The bin that buried three billion years of host age
The claim under dispute is whether the age of a supernova's host galaxy biases how bright the supernova looks, in a way that -- once corrected for -- points to a universe whose expansion is decelerating and whose dark energy changes over time. Such a result would depart from the standard picture of a constant dark-energy density and eternal acceleration. A critique published earlier this year argued the bias was small: combining two supernova samples across redshifts 0.06 to 0.42, after a mass-based dust correction, the age-luminosity slope came out at a shallow -0.007 magnitudes per billion years. Over that redshift range, host galaxies span roughly three billion years of average age -- a fact that turns out to matter more than the headline slope let on.[1]
Chung and colleagues at Yonsei University, whose earlier work this critique targeted, reran the comparison in the narrower 0.06-0.20 redshift band that the field has conventionally used to isolate exactly this effect. The slope came back at -0.034 plus or minus 0.012 magnitudes per billion years -- close to five times steeper than the wide-bin figure. The different number comes from comparing host galaxies of similar age with each other and removing the averaging across galaxies three billion years apart.[1]
Two more levers under separate tests
Two other choices also moved the result. The wide-bin analysis applied a mass-based correction for interstellar dust that assigned unusually low extinction values -- 1.5 to 2.1 -- to massive host galaxies, against the roughly 3.0 to 3.1 seen in directly observed samples. Dropping that correction alone, independent of the bin question, moved the slope to -0.022. Separately, the inferred correction depends on how long a white dwarf waits after its star forms before it detonates; swapping in a different, still-plausible timing model roughly doubled it again.[1]
Reporting each of these levers on its own gives this check a useful sensitivity curve. A result that only survives under one specific bundle of choices is fragile; here the result moves in the same direction as several choices are varied independently -- bin width, dust correction, detonation timing. On that narrow point, the response earns its conclusion.[1]
The independent test
The response supplies a methodological self-audit. The team whose earlier decelerating-universe claim was under attack re-examines the critique and finds the specific step that weakened it. Naming a concrete mechanism -- three billion years of host age averaged into one bin -- gives the answer real analytical weight. Independent confirmation requires a second lab to reach the number from another sample.[1]
The decisive test applies the same narrow-versus-wide comparison to a supernova sample assembled by neither team. The evidence currently supports a narrower conclusion: one analysis choice mattered far more than a headline number implied, and its mechanism is described clearly enough for an independent check. The universe's acceleration remains the larger unresolved claim.[1]