Linearized Kinetic Theory: Mode Approximation

• Mode approximation in linearized kinetic theory is a method that systematically identifies and classifies relaxation and transport modes emerging from the linearization of kinetic equations, distinguishing between hydrodynamic and kinetic contributions.
• It employs spectral decompositions and analytic dispersion relations to rigorously determine the validity and limitations of finite-mode truncations for modeling macroscopic phenomena.
• This framework enables efficient computation of transport coefficients and provides key insights into the multi-scale relaxation processes governing the crossover between hydrodynamic and kinetic regimes.

Mode approximation in linearized kinetic theory refers to the systematic identification, classification, and analytic (or numerical) computation of relaxation and transport modes resulting from the linearization of kinetic equations (e.g., Boltzmann or BGK models) about equilibrium. The modal structure encodes both hydrodynamic (collective, slow) and non-hydrodynamic (kinetic, fast or continuum) excitations, governing macroscopic evolution and relaxation to equilibrium over a range of length and time scales. Modern mode analysis in kinetic theory spans rigorous spectral decompositions, closed-form analytic solutions in special models (notably relaxation time approaches), and asymptotic or numerical methods for more realistic collision operators. Central themes include the regime of validity of finite-mode (hydrodynamic) truncations, the analytic structure (poles, branch cuts), and the emergence or breakdown of effective equations (such as hydrodynamics) across Knudsen regimes (Hu, 2023, Bajec et al., 2024, Perna et al., 2021, Du et al., 2023, Takamoto et al., 2010, Wu, 2013).

1. Linearized Kinetic Equations and Modal Decomposition

Normal mode analysis begins with linearization of the kinetic equation about global or local equilibrium distributions. For the Boltzmann equation describing a classical one-component gas,

$p \cdot \partial_x f(x, p) = C[f],$

where $C[f]$ is the collision operator, the linearized solution $f = f_\text{eq}(1 + \chi)$ leads, under Fourier transformation, to an evolution equation for mode amplitudes $\chi(k, p)$,

$$(-i\omega + \gamma(p)) \chi + i (v \cdot k) \chi = -L[\chi].$$

Here $v^\mu = p^\mu / E_p$, $\gamma(p)$ is a (possibly energy-dependent) relaxation rate, and $L$ is the linearized collision kernel. For the Anderson–Witting relaxation time approximation (RTA), $L[\chi] = \gamma(p) \chi$, which neglects exact conservation laws; the mutilated RTA amends this by projecting out collision invariants to ensure $L[\psi_n] = 0$ for the five invariants $C[f]$0 (associated with particle number and energy-momentum conservation) (Hu, 2023).

Expanding the kinetic fluctuation in these invariants plus non-collective components, the equation is reduced to a finite-dimensional (hydrodynamic) subspace (mode amplitudes $C[f]$1) and an infinite-dimensional nonhydrodynamic sector.

2. Classification of Modes: Hydrodynamic, Nonhydrodynamic, and Kinetic

The modal structure in linearized kinetic theory divides into:

• Hydrodynamic Modes: Isolated, slow-decaying poles linked to conservation laws. In models with energy-independent relaxation time (AW–RTA), exactly five such poles emerge, corresponding to sound, shear, and heat transport sectors.
• Nonhydrodynamic Modes: A continuum of single-particle excitations populating a branch cut (or, in some models, a set of gapped poles), with decay rates $C[f]$2 and frequencies within $C[f]$3.
• Kinetic Modes (in more refined treatments): Further structure within the nonhydrodynamic sector, including fast kinetic roots and, in high-order moment or nonlocal collision kernels, additional weakly damped modes.

For energy-independent $C[f]$4, the separation between hydrodynamic poles and the kinetic continuum is sharp. For energy-dependent relaxation, the spectrum spreads into a region filling all $C[f]$5, obscuring the hydrodynamic/kinetic distinction (“dehydrodynamization”) (Hu, 2023).

3. Dispersion Relations and Analytical Structures

The secular equations (determinantal equations for collective mode existence) in the AW–RTA enable closed-form solutions for the poles: $C[f]$6 with the variables $C[f]$7, $C[f]$8, and $C[f]$9 (Hu, 2023, Bajec et al., 2024, Perna et al., 2021).

For small $f = f_\text{eq}(1 + \chi)$0, hydrodynamic dispersion relations follow: $f = f_\text{eq}(1 + \chi)$1 with $f = f_\text{eq}(1 + \chi)$2 for a massless gas. Nonhydrodynamic roots (gapped or branch-cut) saturate at decay rate $f = f_\text{eq}(1 + \chi)$3 for large $f = f_\text{eq}(1 + \chi)$4, with real parts growing linearly ($f = f_\text{eq}(1 + \chi)$5) (Takamoto et al., 2010, Perna et al., 2021).

Correlation functions (Green’s functions) in these models exhibit a characteristic structure of isolated poles plus branch cuts between $f = f_\text{eq}(1 + \chi)$6, encoding the presence of both collective excitations and a continuum of kinetic relaxation modes (Bajec et al., 2024, Wu, 2013).

