and find a monotonic increase in metallicity with increasing stellar mass, from 12+log(O/H) < 8.2 for galaxies with <M_star> = 2.7e9 Msun to 12+log(O/H) = 8.6 for galaxies with <M_star> = 1e11 Msun. result 1.00 0.98 0.12 asserted 1.00
s3 We use the empirical relation between star formation rate density and gas density to estimate the gas fractions of the galaxies, finding an increase in gas fraction with decreasing stellar mass. method 0.53 0.95 0.02 asserted 1.00
s4 These gas fractions combined with the observed metallicities allow the estimation of the effective yield y_eff as a function of stellar mass; in constrast to observations in the local universe which show a decrease in y_eff with decreasing baryonic mass, we find a slight increase. result 0.99 0.94 0.08 asserted 1.00
s5 Such a variation of metallicity with gas fraction is best fit by a model with supersolar yield and an outflow rate ~4 times higher than the star formation rate. interpretation 0.76 0.84 0.07 asserted 0.84
s6 We conclude that the mass-metallicity relation at high redshift is driven by the increase in metallicity as the gas fraction decreases through star formation, and is likely modulated by metal loss from strong outflows in galaxies of all masses. interpretation 1.00 0.82 0.94 hedged possibility 0.64
s7 There is no evidence for preferential loss of metals from low mass galaxies as has been suggested in the local universe. [Abridged] result 0.91 0.80 0.09 asserted 0.99

What kind of paper the classifier reports this is: empirical study 1.00 · evidence from observation or experiment 0.78.

What was examined, and what was not

sources read
  • arXiv listing abstract v1
not examined
  • full text
  • figures and tables
  • references and citing sentences
  • relations to other papers
levels run
  • P abstract
not run
  • S sections
  • T body sentences
  • C citations
  • R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

Figures
Figure 1. — The [O III]/Hβ vs. [N II]/Hα diagnostic diagram. The four galaxies in our sample for which we have measurements of all four lines are shown by the large red circles. The z = 2.2 galaxy discussed by van Dokkum et al. (2005), which shows evidence of an AGN or shock ionization by a wind, is shown by the green ×. The small grey points represent ∼ 96, 000 objects from the SDSS, and the small black points are the local starburst galaxies studied by Kewley et al. (2001b). The dashed line shows the maximum theo￾retical starburst line of Kewley et al. (2001a); for realistic combina￾tions…
Figure 2. — The composite NIRSPEC spectra of the 87 galaxies in our sample, divided into six bins of 14 or 15 objects each by in￾creasing stellar mass (panels a through f). The spectra are labeled with the mean stellar mass in each bin, and the Hα, [N II], and [S II] lines are marked with dotted lines (left to right respectively). The increase in the strength of [N II] with stellar mass can be seen clearly. The density-sensitive [S II] lines, while weak, indicate a typical electron density of ne ∼ 500 cm−3 , with no significant de￾pendence on mass; this is a value comparable to that seen in…
Figure 3. — The observed relation between stellar mass and oxygen abundance at z ∼ 2 is shown by the large red circles. Each point represents the average value of 14 or 15 galaxies, with the metallicity estimated from the [N II]/Hα ratio of their composite spectrum. Horizontal bars indicate the range of stellar masses in each bin, while the vertical error bars show the uncertainty in the [N II]/Hα ratio. The additional error bar in the lower right corner shows the additional uncertainty in the N2 calibration itself. The dashed blue line is the best-fit mass-metallicity relation of Tremonti et…
Figure 4. — Composite rest-frame UV spectra of galaxies in the two lower (top panel) and two higher (bottom panel) mass bins in our sample. The upper spectrum is the average of 30 LRIS-B spectra of galaxies with mean stellar mass hM⋆i = 5 × 109 M⊙, while 28 spectra contribute to the lower composite for which hM⋆i = 7 × 1010 M⊙. The spectra have been divided by the underlying stellar continuum estimated according to the prescription by Rix et al. (2004). Differences in the stellar, interstellar, and nebular lines between the two composites are discussed in the text. of the galaxies in each…
Figure 6. — The luminosity-metallicity relation at z & 2. We have divided the sample into six bins by rest-frame absolute B magni￾tude, and estimated the metallicity in each bin. The symbols are the same as in
Figure 7. — Absolute rest-frame B magnitude vs. stellar mass, for the individual galaxies in the z ∼ 2 sample (large red circles) and the SDSS (small grey points). The dashed and dotted lines show constant mass-to-light ratios M/L; in the z ∼ 2 sample, the total range in luminosity is narrow and M/L varies by a factor of ∼ 70 at most luminosities, while for most SDSS galaxies, M/L varies by a factor of 2–5 at a given luminosity. large variation in M/L explains the lack of correlation in the z ∼ 2 L-Z relation compared to the local relation
Figure 8. — The variation of stellar, gas, and baryonic mass in each of the six bins with metallicity. Across the observed range in oxygen abundance, stellar mass increases strongly, baryonic mass increases weakly, and gas mass remains approximately constant. We thus see an increase in metallicity with decreasing gas fraction. The strong correlation between stellar mass and age means that metallicity also increases with the age of the stellar population. The error bars at upper left show typical uncertainties in gas and baryonic masses and metallicities; uncertainties in stellar masses are…
Figure 9. — The points in each panel show the mean metallicity Z in each mass bin, plotted against the mean gas fraction µ. Gas fraction decreases from left to right, to show the increase in metallicity with decreasing gas fraction. The curves illustrate the variation of Z with µ for three different values of the true yield: from left to right, y = 0.6 Z⊙ (our observed effective yield), y = Z⊙, and y = 1.5 Z⊙. Within each panel, the curves show varying values of the mass outflow rate M˙ , parameterized as a fraction f of the star formation rate. From top to bottom in each panel, the curves…

Showing 8 of the paper's 9 figures. See all figures in the PDF ↗ Figures reproduced from arXiv:astro-ph/0602473 by the authors.

References
How later papers used it

147 later papers cite it

Evidence: substantial. 126 of the citing papers share no author with it, and 13 of those carry a confident label from Pith's citation classifier.

Pith read 12 of the 195 citing sentences, choosing only those whose citation markers point at this paper alone, and wrote the paraphrases from them. The summary, the counts, the evidence level, the links and the quotes come from code; the summary counts the papers listed below. A later paper is listed under a label when Pith's citation classifier gives one of its sentences that label at probability 0.8 or more (0.95 for support), Pith's reading finds that the citing paper itself did it, and a second check confirms the paraphrase against the sentence. This section lists uses only. It does not list criticism or disagreement, so their absence here says nothing about how the paper has been received. Papers that share an author with this one are listed apart, under "By the same authors". A quote is the citing authors' own sentence, shown when their paper's licence is CC BY, CC BY-SA or CC0. JSON

Uses its data or values 1 of 4 papers with this label
Compares with it 1 of 12 papers with this label
How the literature uses it

17 arXiv papers cite this in 22 sentences

Each citing sentence is taken from the citing paper's LaTeX source and typed by Pith's classifier. A label appears at probability 0.8 or more (0.95 for support), a contest at 0.7 or more; the quoted words are the citing authors'. How each label was checked · JSON

Papers whose main result relates to this one's

Found among the papers closest to this one by title and main claim. Pith's classifier compared each pair's main results and a second model family confirmed the relation; checked against two further families, shown conflicts agreed 118 times in 120 and shown same results 218 times in 239. The quoted sentence is the other paper's main claim.

States a conflicting result 1 paper

Papers citing it · 16 sentences without a confident label

Compares with it 1 paper
Cites it as background 5 papers

Discussion

Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.