On the ampleness of the cotangent bundles of complete intersections (Q1645330): Difference between revisions

Let \(X \subset \mathbb P^N\) be a complete intersection of generic hypersurfaces \(H_1, \ldots, H_c\). By the adjunction formula the canonical bundle \(K_X\) is ample if the degree of the hypersurfaces is sufficiently high; the positivity properties of the cotangent bundle \(\Omega_X\) are much more delicate: if the codimension \(c\) is at most \(N/2-1\) one has \(H^0(X, \text{Sym}^k \Omega_X)=0\) for every \(k \geq 1\) [\textit{P. Brückmann} and \textit{H. G. Rackwitz}, Math. Ann. 288, No. 4, 627--635 (1990; Zbl 0724.14032)], so \(\Omega_X\) is very far from being ample. However a conjecture of \textit{O. Debarre} [Compos. Math. 141, No. 6, 1445--1459 (2005; Zbl 1086.14038)] claims that \(\Omega_X\) is ample if the codimension is at least \(N/2\). In this paper the author proves Debarre's conjecture: let \(X \subset \mathbb P^N\) be a complete intersection of codimension \(c \geq N/2\) of generic hypersurfaces \(H_1, \ldots, H_c\) of degree at least \[ \prod_{l = \lceil N/2 \rceil}^N l^4 (l+1)^{2l}. \] Then \(\Omega_X\) is ample. \newline A similar result was shown independently by \textit{D. Brotbek} and \textit{L. Darondeau} [Invent. Math. 212, No. 3, 913--940 (2018; Zbl 1401.14203)] for complete intersections in any projective manifold, but their bound is not effective (see also the work of \textit{Y. Deng} for effective variants [``Effectivity in the hyperbolicity-related problems'', Preprint, \url{arXiv:1606.03831}]). The basic idea of the proof goes back to the construction of symmetric differentials on complete intersections of Fermat type hypersurfaces by \textit{D. Brotbek} [Math. Ann. 366, No. 1--2, 417--446 (2016; Zbl 1361.32031)]. The author improves this construction by developing the ``moving coefficients method'' and resolving a series of formidable technical challenges. Combined with a result of Y.-T. Siu on Fujita's conjecture, the main theorem also yields an effective bound for a symmetric power \(\text{Sym}^k \Omega_X\) to be very ample.