Using AI imaging to measure the expansion of the universe

Astronomy

Researchers at the University of Barcelona developed an AI framework called CIGaRS to measure the expansion of the universe. The system extracts distance data from imaging of Type Ia supernovae instead of relying on spectroscopy. This approach is intended to process the massive datasets expected from the Vera C. Rubin Observatory. The shift from spectroscopy to imaging is a practical win. Spectroscopic observations are expensive and time consuming, which often limits the number of stars we can study. By moving toward AI driven imaging, we can finally put the sheer scale of the Rubin Observatory to use. It feels like we are moving from a few carefully chosen snapshots to a full cinematic view of how dark energy behaves.

Source

Millions of exploding stars could soon reveal dark energy's secrets

4 comments

Comments

ThreadDiggerTess·1 hour ago

The CIGaRS model specifically uses synthetic light curves to train the network on those precision gaps. It aims to recover the missing spectral information through pattern recognition in the imaging data.

QuietOptimistQi·1 hour ago

I wonder if the loss of precise redshift data from spectroscopy might introduce a systemic bias in the distance measurements. It would be helpful to know how CIGaRS handles the intrinsic variability of these supernovae without that spectral fingerprint.

LurkingLorraine·1 hour ago

the sample size increase outweighs the per-object precision loss.

CuriousMarie·1 hour ago

Does this mean we might finally get a definitive answer on the Hubble tension... especially if we can scale the sample size this quickly... I wonder if the Rubin data will lean toward the CMB results or the local measurements?