Compare · The Dark Universe

wCDM, constant equation of state vs w0waCDM, evolving dark energy

← Back to wCDM, constant equation of state

Standard Cosmological Model· within family

	wCDM, constant equation of state Consensus	w0waCDM, evolving dark energy Consensus
Proposed	1997	2001 / 2003
Key figures	Michael Turner, Martin White, DESI Collaboration	Michel Chevallier, David Polarski, Eric Linder, DESI Collaboration
In one sentence	The simplest extension of plain ΛCDM, treats dark energy's equation of state as one free constant instead of fixing it at exactly -1. Current data say the answer is consistent with a constant.	Lets dark energy evolve in time instead of being a fixed constant. 2024 DESI BAO results give the strongest hint yet that this evolution may be real.
Predictions	If w is not exactly -1, dark energy density changes over cosmic time rather than staying constant, growing denser in the past if w is more negative than -1 and thinning out if w is less negative (formally ρ_de ∝ a^{-3(1+w)}) If w < -1 (phantom dark energy), dark energy density grows with cosmic expansion and the universe ends in a finite-time 'big rip' Distance moduli to high-z supernovae differ from ΛCDM by a few hundredths of a magnitude, within reach of LSST and Roman	Distance-redshift relation departs from ΛCDM by a measurable amount at z ≈ 0.5 to 1 The BAO acoustic feature traces out a distance-redshift curve that bends measurably away from the ΛCDM prediction, detectable by Stage-IV surveys (LSST, Euclid) mapping the feature across many redshifts Stage-IV surveys (LSST, Euclid, Roman) should crisply confirm or refute the trend within the next decade If (w0 > -1, wa < 0) is real, dark energy crosses the phantom barrier (w = -1) toward more negative w in the past, then trends less negative toward today
Where it breaks	Current data don't show meaningful preference for wCDM over plain ΛCDM. The interesting signal is in the evolution (w0waCDM), not in a constant offset wCDM is a parameterization, not a model. No fundamental physics predicts a fluid with constant w ≠ -1 Phantom dark energy (w < -1) violates the dominant energy condition, theoretically awkward but not ruled out by data alone	Efstathiou (2025) argues the signal is a low-redshift systematic in the SN Ia compilations, not real dark energy evolution Significance is sensitive to the choice of SN compilation; the headline 3.9σ uses DES-SN5YR, a single recent sample The CPL parameterization is just a simple linear approximation of how w changes near the present day (w0 + wa·(1-a)), not a fundamental theory. A real signal would still need a physical origin Posteriors can cross the phantom divide (w < -1), which is theoretically awkward for most quintessence models
Key unresolved problem	The no-signal problem: current data show no real preference for this constant-w version over plain ΛCDM, so there is nothing yet that would tell the two models apart.	The interpretation problem: if the DESI hint of changing dark energy is real, no underlying theory predicts that exact pattern, the equation of state w starting above -1 and dropping, without awkward fine-tuning.
Reader vote	No votes yet	No votes yet

wCDM, constant equation of state

1997 · Consensus

w0waCDM, evolving dark energy

2001 / 2003 · Consensus

Proposed

1997

2001 / 2003

Key figures

Michael Turner, Martin White, DESI Collaboration

Michel Chevallier, David Polarski, Eric Linder, DESI Collaboration

In one sentence

The simplest extension of plain ΛCDM, treats dark energy's equation of state as one free constant instead of fixing it at exactly -1. Current data say the answer is consistent with a constant.

Lets dark energy evolve in time instead of being a fixed constant. 2024 DESI BAO results give the strongest hint yet that this evolution may be real.

Predictions

If w is not exactly -1, dark energy density changes over cosmic time rather than staying constant, growing denser in the past if w is more negative than -1 and thinning out if w is less negative (formally ρ_de ∝ a^{-3(1+w)})
If w < -1 (phantom dark energy), dark energy density grows with cosmic expansion and the universe ends in a finite-time 'big rip'
Distance moduli to high-z supernovae differ from ΛCDM by a few hundredths of a magnitude, within reach of LSST and Roman

Distance-redshift relation departs from ΛCDM by a measurable amount at z ≈ 0.5 to 1
The BAO acoustic feature traces out a distance-redshift curve that bends measurably away from the ΛCDM prediction, detectable by Stage-IV surveys (LSST, Euclid) mapping the feature across many redshifts
Stage-IV surveys (LSST, Euclid, Roman) should crisply confirm or refute the trend within the next decade
If (w0 > -1, wa < 0) is real, dark energy crosses the phantom barrier (w = -1) toward more negative w in the past, then trends less negative toward today

Where it breaks

Current data don't show meaningful preference for wCDM over plain ΛCDM. The interesting signal is in the evolution (w0waCDM), not in a constant offset
wCDM is a parameterization, not a model. No fundamental physics predicts a fluid with constant w ≠ -1
Phantom dark energy (w < -1) violates the dominant energy condition, theoretically awkward but not ruled out by data alone

Efstathiou (2025) argues the signal is a low-redshift systematic in the SN Ia compilations, not real dark energy evolution
Significance is sensitive to the choice of SN compilation; the headline 3.9σ uses DES-SN5YR, a single recent sample
The CPL parameterization is just a simple linear approximation of how w changes near the present day (w0 + wa·(1-a)), not a fundamental theory. A real signal would still need a physical origin
Posteriors can cross the phantom divide (w < -1), which is theoretically awkward for most quintessence models

Key unresolved problem

The no-signal problem: current data show no real preference for this constant-w version over plain ΛCDM, so there is nothing yet that would tell the two models apart.

The interpretation problem: if the DESI hint of changing dark energy is real, no underlying theory predicts that exact pattern, the equation of state w starting above -1 and dropping, without awkward fine-tuning.

Reader vote

No votes yet