On the distributional Hessian of the distance function (Q461305)

The paper deals with the distributional Hessian \(\text{Hess}(d_P)\) of the function \(d_P:M\rightarrow\mathbb{R}\) where \(d_P\) is the distance function from a point \(P\) of an \(n\)-dimensional Riemannian manifold \((M,g)\). The precise structure of \(\text{Hess}(d_P)\) is used to discuss various geometrical properties of the cut locus of \(P\) and to compare various weak notions of the Hessian and the Laplacian. The first section of the paper introduces the matters at hand. The second section treats the structure of \(\text{Hess}(d_P)\). There is an appendix dealing with weak definitions of sub solutions and super solutions of partial differential equations.