Scientists created one of the largest simulations of our universe ever – about the size of 500,000 HD movies – Image for illustrative purposes only (Image credits: Pixabay)
Astronomers have made public one of the most expansive datasets ever produced by cosmological simulations, totaling more than 2.3 petabytes of data.[1][2] This massive release from the FLAMINGO project offers a virtual recreation of the universe’s evolution, tracing the interplay of galaxies, dark matter, and other cosmic components over billions of years. The dataset, roughly equivalent in size to hundreds of thousands of high-definition movies, enables scientists to probe fundamental questions about the cosmos with unprecedented detail.[3]
The Sheer Scale of FLAMINGO’s Virtual Universe
FLAMINGO, short for Full-hydro Large-scale structure simulations with All-sky Mapping for the Interpretation of Next Generation Observations, emerged from the Virgo Consortium’s efforts to model the universe comprehensively.[3] Researchers ran these simulations on powerful supercomputers, including the Cosmology Machine at Durham University, to simulate the growth of cosmic structures from near the Big Bang to the present day. The project addressed gaps in earlier dark-matter-only models by incorporating full hydrodynamical physics, such as gas dynamics and feedback processes.
The flagship simulation, labeled L2p8_m9, spanned a cubic volume 2.8 gigaparsecs on each side – equivalent to about 9 billion light-years – and tracked more than 300 billion particles.[4] Baryonic particles represented gas and stars with masses around 1.07 times 10 to the ninth solar masses, while cold dark matter and neutrino particles filled out the ensemble. This setup allowed the model to capture the filamentary cosmic web, galaxy clusters, and the subtle effects of massive neutrinos on large scales.
Key Simulations and Their Parameters
The suite included hydrodynamical runs at three resolution levels: high (m8), intermediate (m9), and low (m10), calibrated to match observed galaxy stellar mass functions and cluster gas fractions at low redshifts.[3][2] Variations explored uncertainties in feedback mechanisms, cosmology, and even the nature of dark matter. For instance, some runs tested jet-based active galactic nucleus feedback or altered neutrino masses.
| Simulation | Box Size (cGpc) | Baryon/CDM Particles | Neutrino Particles | Baryon Mass (M⊙) |
|---|---|---|---|---|
| L2p8_m9 | 2.8 | 5040³ each | 2800³ | 1.07 × 10⁹ |
| L1_m8 | 1.0 | 3600³ each | 2000³ | 1.34 × 10⁸ |
| L1_m9 | 1.0 | 1800³ each | 1000³ | 1.07 × 10⁹ |
| L1_m10 | 1.0 | 900³ each | 500³ | 8.56 × 10⁹ |
This table highlights the primary fiducial runs, showcasing how larger volumes traded resolution for broader coverage of rare structures like massive clusters.[4] The simulations employed the Swift code with SPHENIX for smoothed particle hydrodynamics, ensuring accurate treatment of radiative cooling, star formation, supernovae, and black hole feedback.
A Treasure Trove of Data Products
The public release encompasses snapshots at 13 redshifts from z=5 to z=0, halo and galaxy catalogs generated by the HBT-HERONS finder, and merger trees.[1][2] Researchers can access HEALPix all-sky lightcone maps for observables like weak lensing, kinetic Sunyaev-Zeldovich effect, and X-ray emission, along with particle data from lightcones extending to high redshifts.
- Full and downsampled particle snapshots for detailed analysis.
- Power spectra for matter, gas, stars, and cross-correlations at 123 redshifts.
- Halo-matching catalogs to compare across simulation variants.
- Initial conditions and parameter files for reproducibility.
Due to the dataset’s immensity – exceeding 2.3 petabytes – a web service at the FLAMINGO site facilitates selective downloads via streaming, with Python tools mimicking standard HDF5 access.[2] This approach balances accessibility with storage constraints hosted by the DiRAC facility.
Advancing Cosmology Through Virtual Realities
FLAMINGO’s data empowers comparisons between theory and observations from upcoming surveys like Euclid and the Vera C. Rubin Observatory. By including baryonic effects and neutrino masses, the simulations refine measurements of structure growth and test alternatives to the standard Lambda cold dark matter model, such as decaying dark matter or evolving dark energy.[3] Calibration via machine learning ensured fidelity to real-world data, while variations quantify systematic uncertainties.
Earlier simulations fell short on small scales where gas physics dominates, but FLAMINGO’s volume and resolution bridge that divide. The project also produced gravity-only counterparts for direct contrasts with hydro runs.
As astronomers sift through this digital cosmos, FLAMINGO promises insights into dark matter’s distribution, galaxy formation biases, and the universe’s expansion history. The open release invites global collaboration, potentially reshaping our grasp of cosmic evolution in the years ahead.