The Cosmic Dipole That Refuses to Fit
04 May 2026, Yanjiang
The cosmic dipole measured from infrared and radio surveys consistently disagrees with the CMB’s kinematic dipole, hinting at a fundamental asymmetry in the universe.
Every time astronomers map the sky, they expect the universe to look the same in every direction. Every time they check the data against the most fundamental assumptions of cosmology, they find a quiet but persistent anomaly that suggests it does not.
The assumption that the cosmos is homogeneous and isotropic on large scales—the Cosmological Principle—has been a cornerstone of modern cosmology since Copernicus displaced Earth from the center of the universe. It is why we can apply the same physics to distant galaxies as we do to our own. It is the foundation upon which the expanding universe, the Big Bang, and the standard model of cosmology are built. Over the past decade, that foundation has developed a worrying crack.
The trouble centers on something called the cosmic dipole—a pattern of asymmetry in the sky that should be a simple consequence of our own motion, but which stubbornly refuses to match the prediction.
The simplest explanation
When the Planck satellite mapped the Cosmic Microwave Background (CMB), the faint afterglow of the Big Bang, it found that one half of the sky is slightly warmer than the other. This pattern has a clean and straightforward explanation: the Solar System is moving through the universe at roughly 370 kilometers per second relative to the CMB, and that motion Doppler-shifts the light we see. The direction we are heading toward appears slightly hotter; the direction we are leaving appears slightly colder.
That explanation is settled science. The dipole you measure in the CMB is the kinematic signature of our own motion through the cosmos.
The trouble is that when astronomers measure the dipole using other kinds of sources—radio galaxies, infrared-selected quasars, and distant galaxies—they consistently find a dipole that is larger than the kinematic prediction. Not by a small margin, and not in a way that can be dismissed as a statistical fluctuation. The discrepancy has persisted across multiple surveys, multiple wavelengths, and multiple years of observation.
A team led by Tara Murphy at the University of Sydney has now delivered the most comprehensive statistical accounting yet of just how serious the tension has become. Working with first author Mali Land-Strykowski and Geraint F. Lewis, the team performed a Bayesian analysis comparing four major datasets: the Planck satellite’s CMB observations, the NRAO VLA Sky Survey (NVSS), the Rapid ASKAP Continuum Survey (RACS), and the Wide-field Infrared Survey Explorer catalogue (CatWISE). Their work appears in a preprint (arXiv:2509.18689) on arXiv.
The sky maps of radio and infrared sources reveal a distinct lopsided pattern, with more galaxies clustered in one direction than another. This imbalance challenges the assumption of a uniform universe, hinting at deeper inconsistencies with the cosmic microwave background. (Source: arXiv:2509.18689)
The logic of comparison
The reasoning behind the analysis is simple. If the CMB dipole is purely kinematic—meaning it arises entirely from our motion through space—then every survey of cosmologically distant sources should see exactly the same dipole. A radio galaxy a billion light-years away should show the same directional asymmetry as the CMB. A quasar seen in infrared light should show the same pattern too. The motion that creates the CMB dipole is a fundamental reference frame, and it should affect all light equally, regardless of wavelength or source type.
If all surveys agreed, there would be no puzzle. The Cosmological Principle would be safe, and the kinematic interpretation would be confirmed.
The team’s analysis shows that they do not agree. The disagreement is not subtle, and it is not uniform across surveys—which turns out to be a crucial clue.
What the surveys reveal
The most dramatic tension appears in the infrared. When the team compared Planck with CatWISE—a catalog of millions of infrared-selected quasars and galaxies compiled from the Wide-field Infrared Survey Explorer—they found a discrepancy exceeding five sigma. In the language of statistical significance, that is a discovery-level disagreement. The probability that the two datasets describe the same underlying dipole is vanishingly small.
With RACS, a radio survey conducted by the ASKAP telescope in Western Australia, the tension is strong but slightly less extreme. With NVSS, the classic radio survey conducted at the Very Large Array in New Mexico, the tension is moderate but still present.
