Stability Estimates in Kinetic Wasserstein Distances for the Vlasov–Poisson System with Yudovich Density

Jonathan Junné, Alexandre Rege · 2025

Read Stability Estimates in Kinetic Wasserstein Distances for the Vlasov–Poisson System with Yudovich Density by Jonathan Junné, Alexandre Rege (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-07122.txt.

Introduction

Jonathan Junné and Alexandre Rege prove stability for the Vlasov-Poisson system under rough initial densities, in a 2025 paper. The system governs plasmas and galaxies through a cloud of particles and the field they generate. Measuring how two solutions drift apart with kinetic Wasserstein distances gives uniqueness and control, but earlier results demanded nicely bounded densities.

The authors widen the door to Yudovich-class densities, a rough but structured family. First they bound the gap between two force fields by a crafted quantity tied to how far apart the densities sit. That bound feeds an Osgood-type inequality for the gap between solutions, which closes the proof. Stability now holds for every finite Wasserstein order, joining earlier split results.

Catalog record

CREATORJonathan Junné, Alexandre Rege
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE71 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-07122

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Stability Estimates in Kinetic Wasserstein Distances for the Vlasov–Poisson System with Yudovich Density Jonathan Junné, Alexandre Rege Abstract . We investigate the stability of solutions to the Vlasov–Poisson system using the unifying framework of the kinetic Wasserstein distance, introduced by Iacobelli in (Section 4 in Arch. Ration. Mech. Anal. 244 (2022), no. 1, 27-50). This allows us to treat both macroscopic densities that lie in a Yudovich space, as recently considered by Crippa et al. (Theorem 1.6 in Nonlinearity 37 (2024), no. 9, 095015) for the 1 1 -Wasserstein distance, and higher order Wasserstein distances, for which only bounded macroscopic densities were treated by Iacobelli and the first author (Theorem 1.11 in Bull. Lond. Math. Soc. 56 (2024), 2250-2267). First, we establish an L p L^{p} -estimate on the difference between two force fields in terms of a suitable nonlinear quantity that controls the kinetic Wasserstein distance between their macroscopic densities. Second, we use this estimate in order to derive a closable Osgood-type inequality for the kinetic Wasserstein distance between two solutions. This enables us to prove our main theorem; for 1 ≤ p < + ∞ 1\leq p<+\infty we show the p p -Wasserstein stability of solutions to the Vlasov–Poisson system with macroscopic densities belonging to a Yudovich space. 2020 Mathematics Subject Classification 35Q83, 82C40, 82D10, 35B35 1. Introduction 1.1. The Vlasov–Poisson system We consider the Vlasov–Poisson system for electrostatic or gravitational interaction. This system describes the dynamics of interacting particles, which are modelled by a distribution function f\equiv f(t,x,v) depending on time t\in\mathbb{R}^{+} , position x\in\mathcal{X} denoting either the d -dimensional torus \mathbb{T}^{d} or the Euclidean space \mathbb{R}^{d} , and velocity v\in\mathbb{R}^{d} . It is given by the following set of equations: (VP) \begin{cases}\partial_{t}f+v\cdot\nabla_{x}f-\nabla U_{f}\cdot\nabla_{v}f=0,\\ \sigma\Delta U_{f}=\rho_{f}-1_{\mathcal{X}=\mathbb{T}^{d}},\\ \rho_{f}:=\int_{\mathbb{R}^{d}}f\>dv,\end{cases} where \sigma=\pm 1 encodes either gravitational ( \sigma=1 ) or electrostatic ( \sigma=-1 ) interaction and U_{f}\equiv U_{f}(t;x) is the self-induced gravitational or electrostatic potential. Throughout this paper, we will consider two Lagrangian weak solutions f_{1} and f_{2} to ( VP ) with respective initial data f_{1}(0),f_{2}(0) , which are probability distributions, and respective flows Z_{1}:=(X_{1},V_{1}) and Z_{2}:=(X_{2},V_{2}) . These flows satisfy the system of characteristics (CH-VP) \dot{X}=V,\quad\dot{V}=-\nabla U_{f},\quad X(0;x,v)=x,\quad V(0;x,v)=v. Lagrangian weak solutions are weak solutions that can be written as pushforwards of the initial data by flows solving ( CH-VP ); in this case f_{1}(t)=Z_{1}(t;\cdot,\cdot)_{\#}f_{1}(0) and f_{2}(t)=Z_{2}(t;\cdot,\cdot)_{\#}f_{2}(0) . For notational convenience, unless otherwise stated, all quantities are always evaluated at time t and we denote X_{1}:=X_{1}(t;x,v),\,V_{1}:=V_{1}(t;x,v),\,X_{2}:=X_{2}(t;y,w),\,V_{2}:=V_{2}(t;y,w),\,\nabla U_{f_{i}}(X_{j}):=\nabla U_{f_{i}}(t;X_{j}) for i,j=1,2 . We investigate the stability of solutions to ( VP ) in the Wasserstein distances of order p\geq 1 , which we recall on the phase space \mathcal{X}\times\mathbb{R}^{d} for convenience (see [ 20 , Chapter 6] ): Definition 1.1 (Wasserstein distances) . Let \mu,\nu be two probability measures on \mathcal{X}\times\mathbb{R}^{d} . The Wasserstein distance of order p\geq 1 , between \mu and \nu is defined as W_{p}(\mu,\nu):=\left(\inf_{\pi\in\Pi(\mu,\nu)}\left\{\int_{(\mathcal{X}\times\mathbb{R}^{d})^{2}}\left\lvert x-y\right\rvert^{p}+\left\lvert