Feedback-Driven Core Formation

Updated 25 May 2026

Feedback-driven core formation is a process where repeated, bursty energy injections from supernova feedback irreversibly heat dark matter particles and reduce central densities.
Resonance models and N-body simulations show that impulsive potential fluctuations can redistribute dark matter orbits, creating measurable core radii in low-mass galaxies.
Key factors such as coupling efficiency, burst timing, and scaling relations link observable kinematic and structural signatures to the underlying feedback physics resolving the core–cusp problem.

Feedback-driven core formation refers to a class of physical mechanisms and theoretical models in which repeated or continuous energy input from baryonic processes causes the transformation of an initially cuspy dark matter (DM) halo into a cored one. While the term encompasses a variety of feedback channels, the most prominent astrophysical realization is in low-mass galaxies via bursty supernova (SN) feedback, which can drive potential fluctuations that irreversibly redistribute DM orbits and reduce the central density. The topic also connects to more general feedback-driven transformations in non-ΛCDM contexts and underpins efforts to reconcile observed inner DM profiles with predictions from collisionless CDM N-body simulations.

1. Dynamical Mechanisms: Supernova Feedback and Resonant Heating

In the canonical feedback-driven scenario, rapid energy injection from repeated starbursts heats and expels gas from the central region of a galaxy. As the gas cools and reaccretes, the cycle generates time-dependent oscillations in the gravitational potential. When the frequency of these oscillations matches the local dynamical (orbital) timescale of DM particles, a resonance is established. DM particles at the resonant radius absorb kinetic energy from the varying potential, causing them to migrate to larger orbits and reducing the central DM density (Ogiya et al., 2012).

Analytically, the external potential’s time variation is decomposed into Fourier modes. The resonance condition,

$k v_0 \approx n \Omega,$

matches the pattern speed of the potential ( $\Omega = 2\pi/T$ ) to the natural frequencies of DM particle orbits at radius $r$ [ $T$ is the oscillation period, $n$ an integer]. The critical regime occurs when $T \approx t_{\rm dyn}(r)$ , with $t_{\rm dyn}(r) = \sqrt{\frac{3\pi}{16 G \bar{\rho}(r)}}$ , where $\bar{\rho}(r)$ is the mean enclosed density. This sets the core formation radius.

Numerical $N$ -body simulations validate the resonance model: in halos subjected to oscillating baryonic potentials with periods $T \sim t_{\rm dyn}$ , cusps are rapidly transformed into cores, with the simulated core radius agreeing to within $\Omega = 2\pi/T$ 0 with analytic resonance predictions (Ogiya et al., 2012).

2. Impulsive Versus Continuous Core Formation: Timescales and Feedback Recipes

Feedback-driven DM core formation can be realized through both impulsive (rapid, bursty) and more continuous, adiabatic processes. In isolated dwarf systems, effective core formation is associated with highly bursty star formation histories, yielding impulsive potential changes. Each burst must inject energy on timescales shorter than the local dynamical time to produce a non-adiabatic response in the DM (Teyssier et al., 2012, Burger et al., 2021).

Key features of this regime include:

Bursty star formation with peak-to-trough SFR ratio $\Omega = 2\pi/T$ 15–10 and burst intervals $\Omega = 2\pi/T$ 2, resulting in order-unity gas mass fluctuations in the central kpc.
Impulsive energy injection drives an irreversible increase in DM particle energies, flattening the central cusp to a core.
The process is cumulative: successive episodes progressively increase the core radius, consistent with kinetic energy transfer models.

In contrast, continuous or slowly varying feedback (or non-bursty star formation) fails to provide the necessary impulsive kicks; the potential evolves adiabatically, the DM orbits adjust reversibly, and the cusp remains intact (Burger et al., 2021).

