Floer homology of algebraically finite mapping classes (Q2386899)

For each mapping class group element \(g\) of a compact oriented connected surface \(\Sigma\) of genus bigger than one, there is an associated symplectic Floer homology \(HF_*(g)\) as a \(H^*(\Sigma; {\mathbb Z}_2)\)-module by a result of Seidel. Seidel also showed that \(HF_*(g) \cong H_*(\Sigma, C; {\mathbb Z}_2)\) for a positive Dehn twist along an embedded circle \(C \subset \Sigma\). By the Nielsen--Thurston theory of surface diffeomorphisms, every mapping class without pseudo-Anosov components admits a representative of finite type. The paper under review is to compute \(HF_*(g)\) explicitly for a finite type \(g\) in theorem 1, and for the Milnor fiber \(M \subset \Sigma \) of an isolated plane curve singularity and an mapping class by extending the geometric monodromy trivially to \(\Sigma\), then the author generalizes Seidel's result \(HF_*(g)\cong H_*(\Sigma, M; {\mathbb Z}_2)\) in theorem 2. The proof of theorem 1 is much from the Seidel's computation, and the proof of theorem uses results in the geometric monodromy. Section 2 consists of the monotonicity and the setup of Floer homology, the basic diffeomorphisms of finite type is given in section 3, and the study of the trajectory flow lines is in section 4. Section 5 provides the proof of theorem 1 and section 6 provides the classical results of A'Campo and Le on the geometric monodromy and the proof of theorem 2. Appendix C briefly describes an interesting Floer homology on surfaces with boundary, and appendix A and appendix B give the products of disjoint Dehn twists and decomposition of the monodromy.