Continuity, curvature, and the general covariance of optimal transportation
Abstract
Let M and M be n-dimensional manifolds equipped with suitable Borel probability measures and . Ma, Trudinger & Wang gave sufficient conditions on a transportation cost c ∈ C4(M × M) to guarantee smoothness of the optimal map pushing forward to ; the necessity of these conditions was deduced by Loeper. The present manuscript shows the form of these conditions to be largely dictated by the covariance of the question; it expresses them via non-negativity of the sectional curvature of certain null-planes in a novel but natural pseudo-Riemannian geometry which the cost c induces on the product space M × M. H\"older continuity of optimal maps was established for rougher mass distributions by Loeper, still relying on a key result of Trudinger & Wang which required certain structure on the domains and the cost. We go on to develop this theory for mass distributions on differentiable manifolds -- recovering Loeper's Riemannian examples such as the round sphere as particular cases -- give a direct proof of the key result mentioned above, and revise Loeper's H\"older continuity argument to make it logically independent of all earlier works, while extending it to less restricted geometries and cost functions even for subdomains M and M of Rn. We also give new examples of geometries satisfying the hypotheses -- obtained using submersions and tensor products -- and some connections to spacelike Lagrangian submanifolds in symplectic geometry.
Turn this paper into a lesson
ArcXiv compiles a structured reading guide from this paper's metadata: plain-English importance, contributions, prerequisite concepts, which sections to read first, flashcards, and a quiz. Grounded in the abstract, never invented.