English

A Self-dual Polar Factorization for Vector Fields

Analysis of PDEs 2011-08-12 v2

Abstract

We show that any non-degenerate vector field uu in L(Ω,RN) L^{\infty}(\Omega, \R^N), where Ω\Omega is a bounded domain in RN\R^N, can be written as {equation} \hbox{u(x)=1H(S(x),x)u(x)= \nabla_1 H(S(x), x) for a.e. xΩx \in \Omega}, {equation} where SS is a measure preserving point transformation on Ω\Omega such that S2=IS^2=I a.e (an involution), and H:RN×RNRH: \R^N \times \R^N \to \R is a globally Lipschitz anti-symmetric convex-concave Hamiltonian. Moreover, uu is a monotone map if and only if SS can be taken to be the identity, which suggests that our result is a self-dual version of Brenier's polar decomposition for the vector field uu as u(x)=ϕ(S(x))u(x)=\nabla \phi (S(x)), where ϕ\phi is convex and SS is a measure preserving transformation. We also describe how our polar decomposition can be reformulated as a self-dual mass transport problem.

Keywords

Cite

@article{arxiv.1101.4979,
  title  = {A Self-dual Polar Factorization for Vector Fields},
  author = {Nassif Ghoussoub and Abbas Moameni},
  journal= {arXiv preprint arXiv:1101.4979},
  year   = {2011}
}

Comments

20 pages, Updated version - if any - can be downloaded at http://www.birs.ca/~nassif/

R2 v1 2026-06-21T17:17:11.111Z