GALAH Survey: Chemical Tagging of the Milky Way

Updated 30 August 2025

GALAH Survey is a large-scale, high-resolution spectroscopic program targeting about one million stars to chemically tag dispersed stellar groups.
It utilizes the HERMES spectrograph on the Anglo-Australian Telescope to derive precise multi-element abundance patterns and stellar parameters.
The survey’s findings reveal detailed chemical gradients and kinematic signatures that illuminate the Milky Way’s formation and evolutionary history.

The GALAH Survey (Galactic Archaeology with HERMES) is a comprehensive, high-resolution spectroscopic program targeting approximately one million stars across the Milky Way to reconstruct its assembly and evolutionary history. Utilizing the HERMES spectrograph on the Anglo-Australian Telescope, the survey’s central goal is to perform “chemical tagging”—identifying ancient star formation remnants by their detailed, multi-element abundance patterns, even after these groups have become spatially and kinematically dispersed throughout the Galaxy. The survey’s scope and methodology position it as a keystone project for Galactic archaeology, uniquely enabled by large-N statistics and high-dimensional chemical information.

1. Scientific Objectives and Rationale

The primary objective of the GALAH survey is to unravel the formation and evolutionary history of the Milky Way by identifying and characterizing the fossil remnants of ancient star-forming events (Silva et al., 2015). By reconstructing the “chemical DNA” preserved in stellar atmospheres, the project addresses fundamental questions regarding the assembly of the thin and thick disks, halo, and bulge—such as the roles of dissipative collapse, satellite accretion, and internal processes like radial migration in Galactic evolution.

A central theme is the use of chemical tagging: the identification of dispersed groups of stars that originated in chemically homogeneous clusters, now unidentifiable by spatial, photometric, or kinematic means due to the effects of dynamical heating and radial mixing. Precise multi-element abundance patterns serve as unique formation-site signatures, enabling the reconstruction of the Galaxy’s dissipative and accretive processes far beyond the capabilities of traditional phase-space analysis.

2. Survey Methodology and Target Selection

The survey relies on high-resolution ( $R \sim 28,000$ ) optical spectroscopy of a magnitude-limited sample (typically $12 < V < 14$), with careful handling of excursions beyond this range (Silva et al., 2015, Martell et al., 2016). The target catalogue is compiled using photometric data from 2MASS and APASS, with the Johnson $V$ magnitude for selection synthesized from NIR photometry using the empirical relation:

$V(J, K) = K + 2(J - K + 0.14) + 0.382\exp{\left( \frac{J-K-0.2}{0.5} \right)}$

Selection criteria ensure color-unbiased sampling and high completeness with respect to FGK dwarfs and giants, which provide the optimal combination of spectral features for abundance analysis (Martell et al., 2016). Distance coverage extends to a few kpc for dwarfs and up to $\sim$ 10 kpc for giants, enabling a volume-complete census of the local thin and thick disk, with additional but smaller samples from the bulge and halo.

Observing strategy involves 392 objects per field (across a $2^{\circ}$ diameter) with typical integration of three consecutive 20-minute exposures, optimized for a minimum $S/N \sim 100$ per resolution element, a key threshold for accurate abundance work (Silva et al., 2015, Martell et al., 2016).

3. Instrumentation: The HERMES Spectrograph

HERMES is a state-of-the-art multi-object, fiber-fed echelle spectrograph designed for the survey’s requirements (Silva et al., 2015). Its principal characteristics include:

Four non-contiguous wavelength channels, covering $\sim$ 4713–7887 Å (blue, green, red, and IR arms), with a total coverage of $\sim1000$ Å per observation.
Nominal resolving power $R \sim28,000$ (optionally $R \sim42,000$ at a significant photon throughput tradeoff).
Fiber positioning via the 2dF robot, deploying 400 fibers with 392 dedicated to science targets per configuration.
Custom optics: dichroics, VPH gratings, and tailored CCDs for each spectral arm.

This configuration enables the detection and high-fidelity measurement of up to 29 chemical elements per star, spanning light proton-capture, $\alpha$ -process, odd-Z, iron-peak, and both slow and rapid neutron-capture elements.

4. Data Processing and Chemical Abundance Analysis

The GALAH data pipeline (GAP) integrates robust automation with custom algorithms to handle calibration, extraction, and quality control for the survey’s massive multiplexed datasets (Kos et al., 2016).

