$$\begin{array}{ccc}{sup}_{t\in [0,T[}\parallel \rho {\parallel}_{{L}^{\infty}\left({\mathbb{R}}^{d}\right)}<+\infty .& & \end{array}$$ | (6) |
1 Introduction
$$\begin{array}{ccc}& & {\partial}_{t}f+\xi \cdot {\nabla}_{x}f+\nabla \Psi \cdot {\nabla}_{\xi}f=0,\end{array}$$ | (1) |
$$\begin{array}{ccc}& & \Delta \Psi =\epsilon \rho ,\end{array}$$ | (2) |
$$\begin{array}{ccc}\Psi (t,x)=-\epsilon {\int}_{{\mathbb{R}}^{3}}\rho (t,y)\frac{1}{4\pi |x-y|}dy.& & \end{array}$$ | (3) |
$$\begin{array}{ccc}f{|}_{t=0}={f}^{0}.& & \end{array}$$ | (4) |
$$\begin{array}{ccc}\int f({\partial}_{t}\phi +\xi \cdot {\nabla}_{x}\phi +{\nabla}_{x}\Psi \cdot {\nabla}_{\xi}\phi )=-{\int}_{{\mathbb{R}}^{6}}{f}^{0}\phi {|}_{t=0},& & \end{array}$$ | (5) |
$$\begin{array}{ccc}{sup}_{t\in [0,T[}\parallel \rho {\parallel}_{{L}^{\infty}\left({\mathbb{R}}^{d}\right)}<+\infty .& & \end{array}$$ | (6) |
$$\begin{array}{ccc}& & {\partial}_{t}\rho +\nabla \cdot \left(\rho v\right)=0,\end{array}$$ | (7) |
$$\begin{array}{ccc}& & {\partial}_{t}\left(\rho v\right)+\nabla \cdot (\rho v\otimes v)=\rho \nabla \Psi ,\end{array}$$ | (8) |
$$\begin{array}{ccc}& & \Delta \Psi =\epsilon \rho .\end{array}$$ | (9) |
2 Preliminary results on optimal transportation and Wasserstein distances
$$\begin{array}{ccc}{W}_{2}({\rho}_{1},{\rho}_{2})={\left({inf}_{\pi}{\int}_{{\mathbb{R}}^{d}\times {\mathbb{R}}^{d}}\pi (x,y)|x-y{|}^{2}\right)}^{\frac{1}{2}},& & \end{array}$$ |
$$\begin{array}{ccc}& & {\partial}_{\theta}{\rho}_{\theta}+\nabla \cdot \left({\rho}_{\theta}{v}_{\theta}\right)=0,\end{array}$$ |
$$\begin{array}{ccc}& & \rho {|}_{\theta =1}={\rho}_{1},\rho {|}_{\theta =2}={\rho}_{2},\end{array}$$ |
$$\begin{array}{ccc}& & {\int}_{{\mathbb{R}}^{d}}\left|{v}_{\theta}\right(x\left){|}^{2}d{\rho}_{\theta}\right(x)\equiv {W}_{2}^{2}({\rho}_{1},{\rho}_{2})\forall \theta \in [1,2].\end{array}$$ |
$$\begin{array}{ccc}\forall \theta \in [1,2],\parallel {\rho}_{\theta}{\parallel}_{{L}^{\infty}}\le max\{\parallel {\rho}_{1}{\parallel}_{{L}^{\infty}},\parallel {\rho}_{2}{\parallel}_{{L}^{\infty}}\}.& & \end{array}$$ |
3 Proof of Theorem 1.1
3.1 ${H}^{-1}$ estimates along geodesics
$$\begin{array}{ccc}\Delta {\Psi}_{i}={\rho}_{i}& & \end{array}$$ |
$$\begin{array}{ccc}\Delta {\Psi}_{\theta}={\rho}_{\theta},& & \end{array}$$ | (10) |
$$\begin{array}{ccc}\Delta {\partial}_{\theta}{\Psi}_{\theta}& =& {\partial}_{\theta}{\rho}_{\theta}\end{array}$$ |
$$\begin{array}{ccc}& =& -\nabla \cdot \left({\rho}_{\theta}{v}_{\theta}\right).\end{array}$$ |
$$\begin{array}{ccc}\int |\nabla {\partial}_{\theta}{\Psi}_{\theta}{|}^{2}=-\int {\rho}_{\theta}{v}_{\theta}\cdot \nabla {\partial}_{\theta}{\Psi}_{\theta},& & \end{array}$$ |
