Research
Overview
My work spans cosmic reionization and high-redshift galaxy formation, self-interacting dark matter, the non-ideal magnetohydrodynamics of star and disk formation, and the numerical methods and GPU acceleration that make these large-scale simulations tractable on exascale hardware.
I lead the Lumina simulation project, a new large-volume radiation-hydrodynamic effort to model cosmic reionization and the assembly of galaxies at z > 3.
Learn moreI develop SIDM methods in AREPO and study how velocity-dependent self-interactions reshape halo cores, drive gravothermal collapse, and produce observable signatures.
Learn moreI solved long-standing mesh-noise problems in moving-mesh hydrodynamics, ported AREPO-RT to GPUs for exascale platforms, and co-developed the GADGET-4 code.
Learn moreNon-ideal MHD on a moving mesh, protostellar core collapse, and shearing-box studies of the magnetorotational and gravitational instabilities in accretion disks.
Learn moreLow-Prandtl-number convection in inclined fluid layers; analytical and pseudo-spectral studies of pattern formation in liquid mercury.
Learn moreLumina — Cosmic Reionization & High-z Galaxies
PICosmic reionization is the last major phase transition of the universe — the epoch when light from the first stars and galaxies ionized the intergalactic hydrogen and reshaped subsequent galaxy formation. My goal is to understand this transition self-consistently, modeling galaxies, the radiation they emit, and the intergalactic medium they live in within a single simulation framework.
As principal investigator of the Lumina project, I lead a new generation of cosmological radiation-hydrodynamic simulations targeting reionization, built on the moving-mesh code AREPO with the GPU-accelerated radiation-transport solver I developed during my time at MIT. Lumina evolves a 500 cMpc box from the cosmic dark ages down to z = 3 with nearly half a trillion resolution elements — large enough to capture both the galaxy populations that drive hydrogen reionization and the rare quasars responsible for He II reionization, while still resolving the sources themselves.
Four linked goals shape the project: tracking how galaxies form, grow, and respond to radiation feedback from cosmic dawn to z = 3; measuring when and where hydrogen reionization occurs and which galaxies drive it; modeling AGN-driven He II reionization and its thermal imprint on the intergalactic medium; and producing predictions for high-redshift galaxy surveys, the Lyman-α forest, 21-cm cosmology, line-intensity mapping, and quasar absorption spectroscopy. Lumina also adopts improved initial conditions with separate baryon and dark-matter transfer functions and their relative streaming velocity — physics that single-fluid initial conditions miss at high redshift.
The introduction paper has been accepted for publication in MNRAS, and the first companion studies — an analysis of CMB optical-depth fluctuations from patchy reionization and a census of AGN demographics out to z > 3 — are on arXiv (Zier et al. 2026; Smith, Zier et al. 2026; Shen, Zier et al. 2026).
A complementary analysis combining Lumina and THESAN shows that ionized-gas clumping follows an approximately universal relation, agreeing at the 10–20% level, when parameterized by the global ionized fraction. It also finds that a density-only clumping prescription overpredicts the instantaneous recombination rate by 29% at z ≈ 8 and 84% at z ≈ 5 (Sadain, Smith, Zier et al. 2026). Complementary THESAN-ZOOM work uses more than 35,000 simulated galaxies to connect ionizing-photon escape to bursty star formation and gas clearing, finding that UV-bright galaxies with MUV < −17 provide the dominant contribution to reionization (Summerfield et al. 2026).
Self-Interacting Dark Matter
Cold collisionless dark matter is exquisitely successful on cosmological scales but in tension with several observations on sub-galactic scales — the diversity of dwarf rotation curves, central densities of low-surface-brightness galaxies, and the abundance of strong-lensing substructure. Self-interacting dark matter (SIDM) is a minimal extension that redistributes kinetic energy in halo cores, producing concrete deviations from CDM predictions while remaining consistent with large-scale structure.
I led the development of a new Monte Carlo SIDM implementation in AREPO-2. A dedicated dark-matter neighbor-search tree decouples the scattering solver from gravity, while pairwise communication between MPI tasks tracks multiple scattering events within a timestep and conserves momentum and energy. The implementation supports velocity-dependent cross sections and inelastic interactions, and is substantially faster than the previous AREPO-1 solver except during the late stages of gravothermal collapse (Zier et al. 2026).