3. Scaling Relations, Core Predictions, and Required Energetics

Analytical prescriptions connect the macroscopic energetics of feedback to the resultant core properties. For a halo of fixed mass, the core radius scales with the available feedback energy and the impulsiveness of its deposition. The resonance model yields:

$\Omega = 2\pi/T$ 3

with appropriate choices of halo structural parameters (for NFW: $\Omega = 2\pi/T$ 4) (Ogiya et al., 2012).

Energetic requirements are calibrated empirically: only a small fraction $\Omega = 2\pi/T$ 5 of available SN energy needs to couple dynamically to the DM in order to transform a cusp into a core in nearby SPARC galaxies (Shinozaki et al., 20 Jan 2026). The critical stellar mass needed for core formation in a halo of mass $\Omega = 2\pi/T$ 6 is:

$\Omega = 2\pi/T$ 7

with $\Omega = 2\pi/T$ 8 correction factors reflecting halo profile shape (Shinozaki et al., 20 Jan 2026).

A forbidden region in the $\Omega = 2\pi/T$ 9– $r$ 0 plane emerges: for $r$ 1 or $r$ 2, SN feedback cannot overcome the depth of the halo potential, explaining why ultra-faint dwarfs and clusters remain cuspy. Core formation is optimal in halos of $r$ 3– $r$ 4 (Shinozaki et al., 20 Jan 2026).

4. Observational Consequences, Structural and Kinematic Signatures

Feedback-driven core formation predicts a suite of observable features:

DM Cores: Simulated dwarfs develop inner profiles well-fit by pseudo-isothermal distributions, with core radii $r$ 5500–1000 pc and densities $r$ 60.05–0.1 $r$ 7 pc $r$ 8 (Teyssier et al., 2012).
Hot Stellar Distributions: The stellar $r$ 9 ratio (rotation to velocity dispersion) evolves to $T$ 0 (thick, hot spheroids), contrasting with cold disks ( $T$ 1) in non-feedback models (Teyssier et al., 2012).
Core–core mapping in stellar populations: Extended, kpc-sized stellar cores inherent to the DM potential expansion arise naturally if the initial stellar density slope is shallow ( $T$ 2), matching observations of low-surface-brightness systems (Almeida et al., 3 Sep 2025).
Globular Cluster Distributions: The presence and survival of large GCs (effective radius $T$ 3 pc) with large scatter in size in dwarfs require a cored host, as tidal effects and dynamical friction in a cusp would otherwise destroy or centralize them (Orkney et al., 2019).
Bursty SFH observable via CMDs, HI mapping, integral-field stellar spectroscopy, and comparison of circular-velocity curves provide probes for the impulsive regime and DM core formation (Teyssier et al., 2012).
Universal Scaling Laws: The core–cusp transformation models naturally reproduce the observed near-constant central surface density relation $T$ 4 pc $T$ 5 and the Strigari mass plateau $T$ 6 (Ogiya et al., 2013). The central density traces the formation redshift.

5. Feedback-Driven Cores Beyond Standard Baryonic Processes

Nonstandard feedback-driven scenarios extend this framework:

Self-Interacting Dark Matter (SIDM): In models with $T$ 7– $T$ 8 cm $T$ 9/g, central cores arise from collisional heating and thermalization; the process is continuous and adiabatic, yielding isothermal central velocity profiles (Burger et al., 2021).
Late-Time Annihilating DM: Reactivation of DM annihilation via late-time χ–χ̄ oscillations can flatten the cusp, forming a core whose size is determined by the annihilation rate and energy deposition timescale. This mechanism is distinct from elastic SIDM and can operate in baryon-poor environments (Cline et al., 2020).
Non-CDM and Adiabatic/Slow Core Formation: Some non-CDM models (e.g., thermalization of warm or fuzzy DM) predict gradual core growth, with the expansion of the DM potential adiabatically dragging preexisting stars outward, forming extended stellar cores (Almeida et al., 3 Sep 2025).

