The KBC Supervoid and the Hubble Tension
Two of the most precise measurements in physics — how fast the universe is expanding, taken from the cosmic microwave background and taken from nearby supernovae — disagree by 5σ. One candidate culprit: we may be measuring from inside a 600-million-light-year hole in the cosmos. As of June 2026, the hole appears too small to fully explain the disagreement, and nobody knows what does.
The tension, in numbers
H₀ measures cosmic expansion in km/s per megaparsec. Two measurement chains:
| Method | H₀ (km/s/Mpc) | Probe | Era |
|---|---|---|---|
| Planck CMB (2018) | 67.4 ± 0.5 | Sound horizon at recombination | z ≈ 1100 |
| SH0ES distance ladder (Riess 2022) | 73.0 ± 1.0 | Cepheids → Type Ia supernovae | z < 0.15 |
| TRGB ladder (Freedman 2024) | 69.8 ± 1.7 | Tip of red giant branch | z < 0.05 |
| Megamaser (MCP 2020) | 73.9 ± 3.0 | Direct geometric, NGC 4258 etc. | z < 0.03 |
The early-universe and late-universe numbers refuse to converge. At 5σ this is the threshold the physics community uses to call something a discovery. Either a measurement chain has a hidden systematic, or the ΛCDM standard model is missing physics.
What the KBC void is
In 2013 Ryan Keenan, Amy Barger and Lennox Cowie mapped infrared galaxy counts and reported the Milky Way sits inside a region roughly 20% underdense in galaxies, extending out to 300-600 Mpc (~1-2 billion light-years). Not empty like concept bootes void — just systematically thinner.
The mechanism that could matter for H₀: matter inside an underdensity gets pulled outward toward the denser walls. That outward pull is a peculiar velocity layered on top of the background cosmological expansion. A local astronomer measuring nearby objects can't disentangle the two — they read off an inflated H₀.
Haslbauer, Banik and Kroupa (MNRAS, 2020) modeled this and found a ~20% underdensity out to 300 Mpc inflates the local rate by 11 ± 2%. That number is suspiciously close to the gap between 67.4 and 73.4. The void, they argued, dissolves the tension entirely inside ΛCDM. No new physics required.
The 2025 test that broke the clean story
Boubel et al. (arXiv:2506.10518, MNRAS June 2025) tested the void hypothesis with a different probe. Instead of counting galaxies, they used Tully-Fisher distances from the CosmicFlows-4 catalogue to measure actual peculiar velocity fields — the direction and speed of nearby galaxy flows.
Their finding:
Preferred void size from velocity flows: < 70 Mpc
Void size required by Haslbauer model: ~300 Mpc
Gap: ~4×
Predicted H₀ under various void profiles fit to the velocity data:
| Profile | H₀ predicted | Resolves to 73.4? |
|---|---|---|
| Exponential | 72.1 +0.9/-0.8 | No |
| Gaussian | 70.4 ± 0.4 | No |
| Maxwell-Boltzmann | 70.2 +0.5/-0.4 | No |
All three ease the tension. None close it. The local measurement is still 73.4.
There's a meta-puzzle here: galaxy-count surveys (KBC's original method) find a large void; peculiar-velocity surveys find a small one. They are measuring slightly different things — luminous-tracer density versus actual mass flow — but the disagreement itself needs explaining.
The DESI complication
DESI DR2 (2025) shows 3.1-4.2σ evidence that dark energy may evolve over time (w₀ ≈ -0.77, wₐ ≈ -0.86) rather than being the constant Λ of ΛCDM. This is independent of the void question and points toward new physics.
But evolving dark energy generally makes the Hubble tension worse, not better, in joint fits. The early-universe H₀ inferred from the CMB + BAO + dynamical dark energy moves further from 73, not closer. DESI is a real signal of something interesting; it is not a tension-solver.
A more speculative angle: cosmic backreaction. The averaged expansion of an inhomogeneous universe (voids expanding faster, walls slower) is mathematically not the same as the expansion of a smoothed-out universe. Some cosmologists argue this is large enough to mimic evolving dark energy and partially account for H₀ disagreement. Most working cosmologists find the effect too small. The debate is live.
