On symplectically fat twistor bundles (Q494738): Difference between revisions

In the paper, the authors study symplectically fat associated bundles. Let \(p:P\to B\) be a principal \(G\)-bundle and let \(\theta\) be a connection form on \(P\). Let \(\mathfrak g\) be a Lie algebra of the group \(G\). A vector \(u\in \mathfrak g^*\) is called fat if the 2-form \(u\Theta(X,Y)\) is non-degenerate for all horizontal vectors \(X,Y\) where \(\Theta\) is the curvature form of the connection defined by \(\theta\). Let \((M,\omega)\) be a symplectic manifold with a Hamiltonian action of a Lie group \(G\) and a moment map \(\Psi:M\to\mathfrak g^*\). Let \(P\) be a principal \(G\)-bundle and \(E=P\times_GM\) be an associated bundle. Then, a theorem by Weinstein and Sternberg says that \(E\) admits a symplectic form if there exists a connection on \(P\) such that all vectors in \(\Psi(M)\subset \mathfrak g^*\) are fat. Let \(B=K\slash H\) be a homogeneous space with a \(K\)-invariant \(\mathrm{SO}(2n)\) structure \(P=\mathrm{SO}(B)\) where \(\dim B=2n\). Let us consider the associate bundle \(E=P\times_{\mathrm{SO}(2n)}(\mathrm{SO}(2n)\slash \mathrm{U}(n))\) which is the twistor bundle of \(B\). The homogeneous space \(\mathrm{SO}(2n)\slash \mathrm{U}(n)\) is symplectic. The authors prove that if \(K\slash H\) is one of even-dimensional Grassmanians of maximal rank, i.e., \(K\slash H= \mathrm{SO}(2n+2m)\slash \mathrm{SO}(2n)\times \mathrm{SO}(2m) m,n\neq1,\mathrm{SO}(2n+2m+1)\slash \mathrm{SO}(2n)\times \mathrm{SO}(2m+1), n\neq 1, \mathrm{Sp}(n+m)\slash \mathrm{Sp}(n)\times \mathrm{Sp}(m),\mathrm{U}(n+m)\slash \mathrm{U}(n)\times \mathrm{U}(m)\) and if on \(P=\mathrm{SO}(K\slash H)\) we have the canonical connection then the twistor bundles \(\mathrm{SO}(K\slash H) \times_{\mathrm{SO}(2n)}(\mathrm{SO}(2n)\slash \mathrm{U}(n))\) are symplectically fat.