Convexity Properties of Hamiltonian Group Actions (Q5710173)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 2234780
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Convexity Properties of Hamiltonian Group Actions |
scientific article; zbMATH DE number 2234780 |
Statements
2 December 2005
0 references
Hamiltonian systems
0 references
convex domains matrices
0 references
convex functions
0 references
polytopes
0 references
Convexity Properties of Hamiltonian Group Actions (English)
0 references
0.8080279231071472
0 references
0.7874124646186829
0 references
0.7852664589881897
0 references
0.7826102375984192
0 references
Let \(H_{\lambda}^{n}\) be the set of \(n\times{n}\) matrices whose eigenvalues are the numbers \(\lambda_{1},\dots ,\lambda_{n}\) with \(\lambda_{1}\geqslant\lambda_{2}\geqslant{\dots }\geqslant\lambda_{n}\), and for each \(A\in{H_{\lambda}^{n}}\), let \(a_{11},a_{22},\dots ,a_{nn}\) be the diagonal entries of \(A\). The assignment \(A\mapsto{(a_{11},a_{22},\dots ,a_{nn})}\) defines a mapping \(\Phi:H_{\lambda}^{n}\to\mathbb{R^{n}}\). In the mid of 1950's Horn proved that the image of \(\Phi\) is a convex polytope, i.e., that the image of \(\Phi\) is the convex hull of the vectors \(\lambda_{\sigma}=(\lambda_{\sigma(1)},\dots ,\lambda_{\sigma(n)})\), where \(\sigma\) ranges over the set of permutations of \(\{1,2,\dots ,n\}\). In the early 1970's Kostant showed that this result is a special case of a more general result having to do with coadjoint orbits of Lie groups. More exactly, let \(G\) be a compact connected Lie group, \(T\) the Cartan subgroup of \(G\) and \(\mathfrak{g}\) and \(\mathfrak{t}\) their Lie algebras. The adjoint action of \(G\) on \(\mathfrak{g}\) dualizes to give an action of \(G\) on \(\mathfrak{g}^{*}\) and the coadjoint orbits are by definition the orbits of this action. Let \(\pi:{\mathfrak{g}^{*}}\to{\mathfrak{t}^{*}}\) be the transpose of the inclusion map \(\mathfrak{t}\hookrightarrow{\mathfrak{g}}\). If \(\mathcal{O}\subseteq\mathfrak{g}^{*}\) is a coadjoint orbit, then by restricting \(\pi\) to \(\mathcal{O}\) one gets a map \(\Phi:\mathcal{O}\to{\mathfrak{t}^{*}}\), and Kostant's result asserts that the image of \(\Phi\) is a convex polytope. This result has a nice formulation in terms of symplectic geometry. Namely by a theorem of Kirilov and Kostant, the coadjoint orbits, \(\mathcal{O}\) of \(G\) are symplectic manifolds and the action of \(G\) on \(\mathcal{O}\) preserves the symplectic form. Moreover, if \(G\) is compact and connected, these \(\mathcal{O}'s\) are the only symplectic \(G\)- manifolds on which \(G\) acts transitively. In addition, the action of \(G\) on \(\mathcal{O}\) is Hamiltonian with momentum map \(i:\mathcal{O}\hookrightarrow\mathfrak{g}^{*}\). The restriction of this action to the torus \(T\) is also a Hamiltonian action with momentum map \(\Phi\) as above. Then Kostant's theorem asserts that the image of this momentum map is a convex polytope. In the early 1980 Atiyah, Guillemin and Sternberg prooved that in the symplectic version of Kostant's theorem one can drop the assumption that \(\mathcal{O}\) is a transitive symplectic \(G\)-space, and, in fact, get rid of the role of \(G\) entirely. Their results asserts that if \(M\) is a compact symplectic manifold, \(T\) an \(n\)-torus and \(T\times{M}\to{M}\) a Hamiltonian action of \(T\) on \(M\) and if \(\Phi:M\to{\mathfrak{t}^{*}}\) is the momentum map associated with this action, then \(\Phi(M)\) is a convex polytope. Three years later, Kirwan proved a much deeper non-abelian convexity theorem. Let \(G\) be a compact connected Lie group which is not necessarily abelian. If \(G\) acts in an Hamiltonian fashion on a compact symplectic manifold, then, as above one has a moment map \(\Phi:M\to{\mathfrak{g}^{*}}\), but in general its image is not convex. However, a much more subtle convexity result is true: let \(\mathfrak{g}^{*}/G\) be the orbit space for the coadjoint action of \(G\) on \(\mathfrak{g}^{*}\). If \(W=N(T)/T\) is the Weyl group of \(G\), the action of \(W\) on \(\mathfrak{t}\) gives by duality an action of \(W\) on \(\mathfrak{t}^{*}\), and the orbit spaces \(\mathfrak{g}^{*}/G\) and \(\mathfrak{t}^{*}/W\), are isomorphic. Let us fix a (closed) Weyl chamber, \(\mathfrak{t}^{*}_{+}\), in \(\mathfrak{t}^{*}\). This is a fundamental domain for the action of \(W\) on \(\mathfrak{t}^{*}\), so one has identifications \(\mathfrak{g}^{*}/G\simeq\mathfrak{t}^{*}/W\simeq\mathfrak{t}_{+}^{*}\) and hence a map \(\Phi_{+}:M\to{\mathfrak{t}^{*}_{+}}\). The Kirwan convexity theorem asserts that the image of this map is a convex polytope. The book under review has as a main topic this theorem. The material is divided up into five chapters and has a bibliography of 69 titles. Chapter \(1\), ``The convexity theorem for Hamiltonian \(G\)-spaces'' presents some convexity results which generalize Kirwan's theorem in various non-compact directions: allowing either the group \(G\) or the manifold on which it acts to be noncompact. In particular they discuss a version of Kirwan's theorem, due to Meinrenken and Woodward, for loop groups acting on infinite-dimensional manifolds and a beautiful proof of this theorem by Alekseev, Malkin and Meinrenken in which this result gets recast as a finite-dimensional convexity result. They also discuss briefly some infinite-dimensional versions of the abelian convexity theorem, namely a convexity theorem due to Atiyah and Pressley for loop groups and a convexity theorem for the group of symplectomorphisms of the annulus due to Bloch, Flaschka and Ratiu. Chapter \(2\), ``A constructive proof of the non-abelian convexity theorem'' presents a constructive proof of the Kirwan theorem due to Sjamaar. Chapter \(3\), ``Some elementary examples of the convexity theorem'' deals with some extensions of Sjamaar's result due to Klyachko and others. Chapter \(4\), ``Kähler potential and convexity'' discusses a non-abelian generalization of the convexity theorem of Atiyah's concerning orbits of complex tori acting on Kähler manifolds. Chapter \(5\), ``Applications of the convexity theorem'' makes a review of some applications of Atiyah's theorem in equivariant symplectic geometry. Written by the hands of two masters in the field, the book will be usefull to all those interested in these topics. At last but not the least, the graphical conditions are excelent.
0 references