Mean-field derivation of Landau-like equations
José Antonio Carrillo, Shuchen Guo, Pierre-Emmanuel Jabin
TL;DR
The paper establishes a rigorous mean-field derivation of a class of space-homogeneous Landau-like equations from stochastic particle systems on the torus. Using a relative-entropy framework, it derives a quantitative bound $H_N(t) \le (H_N(0) + C_1/N) e^{C_2 t}$ between the $N$-particle Liouville solution and the tensorized limit, which in turn yields a rate $\|f_{k,N}-f^{\otimes k}\|_{L^{\infty}([0,T],L^1(\mathbb{T}^{dk}))} \le C_3/\sqrt{N}$ for the $k$-particle marginals. This work extends propagation-of-chaos results to Landau-like dynamics via entropy methods, providing explicit, scalable convergence rates. The approach hinges on ellipticity of $a$, the structure $b=\nabla\cdot a$, and a cancellation lemma to control nonlinear interaction terms, with Grönwall-type arguments giving the final quantitative bounds.
Abstract
We derive a class of space homogeneous Landau-like equations from stochastic interacting particles. Through the use of relative entropy, we obtain quantitative bounds on the distance between the solution of the N-particle Liouville equation and the tensorised solution of the limiting Landau-like equation.
