Dark Matter-Dark Radiation Interactions and the Hubble Tension
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.
Vera: I'm Vera.
Jocelyn: And I'm Jocelyn, here with Subrahmanyan.
Vera: We're looking at the paper: "Dark Matter-Dark Radiation Interactions and the Hubble Tension".
Title and authors: Vera: So we're looking at this paper called "Dark Matter-Dark Radiation Interactions and the Hubble Tension." It’s about proposing a way to fix that gap between how fast we think the universe is expanding based on the Cosmic Microwave Background data and what local measurements tell us.
Jocelyn: That sounds complicated, Vera. What exactly are we looking at here? Is this another tweak to our standard model, or something entirely new?
Vera: It's about a subcomponent of dark matter interacting with dark radiation at early times. The idea is that they were in thermal equilibrium and then they decoupled during the epoch when the CMB was forming. This interaction rate scales in a specific way, like Gamma d T over T to the power of two plus n.
Subrahmanyan: So we're talking about modifying how dark matter behaves right around recombination. The paper suggests that how fast this decoupling happens really matters for what we see later on in the universe.
Jocelyn: Right, but what’s the main point they are trying to show? Is it just another way to get a better Hubble constant number?
Vera: They test three different ways this interaction rate can behave—they call them n equals four, n equals two, and n equals zero. The key is that the speed of decoupling affects how matter perturbations grow and what we measure in the matter power spectrum.
Subrahmanyan: So they’re looking at how different decoupling rates change the growth of structure, which then feeds back into the shape of the matter power spectrum. That connects it to structure formation on larger scales.
Jocelyn: And how does that connect to those measurements we have from Planck and Supernovae? What's their actual finding?
Vera: The general finding is that all three scenarios reduce the Hubble tension compared to the standard lambda CDM model, but they say an exponentially fast decoupling rate is the most preferred scenario.
Subrahmanyan: Exponentially fast decoupling implies a very strong interaction right at that early time when dark matter and dark radiation were mixed. That’s a pretty specific physical constraint on the underlying particle physics.
Jocelyn: When you look at the CMB data, how does this decoupling rate affect what we observe about gravitational potentials?
Vera: The evolution of those gravitational potentials changes because of how quickly they transition into a purely cold dark matter scenario. This affects how we calculate the final value of those potentials.
Subrahmanyan: That’s important because the shape and amplitude of those potentials are what eventually get imprinted on the CMB anisotropies we measure.
Title and authors: Jocelyn: So, when you put this into practice, what kind of data are they using to check their models? Are they just looking at one dataset?
Vera: They used a modified version of the CLASS Boltzmann solver and employed MCMC analysis with Cobaya to test against various datasets. They checked Planck data and also included SH0ES data.
Subrahmanyan: Including SH0ES is significant because it brings in those local measurements of how fast the universe is expanding near us, which are what cause the tension in the first place.
Jocelyn: And what did they find when they combined all that data? Did one scenario win out?
Vera: The results show that all three scenarios lower the Hubble tension relative to lambda CDM, but an exponentially fast decoupling is still their top choice. When they added the SH0ES data, the largest H0 value at a ninety-five percent confidence region was seventy-two point five km per second per megaparsec, which reduced the tension to less than three sigma <ref:2511.16554#pg2>.
Subrahmanyan: Less than three sigma is a decent reduction for this kind of modification. It suggests that these extended models are preferred over lambda CDM when you actually include the local SH0ES measurement in your analysis framework.
Jocelyn: That’s a big qualifier there, Vera. So it’s not just about getting a better CMB fit; it depends on those local distance ladder measurements too.
Vera: Exactly, but there’s another point they make about the n equals zero scenario. They found that this model shows a remarkable ability to accommodate higher H0 values while still fitting the Planck data really well.
Subrahmanyan: So they can push the expansion rate higher without destroying the consistency with early universe measurements. That's quite a feat for a new physics scenario.
