Optimal localization for the Einstein constraints

Abstract: We consider asymptotically Euclidean, initial data sets for Einstein's field equations and solve the localization problem at infinity, also called gluing problem. We achieve optimal gluing and optimal decay, in the sense that we encompass solutions with possibly arbitrarily low decay at infinity and establish (super-)harmonic estimates within possibly arbitrarily narrow conical domains. In the localized seed-to-solution method (as we call it), we define a variational projection operator which associates the solution to the Einstein constraints that is closest to any given localized seed data set (as we call it). Our main contribution concerns the derivation of harmonic estimates for the linearized Einstein operator and its formal adjoint which, in particular, includes new analysis on the linearized scalar curvature operator. The statement of harmonic estimates requires the notion of energy-momentum modulators (as we call them), which arise as correctors to the localized seed data sets. For the Hamiltonian and momentum operators, we introduce a notion of harmonic-spherical decomposition and we uncover stability conditions on the localization function, which are localized Poincare and Hardy-type inequalities and, for instance, hold for arbitrarily narrow gluing domains. Our localized seed-to-solution method builds upon the gluing techniques pioneered by Carlotto, Chrusciel, Corvino, Delay, Isenberg, Maxwell, and Schoen, while providing a proof of a conjecture by Carlotto and Schoen on the localization problem and generalize P. LeFloch and Nguyen's theorem on the asymptotic localization problem.