v-w\right\rvert^{p}\>d\pi(x,v,y,w)\right\}\right)^{\frac{1}{p}}, where \Pi(\mu,\nu) is the set of couplings ; that is, the set of probability measures with marginals \mu and \nu . A coupling is said to be optimal if it minimizes the Wasserstein distance. We begin by mentioning the major contribution of Loeper [ 15 , Theorem 1.2] , where the uniqueness and stability of the Vlasov–Poisson system ( VP ), assuming the macroscopic density remains bounded, were established using an optimal transport argument involving the 2-Wasserstein distance. Loeper’s main observation was that the L^{2} -norm of the difference of force fields can be controlled in terms of the 2-Wasserstein distance between their macroscopic densities [ 15 , Theorem 2.9] . The uniqueness criterion was subsequently improved by Miot [ 16 , Theorem 1.1] relying on the second order structure of the characteristic ODE ( CH-VP ) to allow for macroscopic densities satisfying \max_{t\in[0,T]}\sup_{r\geq 1}\frac{\left\lVert\rho_{f}(t)\right\rVert_{L^{r}(\mathbb{R}^{d})}}{r}<+\infty, and succeeded in highlighting unbounded densities that satisfy this condition [ 16 , Theorem 1.2] . These results were later generalized, together with Holding [ 10 , Theorem 1.1] , to provide an effective W_{1} stability estimate allowing for exponential Orlicz macroscopic densities verifying \max_{t\in[0,T]}\sup_{r\geq\alpha>1}\frac{\left\lVert\rho_{f}(t)\right\rVert_{L^{r}(\mathbb{R}^{d})}}{r^{1/\alpha}}<+\infty. Finally, in a very recent contribution [ 4 ] , Crippa, Inversi, Saffirio and Stefani established a W_{1} stability estimate for macroscopic densities within uniformly-localized Yudovich spaces [ 4 , Theorem 1.6] . Their Lagrangian approach was inspired by a previous work of Crippa and Stefani [ 3 ] on the well-posedness of the two-dimensional Euler equations. Similarly to Miot’s work, by using the propagation of moments available for ( VP ), they provided explicit examples of unbounded macroscopic densities in this uniformly localized Yudovich space [ 4 , Theorem 1.7 & Proposition 1.8] . Compared to Loeper’s approach, the aforementioned W_{1} stability estimates not only accommodate a broader class of macroscopic densities for which the uniqueness of ( VP ) is guaranteed, but also provide improved control over the time interval during which two initially close solutions remain close. Generally, the 1 -Wasserstein distance is more flexible, and, the W_{1} stability estimates obtained so far do not rely on optimal transport theorems, whereas Loeper’s stability estimate in the stronger 2 -Wasserstein distance heavily depends on them. To improve control over the time interval in Loeper’s stability estimate and obtain an effective W_{2} stability estimate [ 11 , Theorem 3.1] , Iacobelli introduced the kinetic Wasserstein distance in [ 11 ] , specifically designed to account for the anisotropy between position and momentum that is inherent to kinetic problems (see [ 11 , Section 4] ): Definition 1.2 (kinetic Wasserstein distance) . Let \mu,\nu be two probability measures on \mathcal{X}\times\mathbb{R}^{d} . The kinetic Wasserstein distance of order p , with p\geq 1 , between \mu and \nu is defined as W_{\lambda,p}(\mu,\nu):=\left(\inf_{\pi\in\Pi(\mu,\nu)}\left\{D_{p}(\pi,\lambda)\right\}\right)^{\frac{1}{p}}, where D_{p}(\pi,\lambda) is the unique number s such that s-\lambda(s)\int_{(\mathcal{X}\times\mathbb{R}^{d})^{2}}\left\lvert x-y\right\rvert^{p}\>d\pi(x,v,y,w)=\int_{(\mathcal{X}\times\mathbb{R}^{d})^{2}}\left\lvert v-w\right\rvert^{p}\>d\pi(x,v,y,w), with \lambda:\mathbb{R}^{+}\to\mathbb{R}^{+} a decreasing function. The kinetic Wasserstein distance allowed Iacobelli to obtain the quasi-neutral limit of ( VP ) towards the kinetic isothermal Euler system allowing for W_{2} perturbations of analytic data with almost optimal rate [ 11 , Corollary 3.3 & Remark 3.4] , as a consequence of the enhanced W_{2} stability estimate. Subsequently, with Griffin-Pickering, they obtained the stability of the quasi-neutral limit for the ionic with exponentially small Wasserstein perturbations [ 9 , Theorem 1.1] (see [ 8 ] for a recent account on the subject). The kinetic Wasserstein distance approach also allowed Iacobelli and Lafleche [ 13 , Theorem 2.4] to improve the convergence rate in the semi-classical analogue to the 2 -Wasserstein distance of the semi-classical limit of the quantum Hartree system towards the classical ( VP ). Regarding improved stability estimates, we also mention a recent work of Iacobelli and the first author [ 12 , Theorems 1.9 & 1.11] where Loeper’s approach and the kinetic Wasserstein approach were generalized to the p -Wasserstein setting for 1<p<+\infty and of the second author [ 18 , Theorem 2.1] dealing with the stability of solutions for the magnetized Vlasov–Poisson system ( VPB ), where the kinetic Wasserstein distance was necessary to control the additional terms coming from the varying external magnetic field. In both works, the controls on the time interval match the effective stability estimate in the weaker 1 -Wasserstein distance obtained by Holding and Miot [ 10 , 11

Read the complete text