Arepo

AREPO is a computational astrophysics code used to simulate gas dynamics and gravity in cosmological and galactic contexts. It is built around a moving-mesh finite-volume method that uses a dynamic Voronoi tessellation to define its mesh. The mesh-generating points move with the flow, causing the mesh to adaptively follow gas motions and thereby reduce advection errors and improve resolution of shocks and contact discontinuities compared with fixed-grid methods.

In AREPO, hydrodynamics are solved with a second-order Godunov scheme on the moving cells, typically employing a Riemann solver to compute fluxes across cell interfaces. Gravity is treated with a TreePM approach, combining a Barnes–Hut tree with a particle-mesh solver to efficiently handle long-range forces in cosmological simulations. The code supports cosmological expansion, cooling and heating, and various subgrid physics models for star formation, stellar feedback, and black hole feedback. It also includes magnetohydrodynamics with divergence-cleaning to maintain a solenoidal magnetic field.

Development and impact: AREPO was developed by Volker Springel and collaborators at the Max Planck Institute for Astronomy and has been widely used in major simulation campaigns, most notably the Illustris project and its successor IllustrisTNG, which explored galaxy formation in a cosmological context using the moving-mmesh approach. The code has influenced subsequent moving-mesh and hybrid methods and remains a prominent tool for high-resolution astrophysical simulations.