Jocelyn: Can you tell us more about how that n equals zero model works in this context? Is it fundamentally different from the others?
Vera: The n equals zero case is based on Coulomb-like scattering for the interaction, which is what they call n equals zero. It seems to be particularly good at reducing the tension to less than two sigma even without including those SH0ES data.
Subrahmanyan: That’s a very interesting result because it suggests a mechanism that works well on its own when you only look at the CMB constraints. It effectively reduces the tension below two sigma just using Planck data, which is quite impressive for an interaction model.
Title and authors: Jocelyn: So they found one specific physical scenario where the tension is reduced by just relying on early universe physics and not needing to bring in local measurements?
Vera: Yes, that’s it. The paper "Dark Matter-Dark Radiation Interactions and the Hubble Tension" concludes that these extended models are preferred over lambda CDM only when you include the local SH0ES H0 measurement in your fits.
Subrahmanyan: And they specifically point out that the n equals zero model demonstrates a remarkable ability to accommodate higher H0 values while maintaining an excellent fit to the Planck data, effectively reducing the Hubble tension to less than two sigma without SH0ES data.
Jocelyn: It sounds like they’ve found a specific parameter space where their proposed physics can explain some of the discrepancy without needing immediate confirmation from those local measurements.
Vera: And they also noted that including ACT DR6 data, which measures small-scale anisotropies, actually weakens this preference by tightening the constraints on the effective number of relativistic species.
Subrahmanyan: That means if you want to constrain these interacting dark sector models really tightly, you have to be careful about what small scale measurements you include in your analysis.
Jocelyn: So, where does this leave us? What's the final word on this paper and what should we expect next in the dark matter research community?
Vera: Overall, they show that these interacting dark sector models offer a way to address the Hubble tension, but it really depends on which interaction type you pick and what data you use to constrain it.
Subrahmanyan: The n equals zero model seems like the most promising path for alleviating tension using only early universe constraints, provided we accept those specific interaction dynamics.
Jocelyn: It’s interesting how different scaling behaviors, n equals four versus n equals two, lead to very different predictions for the matter power spectrum features. That might give us a way to predict which physical interaction is most relevant based on what we might eventually see in structure formation data from surveys like DESI.
Vera: Right. The paper gives us clear direction: test these models against multiple datasets, because relying on just one thing won't tell you the full story about how this new physics impacts cosmology.
Subrahmanyan: It’s a good reminder that connecting early universe physics to late-time structure formation is where we need to focus our observational efforts next.
Jocelyn: We'll be keeping an eye on those results, and maybe the next paper will show us what those ACT constraints really look like in practice.
The paper's summary: Jocelyn: So, to recap, this paper is about how dark matter and dark radiation interact early on and whether that interaction can help bridge the gap between our local measurements of how fast the universe is expanding and what we see in the early universe data from Planck.
Vera: Exactly. They looked at different ways this interaction rate could behave—scaling by n equals four, n equals two, or even zero—and they used computer models to see which one fits the data best.
Subrahmanyan: From my side, what’s interesting is that the growth of those initial matter perturbations changes depending on how fast that decoupling happens. It directly feeds into what we measure in the matter power spectrum later on.
Jocelyn: And they found that all three scenarios actually help lower the tension compared to our standard lambda CDM model, but they say an exponentially fast decoupling is the most preferred way for this physics to work.
Vera: That's a big finding because it points toward a very specific type of interaction happening right at recombination, when the CMB was forming. It suggests that if this kind of dark sector interaction is real, it has to be quite strong at that moment.
Subrahmanyan: I agree, and it’s important to note that the paper says you only really prefer these extended models if you include those local measurements from SH0ES data in your analysis. That’s a key caveat for any future work.
Jocelyn: Right, so if you just look at the CMB data alone, the n equals zero model actually shows a remarkable ability to accommodate higher expansion rates while still fitting Planck data very well, even reducing the tension below two sigma by itself.
