Lower envelope of admissible functions on the Grassmannian \(G_{2,4}\mathbb C\) in presence of symmetries (Q2637616)

On a Kähler manifold \((X, \omega)\), a real-valued function \(\varphi\) is said to be \(\omega\)-admissible if \(\omega+ \mathbf{i} \partial \overline{\partial} \varphi\) is also a Kähler metric. The class of \(\omega\)-admissible functions on a Kähler manifold has been intensively studied in various problems involving nonlinear equations on complex manifolds, one of them being the existence of Kähler-Einstein metrics. For (compact) Fano manifolds, \textit{G. Tian} proved in [Invent. Math. 89, 225--246 (1987; Zbl 0599.53046)] that there exists a positive number \(\alpha\) such that the quantity \(\int_X e^{-\alpha \varphi} dV_{\omega}\) remains uniformly bounded as \(\varphi\) runs through all \(\omega\)-admissible functions on \(X\) with zero supremum. If the cohomology class of \(\omega\) is chosen in the first Chern class of \(X\), the supremum of all the possible values of \(\alpha\) is an intrinsic number attached to \(X\), called Tian's invariant of \(X\). In [loc. cit.] Tian proves that if \(\alpha > \frac{m}{m+1}\), where \(\mathrm{dim}\, X=m\), then \(X\) admits a Kähler-Einstein metric. Although this condition if far from being sufficient (as proved by \textit{T. Aubin} [J. Funct. Anal. 88, No. 2, 385--394 (1990; Zbl 0713.53039)] for the projective space), Tian's invariant can be studied for its own sake, and leads to delicate problems in analysis on manifolds. In the papers [Bull. Sci. Math. 126, No. 8, 675--680 (2002; Zbl 1024.53046)] and [Bull. Sci. Math. 130, No. 4, 341--353 (2006; Zbl 1103.53042)], \textit{A. Ben Abdesselem} developed a method to provide an explicit minoration of admissible functions on the projective space that are invariant by a sufficiently large group of isometries \(G\). This minoration allows to estimate a variant Tian's invariant (for \(G\)-admissible functions). In the paper under review, the authors studies another specific case: the complex Grassmannian \(G_{4, 2}(\mathbb{C})\). If no group of invariances is specified, Tian's invariant has been computed by the reviewer in [J. Geom. Anal. 16, No. 3, 523--533 (2006; Zbl 1104.53067)]. Here the authors impose an \(U(2)\)-invariance, as well as the specific point where the supremum is reached.