Rigidity of time changes for horocycle flows (Q909318): Difference between revisions

Let \(G=SL(2,{\mathbb{R}})\) and \(h_ t=\begin{pmatrix} 1 & t \\ 0 & 1 \end{pmatrix}\) for all \(t\in {\mathbb{R}}\). Let \({\mathfrak L}\) be the class of discrete subgroups \(\Gamma\) of G such that \(G/\Gamma\) admits a G-invariant probability measure, say \(\mu_{\Gamma}\). The \((h_ t)_{t\in {\mathbb{R}}}\)-action on \(G/\Gamma\) is called a horocycle flow. The author proved earlier various ``rigidity results'' for horocycle flows [cf. her paper in Contemp. Math. 26, 309-334 (1984; Zbl 0548.58034) and other references therein]; e.g. if \(\Gamma_ 1,\Gamma_ 2\in {\mathfrak L}\) and there exists a measurable conjugacy \(\Psi\) between the corresponding horocycle flows, then \(\Gamma_ 1\) and \(\Gamma_ 2\) are conjugate in G and \(\Psi\) is an affine map. In the paper under review the author considers a larger class of flows obtained by introducing time changes in the horocycle flows as above. An earlier result of the author asserting that the horocycle flows are loosely Bernoulli implies that any two horocycle flows admit (measurable) time changes such that the resulting flows are isomorphic. The thrust of the present paper is to show that, on the other hand, if we restrict to time changes satisfing a certain condition, involving decay at \(\infty\) of the matrix coefficients of the time change function, then appropriate analogues of the above-mentioned and other similar rigidity results hold. It is noted that, thanks to a result of C. Moore proved at the request of the author, the condition needed here is satisfied by any function \(\tau\) such that \(\tau\) and \(1/\tau\) are bounded and \(\tau\) is Hölder continuous with exponent greater than \(1/2\) along the orbits of the rotation subgroup SO(2) of \(G=SL(2,{\mathbb{R}})\).