Vera: That’s a subtle but important detail. It means you don't always need those local measurements to see how these models can alleviate the Hubble tension. The n equals zero scaling seems particularly good at that specific task without needing external input.
Subrahmanyan: However, they do have to be careful because including data from ACT DR6, which measures small-scale anisotropies, actually tightens the constraints on a parameter called Delta Neff in a way that weakens this preference for the n equals zero model.
Jocelyn: So it’s not just about finding *a* solution, but figuring out which specific physical interaction—which 'n' value—is the right one to use depending on what other observational data you decide to include.
Vera: Precisely. The paper sets up a clear roadmap: test these different scaling behaviors against multiple datasets because relying on just one thing won't give you the full story about how this new physics impacts cosmology.
Subrahmanyan: And for me, the real implication is that connecting early universe physics to later structure formation is where we need to focus our observational efforts next, especially with things like DESI surveys.
Jocelyn: We’ll be keeping an eye on those results, and maybe the next paper will show us what those ACT constraints really look like in practice.
The paper's improvements: Tom: So, this paper isn't just proposing one set of models; they’re actually suggesting how to make these dark sector interaction models more robust and testable with real data.
Jocelyn: Right, they’re looking at ways to improve the analysis itself rather than just tweaking the physics a little bit.
Vera: They point out that you need a full likelihood analysis of interacting dark matter and dark radiation models using modified versions of the CLASS code. That means you gotta run the whole simulation through every dataset, not just check one thing in isolation.
Subrahmanyan: That’s smart because it shows how gen-iDM and lambda CDM fits compare against things like Planck, ACT DR6, BAO, DESI, full-shape matter power spectra, supernova data and local H0 measurements all at once.
Jocelyn: And this AI approach lets them see exactly which variant of the interaction model works best when you throw in all that messy real data.
Vera: They’re trying to figure out how to prioritize gen-iDM variants based on whether or not you include the SH0ES data in your analysis. It’s like telling you, "if you want that specific result, then use this specific interaction type."
Subrahmanyan: I agree, and another thing they suggest is using the n equals zero model as a starting point because it seems to handle higher H0 values better while keeping the Planck fit excellent without needing those SH0ES constraints.
Jocelyn: They also look at how different interaction types, like n equals four versus n equals two versus n equals zero, affect the dark acoustic oscillations in the power spectrum. This helps them predict which interaction rate scaling is physically relevant based on what we might eventually see in structure formation data from surveys like DESI.
Vera: And they also address the ACT constraints by looking at how high-precision small-scale anisotropy measurements constrain parameters. They found that ACT data actually doesn't favor a decrease of power at small scales, which leads to tighter constraints on Delta Neff within their model framework.
Subrahmanyan: So the idea is using these observational tests—like the shape of the matter power spectrum or those small-scale ACT features—to constrain new physics in a way that’s more powerful than just looking at one dataset.
Jocelyn: It moves us toward having a more complete picture, where we use different types of probes to pin down exactly what kind of dark sector interaction is happening out there.
Vera: It's a lot of work, but it shows the path forward for testing these extended models against the real universe data.
Conclusion: Jocelyn: So, to wrap up, we’ve seen how this paper on "Dark Matter-Dark Radiation Interactions and the Hubble Tension" tests different interaction rates and finds that exponential decoupling is generally preferred for solving that tension.
Vera: Exactly. The big takeaway is that these extended models only become preferred over lambda CDM when you include those local SH0ES measurements in your fits, although the n equals zero model can be quite effective on its own with just the Planck data.
Subrahmanyan: I think it’s telling us that the choice of interaction dynamics isn't arbitrary; it depends entirely on what observational constraints you decide to prioritize when trying to resolve the Hubble tension.
Jocelyn: It really frames the next steps by showing that we need multi-pronged testing, checking both early universe physics and later structure formation data like DESI or ACT.
