Phase transitions on the Markov and Lagrange dynamical spectra (Q2046461)

Given a surface \(M\), a \(C^2\) diffeomorphism \(\varphi\) on \(M\), and a continuous observable \(f:M\to\mathbb{R}\), the Lagrange value of a point \(x\in M\) is the limit of the suprema of the observable value along the orbit of \(x\), that is, \(\limsup_{n\to\infty}f(\varphi^n(x))\), the Markov value of \(x\) is the suprema of the observable value along the orbit of \(x\), that is, \(\sup_{n\in\mathbb{Z}}f(\varphi^n(x))\). The Lagrange spectrum is the union of the Lagrange values, while the Markov spectrum is the union of the Markov values. Given a diffeomorphism \(\varphi\) with a horseshoe, with Hausdorff dimension bigger than \(1\), there exist a residual subset from an open neighborhood of \(\varphi\) and a residual subset of the smooth observable functions such that the existence of transition points for the Lagrange spectrum and the Markov spectrum can be proved. The Hausdorff dimension of the subsets of the spectrum jumps from values less than one to one.