Skip to content
CosmosExplorer
Chapter 05 · The Dark Universe/Dark Energy Skeptics

Sarkar's Anisotropic Universe

2019 · Jacques Colin, Roya Mohayaee, Mohamed Rameez, Subir Sarkar
Frontier

Cosmic acceleration may not be what it looks like. The standard evidence for dark energy depends on assuming the universe is the same in all directions, and that assumption may be wrong.

Skip 3D content

In one sentence

The universe's apparent acceleration may be a directional effect of our location and motion, not a sign of uniform dark energy filling all of space.

The claim

The evidence for dark energy rests on assuming the universe looks the same in all directions at very large scales. Sarkar and collaborators question this assumption directly. Using large samples of Type Ia supernovae, they find that the strength of apparent acceleration looks stronger in one direction of the sky, roughly aligned with our motion relative to the cosmic microwave background.

If their reading is correct, what we call cosmic acceleration could instead be a local anisotropy. We are part of a large-scale bulk flow, a region where galaxies share a common motion, and that flow biases the inference. The Nobel-winning interpretation that the universe's expansion is accelerating because of dark energy would need revision. The universe might still be expanding, but not in the way ΛCDM assumes.

The family stance

The apparent acceleration of cosmic expansion is not driven by a uniform dark energy filling space. It is a misinterpretation of observational data, caused by our location, motion, or the way measurements are corrected for local effects.

Predictions

  • Measurements of cosmic expansion from supernovae should show a dipole pattern on the sky, stronger in the direction of our motion relative to the CMB and weaker in the opposite direction.
  • As surveys go to higher redshifts, this dipole should gradually decrease, because local bulk-flow effects become less important at great distances.
  • The average inferred acceleration from supernova data, analyzed without forcing isotropy, should be statistically consistent with no global acceleration, removing the need for a dominant dark energy component.
  • Independent measurements of the Hubble constant in different sky directions should show directional differences that exceed quoted errors in local H₀ determinations.

Evidence

  • A maximum-likelihood analysis of 740 Type Ia supernovae in the JLA catalogue finds a strong dipole component aligned with the CMB dipole, with the average acceleration small and consistent with zero at low significance.
  • Observations of galaxy motions show a large local bulk flow that appears faster and extends to larger distances than expected in standard ΛCDM, consistent with the directional signal seen in supernova data.
  • Analyses of the Pantheon+ supernova sample report a dipolar variation in the local Hubble expansion rate, with amplitude larger than the 1% uncertainty quoted by recent local measurements.
  • A separate study by the same group argues that when local bulk-flow effects are handled differently, the so-called Hubble tension is reduced, supporting the case that local anisotropies matter.

Counterpoints

  • Several independent groups analyzing the same or larger supernova datasets find that directional patterns are much weaker than Sarkar's team claims and are consistent with statistical noise once survey biases are handled carefully.
  • The standard ΛCDM model successfully explains CMB anisotropies, baryon acoustic oscillations, large-scale structure, and supernova Hubble diagrams without requiring large-scale anisotropy.
  • Bulk flows and local structure are already modeled in ΛCDM, and simulations suggest such flows are not large enough or coherent enough to mimic a global acceleration signal at the observed level.
  • Sarkar's results may depend sensitively on data selection, redshift cuts, and frame choices. With alternative, more standard corrections for peculiar velocities, the evidence for a strong dipole largely disappears.
  • The nearly isotropic CMB at large scales and the success of ΛCDM fits to Planck data strongly constrain any departure from isotropy on cosmic scales.
0votes
Go deeperTechnical detail with proper terminology

Colin et al. (2019) model the deceleration parameter as a sum of a monopole component plus a dipole component that depends on direction and decays exponentially with redshift on a scale of about 100 Mpc. Fitting this model to JLA data, they find a dipole amplitude of order -8 in the deceleration parameter, aligned with the CMB dipole, indicating that direction matters for inferring acceleration.

Their analysis is performed in the heliocentric reference frame, explicitly undoing the standard transformation to the CMB rest frame used in conventional JLA and Pantheon analyses. The argument is that peculiar velocities and bulk flows do not converge to the CMB frame out to redshifts around 0.1, so assuming isotropy in that frame can bias cosmological inference.

In the later Pantheon+ study, maximum-likelihood estimators are applied in the heliocentric, CMB, and Local Group frames. The result is a scale-dependent Hubble dipole exceeding 1.5 km/s/Mpc in the redshift range 0.023 to 0.15, larger than the quoted error budget of local H₀ measurements and thus relevant to the Hubble tension.

Rameez and Sarkar (2019) analyze how redshift corrections and local flows affect distance-ladder H₀ determinations. They argue that differences between JLA and Pantheon redshifts, and the choice of redshift cuts, can shift inferred H₀ by several km/s/Mpc, complicating claims of a multi-sigma Hubble tension.

The broader theoretical implication is that if the standard FLRW metric with a cosmological constant is not a good description at the scales probed by supernovae, one may need to consider non-Copernican, anisotropic cosmological models, where apparent acceleration is a consequence of our location and motion rather than uniform dark energy.

References

Spotted an error? Have a source to add?

Prefer email?

You can also send a prefilled email with the variant URL already filled in.

Related theories