Source-linked AI summary
The EAGLE simulations of galaxy formation: public release of halo and galaxy catalogues
Stuart McAlpine, John C. Helly, Matthieu Schaller, James W. Trayford, Yan Qu, Michelle Furlong, Richard G. Bower, Robert A. Crain, Joop Schaye, Tom Theuns, Claudio Dalla Vecchia, Carlos S. Frenk, Ian G. McCarthy, Adrian Jenkins, Yetli Rosas-Guevara, Simon D. M. White, Maarten Baes, Peter Camps, Gerard Lemson
TL;DR
The paper addresses the need for accessible tools to explore complex galaxy-formation simulations and their evolution. It releases a queryable relational database built from EAGLE halo, galaxy, and merger-tree catalogues, containing properties for more than 10^5 galaxies. The database supports comparisons with observations and analysis of galaxy assembly histories, while catalogue properties below a stated resolution limit require caution.
Problem
Detailed galaxy-formation simulations contain complex, multiscale data whose scientific use benefits from accessible selection, linking, and evolutionary analysis tools.
Method
The paper constructs and publicly releases a SQL relational database containing EAGLE halo and galaxy catalogues, integrated properties, and merger-tree links across redshifts.
Results
The database contains integrated quantities and merger histories for more than 10^5 EAGLE galaxies, including stellar, gas, photometric, star-formation, metallicity, and image data.
Takeaways & Limitations
The public database supports comparison with observations and investigation of galaxy formation and assembly histories through queryable catalogue data.
Takeaways & Limitations
Many galaxy properties are unreliable below a stellar mass of 10^9 M⊙ in intermediate-resolution simulations, so resolution comparisons are needed.
Abstract
from arXiv · showhide
We present the public data release of halo and galaxy catalogues extracted from the EAGLE suite of cosmological hydrodynamical simulations of galaxy formation. These simulations were performed with an enhanced version of the GADGET code that includes a modified hydrodynamics solver, time-step limiter and subgrid treatments of baryonic physics, such as stellar mass loss, element-by-element radiative cooling, star formation and feedback from star formation and black hole accretion. The simulation suite includes runs performed in volumes ranging from 25 to 100 comoving megaparsecs per side, with numerical resolution chosen to marginally resolve the Jeans mass of the gas at the star formation threshold. The free parameters of the subgrid models for feedback are calibrated to the redshift z=0 galaxy stellar mass function, galaxy sizes and black hole mass - stellar mass relation. The simulations have been shown to match a wide range of observations for present-day and higher-redshift galaxies. The raw particle data have been used to link galaxies across redshifts by creating merger trees. The indexing of the tree produces a simple way to connect a galaxy at one redshift to its progenitors at higher redshift and to identify its descendants at lower redshift. In this paper we present a relational database which we are making available for general use. A large number of properties of haloes and galaxies and their merger trees are stored in the database, including stellar masses, star formation rates, metallicities, photometric measurements and mock gri images. Complex queries can be created to explore the evolution of more than 10^5 galaxies, examples of which are provided in appendix. (abridged)
1. Introduction
Galaxy formation requires detailed modelling across many physical scales, and EAGLE provides a validated simulation resource alongside a public database for exploring galaxy evolution. The database enables SQL queries over galaxy properties and merger histories without requiring users to handle the raw simulation data.
- Motivation: Galaxy formation couples dark matter clustering, gas cooling, outflows, star formation, and black hole formation across a wide range of scales.Hydrodynamical simulations are one of the two extensively developed approaches used to model this complexity.
- The EAGLE resource: EAGLE reproduces key observations including the present-day stellar mass function, black hole–stellar mass relation, and galaxy-size dependence on stellar mass.The simulations also match many other observed properties at present and earlier epochs.
- Database contribution: The paper introduces a public SQL relational database containing integrated galaxy quantities for more than 10^5 simulated galaxies.Stored quantities include stellar masses, star formation rates, metallicities, and luminosities.
- Database contribution: Merger-tree indexing allows individual galaxies to be followed through their evolution across cosmic time.The database is intended to support exploration of halo and galaxy catalogues from the main EAGLE simulations.
- Database contribution: Relational storage makes it faster to select samples by multiple properties, connect galaxies to haloes, and analyse smaller subsets without copying raw simulation data.Multiple indexing reduces data-transfer volumes and supports analysis with simple scripting languages.
2. The EAGLE simulation suite
EAGLE comprises cosmological hydrodynamical simulations spanning multiple volumes and resolutions, with subgrid physics calibrated against selected z = 0 galaxy observations. The public database organizes catalogues, aperture measurements, merger trees, and simulation variants for querying galaxy evolution.