What's contested
- How big is the KBC void really? Galaxy counts say 300-600 Mpc; velocity fields say under 70 Mpc. Both teams stand by their methods.
- Does backreaction matter? A small camp (Buchert, Wiltshire) says yes; mainstream ΛCDM says it's a sub-percent effect.
- Is the SH0ES Cepheid ladder right? Freedman's TRGB-only ladder gives 69.8, midway between the camps. The Cepheid metallicity correction is the most-attacked single step in late-universe H₀ work.
- Modified gravity at Gpc scales? Haslbauer's group reads the void+tension combination as evidence for MOND-like dynamics on supercluster scales. This is a minority view.
The tension itself is established. Every plausible cause is contested.
Why this has to do with other realms
The KBC void turns cosmology into an epistemology problem. If our local large-scale structure systematically biases what we measure, then the universe's "true" expansion rate is something we cannot read off directly from our vantage point — we have to model it. This is the same shape of problem concept anthropic principle raises about why physical constants look life-friendly: the observer's location is part of the data.
It also recurs in operator life. Any metric measured from inside a system inherits the system's biases. A buying team measuring "what customers want" from store sales misses the customers who never walked in. A model trained on Western internet text measures "language" as a 4% slice of human linguistic production. Cosmology is the limit case: there is no outside to measure from.
An open question
If a 70 Mpc local void explains 30% of the H₀ gap, and DESI's evolving dark energy explains another fraction but worsens the joint fit, what kind of two-effect cancellation would fit all the data? Nobody has written that paper yet. Euclid's first cosmology results (expected late 2026) will either constrain or open this possibility.
Key sources
- Riess et al. (2022) "A Comprehensive Measurement of the Local Value of the Hubble Constant", ApJL — the SH0ES 73.0 ± 1.0 result.
- Planck Collaboration (2020) "Planck 2018 results. VI. Cosmological parameters", A&A — the 67.4 ± 0.5 CMB result.
- Keenan, Barger, Cowie (2013) "Evidence for a ~300 Mpc Scale Under-density in the Local Galaxy Distribution", ApJ — original KBC void detection.
- Haslbauer, Banik, Kroupa (2020) "The KBC void and Hubble tension contradict ΛCDM on a Gpc scale", MNRAS — the void-as-resolution argument.
- Boubel et al. (2025) "Constraints on the local void from peculiar velocity surveys", MNRAS, arXiv:2506.10518 — the velocity-field test that shrank the void.
- DESI Collaboration (2025) "DESI 2024 VI: Cosmological constraints from the BAO measurements", to verify exact citation — the evolving dark energy signal.
Further reading
- concept cosmic distance ladder — the chain of measurements that produces the late-universe H₀; understanding the rungs is the only way to evaluate the systematics debate.
- concept dark energy — the DESI signal in context.
- The First Three Minutes by Steven Weinberg (1977) — still the cleanest plain-English introduction to how early-universe cosmology became precision physics.
- Adam Riess's Nobel lecture (2011) on YouTube — the discovery of accelerating expansion, by the person now at the center of the H₀ fight.
- arxiv.org/abs/2506.10518 — the Boubel et al. paper directly; readable for someone with first-year cosmology.
Abhishek's take
The KBC story is the rare cosmology problem that turns on a craft question rather than a theory question: how do you measure mass distribution in a region you live inside? The galaxy-count people and the velocity-field people are using different instruments and getting different answers about the same patch of sky, and that disagreement is more interesting than the void itself. My bet is the tension's eventual resolution will be unglamorous — a half-percent Cepheid systematic plus a real-but-modest void effect plus a small DESI-flavored dark energy drift, none alone sufficient. The field wants one clean cause. The data is going to give it three messy ones.
See Also
- concept dark energy — the DESI DR2 result and its implications.
- concept cosmic distance ladder — the rung-by-rung chain whose systematics are now the central H₀ battleground.
- concept bootes void — the canonical supervoid for scale comparison.
- concept anthropic principle — observer-location problems in physics, a structural cousin of the void-bias issue.
- concept void fermi paradox — void location as a filter on what civilizations can infer about cosmology.
- concept dark matter — the other large invisible component of ΛCDM whose properties feed into every H₀ chain.
Tags: #cosmology #hubble-tension #supervoid #dark-energy #desi #expansion-rate #lcdm