Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Gradient estimates for nonlinear kinetic Fokker-Planck equations

Kyeongbae Kim, Ho-Sik Lee, Simon Nowak

TL;DR

The paper develops a comprehensive gradient regularity theory for nonlinear kinetic Fokker-Planck equations of the form $\partial_t f+v\cdot \nabla_x f-\nabla_v\cdot a(t,x,v,\nabla_v f)=\mu-\nabla_v\cdot G$, by translating nonlinear potential theory into the kinetic setting. It introduces kinetic Riesz potentials and fractional maximal functions to obtain sharp, pointwise gradient estimates for $\nabla_v f$, and proves gradient Hölder regularity for homogeneous problems with constant coefficients, along with robust comparison and decay estimates to handle general data. The results include Dini-continuous and Lorentz-space data criteria, Calderón–Zygmund-type $L^q$ bounds, and VMO/BMO-type control, applicable to both nondivergence and divergence data. A key novelty is extending gradient potential estimates to nonlinear, hypoelliptic kinetic equations, including new spatial differentiability mechanisms and a nonlinear atomic decomposition to manage the $x$-dependence. Collectively, the work advances fine gradient regularity for kinetic equations, with potential implications for kinetic theory models and related PDEs such as the Landau equation.

Abstract

In this work, we provide a comprehensive gradient regularity theory for a broad class of nonlinear kinetic Fokker-Planck equations. We achieve this by establishing precise pointwise estimates in terms of the data in the spirit of nonlinear potential theory, leading to fine gradient regularity results under borderline assumptions on the data. Notably, our gradient estimates are novel already in the absence of forcing terms and even for linear kinetic Fokker-Planck equations in divergence form.

Gradient estimates for nonlinear kinetic Fokker-Planck equations

TL;DR

The paper develops a comprehensive gradient regularity theory for nonlinear kinetic Fokker-Planck equations of the form , by translating nonlinear potential theory into the kinetic setting. It introduces kinetic Riesz potentials and fractional maximal functions to obtain sharp, pointwise gradient estimates for , and proves gradient Hölder regularity for homogeneous problems with constant coefficients, along with robust comparison and decay estimates to handle general data. The results include Dini-continuous and Lorentz-space data criteria, Calderón–Zygmund-type bounds, and VMO/BMO-type control, applicable to both nondivergence and divergence data. A key novelty is extending gradient potential estimates to nonlinear, hypoelliptic kinetic equations, including new spatial differentiability mechanisms and a nonlinear atomic decomposition to manage the -dependence. Collectively, the work advances fine gradient regularity for kinetic equations, with potential implications for kinetic theory models and related PDEs such as the Landau equation.

Abstract

In this work, we provide a comprehensive gradient regularity theory for a broad class of nonlinear kinetic Fokker-Planck equations. We achieve this by establishing precise pointwise estimates in terms of the data in the spirit of nonlinear potential theory, leading to fine gradient regularity results under borderline assumptions on the data. Notably, our gradient estimates are novel already in the absence of forcing terms and even for linear kinetic Fokker-Planck equations in divergence form.

Paper Structure

This paper contains 19 sections, 38 theorems, 314 equations.

Table of Contents

  1. Introduction
  2. Nonlinear kinetic Fokker-Planck equations
  3. Schauder-type velocity gradient regularity for homogeneous kinetic Fokker-Planck equations
  4. Setting
  5. Fine gradient regularity
  6. Related previous results
  7. Technical approach and novelties
  8. Outline
  9. Preliminaries
  10. Covering-type lemmas
  11. Embeddings via fractional maximal functions and Riesz potentials
  12. Technical lemmas
  13. Zero-order regularity results for solutions to kinetic equations
  14. Gradient Hölder regularity for homogeneous nonlinear kinetic equations with constant coefficients
  15. Comparison estimates
  16. ...and 4 more sections

Key Result

theorem 1

Let $f$ be a weak solution of where $a$ satisfies Assumption assump. Then there exists some $\alpha=\alpha(n,\Lambda)\in(0,1)$ such that if $a$ is Hölder continuous with exponent $\beta \in (0,\alpha)$ in $W \times V$ in the sense of Definition def:DH, then $\nabla_v f\in C^{\beta}_{\textnormal{kin}}(W \times V)$.

Theorems & Definitions (77)

  • theorem 1: Gradient regularity for nonlinear homogeneous kinetic equations
  • remark 1: Autonomous case
  • theorem 2: Gradient regularity for linear homogeneous kinetic equations
  • definition 1: Weak solutions
  • remark 2
  • definition 2: Kinetic Hölder spaces
  • definition 3: Dini and Hölder coefficients
  • theorem 3: Kinetic gradient potential estimates
  • remark 3: Consistency with parabolic case
  • corollary 1: Gradient continuity via potentials
  • ...and 67 more