$$\begin{array}{ccc}\parallel {\partial}_{\theta}\nabla {\Psi}_{\theta}{\parallel}_{{L}^{2}\left({\mathbb{R}}^{d}\right)}& \le & \parallel {\rho}_{\theta}{v}_{\theta}{\parallel}_{{L}^{2}\left({\mathbb{R}}^{d}\right)}\end{array}$$ |
$$\begin{array}{ccc}& \le & \parallel {\rho}_{\theta}{\parallel}_{{L}^{\infty}\left({\mathbb{R}}^{d}\right)}^{1/2}{W}_{2}({\rho}_{1},{\rho}_{2}).\end{array}$$ |
3.2 Lagrangian formulation of the Vlasov-Poisson system
$$\begin{array}{ccc}& & \dot{X}=\xi ,\end{array}$$ | (11) |
$$\begin{array}{ccc}& & \dot{\xi}=\nabla \Psi (t,X).\end{array}$$ | (12) |
$$\begin{array}{ccc}& & \forall t\in [0,T[,\forall (x,y)\in {\mathbb{R}}^{3}\times {\mathbb{R}}^{3},|x-y|\le \frac{1}{2},\end{array}$$ |
$$\begin{array}{ccc}& & |\nabla \Psi (t,x)-\nabla \Psi (t,y\left)\right|\le C|x-y|log\frac{1}{|x-y|},\end{array}$$ | (13) |
$$\begin{array}{ccc}\Xi (t,x,v)=(X,\xi )(t,x,v)& & \end{array}$$ |
$$\begin{array}{ccc}\dot{\gamma}\left(t\right)=u(t,\gamma (t\left)\right),& & \end{array}$$ |
$$\begin{array}{ccc}{\partial}_{t}\mu (t,x)+\nabla \cdot \left(\mu \right(t,x\left)u\right(t,x\left)\right)=0.& & \end{array}$$ |
$$\begin{array}{ccc}f\left(t\right)=\Xi (t,\cdot ,\cdot {)}_{\#}{f}^{0},& & \end{array}$$ | (14) |
$$\begin{array}{ccc}\rho \left(t\right)=X(t,\cdot ,\cdot {)}_{\#}{f}^{0}.& & \end{array}$$ | (15) |
3.3 Final estimate
$$\begin{array}{ccc}Q\left(t\right)\ge \frac{1}{2}{W}_{2}^{2}\left({f}_{1}\right(t),{f}_{2}(t\left)\right).& & \end{array}$$ |
$$\begin{array}{ccc}\frac{d}{dt}Q\left(t\right)& =& {\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi )\left({\Xi}_{1}\right(t,x,\xi )-{\Xi}_{2}(t,x,\xi \left)\right)\cdot {\partial}_{t}\left({\Xi}_{1}\right(t,x,\xi )-{\Xi}_{2}(t,x,\xi \left)\right)\end{array}$$ |
$$\begin{array}{ccc}& =& {\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi )\left[\left({X}_{1}\right(t,x,\xi )-{X}_{2}(t,x,\xi \left)\right)\cdot \left({\xi}_{1}\right(t,x,\xi )-{\xi}_{2}(t,x,\xi \left)\right)\right]\end{array}$$ |
$$\begin{array}{ccc}& +& {\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi )[\left({\xi}_{1}\right(t,x,\xi )-{\xi}_{2}(t,x,\xi \left)\right)\cdot \end{array}$$ |
$$\begin{array}{ccc}& & (\nabla {\Psi}_{1}(t,{X}_{1}(t,x,\xi ))-\nabla {\Psi}_{2}(t,{X}_{2}(t,x,\xi )\left)\right)].\end{array}$$ |
$$\begin{array}{ccc}& & {Q}^{1/2}\left(t\right){\left({\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi ){\left|\nabla {\Psi}_{1}(t,{X}_{1}(t,x,\xi \left)\right)-\nabla {\Psi}_{2}(t,{X}_{2}(t,x,\xi \left)\right)\right|}^{2}\right)}^{1/2}\end{array}$$ |
$$\begin{array}{ccc}& \le & {Q}^{1/2}\left(t\right)({\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi ){\left|\nabla {\Psi}_{2}(t,{X}_{1}(t,x,\xi \left)\right)-\nabla {\Psi}_{2}(t,{X}_{2}(t,x,\xi \left)\right)\right|}^{2}\end{array}$$ |