6. Distinguishing Feedback Mechanisms, Degeneracies, and Observational Diagnostics

Core formation driven by feedback displays degeneracies with non-baryonic processes. For fixed core radii, both bursty SN feedback and SIDM can reproduce flat central profiles and similar circular-velocity curves. However, they produce different structural and kinematic imprints:

SIDM: Generates spatially extended, isothermal stellar and gas distributions with shallow or negative stellar age gradients.
Impulsive Feedback: Produces hot, centrally concentrated stellar populations, positive stellar age gradients, and non-isothermal gas kinematics (Burger et al., 2021).

Discriminating between these scenarios requires joint analysis of galaxy size, gas kinematics, age-metallicity gradients, and resolved phase-space distributions at high spatial and spectral resolution.

Bayesian reliability analysis informs on which galaxies (e.g., low surface density, extended RCs) offer robust core–cusp discrimination given current data limitations; only $n$ 021 out of 128 SPARC galaxies provide $n$ 175% reliable cusp-versus-core inference (Manju et al., 2023).

7. Theoretical, Computational, and Model-Building Implications

Core formation by feedback-driven mechanisms critically constrains galaxy formation models. The empirically inferred feedback-to-DM energy coupling efficiency ( $n$ 2) sets a calibration for subgrid feedback prescriptions in simulations and demarcates the mass scales for core viability (Shinozaki et al., 20 Jan 2026).

Simulation convergence depends on adopting high-density thresholds for star formation ( $n$ 3 cm $n$ 4), spatial resolution sufficient to avoid artificial gas concentrations and contraction, and feedback recipes that yield sufficiently bursty SFHs in sub- $n$ 5 halos (Dutton et al., 2020).

Open issues include resolving observed diversity in core properties at fixed mass, mapping the role of initial orbital structure and baryonic microphysics, and unbiased inference of DM profiles amid baryon-dominated central regions.

Summary Table: Core Formation Regimes and Diagnostics

Mechanism	Physical Trigger	Timescale (Impulsive/Continuous)	Kinematic Signature	Core Size Constraints
Supernova feedback (bursty)	Recurrent, fast SN-driven gas outflows	Impulsive ( $n$ 6)	$n$ 7, positive age gradient	$n$ 80.5–1 kpc, set by resonance condition
Self-interacting DM (SIDM)	Elastic DM self-scattering	Continuous (Adiabatic)	Isothermal $n$ 9, shallow/negative age gradient	$T \approx t_{\rm dyn}(r)$ 01 kpc for $T \approx t_{\rm dyn}(r)$ 11 cm $T \approx t_{\rm dyn}(r)$ 2/g
Late-time DM annihilation	$T \approx t_{\rm dyn}(r)$ 3– $T \approx t_{\rm dyn}(r)$ 4 oscillation-induced annihilation	Impulsive	Cooled, depleted central regions	$T \approx t_{\rm dyn}(r)$ 51–3 kpc (dwarfs), up to 200 kpc (clusters) depending on DM parameters
Globular cluster crossings	Repeated GC–halo encounters	Impulsive ( $T \approx t_{\rm dyn}(r)$ 6)	Enlarged GC radii, debris	$T \approx t_{\rm dyn}(r)$ 7100–400 pc with repeated crossings
Adiabatic DM core growth	Thermalization (non-CDM or slow feedback)	Slow (Adiabatic)	Extended, isotropic stellar cores	$T \approx t_{\rm dyn}(r)$ 8– $T \approx t_{\rm dyn}(r)$ 9

Feedback-driven core formation provides a physically grounded, simulation-tested solution to the core–cusp problem over a broad range of galaxy masses. The key ingredients—rapid, bursty feedback with sufficient energy coupling, resonance or impulsivity matching the local dynamical time, and cumulative heating—govern both the efficacy and structural signatures of core creation in cosmological and isolated galaxy contexts (Ogiya et al., 2012, Teyssier et al., 2012, Burger et al., 2021, Shinozaki et al., 20 Jan 2026).