Extensions of real bounded symmetric domains (Q2192365)

Based on results on the complex differential geometry of Riemannian symmetric spaces of non-compact types, in particular their associated Akhiezer-Gindinkin domains (due partially to the second author [\textit{B. Krötz} and \textit{R. J. Stanton}, Ann. Math. (2) 159, No. 2, 641--724 (2004; Zbl 1053.22009); Geom. Funct. Anal. 15, No. 1, 190--245 (2005; Zbl 1078.22009)]), the paper provides new achievements and alternative approaches concerning the fine structure of the boundary and some naturally associated fiber bundles to real bounded symmetric domains in a rather detailed and heuristic treatment illustrated with examples. For a given real finite dimensional bounded symmetric domain \(\mathcal{D}\), the classical realization of the form \(\mathcal{D}\simeq G/K\) is used as totally real submanifold within a complex bounded domain \(G_h/K_h\) where \(G_h\) is a semisimple Hermitian Lie group of noncompact type with maximal compact subgroup \(K_h\). By construction, \(G\) is the fixed point set of an involution \(\theta\) of \(G_h\) which commutes with the Cartan involution corresponding to \(K_h\) such that \(\theta^2_h=\mathrm{id}\) and \(G_h^{\theta_h}=K_h\). Throughout the paper, the real and complex cases are studied in parallel presentation. The structure group \(\mathbf{G}_h\) with Lie algebra \(\mathfrak{g}^\mathbf{C}_h\) (the complexification of the Lie algebra \(\mathfrak{g}_h\) of \(G_h\)) is applied as covering object during the investigations. A crucial role is played by the conjugate linear extension of the induced involution \(\dot{\tau}\) by \(\tau\) on \(\mathfrak{g}_h\) (denoted by \(\eta\) in the paper), whose fixed point sets \(G_c\) resp. \(L_c\) corresponding to \(G_h\) resp. \(K_h\) are semisimple split-Hermitian Lie groups giving rise to the split-complex pseudo Riemannian symmetric space \(G_c/L_c\). The paper is divided into seven sections. In Section 1 the description of the boundary of \(G_h/K_h\) as a finite union of \(G_h\) is recalled from original view points related also to the Harish-Chandra realization, the holomorphic extension of the Iwasawa decomposition and the Akhiezer-Gindikhin crown. Section 2 gives an alternative description for the boundary orbits in the real case \(G/K\) using the involution \(\eta\) while Section 3 is devoted to the fine structure of the stabilizer of the orbit \(E_\kappa=\dot{\kappa}(e)\) with basepoint \(e\) where \(\dot{\kappa}\) is a standard homomorphism \(\mathfrak{sl}(2,\mathbf{C}) \to \mathfrak{g}_h^\mathbf{C}\). This geometric description serves a crucial tool to describe the decomposition of a natural holomorphic extension of homogeneous vector bundles to the boundary along these \(G_h\) orbits. Sections 4-6 are devoted to the fine structure of extensions of homogeneous vector bundles over \(G/K\). The treatment starts with a classification based Lie algebraic discussion of the natural lifting \(\mathfrak{k}\to \mathfrak{l}_c^\prime\) continued with an extension procedure from \(K\) to \((L^\prime_c)_0\) and then to \(L^\prime_c\) concluded with a nice generalization the real setting of results of results concerning extensions to natural compactifications of homogeneous holomorphic vector bundles and their restrictions to boundary orbits. In Section 7, the paper is completed with an alternative approach to Matsuki's crown theorems, and using this, one establishes the existence of an analytic extension of K-finite matrix coefficients of irreducible representations of \(G\) to \(D=L_c \exp i\Omega^{\dot{\eta}}_h G\) in terms of the domain \(\Omega_h\) obtained with the holomorphic extension of the Iwasawa decomposition. The deep technical background involves typically finite dimensional features as the classification of finite dimensional semisimle complex Lie algebras and the equivalence of Lie groups with Lie algebras, thus a continuation toward the infinite dimensional setting can be an interesting future task.