LEGA-C Spectroscopic Survey

• LEGA-C Spectroscopic Survey is a deep, high signal-to-noise spectroscopic program on the VLT that targets a K-band–selected, mass-complete sample of intermediate-redshift galaxies.
• The survey employs multi-object spectroscopy with ~20 hr exposures using VIMOS to accurately determine stellar ages, metallicities, element abundances, star formation histories, and kinematics.
• Its extensive dataset and robust analysis pipelines provide definitive benchmarks for understanding galaxy assembly, quenching evolution, and chemo-dynamical scaling relations at z ~0.7–1.0.

The Large Early Galaxy Astrophysics Census (LEGA-C) Spectroscopic Survey is a high-signal-to-noise, deep multi-year spectroscopic program on the Very Large Telescope (VLT) using VIMOS, designed to recover the stellar population and dynamical properties of massive galaxies at intermediate redshift ($0.6 < z < 1.0$). The survey uniquely targets a $K$-band–selected, mass-complete sample, enabling simultaneous determination of stellar ages, metallicities, element abundance patterns, star formation histories (SFHs), and stellar kinematics for thousands of galaxies during an epoch when the Universe was roughly half its current age. LEGA-C’s high data quality, broad ancillary coverage, and robust physical parameter pipeline collectively provide a definitive benchmark for the physics of galaxy assembly at $z\sim0.7$–$1$.

1. Survey Design, Motivation, and Sample Construction

LEGA-C comprises $\sim4200$ primary galaxies in the COSMOS field observed with VIMOS/VLT, drawn from the UltraVISTA $K$-band catalog with a redshift and $K$ magnitude cut ensuring a uniform stellar mass limit approximately $M_\ast\gtrsim10^{10}\,M_\odot$ across $0.6 < z < 1.0$ and including both quiescent and star-forming systems in the appropriate volume fractions (Wel et al., 2016, Straatman et al., 2018, Gallazzi et al., 8 Dec 2025).

The 1.6 deg$^2$ COSMOS footprint, with extensive multiwavelength photometry and HST/ACS imaging, underpins the structural measurement pipeline, dynamical modeling, and per-galaxy environmental metrics. Each target received $\sim20$ hr integration at $R\sim2500$–3500, delivering S/N $\sim$20\,\AA$^{-1}$ (continuum), sufficient to resolve both age and metallicity-sensitive indices and kinematical features for galaxies individually, rather than in stacks.

Sample selection relies primarily on $K$-band photometry for mass representativeness. All colors, morphologies, and AGN types are included, but analyses targeting quiescent galaxies typically apply additional selection using rest-frame UVJ diagrams, emission-line EW cuts, and morphological flags to minimize contamination (2407.12704, Bevacqua et al., 2023).

Parameter	Value
Redshift Range	0.6 < z < 1.0
K-band Limit	$K<20.7-7.5\log((1+z)/1.8)$
Mass-Completeness	$M_\ast\gtrsim10^{10}-10^{10.4}M_\odot$
S/N (continuum)	15–30 $\rm{\AA}^{-1}$ (DR3; median $\sim$20)
Sample Size	Primary $\sim$3200–4000; science-ready $\sim$2000–4000

2. Observational Strategy, Data Reduction, and Measurement Pipeline

VIMOS was operated in multi-object mode (HR_red grating, $R\sim2500$) with 1″-wide, N–S oriented slits and a typical wavelength range of 6300–8800 Å (rest frame $\sim$3600–5200 Å at $z\sim0.7$), covering classical age- and metallicity-sensitive features (e.g., $D_n4000$, H$\delta$, H$\gamma$, H$\beta$, Fe5270, Mgb) (Straatman et al., 2018).

Data reduction combines the ESO pipeline with a custom LEGA-C pipeline: precise slit tracing, optimal extraction, Gaussian spatial modeling for sky/object decomposition, per-OB and global telluric correction, and broadband flux calibration via UltraVISTA SED matching. Repeat exposures and overlapping slits are used to empirically correct errors and noise estimates (Straatman et al., 2018).

