🔭 astrophysics

An Enhanced Isothermal Jeans Approach to Constraining Dark Matter Self-Interactions from Galactic Kinematics

This paper presents an enhanced semi-analytical isothermal Jeans model incorporating velocity-dependent cross sections and core-collapse treatments to constrain self-interacting dark matter using SPARC galaxy rotation curves, revealing that velocity-dependent models with best-fit parameters $\sigma_0 \simeq 5\,{\rm cm}^2$ /g and $\omega \simeq 250\,$ km/s successfully explain small-scale structure diversity and outperform standard CDM profiles without relying on feedback mechanisms.

Original authors: Zixiang Jia, Fangzhou Jiang, Shubo Li, Ran Li, Jing Wang, Ling Zhu

Published 2026-06-10

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Zixiang Jia, Fangzhou Jiang, Shubo Li, Ran Li, Jing Wang, Ling Zhu

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: A Cosmic Mystery

Imagine the universe is a giant city. In this city, there are invisible "ghosts" called Dark Matter that hold everything together. Without these ghosts, the stars in galaxies would fly apart because there isn't enough visible stuff (stars and gas) to keep them in orbit.

For a long time, scientists thought these ghosts were "Cold Dark Matter" (CDM)—meaning they are shy, don't talk to each other, and just float around. But when they looked at the smallest galaxies (dwarf galaxies), the CDM model didn't quite fit the data. It was like trying to fit a square peg in a round hole.

This paper proposes a new idea: Self-Interacting Dark Matter (SIDM). In this version, the ghosts aren't shy; they bump into each other, like people in a crowded dance hall. These "bumps" (collisions) change how the ghosts are arranged, which might explain why those small galaxies look the way they do.

The New Tool: A Better Map

To test this idea, the authors built a new, improved "map" (a mathematical model) to predict how these dancing ghosts arrange themselves inside a galaxy.

The Old Map: Previous models were like a rough sketch. They assumed the ghosts never bumped into each other at high speeds and couldn't handle the messy parts of a galaxy where stars are crowded.
The New Map: The authors upgraded their sketch in three ways:
1. Speed Matters: They realized that how often ghosts bump depends on how fast they are moving. Fast ghosts bump less; slow ghosts bump more.
2. The "Crunch" Phase: They figured out how to model what happens when the ghosts get too crowded in the center. Eventually, the center can collapse inward (like a crowd surging toward a stage). The old maps broke down here; the new one handles it.
3. Better Math: They made the computer code more robust so it doesn't get confused when looking at galaxies with heavy star centers.

The Experiment: Fitting the Puzzle Pieces

The team took 68 galaxies from a famous database (SPARC) and tried to fit their new SIDM map to the actual rotation speeds of stars in those galaxies.

The "Two Faces" Discovery:
When they tried to fit the map, they found something weird for about 1 out of every 6 galaxies. The data fit two completely different scenarios equally well:

The "Growing Core" Scenario: The center is fluffy and expanding (like a balloon inflating).
The "Collapsing Core" Scenario: The center is dense and shrinking (like a balloon deflating).

It's like looking at a photo of a person and not knowing if they are walking toward you or walking away; both movements look the same in a single snapshot. This means we can't be 100% sure which stage of life these specific galaxies are in just by looking at their current speed.

The Results: Finding the "Sweet Spot"

By analyzing all 68 galaxies together, the authors tried to find the "rules of the dance" (the collision rate) that fit the whole group.

The L-Shaped Clue: They found that the data doesn't point to just one single rule. Instead, it points to an "L-shaped" zone of possibilities.
- Option A: The ghosts bump a little bit all the time, regardless of speed (like a slow, steady handshake).
- Option B: The ghosts bump a lot when they are slow, but stop bumping when they get fast (like a sticky trap that only catches slow movers).
The Best Guess: If you have to pick the most likely "average" rule, it's somewhere in the middle: a moderate amount of bumping that slows down as speed increases.

Why This Matters: Not Just "Feedback"

A common counter-argument is: "Maybe the dark matter isn't interacting; maybe the stars and gas are just pushing the dark matter around with their energy (feedback)."

The authors checked this by looking for a pattern. If stars were pushing the dark matter, the "bumpiness" of the dark matter should be directly linked to how many stars are in the galaxy.

The Finding: They found no link. The way the dark matter behaves doesn't seem to depend on how many stars are there. This suggests that the dark matter is doing its own thing (bumping into itself) rather than just reacting to the stars.

The Bottom Line

It works: The new model with "bumping" dark matter explains the shapes of these galaxies just as well as, or better than, the old "no-bump" models.
It's distinct: This behavior seems to be a unique property of dark matter, not just a side effect of star formation.
The Catch: We still can't tell for sure if a specific galaxy is in its "expanding" phase or its "collapsing" phase. To solve this, we need better data on the tiniest, faintest galaxies (the "ultra-faint dwarfs") where the differences are most obvious.

In short, the authors have built a better telescope for the invisible universe, and it suggests that dark matter is a bit more social and dynamic than we previously thought.

Technical Summary: An Enhanced Isothermal Jeans Approach to Constraining Dark Matter Self-Interactions from Galactic Kinematics

Problem Statement
The standard $\Lambda$ CDM cosmological paradigm faces challenges on subgalactic scales, particularly regarding the structural diversity of dwarf galaxies. While baryonic feedback can explain some discrepancies, Self-Interacting Dark Matter (SIDM) offers an alternative mechanism where dark matter particles undergo frequent self-scattering, driving a two-stage evolution: initial core formation followed by potential gravothermal core collapse. Previous attempts to constrain the SIDM self-scattering cross section using galactic rotation curves have been limited by computational costs (relying on full $N$ -body or 1D fluid simulations) or by the inability of semi-analytical models to capture the core-collapse phase. Furthermore, earlier implementations of the isothermal Jeans model often excluded systems with baryon-dominated centers due to numerical robustness issues.

