The cohomology of the general stable sheaf on a \(K3\) surface (Q6103492)

In this extremely well written and interesting paper the authors study the cohomology of general stable sheaves on \(K3\) surfaces. Let \(X\) be a \(K3\) surface such that \(\mathrm{Pic}(X) \cong \mathbb{Z} H\) with \(H^2 = 2n\). Let \(\mathbf{v}\) be a Mukai vector with \(\textbf{v}^{2}\geq -2\) and let \(M_{H}(\textbf{v})\) denote the moduli space parameterizing \(S\)-equivalence classes of Gieseker semistable sheaves on \(X\) with the Mukai vector \(\textbf{v}\). We say that \(\textbf{v}\) satisfies weak Brill-Noether if the general sheaf \(E \in M_{H}(\textbf{v})\) has at most one non-zero cohomology group. If \(\textbf{v}\) satisfies weak Brill-Noether, then the Euler characteristic and the slope completely determine the cohomology of the general sheaf in \(M_{H}(\textbf{v})\). The first main qualitative result delivered by the authors goes as follows. Theorem A. Let \(X\) be a \(K3\) surface such that \(\mathrm{Pic}(X) \cong \mathbb{Z} H\) with \(H^2 = 2n\). Let \(\textbf{v} = (r,dH,a)\) be a Mukai vector with \(\textbf{v}^{2} \geq -2\), \(r\geq 2\), and \(d>0\). i) For each \(r\geq 2\), there exists a finite set of tuples \((n,r,d,a)\) for which \(\textbf{v}\) fails to satisfy weak Brill-Noether. ii) If \(n\geq r\), then \(\textbf{v}\) satisfies weak Brill-Noether. iii) If \(a\leq 1\), then \(\textbf{v}\) satisfies weak Brill-Noether. iv) If \(d\geq r\lfloor r/n\rfloor +2\), then \(\textbf{v}\) satisfies weak Brill-Noether. Next, the authors classify boundary cases in Theorem A. Theorem B. Let \(X\) be a \(K3\) surface such that \(\mathrm{Pic}(X) = \mathbb{Z} H\) with \(H^{2} = 2n\). Let \(\textbf{v} = (r,dH,a)\) be a Mukai vector such that \(n < r \leq 3n\) and \(d > 0\). Then \(\textbf{v}\) fails to satisfy weak Brill-Noether if and only if \(\textbf{v}\) belongs to one of the following three cases: i) \(\textbf{v} =(n+r_{1}^{2}, (((n+1)/r_{1})+r_{1})H, ((n+1)/r_{1})^2+n)\), where \(r_{1}\, |\, n+1\) and \(1\leq r_{1} < \sqrt{2n}\); ii) \(\textbf{v} = (r,(r+1)H, nr+2n)\) with \(2n < r \leq 3n\); iii) \(\textbf{v} = (3n, (3n+2)H, 3n^{2}+4n+1)\) with \(n>1\). As a consequence of Theorem A, the authors obtain a classification of stable Ulrich bundles on \(K3\) surfaces of Picard rank 1. More precisely, we have the following. Proposition. Let \(X\) be a \(K3\) surface with \(\mathrm{Pic}(X) = \mathbb{Z} H\). There exists an Ulrich bundle of rank \(r\) with respect to \(mH\) if and only if \(2 \, | \, rm\). Moreover, when an Ulrich bundle of rank \(r\) exists, it has Mukai vector \(\textbf{v} =(r, ((3rm)/2)H, r(2m^{2}n-1))\). In particular, there exists an Ulrich bundle of any rank \(r\geq 2\) with respect to \(2H\). Then the author pass to the globally generated case. More precisely, the question is when the general sheaf in \(M_{H}(\textbf{v})\) is globally generated. Theorem C. Let \(X\) be a \(K3\) surface such that \(\mathrm{Pic}(X) \cong \mathbb{Z} H\) with \(H^{2}=2n\). Let \(\textbf{v}=(r,dH,a)\) be a Mukai vector with \(\textbf{v}^{2} \geq -2\), \(r\geq 2\), \(d>0\), and \(a\geq 2\). i) If \(n\geq 2r\), then the generic sheaf in \(M_{H}(\textbf{v})\) is globally generated. ii) If \(n>1\) and \(d\geq r\lfloor (2r)/n \rfloor +2\), then the generic sheaf in \(M_{H}(\textbf{v})\) is globally generated. In the ample case, as a consequence of Theorem C, the authors provide the following characterization. Theorem D. Let \(X\) be a \(K3\) surface such that \(\mathrm{Pic}(X) \cong \mathbb{Z}H\) with \(H^{2}=2n\). Let \(\textbf{v}=(r,dH,a)\) be a Mukai vector with \(\textbf{v}^{2}\geq -2\), \(r\geq 2\), \(d<0\), and \(a\geq 2\). Let \(\textbf{v}_{H} = (r,(d+r)H,a+(2d+r)n)\). i) If \(n \geq 2r\), then the generic sheaf in \(M_{H}(\textbf{V}_{H})\) is ample. ii) If \(n > 1\) and \(d\geq r\lfloor (2r)/n \rfloor +2\), then the generic sheaf in \(M_{H}(\textbf{v}_{H})\) is ample.