Smithsonian Astrophysicists Model 1.5 Billion Light-Year Cosmic Volume in Lumina Simulation

Astrophysicists from the Center for Astrophysics | Harvard & Smithsonian and their colleagues have developed Lumina, a cosmological radiation-hydrodynamic simulation tracing the first 2 billion years of the universe down to redshift z=3 oup.comMonthly Notices of the Royal Astronomical Societyoup.comOpen source ↗. The work, published in Monthly Notices of the Royal Astronomical Society, reproduces cosmic evolution inside a comoving cubic volume with an edge length of 500 cMpc (approximately 1.5 billion light-years). Calculations were executed on the Frontier supercomputer at Oak Ridge National Laboratory using over 15,000 GPUs to simultaneously capture large-scale structure and discrete luminous sources.

Computational Architecture and Resolution

The simulation resolves cosmic evolution using the moving-mesh hydrodynamic code AREPO, combining the IllustrisTNG galaxy formation physics with a GPU-accelerated M1 radiative transfer solver across six frequency bins. Lumina contains 2 × 6000³ resolution elements, providing a baryonic mass resolution of 3.6 × 10⁶ solar masses and a dark matter mass resolution of 1.9 × 10⁷ solar masses. The initial conditions incorporate separate transfer functions for baryons and dark matter, accounting for their relative streaming velocity. This resolution allows the system to follow the coupled evolution of the intergalactic medium, galaxies, and active galactic nuclei.

Reionization Dynamics and Constraints

Lumina models the timeline and mechanics of cosmic reionization self-consistently across the full simulated volume. The simulation predicts late hydrogen reionization powered primarily by stars, with the median sub-volume becoming fully ionized at z ≈ 5.2 and residual neutral hydrogen patches remaining until z ≈ 4.75. Helium (He II) reionization is driven by active galactic nuclei and reaches near completion by z = 3, causing a distinct late-time thermal boost in the intergalactic medium. The resulting Thomson scattering optical depth matches data from the Planck observatory, while the thermal history and photoionization background align with empirical astrophysical constraints.