Vera: That’s right, it suggests that if we want to pin down these dark sector interactions, we have to be careful about what small scale measurements you include because they tighten constraints on things like Delta Neff.
Subrahmanyan: So the paper lays out a clear path: test these models against multiple datasets because relying on just one thing won't give you the full story about how this new physics impacts cosmology.
Jocelyn: It’s a good reminder that connecting early universe physics to late-time structure formation is where we need to focus our observational efforts next.
Vera: That’s all we have on today’s look at "Dark Matter-Dark Radiation Interactions and the Hubble Tension."
Subrahmanyan: It's an interesting piece, showing how specific interaction rates can lead to very different predictions for the matter power spectrum features.
Jocelyn: We’ll be watching what comes next, especially how those ACT constraints really look in practice when we try to test these scaling behaviors.
Maryland Center for Fundamental Physics University of Maryland Department of Physics Dual CP Institute of High Energy Physics Weinberg Institute for Theoretical Physics University of Texas at Austin Department of Physics and Astronomy University of Utah Laboratory for Elementary Particle Physics Cornell University School of Physics Korea Institute for Advanced Study
astro-ph.CO, hep-ph
Submitted: 2025-11-20
Updated: 2026-11-18
Comments: 44 pages, 19 figures: Minor changes. Matches published version
Code: https://github.com/lesgourg/class
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
The gist: Models in which a subcomponent of dark matter interacts with dark radiation have been proposed as a solution to the Hubble tension.
Key concepts
- Dark Matter-Dark Radiation Interaction
- This framework explores a scenario where a small piece of cold dark matter interacts with dark radiation early on. This interaction affects how the dark matter and radiation evolve together, potentially altering cosmological predictions compared to standard models.
- Decoupling Rate ($\Gamma_d(T)$)
- This describes how quickly the interaction between dark matter and dark radiation stops during the early universe. The rate is modeled as $\Gamma_0 d^T / T_0^{2+n}$, and this rate is crucial because it dictates how much the interactions affect structure growth later on.
- Hubble Tension
- This refers to a significant discrepancy between the value of the Hubble constant ($H_0$) measured in the early universe (e.g., from Planck data) and the value measured locally using nearby galaxies (e.g., SH0ES data). The paper explores models that can bridge this gap.
- n=0 Model
- This specific interaction scenario, where $n=0$, is highlighted as a particularly effective solution. It allows the model to accommodate higher local H0 values while still fitting Planck CMB data well, significantly reducing the tension without requiring additional local measurements.
Terminology
Summary
Models in which a subcomponent of dark matter interacts with dark radiation have been proposed as a solution to the Hubble tension. p. 0 The paper studies these general classes of models and evaluates their fit to cosmological data, focusing on how the rate at which dark matter decouples from dark radiation affects the results. p. 0
Framework Description
The framework considers a scenario where a subcomponent of cold, noninteracting dark matter is in thermal equilibrium with dark radiation at early times and eventually decouples during the CMB epoch. This rate scales phenomenologically as Γd(T) = Γ0 d T / T0(2+n). p. 3 These scenarios map into well-motivated physical models. p. 3
Impact of Decoupling Rate
The primary difference between the benchmark models is how quickly the DM-DR interactions decouple, which affects the growth of matter perturbations and consequently impacts the matter power spectrum (MPS). The impact on the CMB comes from the evolution and final value of gravitational potentials, which is affected by how quickly these potentials transition to a purely CDM scenario. p. 7
Numerical Analysis and Findings
The authors implemented their models into a modified version of the CLASS Boltzmann solver and used MCMC analysis with Cobaya to determine the fits against various cosmological datasets. The results show that all three scenarios reduce the Hubble tension relative to ΛCDM, with an exponentially fast decoupling being the most preferred. p. 0 When including the SH0ES data, the largest H0 value at 95% confidence region is 72. p. 05 km/s/Mpc, corresponding to a reduction in tension to less than 3 σ. p. 0
Conclusion
The analysis reveals that the extended models are preferred over ΛCDM only when the local SH0ES H0 measurement is included in the fits. The n = 0 model demonstrates a remarkable ability to accommodate higher H0 values while maintaining an excellent fit to the Planck data, effectively reducing the Hubble tension to less than 2 σ without SH0ES data. p. 0 However, the inclusion of ACT DR6 data weakens this preference by tightening constraints on ∆Neff. p. 2 Overall, the n = 0 model can substantially reduce the tension between the SH0ES result and Planck CMB data. p. 0
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved AI can perform full likelihood analysis of interacting dark matter and dark radiation models using modified versions of CLASS code to determine how gen-iDM and Lambda CDM fits compare against various cosmological datasets like Planck, ACT DR6, BAO, DESI, full-shape MPS, supernova data, and local H0 measurements.
