An international team of researchers has unveiled an AI-powered framework that could dramatically improve how scientists measure the expansion of the universe, extracting up to four times more information from supernova observations without the need for costly spectroscopic data.
A Unified Model for the Universe
The method, called CIGaRS (Combined Inference and Galaxy-Related Standardisation), was published May 6 in Nature Astronomy. Developed by Konstantin Karchev and Roberto Trotta of the International School for Advanced Studies (SISSA) in Trieste, together with Raúl Jiménez of the University of Barcelona’s Institute of Cosmos Sciences (ICCUB), the framework uses neural networks to analyze Type Ia supernovae — stellar explosions used as “standard candles” for gauging cosmic distances.
Rather than correcting for environmental effects on supernova brightness through piecemeal adjustments, CIGaRS models everything simultaneously: galaxy evolution, interstellar dust, the rate at which supernovae appear over cosmic time, and the observable properties of the explosions themselves. This holistic approach allows the system to disentangle intrinsic supernova properties from environmental distortions using photometric (imaging) data alone.
Preparing for Millions of Supernovae
The framework arrives at a critical moment. The Vera C. Rubin Observatory in Chile will soon begin a decade-long sky survey expected to detect millions of supernova candidates, roughly 99 percent of which will be observed only through images rather than spectroscopy.
“Unlike other frameworks, which require analytic simplifications, our no-compromise end-to-end simulation-based inference approach is uniquely capable of extracting the full cosmological and astrophysical information from the Rubin Observatory’s hard-earned data, while avoiding the pitfalls of selection and modelling biases,” said Karchev, the study’s lead author.
The researchers found that CIGaRS can improve cosmological constraints by up to a factor of four compared to traditional methods that rely only on a small, spectroscopically observed subset of supernovae. Its precision in estimating galaxy distances from images alone rivals that of spectroscopic measurements.
A Wider Quest to Understand Dark Energy
The work feeds into a broader effort to resolve fundamental questions about the cosmos. Scientists remain divided over the precise rate of cosmic expansion — a disagreement known as the Hubble tension — and recent data from the Dark Energy Survey and the Dark Energy Spectroscopic Instrument have added new complexity to debates over whether dark energy itself may be evolving over time.
“A powerful way of modeling the Universe is to simulate it ab initio in the computer using Bayesian inference,” said Jiménez. “This provides a way to vary all possible parameters at the same time to predict what Universe we live in”.
