Marked length spectrum rigidity for surface amalgams

From MaRDI portal





This article is about marked length spectrum rigidity of thick, negatively curved, 2-dimensional \(P\)-manifolds. These are, roughly speaking, 2-dimensional spaces obtained by glueing together a collection of negatively curved compact surfaces along their boundaries, such that each glueing involves at least three of the surfaces. Marked length spectrum rigidity is a classical conjecture that states that the marked length spectrum -- the map that associates to each homotopy class of closed curves the length of its shortest representative -- determines the space up to isometry. The main result of this paper is that, if \((X, g_1)\) and \((X, g_2)\) are both simple, thick, negatively curved, 2-dimensional \(P\)-manifolds such that \(g_1\) and \(g_2\) induce the same marked length spectrum, then there exists an isometry \((X, g_1) \rightarrow (X, g_2)\) that is isotopic to the identity. In order to prove this, it is also proven that the geodesic flow on hyperbolic \(P\)-manifolds is Anosov with respect to the Bowen-Margulis measure; a result that will be useful for future work on hyperbolic \(P\)-manifolds.\N\NMarked length spectrum rigidity is a famous open problem in the geometry of negatively curved (or Anosov) spaces. The earliest results go back to 1990, when two independent proofs [\textit{C. B. Croke}, Comment. Math. Helv. 65, No. 1, 150--169 (1990; Zbl 0704.53035); \textit{J.-P. Otal}, Ann. Math. (2) 131, No. 1, 151--162 (1990; Zbl 0699.58018)] were given for rigidity of negatively/non-positively curved, closed surfaces. For higher-dimensional manifolds, progress has been very limited, with a positive result for locally symmetric spaces in the 1990s [\textit{U. Hamenstädt}, Geom. Funct. Anal. 9, No. 1, 90--140 (1999; Zbl 0951.37007)] and a very interesting `local' result in the late 2010s [\textit{C. Guillarmou} and \textit{T. Lefeuvre}, Ann. Math. (2) 190, No. 1, 321--344 (2019; Zbl 1506.53054)]. However, more progress has been made for 2-dimensional spaces that are not closed manifolds. There are various positive results for compact manifolds with boundary, or with singularities. This article adds a natural class of spaces to the class of spaces for which Marked Length Spectrum Rigidity is known.\N\NThe strategy of the proof is as follows: The authors first consider \(P\)-manifolds whose chambers (the pieces that are glued together) all admit an inclusion into a closed surface that is contained in \(X\). They then consider the universal covering \(\tilde{X}\) of \(X\). Since \((X, g_1)\) and \((X, g_2)\) have the same marked length spectrum, the identity map on \(\tilde{X}\) induces a Möbius homeomorphism between the visual boundaries. We now have to show that this Möbius homeomorphism can be extended to an isometry between the interior spaces. This is a classical strategy to prove marked length spectrum rigidity. In general, such a Möbius homeomorphism can be extended to a \((1,C)\)-quasi-isometry, meaning that it is an isometry up to an additive error. However, for surfaces this additive error can be removed, using Anosov-properties of the geodesic flow and properties of surfaces. The authors use this to produce isometries defined in different parts of \(\tilde{X}\) and show that these isometries can be patched together to form an isometry on the entire space. They then reduce the general case to this special one.



Cites work









This page was built for publication: Marked length spectrum rigidity for surface amalgams

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6931136)