On the Kähler-Ricci flow on Fano manifolds (Q2671280)

The (normalized) Kähler-Ricci flow on a Fano manifold \(M\), is the geometric flow defined by the equation \[ \frac{\partial \omega}{\partial t} = -\operatorname{Ric}(\omega) + \omega, \] with initial data \(\omega|_{t=0}=\omega_0\) a Kähler form in \(c_1(M)\). There exists a solution for all time by the fundamental work of [\textit{H.-D. Cao}, Invent. Math. 81, 359--372 (1985; Zbl 0574.53042)]. The limiting behavior of this flow, which is the object of the article under review, has since then been studied in various degrees of generality. In order to say more, let us recall that a Kähler-Ricci soliton on a Fano manifold \(M\) is a solution to a Monge-Ampère equation of the form \[ \operatorname{Ric}(\omega) = \omega +L_X\omega \] where \(X\) is a holomorphic vector field and \(L_X\) denotes the Lie derivative. In the article under review, the authors propose a new proof of the following result: if \(M\) admits a Kähler-Ricci soliton with vector field \(X\), then the Kähler-Ricci flow starting from an \(\operatorname{Im} X\)-invariant initial form converges smoothly and exponentially fast to a Kähler-Ricci soliton, up to the action of the one-parameter subgroup generated by \(X\). This result was first announced by Perelman in 2002 for the case when \(X=0\), that is, when there exists a Kähler-Einstein metric, and by Tian and Zhu for general vector fields [\textit{G. Tian} and \textit{X. Zhu}, J. Am. Math. Soc. 20, No. 3, 675--699 (2007; Zbl 1185.53078)], with various subsequent developments, see [\textit{R. Dervan} and \textit{G. Székelyhidi}, J. Differ. Geom. 116, No. 1, 187--203 (2020; Zbl 1447.53081)] for the most general statement. The new proof relies on recent techniques introduced by the authors in [\textit{B. Guo}, \textit{D. H. Phong} and \textit{J. Sturm}, ``Kähler-Einstein metrics and eigenvalue gaps'', Preprint, \url{arXiv:2001.05794}]. Let us finally note that the limiting behavior of the Kähler-Ricci flow when there exists no Kähler-Ricci solitons is now quite well understood: the metrics should converge to a Kähler-Ricci soliton on a degeneration of \(M\) [\textit{X. Chen} et al., Geom. Topol. 22, No. 6, 3145--3173 (2018; Zbl 1404.53058)].