Scientists Find a Smarter Way To Measure the Universe Using Exploding Stars

Researchers led by the University of Barcelona’s Institute of Cosmos Sciences developed CIGaRS, an AI framework for supernova cosmology published in Nature Astronomy.

CIGaRS combines imaging, supernova physics, host-galaxy effects, dust, and cosmic expansion in a single model, using simulation-based inference and neural networks.

The system can estimate key parameters such as galaxy redshift from images alone, reducing dependence on spectroscopy.

It could make Type Ia supernova distance measurements more precise and scalable for future surveys like the Vera C. Rubin Observatory, sharpening studies of dark energy.