Spectral invariants, analysis of the Floer moduli space, and geometry of the Hamiltonian diffeomorphism group (Q816342): Difference between revisions

In this article which is a sequel to [Asian J. Math. 6, No. 4, 579--624 (2002; Zbl 1038.53084)] and [Construction of spectral invariants of Hamiltonian paths on closed symplectic manifolds. The breadth of symplectic and Poisson geometry. Festschrift in honor of Alan Weinstein. Boston, MA: Birkhäuser. Progress in Mathematics 232, 525--570 (2005; Zbl 1084.53076)] the author applies spectral invariants, of Hamiltonian diffeomorphisms, constructed in the previous papers to study the geometry of the Hamiltonian diffeomorphism group \(\mathrm{Ham}(M,\omega)\) of a symplectic manifold \((M,\omega)\). These invariants are the spectral norm \(\gamma\) and the homological area \(A(\cdot;1)\). It is shown that the norm \(\| \phi\| \) is nondegenerate. Moreover it is proved that \(\| \phi\| \geq \gamma(\phi) \geq A(\phi;1)\) where \(\| \phi\| \) is the Hofer norm of \(\phi\) and these inequalities become equalities for \(\phi\) sufficiently \(C^1\)-small. In particular, the former implies a theorem by McDuff on length minimizing paths in the Hamiltonian diffeomorphism group. The proofs use complicated geometrical structures which are interesting by their own. In particular, the proofs of the nodegeneracy and of the equalities of norms for \(C^1\)-small diffeomorphisms use the adiabatic degeneration and the thick-thin decomposition of Floer moduli spaces. For simplicity, throughout this paper it is assumed that the symplectic manifold \((M,\omega)\) is strongly semipositive however in the preface it is mentioned that ``the scheme of the proof should go through on arbitrary closed symplectic manifolds in the general framework of virtually moduli spaces''.