The Evidence for Dark Energy Has a 3.9-Sigma Problem That Nobody Is Talking About
A peer-reviewed reanalysis of the supernova data that won the Nobel Prize found directional anisotropy inconsistent with universal acceleration. This finding has not been refuted.
In 1998, two independent teams of astronomers studying Type Ia supernovae made an announcement that changed cosmology. The universe’s expansion was not slowing down. It was accelerating. Something was pushing galaxies apart, something called dark energy. The Nobel Prize in Physics was awarded for this discovery in 2011.
In 2019, a peer-reviewed paper published in Astronomy and Astrophysics reported a finding that should have generated enormous scientific debate. It has been largely ignored.
The finding: the evidence for cosmic acceleration may be an artefact of our position in the universe, not a property of the universe itself.
What Colin et al. (2019) Found
Jacques Colin, Roya Mohayaee, Mohamed Rameez, and Subir Sarkar analysed the Joint Light-curve Analysis catalogue , 740 Type Ia supernovae, the largest and most comprehensive dataset of its kind. They examined the deceleration parameter q₀, the quantity that, when negative, indicates acceleration.
They found that q₀ is not uniform across the sky. It exhibits a significant dipole component, higher in one direction, lower in the opposite direction, aligned with the CMB dipole. The statistical significance of this anisotropy is 3.9 sigma.
Their conclusion: the cosmic acceleration deduced from supernovae may be an artefact of our being non-Copernican observers, instead of evidence for a dominant component of dark energy in the universe.
This paper has not been refuted in the peer-reviewed literature.
What the CMB Dipole Tells Us
The CMB dipole, the slight temperature asymmetry across the sky, is interpreted as evidence that our local region is moving at approximately 550 km/s relative to the large-scale background. This is a bulk flow. We are not stationary observers. We are embedded in a moving structure.
A moving observer measuring recession velocities of distant objects will see asymmetry. Objects in the direction of motion will appear to recede more slowly. Objects in the opposite direction will appear to recede faster. This produces exactly the kind of dipole in the deceleration parameter that Colin et al. found.
In the Big Flare-Up Theory, this bulk flow is the natural consequence of gravitational sorting at large scales. Our local group of galaxies is moving in a specific direction, not because of expansion, but because of the gravitational dynamics of the sorted population in our region of the infinite universe.
The Simulation Confirmation
BFUT includes a dedicated simulation of the dark energy illusion. A simulated observer embedded in a bulk-flowing region of an infinite universe, with no dark energy in the simulation physics, produces a dipole signal in the recession measurements. The 3.9-sigma signal reported by Colin et al. is fully reproduced by observer bulk motion alone. When the bulk flow is set to zero, the entire signal vanishes. The simulation is available at vijayshankarsharma.com/acceleration.
Three Values of the Hubble Constant
The Hubble tension provides independent support for this interpretation. If the universe were uniformly accelerating under dark energy, all measurements of the expansion rate should converge on the same value. Instead, three independent methodologies yield three distinct values: 63, 68, and 73 km/s/Mpc.
Local measurements, those probing smaller scales and more recent epochs, consistently return lower values. The Wagner, Benisty and Karachentsev (2026) measurement of 63 +/- 6 km/s/Mpc from galaxy group dynamics also reports that those groups are fully explained by visible baryonic mass, without requiring a dark matter halo.
The directional trend in Hubble constant measurements, combined with the directional anisotropy in the supernova data, points consistently toward the same conclusion. The apparent acceleration is a local observational artefact. Dark energy is the explanation that was needed when the data was assumed to be isotropic. The data is not isotropic.
The Nobel Prize Controversy
The 2011 Nobel Prize in Physics was awarded to Perlmutter, Schmidt, and Riess for the discovery of cosmic acceleration. The prize citation described it as one of the most remarkable observations in the history of cosmology. It was awarded on the basis of the supernova data that Colin et al. subsequently found to exhibit 3.9-sigma directional anisotropy. This creates an uncomfortable situation. If the Colin et al. finding is correct, if the acceleration signal is a bulk flow artefact, then the Nobel Prize was awarded for an observational artefact instead of a genuine cosmological discovery. The prize was not awarded in error in any procedural sense. The data was analysed correctly using the methods standard at the time. The error, if it is one, is the failure to account for the observer’s bulk flow velocity when analysing the full sky dataset. This is a correctable methodological limitation, not a fraud or fabrication. The scientific question, whether the acceleration is real or an artefact, has not been settled. The fact that it has not been settled eight years after a peer-reviewed paper raised it at 3.9 sigma is itself noteworthy. The standard response in physics to a 3.9-sigma anomaly in a foundational dataset is intensive investigation and reanalysis. The response in cosmology has been muted.
The Bulk Flow Measurement
The CMB dipole velocity of approximately 550 km/s has been measured with high precision. It is interpreted as the motion of our local group relative to the large-scale CMB rest frame. This motion is real and confirmed. What has not been consistently applied is the correction for this motion when analysing supernova recession velocities across the full sky. At a distance of 1 billion light years, a bulk flow velocity of 550 km/s produces a fractional recession velocity asymmetry of approximately 550/(c × 0.07) ≈ 0.003, or 0.3%. This is small but not negligible at the precision required to measure the deceleration parameter. The Colin et al. analysis showed that this uncorrected asymmetry is sufficient to produce the observed acceleration signal. The BFUT acceleration simulation reproduces the 3.9-sigma signal with bulk flow as the only input and no dark energy in the physics.
DESI and Future Tests
The Dark Energy Spectroscopic Instrument (DESI) has been mapping galaxy redshifts across a large fraction of the sky since 2021. Its early results, published in 2024, suggested that dark energy may not be a simple cosmological constant but may vary with time, a w₀wₐ dark energy model. This finding, if confirmed, would create additional problems for the standard model instead of resolving them, because a time-varying dark energy requires a physical mechanism that LCDM does not provide. From the BFUT perspective, the DESI finding is consistent with the bulk flow interpretation. A bulk flow artefact would produce apparent time variation in the dark energy equation of state, because different epochs of observation probe different regions of the sorted galaxy population with different bulk flow components. The apparent variation is not a property of dark energy. It is a property of the observer’s changing line of sight through an inhomogeneous sorted population. DESI’s larger dataset will eventually provide the directional resolution to test the Colin et al. anisotropy at higher significance.