The analysis combined DESI DR2 baryon acoustic oscillations with three Type Ia supernova compilations and a compressed CMB likelihood across eight dark-energy parameterizations. Several models favored the w0 > -1, wa < 0 region and crossed the phantom divide near z about 0.5. However, the preference was driven mainly by the limited LRG1-2 tracers. Mirage-model Bayes factors were only inconclusive to moderate, and no model was robustly preferred by late-time data alone.
Key findings
- Models tended toward w0 > -1 and wa < 0, but significance was only about 1.1-2.3 sigma. Mirage received inconclusive-to-moderate support, and preference weakened without LRG1-2. No robust evidence for dynamical dark energy emerged.
Why this matters globally
The work identifies which datasets and tracers drive cosmological tension and helps target systematic checks in future DESI analyses. Persistence with larger datasets would have major implications for the standard cosmological model.
Thai researcher contribution
G. Mustafa of Walailak University participates in the theoretical and statistical analysis, linking a Thai institution to a frontier cosmology question. Individual author contributions are not detailed in the abstract.
Limitations to consider
Results are sensitive to LRG1-2 and parameterization choice, with few observables per tracer. Significance is far below discovery standards, compressed CMB information is incomplete, and multiple-model testing increases the chance of apparently notable fluctuations.