4. Regimes of Validity, Onset Transitions, and Dehydrodynamization

Finite-mode (hydrodynamic) approximations are asymptotically valid for small wave numbers $f = f_\text{eq}(1 + \chi)$7 (long-wavelength regime). As $f = f_\text{eq}(1 + \chi)$8 increases, hydrodynamic poles move toward and eventually merge with the nonhydrodynamic continuum—the “onset transition.” This threshold is sharply quantified: e.g., for sound modes, hydrodynamics persists only for $f = f_\text{eq}(1 + \chi)$9, beyond which the isolated pole ceases to exist and is absorbed by the continuum (Hu, 2023). Similar thresholds hold for heat and shear.

In models with energy-dependent relaxation, there is no spectral gap separating hydrodynamic from kinetic modes: nonhydrodynamic excitations always overlap or surpass any given hydrodynamic pole in decay rate, even at arbitrarily small $\chi(k, p)$0. The loss of coherent pole dominance—a scenario termed “dehydrodynamization”—implies that truncation to finitely many modes is not consistent, and the full kinetic spectrum must be retained (Hu, 2023).

5. Comparisons: Moment Expansions, Hydrodynamics, and Universality

Moment truncations (e.g., Grad’s or Chapman–Enskog expansions) admit mode-approximation by projecting onto a finite-dimensional subspace. For small $\chi(k, p)$1, the leading-order modal content exactly reproduces Navier-Stokes (diffusive) behavior. However, these truncations fail at large $\chi(k, p)$2, where kinetic theory’s full spectrum is essential to capture the crossover to kinetic transport (Knudsen regime) (Takamoto et al., 2010, Perna et al., 2021, Wu, 2013).

Second-order (and higher) hydrodynamic theories such as Israel–Stewart (IS) hydrodynamics mimic the main hydrodynamic poles and predict relaxation modes (at $\chi(k, p)$3), but generically lack the kinetically-derived propagating tensor modes and full cut structure revealed by the kinetic analysis (Perna et al., 2021). Mode analysis reveals a surprising universality: when properly rescaled, hydrodynamic sound mode frequency and attenuation agree quantitatively across distinct weakly-coupled kinetic theories (RTA, scalar $\chi(k, p)$4, SU(3) Yang-Mills, QCD), far beyond the naïve hydrodynamic regime, up to $\chi(k, p)$5 (Du et al., 2023).

6. Modal Decomposition on Bounded Domains and Pointwise Construction

On bounded domains (e.g., torus), the mode approximation strategy decomposes solutions into spatial Fourier modes. Long-wavelength modes are dominated by five hydrodynamic eigenvalues (with explicit decay rates), while kinetic-like “microscopic” waves are constructed by iterative application of the damped transport operator and higher-order corrections via Picard iteration. The “mixture lemma” provides rigorous bounds, ensuring controlled transfer of regularity between velocity and position variables (Wu, 2013).

The pointwise solution structure is thus

$\chi(k, p)$6

where the first term sums the five fluid modes, the second represents $\chi(k, p)$7 dominant kinetic modes, and the final term is an exponentially decaying residual.

7. Algorithmic and Physical Implications

From an algorithmic perspective, mode approximation provides the foundation for both analytic understanding and practical computation in kinetic theory. In RTA-type models, exact modal decompositions enable efficient evaluation of Green’s functions and explicit expressions for transport coefficients (shear viscosity, diffusion, etc.) (Bajec et al., 2024, Takamoto et al., 2010). In numerical simulations, modal truncations or low-rank expansions trade off between efficiency and physical fidelity: the choice of truncation governs which macroscopic and microscopic relaxation processes are captured (Wu, 2013).

Physically, the mode structure governs the emergence and eventual breakdown of hydrodynamics, the influence of Knudsen-number effects, and the role of fast kinetic processes in phenomena such as “hydrodynamization” and “dehydrodynamization.” The mode analysis framework generalizes across weak and strong coupling, with the latter typically exhibiting only finitely many poles (cf. holographic CFTs), in contrast to the kinetic cutoff (Bajec et al., 2024).

---

Key model features and parameters in mode analysis:

Kinetic Model	Hydrodynamic Modes	Nonhydrodynamic Structure
AW–RTA, constant τ_R	5 isolated poles (sound, shear, heat)	Branch cut in Im ω = –γ_s
Mutilated RTA, energy-dep. τ_R	Poles + strip of cuts; blurred hydrodynamic/kinetic split	Continuum for all Im ω < 0
BGK (nonrelativistic)	Shear, sound, thermal	Fast kinetic roots at –i/τ
IS/Second-order hydro	Sound, shear, relaxation mode	No propagating tensor pole

The mode approximation in linearized kinetic theory thus provides a comprehensive lens onto the multi-scale relaxation and transport properties of kinetic systems, offering analytic clarity and practical tools for probing the limitations of hydrodynamics and the full content of kinetic formulations (Hu, 2023, Bajec et al., 2024, Perna et al., 2021, Takamoto et al., 2010, Du et al., 2023, Wu, 2013).