$$\begin{array}{ccc}& & +{{\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi ){\left|\nabla {\Psi}_{2}(t,{X}_{1}(t,x,\xi \left)\right)-\nabla {\Psi}_{1}(t,{X}_{1}(t,x,\xi \left)\right)\right|}^{2})}^{1/2}\end{array}$$ |
$$\begin{array}{ccc}& =& {Q}^{1/2}\left(t\right){\left({T}_{1}\left(t\right)+{T}_{2}\left(t\right)\right)}^{1/2}.\end{array}$$ |
$$\begin{array}{ccc}\frac{d}{dt}Q\left(t\right)\le Q\left(t\right)+{Q}^{1/2}\left(t\right){\left({T}_{1}\left(t\right)+{T}_{2}\left(t\right)\right)}^{1/2},& & \end{array}$$ | (16) |
$$\begin{array}{ccc}{T}_{2}\left(t\right)& =& {\int}_{{\mathbb{R}}^{3}}{\rho}_{1}(t,x){\left|\nabla {\Psi}_{1}(t,x)-\nabla {\Psi}_{2}(t,x)\right|}^{2}\end{array}$$ |
$$\begin{array}{ccc}& \le & C{W}_{2}^{2}\left({\rho}_{1}\right(t),{\rho}_{2}(t\left)\right),\end{array}$$ |
$$\begin{array}{ccc}{W}_{2}\left({\rho}_{1}\right(t),{\rho}_{2}(t\left)\right)\le {\left({\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi ){\left|{X}_{1}(t,x,\xi )-{X}_{2}(t,x,\xi )\right|}^{2}\right)}^{\frac{1}{2}}.& & \end{array}$$ |
$$\begin{array}{ccc}{W}_{2}^{2}({\rho}_{1},{\rho}_{2})& \le & {\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi ){\left|{X}_{2}(t,x,\xi )-{X}_{1}(t,x,\xi )\right|}^{2}\end{array}$$ |
$$\begin{array}{ccc}& \le & {\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi ){\left|{\Xi}_{1}(t,x,\xi )-{\Xi}_{2}(t,x,\xi )\right|}^{2}=2Q\left(t\right),\end{array}$$ |
$$\begin{array}{ccc}{T}_{1}& =& {\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi ){\left|\nabla {\Psi}_{2}(t,{X}_{1}(t,x,\xi \left)\right)-\nabla {\Psi}_{2}(t,{X}_{2}(t,x,\xi \left)\right)\right|}^{2}\end{array}$$ |
$$\begin{array}{ccc}& \le & C{\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi )\left(|{X}_{1}-{X}_{2}{|}^{2}{log}^{2}\frac{1}{|{X}_{1}-{X}_{2}|}\right)(t,x,\xi )\end{array}$$ |
$$\begin{array}{ccc}& =& \frac{C}{4}{\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi )\left(|{X}_{1}-{X}_{2}{|}^{2}{log}^{2}(|{X}_{1}-{X}_{2}{|}^{2})\right)(t,x,\xi ).\end{array}$$ |
$$\begin{array}{ccc}& & {T}_{1}\left(t\right)\end{array}$$ |
$$\begin{array}{ccc}\le & & \frac{C}{4}\left[{\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi )|{X}_{1}-{X}_{2}{|}^{2}(t,x,\xi )\right]{log}^{2}\left[{\int}_{{\mathbb{R}}^{6}}{f}^{0}(x,\xi )|{X}_{1}-{X}_{2}{|}^{2}(t,x,\xi )\right]\end{array}$$ |
$$\begin{array}{ccc}\le & & \frac{C}{2}Q\left(t\right){log}^{2}\left(2Q\right(t\left)\right).\end{array}$$ |
$$\begin{array}{ccc}\frac{d}{dt}Q\left(t\right)\le CQ\left(t\right)(1+log\frac{1}{Q\left(t\right)}),& & \end{array}$$ |