Gevrey Regularity of Mild Solutions

• The paper demonstrates that Gevrey regularity can be achieved for mild solutions using Duhamel formulations and an induction-on-derivatives approach.
• It employs subelliptic and hypoelliptic estimates alongside macro-micro decomposition to control derivative growth up to factorial rates.
• The approach provides both local and global estimates, ensuring uniform analyticity and smoothing effects in kinetic and fluid systems.

Gevrey regularity of mild solutions refers to the property that mild (i.e., integral/Duhamel-form) solutions to PDEs or kinetic equations exhibit control of derivatives with factorial growth in a generalized sense—placing them in a Gevrey class, a hierarchy between $C^\infty$-smoothness and real-analyticity. The Gevrey class $G^r$ consists of functions whose derivatives grow at most as $(\alpha!)^r$, and for $r=1$ recovers analyticity. The study of Gevrey regularity of mild solutions has been central in kinetic theory, fluid mechanics, and mathematical analysis of PDEs, particularly in quantifying smoothing and regularization effects for equations with rough initial data or degenerate dissipation mechanisms.

1. Functional Framework and Definition of Mild Solution

For kinetic equations such as the non-cutoff Boltzmann equation with hard potentials, the mild solution is constructed as a perturbation around the global Maxwellian equilibrium $\mu(v) = (2\pi)^{-3/2}e^{-|v|^2/2}$. The perturbation $f(t,x,v)$ satisfies the evolution equation

$$\partial_t f + v\cdot\nabla_x f - L f = \Gamma(f,f), \quad f|_{t=0}=f_0,$$

where $L$ is the linearized collision operator and $\Gamma$ encodes the nonlinear quadratic effects of collisions.

A mild solution is any $f$ for which the Duhamel formula holds in the space $L^1_k L^\infty_T L^2_v$, i.e., for each spatial Fourier mode $k \in \mathbb{Z}^3$

$$\hat{f}(t,k) = e^{t(-L - i v \cdot k)} \hat{f}_0(k) + \int_0^t e^{(t-s)(-L - i v \cdot k)} \widehat{\Gamma}(f, f)(s,k)\,ds,$$

with $f$ uniformly bounded in $L^1_k L^\infty([0,T];L^2_v)$. Global existence and uniqueness of such mild solutions for small enough initial data are established for the non-cutoff Boltzmann equation in (Li et al., 20 Jan 2026).

The Gevrey spaces for such systems are defined (cf. Definition 1.1 in (Li et al., 20 Jan 2026)) by the norm

$$f\in\mathcal{G}^r \iff \exists C>0: \forall\text{ multi-indices } \alpha, \beta, \quad \|\partial_x^\alpha \partial_v^\beta f\|_{L^1_k L^2_v} \leq C^{|\alpha|+|\beta|+1}\left[(|\alpha|+|\beta|)!\right]^r.$$

2. Local-in-Time Gevrey Regularity and Sharp Analyticity Radii

For the non-cutoff Boltzmann equation with angular singularity index $0$\tau = \max\{1,1/(2s)\}$, local-in-time estimates give for all $0<t\leq 1$

$$\int_{\mathbb{Z}^3} \sup_{0<t\leq 1} t^{((1+2s)/(2s))|\alpha| + (1/(2s))|\beta|} \|\mathcal{F}_x(\partial_x^\alpha \partial_v^\beta f)(t,k)\|_{L^2_v} d\Sigma(k)$$

$$\leq C^{|\alpha|+|\beta|+1}[(|\alpha|+|\beta|)!]^\tau.$$

This implies that, for each time $t>0$, $f(t,\cdot,\cdot)$ is in the Gevrey class $\mathcal{G}^{\tau}$, with a radius in $x$ of order $\gtrsim t^{(1+2s)/(2s)}$ and in $v$ of order $\gtrsim t^{1/(2s)}$. In the analytic case $\tau = 1$, this matches the standard radius scaling $t^{(1+2s)/(2s)}$ for strong smoothing effects. Such sharp local-in-time Gevrey regularity is obtained using hypoelliptic propagation combined with micro-macro decomposition and induction on derivatives (Li et al., 20 Jan 2026).

3. Global-in-Time Gevrey Radius Estimates

For long times $t \geq 1$, the radius of analyticity/Gevrey class in the spatial variable actually grows. The main global-in-time result states that for each multi-index $\alpha$,

$$\int_{\mathbb{Z}^3} \sup_{t \geq 1} t^{((1+2s)/(2s))|\alpha|} \|\partial_x^\alpha \hat{f}(t,k)\|_{L^2_v} d\Sigma(k) \leq C^{|\alpha|+1}(|\alpha|!)^{(1+2s)/(2s)}.$$

This yields a uniform positive lower bound for the Gevrey radius, precluding loss of analyticity at large times. The mechanism is that the diffusive or fractional hypoelliptic effects dominate in the large time regime, while the nonlinearity is absorbed by the smoothing transforms (Li et al., 20 Jan 2026).