Key features:

Parallelized data reduction, with median bias subtraction, flat-fielding with “flap flats,” cosmic ray removal (LaCosmic variant), and optimized fiber trace/PSF correction using Chebyshev polynomials.
Extraction includes scattered light subtraction, fiber cross-talk correction (using “phantom aperture” calibration), and 2D wavelength calibration leveraging ThXe arc lamps.
Sky subtraction employs 25 dedicated sky fibers and spatial modeling, while telluric absorption is mitigated via the Molecfit package.
Propagation of Poisson uncertainties yields full error spectra for downstream parameter estimation.

Atmospheric parameters ( $T_\mathrm{eff}$ , $\log g$ , [Fe/H]) and multi-element abundances are determined through a hybrid multi-step approach (Buder et al., 2018):

A subset of stars is modeled with the physics-driven SME (Spectroscopy Made Easy), using 1D MARCS or ATLAS9 atmospheres and a carefully curated linelist.
SME products, processed with non-LTE corrections for key elements (e.g. Li, O, Na, Mg, Al, Si, Fe), train The Cannon, a data-driven, quadratic-parameterized model that maps the observed spectra onto the full stellar sample for homogeneous label propagation.
Only high-fidelity diagnostic lines are considered, and validation with Gaia benchmark stars, cluster members, and asteroseismic calibrators is routinely performed.

5. Scientific Results and Key Findings

Results to date from GALAH have yielded highly precise abundance and kinematic information for several hundred thousand stars (Buder et al., 2018, Buder et al., 2020). Early outcomes include:

Demonstration of the resolving power and S/N, with derived atmospheric parameters for Sun-like stars at the level of $T_\mathrm{eff}\sim5750$ K, $\log g \sim 4.5$ , [Fe/H] uncertainties $\lesssim \pm 0.05$ dex.
Measurement of precise radial velocities (typical accuracy better than 0.1 km/s) after correction for convective blueshift and gravitational redshift, benchmarked against synthetic 3D convective models and Gaia DR2 (Zwitter et al., 2018, Zwitter et al., 2020).
Recovery of vertical metallicity and $\alpha$ -element gradients, with the thin disk showing $d[\mathrm{M}/\mathrm{H}]/dz = -0.18 \pm 0.01$ dex/kpc and the thick disk displaying much flatter gradients (Duong et al., 2018), in agreement with scenarios of radial migration and “born thick” early disk formation.
Chemodynamical sequencing (in combination with Gaia) showing that high- $\alpha$ stars are predominantly older and have lower angular momentum than the Sun, implying an origin in the inner disk and a continuous evolutionary sequence without the need to invoke distinct “high- $\alpha$ metal-rich” components (Buder et al., 2018).

6. Challenges, Limitations, and Future Prospects

Multiple technical and astrophysical challenges are integral to the GALAH project:

Precision chemical tagging demands abundance uncertainties $\ll0.1$ dex, placing stringent requirements on S/N, line selection, and non-LTE corrections.
Identification of clusters in high-dimensional (chemical + spatial) space is sensitive to sample completeness and biases, as well as to the mixing factor $f_\mathrm{mix}$ describing how cluster debris populates the survey volume (Bland-Hawthorn et al., 2015).
Degeneracies in spectral fitting among $T_\mathrm{eff}$ , [Fe/H], and continuum level are amplified in cool stars; robust photometric priors and complete line lists are necessary to mitigate systematic effects (Kos et al., 10 Jan 2025).
Cross-survey calibration (with projects like APOGEE, Gaia-ESO, RAVE) and integration of asteroseismic, astrometric, and photometric data streams are crucial for maximizing accuracy and for absolute abundance scale alignment (Martell et al., 2016).

Looking forward, the statistically homogeneous, multi-dimensional datasets from GALAH, when fully coupled with high-precision Gaia astrometry and complementary spectroscopic surveys, will enable reconstruction of thousands of dispersed star clusters, probe the internal kinematics and chemistry of the thin and thick disks, and illuminate substructures resulting from minor mergers and radial migration. Systematic error characterization and correction using cluster benchmarks and spline fitting methods, as explored in later data releases (Kos et al., 10 Jan 2025), further enhance the reliability and legacy value of the dataset.

7. Broader Impact and Integration in Galactic Archaeology

GALAH’s high-fidelity chemical tagging represents a major advance in near-field cosmology, enabling the deconstruction of the Milky Way into its foundational star formation units. Synergy with Gaia provides full 6D phase-space data for most targets—allowing true chrono-chemodynamical (age–chemistry–kinematics) analysis at the scale of the Galactic disk. The survey is positioned as both a standalone resource and a critical complement to other ongoing and future surveys, establishing a multidimensional “fossil record” that will inform theoretical models of galaxy formation, the lifecycle of star clusters, and the chemical evolution of the Milky Way far into the future (Martell, 2015).