Prompt cusps form by the direct collapse of primordial overdensities in the early universe and reside at the center of every dark-matter halo, with profiles steeper than the NFW form. In recent work, we found that more prominent prompt cusps delay the early core-formation phase by a factor of about two. During most of the subsequent core-collapse phase, the halo evolution follows a common trajectory after rescaling density, radius, and velocity dispersion, with deviations of about 5% late in the collapse. The results highlight how both the inner concentration and the outer-halo temperature regulate heat transport (Tran, Gilman, Delos, Shen, Zier et al. 2026). In parallel we derived a compact analytic profile for the isothermal cores predicted by SIDM models, simple enough to fit observed halos but accurate where standard NFW-based fits fail (Tran et al. 2026).
AREPO & Numerical Methods
The science I care about is rate-limited by the codes available to do it. Reionization, SIDM, and small-scale MHD all demand simulations that are simultaneously high-resolution, large-volume, and physically rich — a combination that is only achievable on exascale hardware, with carefully designed numerical methods underneath.
For reionization specifically, the radiation-transport step is the bottleneck. I rewrote AREPO-RT to run on GPUs and developed a communication strategy that avoids per-step host–device synchronization, demonstrating efficient strong scaling on pre-exascale systems and opening the door to the next generation of radiation-hydrodynamic simulations (Zier et al. 2024).
Earlier I traced a long-standing source of spurious noise in moving-mesh shear flows to the first-order flux integration along cell interfaces and replaced it with higher-order Gauss–Legendre quadrature, eliminating the artifact; optimized vector kernels limit the overhead to ≈30% for ideal MHD (Zier & Springel 2022). I am also one of the co-developers of GADGET-4 — the open-source successor to GADGET-2 — where I contributed the SPH implementation (density- and pressure-based formulations with time-dependent artificial viscosity) and vectorized gravity and SPH kernels (Springel, Pakmor, Zier & Reinecke 2021).
Non-ideal MHD & Accretion Disks
In dense, weakly ionized gas the ideal-MHD approximation breaks down and three non-ideal effects — Ohmic resistivity, ambipolar diffusion, and the Hall effect — control how magnetic fields couple to the gas. Together they regulate the magnetic-flux problem of star formation, outflow launching, and disk fragmentation. I implemented all three effects on AREPO's moving mesh for the first time, using a least-squares gradient reconstruction on cell interfaces (Zier, Springel & Mayer 2024), and showed that the Hall solver can be stabilized at high resolution by imposing a minimum artificial Ohmic resistivity of one-fifth the Hall diffusion coefficient, yielding second-order convergence in shock tests (Zier, Mayer & Springel 2024).
Building on this framework, we performed the first moving-mesh simulations of cloud-core collapse to protostars that resolve all three non-ideal MHD effects simultaneously, studying magnetic-flux redistribution and outflow launching in a realistic protostellar environment (Mayer, Zier et al. 2025).
The same moving-mesh infrastructure powers my work on accretion disks — the workhorses through which gas accumulates onto protostars and black holes, with their angular-momentum transport set by disk-scale instabilities. I implemented the shearing-box approximation in AREPO — Coriolis and tidal forces, shearing-periodic boundaries, second-order convergence — and used it for convergence studies of the magnetorotational instability across magnetic-field configurations, comparing the moving-mesh results to established static-grid codes and characterizing the numerical Prandtl-number behavior (Zier & Springel 2022).
For self-gravitating disks I added a TreePM solver for the shearing box and ran an extensive study of fragmentation and gravito-turbulence across cooling strengths, box sizes, and resolutions, mapping out where cool disks settle into a quasi-steady turbulent state versus fragmenting into bound clumps (Zier & Springel 2023).
Earlier work
Before astrophysics my background was in non-equilibrium fluid dynamics. In my bachelor's thesis I analyzed the stability of an inclined layer of liquid mercury to convection, using Galerkin methods and pseudo-spectral simulations to characterize linear instability thresholds and the nonlinear pattern formation that follows in the low-Prandtl-number regime (Zier, Zimmermann & Pesch 2019).