4. Analytic and Gevrey Estimates via Hypoelliptic and Macro-Micro Techniques

The derivation of Gevrey regularity exploits a subelliptic estimate for the linearized Cauchy problem: $$\partial_t h + v\cdot\partial_x h - L h = g, \quad h(0) = h_0.$$ For each Fourier mode, a quantitative hypoelliptic estimate is established, involving a multiplier $M = 1 + c \lambda_k(D_v)$ tuned so that the commutator with $v\cdot k$ creates a positive coercive term of size $k^{2s/(1+2s)}\|h\|_{L^2_v}^2$.

Simultaneously, a macro-micro decomposition is carried out: $f = Pf + (I - P)f,$ where $P$ projects onto the null-space of $L$ (corresponding to local hydrodynamic moments). The macroscopic part $Pf$ satisfies a closed finite-dimensional fluid system, while $(I-P)f$ is controlled by subelliptic coercivity and symbolic hypoelliptic gain.

The nonlinear energy functionals are crafted so that, when summed over all frequencies and differentiated repeatedly (with factorial weights), the commutator and error terms can be absorbed via an induction-on-order scheme, closing the Gevrey estimates both in short time and globally (Li et al., 20 Jan 2026).

5. Relations to Other Models and Broader Context

Gevrey regularity of mild solutions is a robust principle observed across diverse kinetic and fluid systems. In non-cutoff Boltzmann theory, local and global Gevrey regularity is now fully characterized for linear and nonlinear perturbative settings (Zhang et al., 2013, Duan et al., 2021, Li et al., 20 Jan 2026). The precise Gevrey index crucially reflects the angular singularity: for singularity of order $s$, the optimal Gevrey class is $G^{(1+2s)/(2s)}$.

For the Navier–Stokes equations, Gevrey smoothing is established globally for small initial data in various critical spaces, including Triebel–Lizorkin–Lorentz scales, which properly contain the conventional Besov and Lebesgue spaces. The Gevrey indices can be strictly greater than $1$ (i.e., strictly sub-analytic) and capture the decay of high-order derivatives with time (Yang et al., 19 Sep 2025, Danchin, 2023).

Analogous Gevrey propagation results appear for vortex/transport PDEs (supercritical quasi-geostrophic, MHD, Boussinesq, Euler), with adaptively chosen Gevrey norms reflecting linear optimality, hypoellipticity, or subellipticity of the underlying operators (Biswas, 2013, Cheng et al., 2017, Melo et al., 9 Dec 2025).

6. Summary Table of Gevrey Regularity for Mild Solutions (Selected Models)

Equation / Model	Gevrey Index (Optimal)	Key Mechanism / Method
Non-cutoff Boltzmann (hard)	$\max\{1, 1/(2s)\}$	Hypoelliptic subellipticity, macro-micro decomposition (Li et al., 20 Jan 2026)
Navier–Stokes in critical Triebel–Lizorkin–Lorentz	$\sigma > 1$, strictly sub-analytic	Paraproduct/Fourier-multiplier estimates, wavelet and Lorentz techniques (Yang et al., 19 Sep 2025)
Supercritical QG ($0	Near optimal $a<k$	Gevrey commutator, Littlewood–Paley analysis (Biswas, 2013)
Ideal MHD / Euler	$s\ge 1$ (analyticity, Gevrey-$s$)	Fourier-multiplier Gevrey smoothing, Riccati-type control (Cheng et al., 2017)
Fractional Boussinesq	$\sigma>1$	Contraction in Sobolev–Gevrey norm, precise blow-up lower bounds (Melo et al., 9 Dec 2025)

7. Influence and Open Directions

Refined understanding of Gevrey regularity of mild solutions underpins advances in hypoelliptic theory, multiscale analysis, and dissipation enhancement in degenerate kinetic models. Current work builds upon micro-macro analysis, symbolic calculus, and careful commutator estimates to optimize thresholds for both regularity and singularity formation. Quantitative control of the Gevrey radius is central to probing analyticity loss mechanisms, exponential norm inflation in finite-time blow-ups, and the effect of angular singularity in inhomogeneous kinetic theory. Extensions to nonlocal/fluid-structure models, inhomogeneous kinetic equations with non-cutoff kernels, and critical initial data regimes remain active areas of high technical interest (Li et al., 20 Jan 2026, Melo et al., 9 Dec 2025, Avalos et al., 2024).