But here is where the story takes an unexpected turn. When the team compared CatWISE with NVSS directly—not each against Planck, but against each other—they found something remarkable: the two surveys agree. Their dipoles point in similar directions and have similar amplitudes. This concordance strongly suggests that the dipole measured in these surveys is not random noise or a statistical fluctuation, but a real astrophysical signal.
Think of it like two compasses that agree with each other while a third, supposedly more authoritative compass, points somewhere else. The two that agree are probably reading something real—a genuine asymmetry in the distribution of matter across the sky. Unlike dinner guests who might coincidentally order the same meal, these surveys operate at completely different wavelengths and detect completely different populations of sources. Infrared quasars and radio galaxies are not the same objects. Their agreement points to something fundamental.
The team’s analysis of the posterior distributions—the probability maps of where the dipole might actually point—reveals this concordance clearly. When plotted together, the marginalized posteriors of CatWISE and NVSS overlap substantially, while both sit distinctly apart from the Planck posterior. This is not a subtle effect visible only to experts. It is a clean, visual statement that something is systematically different between the CMB dipole and the dipole seen by surveys of distant sources.
Planck’s cosmic microwave background dipole does not align with dipoles from radio and infrared galaxy surveys. This tension hints that our motion through space may not be the sole cause of the cosmic dipole, challenging a key prediction of standard cosmology. (Source: arXiv:2509.18689)
A warning about RACS
The RACS survey tells a different story. It does not agree well with either CatWISE or NVSS. The team suspects this may be due to systematic effects in the RACS catalog itself—perhaps related to how the survey handles the bright radio emission of the Milky Way’s galactic plane, or how it corrects for variations in telescope sensitivity across the sky. The strong discordance between RACS and both CatWISE and NVSS, the team notes, “indicates a possible systematic difference in the RACS catalogue itself.”
This is an important finding in its own right. It means that not every radio survey is equally reliable for dipole measurements. Some catalogs contain systematics that can mimic or mask a dipole signal, and identifying which surveys can be trusted is a crucial step toward resolving the anomaly.
How to settle the question
The team does not claim to have solved the cosmic dipole puzzle. Instead, they have done something arguably more valuable: they have shown exactly what is needed to settle it, and how close we are to getting there.
Their calculations indicate that roughly a million radio sources are required to measure the dipole tension at a significance of five sigma. That is a staggering number—far beyond what any existing survey can provide. The current generation of radio catalogs contains tens of thousands to a few hundred thousand sources. The team demonstrated this by generating synthetic samples at different source counts, showing that the observed tension fluctuates wildly at low counts but stabilizes only when the source count reaches into the millions.
The Square Kilometre Array (SKA), currently under construction in South Africa and Australia, will be the world’s largest and most sensitive radio telescope. When fully operational, it is expected to detect tens of millions of radio sources—more than enough to resolve the dipole question.
“We are on the cusp of disentangling the anomaly of the cosmic dipole,” the authors write.
What is at stake
The Cosmological Principle is not a minor assumption. It is the foundation of the standard model of cosmology. If the universe is not homogeneous and isotropic on the largest scales, then the entire framework of modern cosmology may need to be revised.
That is a dramatic statement, and the team is careful not to make it lightly. The dipole tension could still have a mundane explanation: an unknown systematic effect in the surveys, a local large-scale structure that distorts the measurement, or a statistical fluctuation that will disappear with more data.
But the pattern is persistent. It appears across multiple surveys, at multiple wavelengths, and it refuses to go away as the data improve. That is not the behavior of a systematic error. It is the behavior of a genuine physical signal—or a genuine crack in the foundation of cosmology.
For now, the tension remains—a quiet, steady anomaly that continues to grow as the data accumulate. The SKA will not begin full science operations for several years, but when it does, the question may finally be answered. Whether that answer confirms the Cosmological Principle or forces us to rethink it, one thing is clear: the cosmic dipole is not going away.
Yanjiang is an online editor of Loom Science
References
- Mali Land-Strykowski et al., Cosmic dipole tensions: confronting the cosmic microwave background with infrared and radio populations of cosmological sources, arXiv:2509.18689