- Simulation design: 25–100 cMpc cubic volumes track baryonic and dark-matter elements from z = 127 to the present using a modified GADGET-3 SPH code.The suite adopts a flat ΛCDM cosmology and includes full gravity and hydrodynamics.
- Subgrid model: Subgrid prescriptions model unresolved stellar mass loss, radiative cooling and heating, pressure support, star formation, black-hole growth, and AGN feedback.These processes are implemented through local hydrodynamic source and sink terms.
- Calibration and validation: Feedback parameters are calibrated against the z = 0 galaxy stellar mass function, galaxy sizes, and black-hole–stellar-mass relation, then assessed against additional observables.Reported validation targets include scaling relations, gas and metal distributions, luminosity functions, and higher-redshift properties.
- Simulation variants: The database contains multiple simulation variants whose box sizes, particle masses, softening lengths, and varying subgrid parameters determine available dynamic range and resolution.An AGNdT9 variant changes AGN heating and black-hole accretion viscosity, improving diffuse group gas while having only a small effect on galaxy properties.
- Halo and galaxy catalogues: FoF and SUBFIND identify haloes, self-bound subhaloes, and galaxies, while aperture-based measurements address diffuse intra-group or intra-cluster material.The catalogue also flags occasional spurious internal subhalo identifications.
- Merger trees: D-Trees links galaxies across outputs using the most-bound particles, and branch mass defines the main progenitor to reduce main-branch swapping during similar-mass mergers.Depth-first ordering and GalaxyID indexing support traversal of progenitors and descendants.
3. Use of the database
The database provides a web interface and SQL tables for querying EAGLE galaxy and halo properties, visualising results, and navigating merger trees. Its indexed identifiers and linked tables support efficient selection, cross-table analysis, and tracking galaxies across cosmic time.
- Database interface: The web interface supports SQL queries through browser and stream modes, with browser results enabling quick syntax checks and stream results returning downloadable CSV files.Browser queries return limited HTML-table results, whereas stream queries return all rows in a separate window.
- Galaxy merger-tree traversal: Merger trees use depth-first ordering and GalaxyID-based fields to identify main-branch progenitors, all progenitors, and descendants.TopLeafID delimits the main progenitor branch, LastProgID covers all progenitor branches, and DescendantID identifies the unique descendant.
- Content of the database: The database distributes each simulation’s information across five SQL tables linked through GalaxyID or GroupID.SubHalo stores galaxy properties; Aperture, Magnitudes, and Sizes add galaxy measurements, while FOF stores halo properties.
- Content of the database: The FOF and SubHalo tables include uniformly distributed random numbers in [0, 1), enabling unbiased galaxy or halo subsamples.These fields are intended for generating random subsets directly from the catalogues.
- Query examples: Example queries demonstrate plotting z = 0 relations, joining tables for colour-magnitude diagrams, and following individual galaxies through time.The examples use simulation, snapshot, aperture, and galaxy-property fields to retrieve and combine catalogue data.
4. Recommendations, caveats and credits
The authors outline practical limitations affecting database queries, simulated galaxy properties, images, and interpretation of EAGLE outputs. Users should account for query timeouts, finite resolution and volume, measurement definitions, and known simulation–observation discrepancies.
- Finite resolution: Galaxy properties below 10^9 M⊙ are generally unreliable in intermediate-resolution simulations because adequately sampling galaxy formation requires many particles.Reliability for a given quantity can be assessed by comparing simulations with different resolution.
- Finite volume: The main simulation volume is 10^-3 cGpc^3, so rare objects are unlikely to be represented and cluster-like analyses are limited.Only a handful of haloes above 10^14 M⊙ are present in the main simulation.
- Black hole quantities: BlackHoleMass sums all black holes assigned to a subhalo and, because outputs are coarsely sampled, database accretion rates cannot capture high temporal variability.BlackHoleMassAccretionRate should therefore be treated with great care.
- Stellar velocity dispersion and morphology: StellarVelocityDispersion measures stellar kinetic energy through σ = 2EK/3M, not the division between dispersion and rotation.It cannot distinguish rotationally supported spirals from dispersion-supported ellipticals.
- Galaxy images and magnitude tables: Database images include only particles within a subhalo, so satellites or merging partners may be absent, while images are redshifted to z = 0.1 and magnitudes are rest-frame.The differing image and magnitude conventions matter when comparing these products.
- Simulation caveats: The simulated stellar mass density is approximately 20% lower than inferred from observations despite calibration to the z = 0.1 stellar mass function.The authors relate this missing mass to a slight undershoot of the simulated stellar mass-function knee.
- Simulation caveats: Specific star formation-rate evolution follows observed trends but has a redshift-dependent normalisation lower by 0.3 - 0.5 dex.The authors note good agreement with a recent recalibration of star-formation indicators.
