Cosmology@Home

Cosmology@Home is a volunteer computing project that uses the BOINC platform to run cosmological simulations and model the universe's evolution, aiming to improve our understanding of dark matter, dark energy, and the fundamental parameters that shape our cosmos. The project is hosted by the University of Illinois at Urbana-Champaign and was launched in 2007.¹ It leverages the idle processing power of personal computers worldwide to perform computationally intensive calculations that would otherwise require supercomputing resources.

The primary scientific objective is to constrain cosmological parameters by comparing simulated universes with observational data from surveys such as the Sloan Digital Sky Survey and the Planck satellite. By exploring a wide range of parameter combinations, the project helps identify which models best match the observed large-scale structure of the universe, including the distribution of galaxies and the cosmic microwave background.

Cosmology@Home uses the BOINC (Berkeley Open Infrastructure for Network Computing) framework, which allows volunteers to download work units containing simulation parameters and run them on their own computers. Each work unit represents a specific set of cosmological parameters, such as the density of dark matter, the equation of state of dark energy, and the amplitude of primordial fluctuations. The software then runs a simulation of the universe's evolution, typically using N-body methods or semi-analytic models, and returns the results to the project's servers.

The project employs a Markov Chain Monte Carlo (MCMC) technique to efficiently sample the parameter space, focusing computational effort on regions that are most likely to match observations.² This approach reduces the total number of simulations needed while still providing robust constraints. Volunteers can also participate in the project's forum and contribute to the development of the software, making it a collaborative effort between scientists and the public.

Cosmology@Home has produced several notable scientific results. One of its key contributions is the development of the "CosmicFitting" method, which uses machine learning to accelerate the parameter estimation process.³ This method has been used to analyze data from the Planck satellite, providing independent constraints on cosmological parameters that are consistent with those from the official Planck analysis.

The project has also explored the possibility of non-standard cosmological models, such as those with a time-varying dark energy equation of state or modified gravity theories.⁴ These studies have helped rule out some exotic models and have strengthened the case for the standard ΛCDM model, though they also highlight the need for more precise data to distinguish between competing theories.

Beyond its primary mission, Cosmology@Home has been involved in educational outreach, providing a platform for students and enthusiasts to learn about cosmology and computational science. The project's website includes tutorials and resources that explain the science behind the simulations, making it a valuable educational tool.

One lesser-known aspect is the project's use of "emulators" — surrogate models that approximate the results of full simulations. These emulators are trained on a small set of simulations and can predict the outcome for new parameter sets in milliseconds, drastically speeding up the MCMC process.³ This technique is now widely used in cosmology, but Cosmology@Home was among the first to apply it in a volunteer computing context.

Another edge case is the project's handling of "degenerate" parameter sets, where different combinations of parameters produce nearly identical observable predictions. The MCMC algorithm must carefully navigate these degeneracies to avoid biased results, a challenge that the project has addressed through careful design of the likelihood function.²

Cosmology@Home continues to evolve, with plans to incorporate data from next-generation surveys such as the Dark Energy Spectroscopic Instrument (DESI) and the Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST).⁵ These surveys will provide unprecedented precision, requiring even more sophisticated simulations and analysis techniques.

The project also aims to expand its scope to include simulations of the epoch of reionization and the formation of the first galaxies, which are key to understanding the early universe.⁵ By engaging the public in these cutting-edge research areas, Cosmology@Home hopes to maintain its role as a bridge between advanced cosmology and citizen science.

BOINC

Berkeley Open Infrastructure for Network Computing, a platform for volunteer computing.

Dark energy

A mysterious force causing the accelerated expansion of the universe.

Dark matter

Non-luminous matter that exerts gravitational effects, inferred from observations.

ΛCDM model

The standard cosmological model with a cosmological constant (Λ) and cold dark matter (CDM).

MCMC

Markov Chain Monte Carlo, a statistical method for sampling parameter spaces.

N-body simulation

A simulation of the motion of particles under gravitational forces.