Simple geodesics on a punctured surface (Q857045)

From MaRDI portal





scientific article
Language Label Description Also known as
English
Simple geodesics on a punctured surface
scientific article

    Statements

    Simple geodesics on a punctured surface (English)
    0 references
    0 references
    0 references
    14 December 2006
    0 references
    Let \(F\) be a complete, finite area hyperbolic surface with at least one cusp. \textit{G. McShane} proved that \[ \sum_{\alpha,\beta} \frac{1}{1+e^{(l(\alpha)+l(\beta))/2}}=1/2 \] where the sum is taken over all unordered pairs of simple closed geodesics \(\alpha\), \(\beta\) on \(F\) which bound together with a distinguished cusp \(c_0\) of \(F\) an embedded pair of pants [Invent. Math. 132, 607--632 (1998; Zbl 0916.30039)]. This generalizes the identity he had earlier obtained in his thesis in Warwick for a hyperbolic torus with one cusp. In the paper under review, the authors give a simple proof of McShane's result. To simplify the exposition, they only do this in the case where \(F\) contains exactly one cusp, although the argument works for the general case. The basic outline of the proof is similar to McShane's original proof although this exposition should be useful to readers who would like a more accessible proof of this remarkable identity, which has recently found some striking applications in the work of M. Mirzakhani. The basic idea is to show that the set of simple geodesics emanating from the cusp has a Cantor set-like structure, and that each gap in the complement of this set contains exactly one bi-cuspidal simple geodesic (that is, emanating and ending at the same cusp). Furthermore, the boundary of the gaps corresponds to simple geodesics emanating from the cusp which spiral asymptotically to the closed geodesics \(\alpha\) and \(\beta\) in the identity. The width of the gaps can then be calculated using hyperbolic geometry and the identity then follows from the Birman-series result that the set of simple geodesics has Lebesgue measure zero. The authors do this by working in the equivariant setting in the hyperbolic plane \({\mathbb H}\), which is the universal covering space of \(F\). Generalizations and variations of the identity have also been obtained previously by several authors, including B. Bowditch; H. Akiyoshi, H. Miyachi and M. Sakuma; S. P. Tan, Y. L. Wong and Y. Zhang; and M. Mirzakhani.
    0 references
    hyperbolic surfaces
    0 references
    McShane's identity
    0 references
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references