DESI DR1 versus DR2: BAO release dependence
Individual run · T-f48a3f2c5189
Public summary of an internal test, not yet independently validated or published.
Question investigated
Does replacing DESI DR2 BAO with DR1 change the preference for dynamical dark energy in DES-SN5YR and Union3, with Ωm = 0.33 ± 0.01 and BAO scale parameter = 30 ± 8?
Test conditions
Replace DESI DR2 BAO data with DR1 while retaining DES-SN5YR and Union3, the other assumptions and the comparison criteria. Repeat the fits after separately excluding the 3 groups at nominal redshifts z = (0.295, 0.51, 0.706); (0.93, 1.317, 1.491); (2.33). Adopt Ωm = 0.33 ± 0.01. Set the BAO scale-parameter prior to 30 ± 8 in this comparison of prior widths.
Result
The group exclusions produced unequal responses in the fit. The comparison established neither independence of the preference from every group nor exclusive dependence on any one group.
Physical interpretation
The comparison assesses sensitivity to the BAO data release and to the exclusion of tracer groups. By itself, it does not identify a specific systematic effect.