5. Conclusions
The paper introduces a public SQL database containing EAGLE halo and galaxy catalogues, integrated properties, and merger histories for more than 10^5 galaxies. It is intended to support comparisons with observations and investigations of galaxy formation physics, while users should account for finite simulation resolution.
- The public SQL database contains integrated quantities and merger histories for more than 10^5 galaxies from the EAGLE hydrodynamic simulations.
- The database includes galaxies from the largest EAGLE simulation and smaller volumes with varied resolution and AGN models.
- Galaxy records provide halo and galaxy properties, metal abundances, photometry, sizes, and mock gri images, while merger trees enable tracking assembly histories through time.
- The released data are intended for comparison with observations and for gaining physical insight into galaxy formation.
- Users should exercise caution because the simulations have finite resolution; additional products may later include dust-attenuated photometry, morphology, higher-time-resolution trees, and raw particles.
Appendix A. Examples of more complex queries
The appendix demonstrates database queries and Python workflows for selecting galaxy samples, joining galaxy and halo tables, and tracing progenitors. Examples reproduce published relations and compute derived quantities from indexed catalogue fields.
- Python and SQL examples query galaxy stellar masses and reproduce the EAGLE galaxy stellar mass function at z = 0.1.The example uses 30 pkpc apertures and compares three EAGLE simulations.
- The examples connect to the database with eagleSqlTools, iterate over simulations, bin stellar masses, and normalize counts by simulation volume and bin width.
- A black hole mass–stellar mass query joins SubHalo and Aperture tables through GalaxyID and selects z = 0.1 galaxies with positive masses within 30 pkpc.
- A galaxy-size query joins SubHalo, Aperture, and Sizes tables through GalaxyID to compare half-mass radius with stellar mass at z = 0.1.
- A halo–galaxy query joins FOF and SubHalo tables using GroupID and SnapNum to calculate the offset between galaxy and halo potential centres.
- A progenitor query selects a random subset of Milky Way-like galaxies and retrieves progenitor locations and specific star formation rates above a stellar-mass threshold.
Appendix B. Description of all fields contained in the database
The database documentation describes tables for galaxy, halo, size, aperture, and magnitude data, together with the identifiers and measurement conventions needed to interpret and join them.
- The main SubHalo table stores galaxy properties and navigation indices for side tables and merger trees, covering dark matter, gas, stars, and black holes.
- The SubHalo documentation continues across additional table pages, with the continuation caption providing no separate field description.
- The FOF halo table contains halo properties and links to SubHalo through the unique GroupID identifier.
- The Sizes table provides stellar half-mass sizes from spherical apertures and includes only galaxies with total stellar mass M∗ > 10^8 M⊙.
- The Aperture table stores measurements within spherical apertures centered on each galaxy’s minimum gravitational potential, with each row representing one galaxy and aperture size.
- The Magnitudes table contains absolute rest-frame magnitudes without dust attenuation for galaxies with M∗ > 10^8.5 M⊙.
Appendix C. List of snapshot output times
The database provides snapshot outputs spanning cosmic history, and its documentation identifies snapshot number as the convenient query index while also listing redshifts and lookback times.
- The snapshot-time table lists all output redshifts for simulations in the database, with snapshot numbers simplifying SQL queries and lookback times reported in Gigayears.
Appendix D. Detailed expressions for quantities in the database
Appendix D defines the equations, coordinate conventions, and database quantities used to describe the EAGLE simulations and their halo and galaxy properties.
- Equations and variables: The appendix states the fundamental equations solved by the simulations to clarify the meaning of database entries.These include the continuity, Euler, energy, and Poisson equations for a gravitating fluid.
- Equations and variables: The fluid equations use gas density, collisionless-matter density, effective pressure, gravitational potential, and cosmological constant as key variables.The thermal energy per unit mass includes the ratio of specific heats and mean molecular weight, while C(T)ρ represents radiative cooling and heating.
- Equations and variables: The equations are formulated using proper time, position, and velocity, with d/dt defined as the Lagrangian derivative and ∇ as the spatial derivative.The appendix distinguishes time derivatives at constant position from total derivatives following the flow.
- Coordinate system: The simulations solve the equations in expanding coordinates governed by the scale factor a(t), applying periodic boundary conditions.The comoving formulation simplifies integration in the expanding reference frame.
- Coordinate system: Database locations are recorded in comoving coordinates, while halo or galaxy velocities refer to peculiar velocities and other variables are generally physical.The appendix distinguishes comoving and physical distance units and defines centre-of-mass and centre-of-potential positions with periodic boundaries accounted for.
- Particle and object quantities: Star particles have current and birth masses because stellar mass loss transfers mass to gas, while black holes have particle and subgrid masses.The black-hole subgrid mass enters the accretion-rate equations, and the two masses converge closely once the black hole is well above its seed mass.