Projection of Markov measures may be Gibbsian (Q1810994)

Let \((A_M,\sigma)\) be a topological Markov chain (TMC) where \(A\) stands for the (finite) alphabet, \(M\) for the incidence matrix and \(\sigma\) denotes the shift map. This means that \(M\) is a 0-1-matrix selecting a subset of all possible infinite sequences \(a=a(0)a(1)...\) drawn from the alphabet \(A\): \(A_M=\{a\in A^N:M(a(i),a(i+1))=1, \forall i \in N\}.\) The shift transformation \(\sigma\), \(\sigma A_M=A_M,\) is defined as follows: \((\sigma a)(i):=a(i+1)\) for any \(a\in A_M.\) The paper is concerned with the nature of the ``projection'' of a Markov measure, supported by a TMC, obtained by the action of a factor map mapping the original TMC onto another one. Sufficient conditions for the induced measure to be a Gibbs measure (in the sense of Bowen) when the factor system is again a topological Markov chain are given. This amounts to construct, when it exists, the induced potential and to prove its Hölder continuity. This is achieved through a matrix method.