New results on noncompact harmonic manifolds (Q439893): Difference between revisions

A complete Riemannian manifold \(X\) is called harmonic if all harmonic functions on \(X\) satisfy the mean value property. Equivalently, in normal coordinates centered at \(p\in X\) the square of the volume density \(\theta_p(q)^2=\det[g_{ij}(q)]\) depends only on dist\((p,q)\). The Euclidean and rank 1 symmetric spaces are harmonic. The Lichnerowicz conjecture of 1944 asserts the converse, namely that all harmonic manifolds are either flat or locally symmetric spaces of rank 1. This conjecture was proven in 1990 by Z. I. Szabó for manifolds with compact universal cover. In 1992, E. Damek and F. Ricci showed that generally in the noncompact case the conjecture is wrong. In 1995 G. Besson, G. Courtois and S. Gallot proved the conjecture for negatively curved manifolds with a compact quotient. The first main result of the paper is the equivalence of the following properties for a simply connected nonflat noncompact harmonic manifold \(X\): (1) the volume growth is purely exponential; (2) the rank of \(X\) is one; (3) the geodesic flow is Anosov with respect to the Sasaki metric; (4) \(X\) is Gromov hyperbolic. The second main result of the paper is that the geodesic flow is Anosov if \(X\) is a nonflat harmonic manifold with no focal points. Finally, as a combination of the main results of this paper and the rigidity theorems of G. Besson, G. Courtois, S. Gallot, and that of Y. Benoist, P. Foulon, F. Labourie, the author deduces the Lichnerowicz conjecture for all compact manifolds without focal points (in particular, for nonpositively curved spaces) and for compact manifolds with Gromov hyperbolic fundamental groups.