-
Improved AI can identify that
the extended models are preferred over Lambda CDM only when the local H0 measurement is included in the fits,
which means it can prioritize gen-iDM variants based on whether the SH0ES data is part of the analysis. -
Improved AI can determine that
the n = 0 model demonstrates a remarkable ability to accommodate higher H0 values while maintaining an excellent fit to the Planck data, effectively reducing the Hubble tension to less than 2 sigma,
indicating its capability to find specific physical scenarios that alleviate tension without sacrificing CMB fit quality. -
Improved AI can analyze the impact of different interaction types (n = 4, n = 2, and n = 0) by noting that
the faster decoupling scenario leads to much larger amplitudes for the dark acoustic oscillations,
allowing it to predict which interaction rate scaling is most physically relevant based on observed power spectrum features. -
Improved AI can assess how high-precision small-scale anisotropy measurements from ACT constrain parameters by observing that
ACT data does not favor a decrease of power at small scales, leading to tighter constraints on Delta Neff,
thus providing a mechanism for constraining new physics within the model framework.
Sources
- Planck 2018 results. VI. Cosmological parameters
- A Comprehensive Measurement of the Local Value of the Hubble Constant with 1 km/s/Mpc Uncertainty from the Hubble Space Telescope and the SH0ES Team
- Measurements of the Hubble Constant: Tensions in Perspective
- The Hubble Tension and Early Dark Energy
- Status Report on the Chicago-Carnegie Hubble Program (CCHP): Measurement of the Hubble Constant Using the Hubble and James Webb Space Telescopes
- Non-Abelian dark matter and dark radiation
- Evidence for dark matter interactions in cosmological precision data?
- Interacting Dark Sector and Precision Cosmology
- Dynamical dark energy in light of the latest observations
- Crack in the cosmological paradigm
- Early Dark Energy Can Resolve The Hubble Tension
- Oscillating scalar fields and the Hubble tension: a resolution with novel signatures
- Acoustic Dark Energy: Potential Conversion of the Hubble Tension
- Axion-Dilaton Destabilization and the Hubble Tension
- Rock 'n' Roll Solutions to the Hubble Tension
- A CMB Search for the Neutrino Mass Mechanism and its Relation to the $H_0$ Tension
- Thermal Friction as a Solution to the Hubble Tension
- New physics in light of the $H_0$ tension: an alternative view
- Is the Hubble tension a hint of AdS phase around recombination?
- Flavor-specific Interaction Favors Strong Neutrino Self-coupling in the Early Universe
Related papers
- 21cmEMUv3: a hybrid diffusion-LSTM emulator of 21cmFAST summary observables
- Connecting field-level and n-point function measurements of galaxy bias
- The Impact of Cosmic Variance and Satellites on JWST Clustering Measurements at Redshift around 6
- Universality of Primordial Anisotropies in Gravitational Wave Background
- KiDS-1000 cosmic shear reanalysis using MetaCalibration
- 2Flow2SN: Dusty Supernova Cosmology with Photometric Redshifts and Contamination using Neural Simulation-Based Inference