An international team led by researchers at Australia's University of Queensland has assembled 2,884 likely Type Ia supernovae to test whether dark energy changes over time. The work received fresh coverage on October 22, 2026, while the underlying papers appeared on arXiv in September. The results suggest a possible departure from constant dark energy, without establishing a discovery.
The compilation, called Unite, combines Pantheon+ observations with the Dark Energy Survey's five-year sample. These stellar explosions help astronomers measure distances and reconstruct cosmic expansion. Ryan Camilleri and colleagues applied consistent modelling, selection and bias corrections so measurements collected by different surveys could be compared within one framework.
The authors report tension when supernova and galaxy-distribution measurements are combined with relic radiation from the early Universe under a model with an unchanging dark-energy equation of state. Allowing that property to evolve reduces the disagreement. The reported preference reaches about 3.1 to 3.3 standard deviations in frequentist tests, but Bayesian analysis finds only weak support for evolving dark energy.
A companion paper led by Jaemyoung Lee examines a crucial source of uncertainty: the masses of galaxies hosting the explosions. The team remeasured stellar masses for more than 98% of the sample consistently. Because corrected supernova brightness is linked to host-galaxy mass, differences in these estimates can shift cosmological conclusions, making the treatment of systematic effects essential.
The arXiv records describe journal submissions rather than confirmed peer-reviewed publications. The authors plan to release the analysis products after acceptance. Further observations, including nearby supernovae from the DEBASS programme, should extend the dataset. Whether dark energy truly evolves remains open; the findings sharpen a test of the standard cosmological model.
Comments