On Fujita's log spectrum conjecture (Q330856)

Let \((X, \Delta)\) be a pair of data with a normal variety \(X\) and a boundary \(\Delta\) which is an effective \(\mathbb{R}\)-Weil divisor such that \(K_X+ \Delta\) is an \(\mathbb{R}\)-Cartier \(\mathbb{R}\)-divisor. If \((X, \Delta)\) is a lc pair and \(M\) is a big divisor on \(X\), then the Kodaira energy of \((X, \Delta, M)\) which was introduced by Fujita, is defined by \[ \kappa \epsilon (X, \Delta, M)= - \inf \{ t\in {\mathbb{Q}}_{\geq 0} |\kappa(K_X+ \Delta +tM) \geq 0\}. \] Fujita proposed the spectrum conjecture: if \(\Delta =0\) and \(M\) is ample then the set of negative Kodaira energies is finite away from zero. In this paper, the author proves an analogue of the spectrum conjecture for pairs. Theorem 1. Let \(S_n^{ls}\) be the set of Kodaira energies where \((X, \Delta)\) is a pair with \(X\) smooth, \(\Delta\) is a reduced divisor with simple normal crossing support and \(M\) is an ample Cartier divisor on \(X\). Then for any real number \(x<0\) the set \(\{ t\in S_n^{ls} | x< t< x+ \epsilon \}\) is finite, for \(\epsilon\) small enough. Let \((X, \Delta)\) be a lc pair and \(M\) a \(\mathbb{R}\)-Weil divisor on \(X\) such that \(K_X+ \Delta + tM\) is pseudo-effective for some \(t\geq 0\). The pseudo-effective threshold \(\tau (X, \Delta, M)\) of \(M\) with respect to \((X, \Delta)\) is defined by \[ \tau (X, \Delta, M) = \inf \{ t\in {\mathbb{R}}_{\geq 0} | K_X+ \Delta + tM\text{ is quad pseudo-effective} \}. \] Fix a positive integer \(n\) and two subsets \(I\subset [0, 1]\) and \(J\subset \mathbb{R}\). Let \({\mathcal{D}}_n(I, J)\) be the set of all lc pairs \((X, \Delta)\) with an effective \(\mathbb{R}\)-Weil divisor \(M\) on \(X\) such that \(X\) is a projective variety of dimension \(n\), the coefficients of \(\Delta\) (resp. \(M\)) are in \(I\) (resp. \(J\)) with \((X, \Delta+tM)\) lc and \(K_X+ \Delta + tM\) pseudo-effective for some \(t\geq 0\). Let \[ {\mathcal{T}}_n (I, J)=\{ \tau (X, \Delta, M) | (X, \Delta, M) \in {\mathcal{D}}_n(I, J) \} \] \(I\) satisfies the ACC (ascending chain condition) if every non-decreasing sequence \(\{x_i\}\subset I\) is eventually constant. \(I\) satisfies the DCC (descending chain condition) if every non-increasing sequence \(\{x_i\}\subset I\) is eventually constant. The main theorem in this paper is the following. Theorem 2. Fix a positive integer \(n\) and two subsets \(I\subset [0, 1]\) and \(J\subset \mathbb{R}\). If \(I\) and \(J\) satisfy the DCC then \({\mathcal{T}}_n (I, J)\) satisfies the ACC. The author proves these two theorems by the techniques from minimal model program and the recent results of \textit{C. D. Hacon} et al. [Ann. Math. (2) 180, No. 2, 523--571 (2014; Zbl 1320.14023)]. The main ideas are inspired by the papers of \textit{J.-P. Demailly} et al. [Acta Math. 210, No. 2, 203--259 (2013; Zbl 1278.14022)] and \textit{Y. Gongyo} [Math. Res. Lett. 18, No. 5, 991--1000 (2011; Zbl 1246.14026)].