Methodology
The authors present an improved semi-analytical model based on the isothermal Jeans approach, building upon frameworks established by Jiang et al. (2023) and Yang et al. (2024). The methodology involves three key technical upgrades:

Velocity-Dependent Cross Sections: The model incorporates elastic Rutherford scattering via a Yukawa potential, replacing the constant cross-section assumption with an effective cross section ( $\sigma_{\text{eff}}$ ) that depends on the halo's maximum circular velocity ( $V_{\text{max}}$ ). This is parameterized by a low-velocity normalization ( $\sigma_0$ ) and a velocity scale ( $\omega$ ).
Empirical Core-Collapse Treatment: To address the breakdown of the isothermal assumption during gravothermal collapse, the authors adopt an empirical workaround. They utilize the high-density solution of the Jeans equations, which is typically discarded, and mirror it around the merger time ( $t_{\text{merge}}$ ) where low- and high-density solutions converge. This allows the model to approximate post-collapse density profiles.
Enhanced Numerical Robustness: The authors refine the "stitching" procedure (matching the isothermal core to the CDM-like outer halo) by implementing stringent convergence thresholds ( $\delta^2 = 10^{-5}$ ) and partitioning search intervals. This enables the reliable identification of both low-density (core-growth) and high-density (core-collapse) solutions, even for galaxies with baryon-dominated centers.

The model is applied to a sample of 68 galaxies from the SPARC database. The fitting process involves:

Modeling baryonic distributions (gas and stars) using a Hernquist profile, with stellar mass-to-light ratios derived from previous multi-profile fits.
Using Markov Chain Monte Carlo (MCMC) methods to infer halo virial mass ( $M_{\text{vir}}$ ), concentration ( $c$ ), and the degenerate product $p \equiv \sigma_{\text{eff}} t_{\text{age}}$ .
Converting $p$ to $\sigma_{\text{eff}}$ by estimating halo ages via extended Press–Schechter theory.
Performing a joint analysis to constrain the global parameters $\sigma_0$ and $\omega$ by minimizing the disagreement between model predictions and observed rotation curves.

Key Contributions and Results

Sample Expansion: The improved numerical stability allows the inclusion of galaxies with baryon-dominated centers, significantly expanding the sample size compared to previous studies using similar semi-analytical methods.
Bimodality and Evolutionary States: Approximately 1/6 of the sample (11 galaxies) exhibits bimodal posteriors, where both a core-growth solution and a core-collapse solution provide equally good fits. This degeneracy arises because halos at different evolutionary stages can possess nearly identical central densities. The remaining galaxies are predominantly in the core-growth phase or near the maximal-core stage.
Velocity-Dependent Constraints: Joint constraints across the sample reveal a clear velocity dependence in the allowed parameter space. The results form an "L-shaped" degeneracy in the $(\sigma_0, \omega)$ $(σ_{0}, ω)$ plane, indicating that two distinct regimes are viable:
- Nearly constant, low cross sections: $\sigma_0 \sim 2 \, \text{cm}^2/\text{g}$ with $\omega \gtrsim 500 \, \text{km/s}$ .
- Strongly velocity-dependent models: $\sigma_0 \sim 100 \, \text{cm}^2/\text{g}$ with $\omega \sim 60 \, \text{km/s}$ .
- Adopting the core-growth interpretation yields best-fit values of $\sigma_0 \simeq 5 \, \text{cm}^2/\text{g}$ and $\omega \simeq 250 \, \text{km/s}$ .
Comparison with CDM: When compared to Cold Dark Matter (CDM) models, the SIDM model outperforms simple adiabatic-contraction profiles and rivals empirical feedback-based CDM profiles (specifically the Di Cintio et al. 2014 model) in fitting rotation curves. However, unlike baryonic feedback models, the inferred SIDM cross sections show no significant correlation with the stellar-to-halo mass ratio, suggesting SIDM provides a distinct mechanism for structural diversity.
Scaling Relations: The study finds that SIDM mass models result in systematically lower halo concentrations compared to standard CDM relations (e.g., Dutton & Macciò 2014). Additionally, while some galaxies in the SIDM fits exhibit high star-formation efficiencies (approaching unity), this trend is also observed in previous CDM-based studies, indicating that SIDM does not uniquely resolve tensions in the stellar-to-halo mass relation.

Significance
The paper claims that the SIDM paradigm remains a compelling framework for explaining small-scale structure diversity. The derived constraints are remarkably consistent with independent probes, including stellar kinematics in ultra-faint dwarfs, strong lensing, and the mass functions of supermassive black holes in the early Universe (specifically "Little Red Dots"). The authors emphasize that while the L-shaped degeneracy prevents a single unique cross-section model from fitting the entire galaxy sample, the results highlight a specific region of parameter space that simultaneously satisfies orthogonal constraints from local galaxy kinematics and high-redshift black hole demographics. The study concludes that future low-mass kinematic data, particularly for ultra-faint dwarfs ( $V_{\text{max}} \lesssim 30 \, \text{km/s}$ ), will be crucial for breaking the current degeneracies in velocity-dependent cross-section models.