Key measurement products per galaxy include:

• 1D and 2D sky-subtracted, flux- and wavelength-calibrated spectra
• Stellar velocity dispersions (spatially integrated and resolved)
• Individual Lick/IDS absorption-line indices with uncertainty scaling
• Emission line fluxes/EWs for key nebular lines
• HST/ACS-derived structural parameters (Sérsic index $n$, semi-major/effective radii, axis ratio)
• Quality flags for spectrum, morphology, extraction, and pPXF fit fidelity (Gallazzi et al., 8 Dec 2025, Straatman et al., 2018)

3. Stellar Population Analysis: Ages, Metallicity, and Chemical Abundances

Robust stellar population analysis is central to LEGA-C. Full-spectrum fitting (e.g., pPXF, BAGPIPES, BaStA) and classical Lick index inference are both used—typically employing high-resolution templates (E-MILES, CB19, BC03, FSPS) coupled to Bayesian modeling.

For DR3, absorption indices (e.g., H$\beta$, H$\gamma_F$, H$\delta_F$, Fe5270, Fe5335, Mgb, [MgFe]$^{\prime}$, Mg$\langle$Fe$\rangle$) are measured after emission-line subtraction (Gallazzi et al., 8 Dec 2025).

Light-weighted and mass-weighted mean ages ($\langle t_L\rangle$, $\langle t_M\rangle$) and metallicities ($\langle Z_L\rangle$, [M/H]) are derived by fitting extended index+photometry vectors to large stochastic model libraries that incorporate variable SFHs, metallicity enrichment (e.g., leaky-box chemical models), and dust attenuation (Gallazzi et al., 8 Dec 2025). Typical uncertainties from BaStA are $\sim0.15$ dex (age), $\sim$0.25 dex (Z).

Alpha-element abundances [α/Fe] are inferred using Mgb and Fe4383 indices, compared to SSP grid predictions, with Monte Carlo error propagation. For $\sim$180 massive quiescent galaxies at $z\approx0.7$, the mean is $\langle$[α/Fe]$\rangle=+0.24\pm0.01$ dex; 91% are α-enhanced, and there is no evolution in the average [α/Fe] from $z\sim0.75$ to $z=0$ for $M_\ast\gtrsim10^{11}M_\odot$ (Bevacqua et al., 2023).

4. Scaling Relations and the Mass–Metallicity Relation

LEGA-C enables direct assessment of chemo-archaeological scaling relations at $z\sim0.7$. Using absorption indices and Bayesian libraries, Gallazzi et al. (Gallazzi et al., 8 Dec 2025) demonstrate the following key relations:

• Ages: Bimodal distribution (young, $\sim$1–2 Gyr, and old, $\sim$4–6 Gyr) crossing at $M_\ast\sim10^{11}\,M_\odot$.
• Metallicities: [Z/H]–mass relation is steep below $M_\ast\approx10^{10.8}\,M_\odot$ and flattens at higher mass (saturating at solar/super-solar Z).

A mass–metallicity relation (MZR) for quiescent galaxies is quantified over $10.4\leq\log(M_\ast/M_\odot)<11.7$ using full-spectrum mass-weighted M/H. The MZR is characterized by a lower exclusion boundary, the "Metallicity–Mass Exclusion (MEME) relation,” whereby

$$[\mathrm{M/H}]_\mathrm{min}(M_\ast) = -0.16 + 0.66\times[\log_{10}(M_\ast/M_\odot)-11.0]$$

No massive ($\log M_\ast\gtrsim11$) quiescent galaxies are metal-poor; low-mass quiescent galaxies exhibit large scatter, fully spanning the metallicity range seen in more massive objects. The distribution is also observed using the model-independent index MgFe.

Among additional archeological scaling relations, metallicity shows a sharper transition with velocity dispersion ($\sigma$) than mass, while ages at fixed $\sigma$ display clearer bimodality, suggesting tight coupling between kinematics and star formation quenching (Gallazzi et al., 8 Dec 2025).

5. Kinematic Analysis, Dynamical Modeling, and Environmental Dependence

LEGA-C provides the first large, spatially-resolved sample of stellar kinematic maps for massive galaxies at $z\sim0.8$ (Houdt et al., 2021, Straatman et al., 2018). Using axisymmetric Jeans modeling (JAM, MGE), the dynamical mass, dark matter fraction, anisotropy, and the degree of rotational support ($V/\sigma$) are quantified for $\sim$800 galaxies.

