A new review argues that modern numerical relativity can push cosmology past its usual shortcuts and test ideas about what, if anything, came before the Big Bang.

The approach is simple to state and hard to execute: solve Einstein’s equations on a computer where pencil and paper symmetry tricks fail.

“This review is an up to date account of the use of numerical relativity to study dynamical, strong gravity environments in a cosmological context and we suggest directions in which future work can be fruitfully pursued,” said the authors.

Understanding numerical relativity

Numerical relativity is basically the way scientists use computers to “do” Einstein’s math when it gets too complicated to solve with pencil and paper.

Einstein’s equations explain how space and time bend when there’s matter or energy around. But the equations are messy, and in most real situations – like when two black holes spiral toward each other – you can’t find a neat, exact solution.

Instead, physicists chop space and time into a grid, feed in the equations, and let supercomputers crunch through the steps to show how the system evolves moment by moment.

Running relativistic simulations

The authors map out how to run relativistic simulations that do not assume the universe is perfectly smooth in space, then show what those simulations have already taught us and where they can go next.

They separate the story into early universe beginnings, the hot Big Bang through recombination, and the late universe of large scale structure.

They also spell out the technical roadblocks that trip people up, from setting consistent initial data to choosing stable coordinate gauges as the code evolves in time.

The point is not just to advocate computation, it is to standardize methods so cosmologists and relativists can understand one another’s results.

Why numerical relativity matters

Fifty years of work turned numerical relativity from a niche craft into the engine behind precise gravitational waves templates, beginning with the first stable evolutions of binary black holes in 2005.

When detectors like LIGO finally picked up those waves in 2015, the signals matched the computer predictions almost perfectly.

That’s how we knew the detection was real and how we turn Einstein’s theory into something we can actually test.

Cosmology’s frontier problems share the same strong gravity DNA. Near singularities, or in violent phases of reheating and collapse, analytic symmetry assumptions break, so the only honest way forward is to evolve the full, non linear system on a grid.

From black holes to the Big Bang

Walk time backward and the standard equations hit a wall, densities and temperatures diverge while the usual homogeneous, isotropic picture loses validity.

Fully relativistic simulations let researchers ask how matter, curvature, and anisotropy actually interact when no simplifications are allowed.

That matters for questions such as whether a cosmological singularity is local and oscillatory, or whether new physics softens it into a bounce.

It also matters for identifying observables we can trust, since averages over space are not the same thing as what real observers on timelike or lightlike paths would measure.

Testing inflation honestly

Cosmic inflation explains why the cosmos looks nearly the same in all directions and why primordial fluctuations follow a simple pattern, and it remains consistent with the high precision constraints set by the Planck mission.

But most analytic studies of inflation start by assuming the very smoothness that inflation is meant to produce.

Relativistic simulations remove the need for oversimplified assumptions. They let you begin with large irregularities and complex motions, even allowing black holes to form.

From there, you can still explore whether inflation occurs at all, how long it lasts, and how much the outcome depends on the shape of the potential rather than just on symmetry.

Numerical relativity beyond our universe

The same toolkit reaches hypotheses that used to be out of reach. Hypothetical strings are one example, defects that could shed energy into a background of gravitational waves with a spectrum shaped by their network dynamics.

With full relativity, one can test how those waves depend on curvature, expansion, and string interactions without linearizing the metric.

Another is the multiverse, where bubbles of space with different vacuum energies collide and, in principle, paint subtle signatures on the cosmic microwave background.

Numerical relativity is already being used to compute realistic collision spacetimes and to translate them into curvature and temperature perturbations that observers could look for.

What to watch next

The review highlights several technical roadblocks that remain active research challenges.

These include finding better boundary conditions for expanding universes and developing gauges that stay stable when regions of collapse happen alongside overall expansion.

It also emphasizes the need for shared community tools that can generate consistent initial data for cosmological simulations, moving beyond the standard conformally flat setups.

While computing power is critical, progress won’t come from raw speed alone.

Smarter formulations, rigorous convergence tests, and geodesic-based diagnostics are key to making the next generation of cosmological simulations more precise and directly comparable across research groups.

That’s what transforms ideas about the universe’s first fractions of a second into testable science.

The study is published in Living Reviews in Relativity.

—–

Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates.

Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.

—–