Key findings:

• Almost all star-forming and $\sim$50% of quiescent systems are rotation-dominated ($V/\sigma\sim1$–2) (Houdt et al., 2021).
• The most massive quiescent galaxies are slow rotators with little angular momentum ($|V_5|/\sigma_0<0.2$ for $M_\ast>2\times10^{11}M_\odot$), while lower-mass quiescents show significant rotation (Bezanson et al., 2018).
• Rotational support at $z\sim0.8$ is almost twice as high as in local quiescent galaxies at fixed $\sigma_0$ (Bezanson et al., 2018).
• There is no universal environmental trend in rotational support at fixed mass or $\sigma$ (Cole et al., 2020), except that ultra-massive slow rotators are found predominantly in the densest environments, consistent with hierarchical assembly via minor mergers at early times.

1D dynamics show that stellar and (when available) gas velocity dispersions agree on average (mean log ratio $-0.003$ dex), but individual galaxies show substantial scatter, resulting in $\sim$0.24 dex additional uncertainty in dynamical mass when using $\sigma_\mathrm{gas}$ as a proxy (Bezanson et al., 2018).

Calibrated virial mass estimators, empirically tied to JAM dynamical masses, are established:

$$M_\mathrm{vir} = [8.87-0.831\,n+0.0241\,n^2][0.87+0.38\,e^{-3.78(1-q)}]^2\,\sigma'^2_\star\,R_\mathrm{sma}/G$$

where $n$ is Sérsic index, $q$ is axis ratio, $R_\mathrm{sma}$ is the major-axis effective radius, and $\sigma'_\star$ is the integrated stellar velocity dispersion (Wel et al., 2022).

6. Star Formation Histories, Quenching Evolution, and Physical Pathways

LEGA-C reveals diversity and memory in star-formation histories at $z\sim1$ (Chauke et al., 2018). Using non-parametric full-spectrum fitting, quiescent galaxies predominantly formed >80% of their mass at $z>2$, while most star-forming galaxies peaked in SFR at $z<1.5$. Quiescent galaxies can show signatures of rejuvenation (e.g. $\sim$10% late-time starburst).

Key aspects:

• Stellar mass-weighted age correlates tightly with $\sigma$; high-$\sigma$ galaxies formed earlier and more rapidly.
• Light-weighted ages show only a weak, shallow trend with size at fixed mass ($\Delta\,t\lesssim0.5$ Gyr), but strong Balmer absorption post-starburst galaxies tend to be compact, indicating multiple physical quenching channels (“progenitor bias” and rapid, merger-driven transformation) (Wu et al., 2018).
• At the epoch probed, downsizing—quenching of the most massive galaxies first—was already established, with both age and metallicity scaling with mass and $\sigma$ comparably to the local universe (Gallazzi et al., 8 Dec 2025).

7. Cosmological and Theoretical Applications

The precision and systematic control of LEGA-C’s stellar population parameters allow new cosmological probes. The cosmic-chronometer method applied to the full-spectrum-fitted stellar ages of passive galaxies yields $$H(z=0.8)=113.1\pm15.1\,(\mathrm{stat})^{+29.1}_{-11.3}(\mathrm{syst})$$ km s$^{-1}$ Mpc$^{-1}$, in good agreement with Planck $\Lambda$CDM and other OHD measurements (Jiao et al., 2022). The differential ages are robust to the modeling framework, as confirmed by independent Lick-index-based methods.

Comparisons with state-of-the-art hydrodynamical (TNG50) and semi-analytical (GAEA) simulations show reasonable reproduction of the observed MZR envelope and massive galaxy metallicity ceiling, though with details depending on model parameterizations (2407.12704).

8. Legacy Value and Future Prospects

LEGA-C, via public data releases (e.g. DR3), provides the community with an SDSS-like, high-S/N, mass-complete spectroscopic database at $z\sim0.7$–1.0. Its unique combination of sample size, depth, broad value-added catalogs, and ancillary data has already yielded benchmarks for galaxy chemical and dynamical evolution, direct scaling-relations at lookback times $\sim6$–8 Gyr, and critical tests for hierarchical and chemo-dynamical galaxy formation models.

Continued analysis, including expanded catalogs of revised absorption indices and stellar population parameters, will further advance our empirical understanding of quenching pathways, kinematic transformation, environmental assembly, and the interplay between physical processes during the epoch of peak cosmic star formation (Gallazzi et al., 8 Dec 2025).