Stability conditions and \(\mu\)-stable sheaves on \(K3\) surfaces with Picard number one (Q1951960): Difference between revisions

From MaRDI portal
Added link to MaRDI item.
Import240304020342 (talk | contribs)
Set profile property.
Property / MaRDI profile type
 
Property / MaRDI profile type: MaRDI publication profile / rank
 
Normal rank

Revision as of 06:20, 5 March 2024

scientific article
Language Label Description Also known as
English
Stability conditions and \(\mu\)-stable sheaves on \(K3\) surfaces with Picard number one
scientific article

    Statements

    Stability conditions and \(\mu\)-stable sheaves on \(K3\) surfaces with Picard number one (English)
    0 references
    0 references
    24 May 2013
    0 references
    In the paper under review, the author studies stable sheaves on projective \(K3\) surfaces with Picard number 1 with respect to \(\mu-\)stability (slope-stability) and with respect to \(\sigma-\)stability for some stability condition \(\sigma\) on the bounded derived category \(D(X)\) of coherent sheaves on \(X\). Stability conditions are defined on an arbitrary \(\mathbb{C}-\)linear triangulated category \(\mathcal{D}\) as a pair \((Z,\mathcal{A})\) given by the heart \(\mathcal{A}\) of a bounded \(t-\)structure on \(\mathcal{D}\) (hence \(\mathcal{A}\) is an abelian category) and by a group homomorphism \(Z:K(\mathcal{A})\longrightarrow\mathbb{C}\) such that for every \(E\in\mathcal{A}\) we have \(|Z(E)|>0\) and \(\arg(Z(E))\in(0,\pi]\). The space of stability conditions on \(\mathcal{D}\) is denoted \(\mathrm{Stab}(\mathcal{D})\). Suppose that \(\mathcal{D}=D(X)\) for a \(K3\) surface \(X\): among all stability conditions on \(D(X)\), those which are locally finite (for which there are Jordan-Hölder filtrations) and numerical (for which \(Z\) factors through the Neron-Severi lattice \(N(X)\) of \(X\)) play a major role, and form a space \(\mathrm{Stab}(X)\) having the structure of complex manifold. It is in general very difficult to produce examples of stability conditions on the bounded derived category \(D(X)\) of a compact complex manifold, but at least for \(K3\) surfaces this is done by \textit{T. Bridgeland} in [Duke Math. J. 141, No. 2, 241--291 (2008; Zbl 1138.14022)]. More precisely, to every pair \((\beta,\omega)\) with \(\beta\) a real Neron-Severi class and \(\omega\) a real ample class verifying some numerical conditions, one associates a locally finite numerical stability condition \(\sigma_{(\beta,\omega)}=(Z_{(\beta,\omega)},\mathcal{A}_{(\beta,\omega)})\). The set of all these stability conditions is denoted \(V(X)\), and by \(V(X)_{>2}\) we mean all the stability conditions \(\sigma_{(\beta,\omega)}\) such that \(\omega^{2}>2\). The orbits of \(V(X)\) and \(V(X)_{>2}\) under the action of the unversal covering space \(\widetilde{\mathrm{GL}}^{+}(2,\mathbb{R})\) of \(\mathrm{GL}^{+}(2,\mathbb{R})\) are denoted \(U(X)\) and \(U(X)_{>2}\). The main result of the paper is the following: let \(X\) be a projective \(K3\) surface of Picard number 1, \(L\) the ample generator of \(\mathrm{Pic}(X)\), of degree \(2d\), \(E\) a torsion free semi-rigid sheaf of rank \(\mathrm{rk}(E)\leq\sqrt{d}\) and \(\sigma_{(\beta,\omega)}\in V(X)_{>2}\). If \(E\) is stable and \(\beta\omega<c_{1}(E)\cdot\omega/\mathrm{rk}(E)\), then \(E\) is \(\sigma_{(\beta,\omega)}-\)stable. If \(E\) is \(\mu-\)stable, locally free and \(\beta\cdot\omega\geq c_{1}(E)\cdot\omega/\mathrm{rk}(E)\), then it is \(\sigma_{(\beta,\omega)}-\)stable. In particular any \(\mu-\)stable locally free sheaf of rank at most \(\sqrt{d}\) is \(\sigma-\)stable for every \(\sigma\in U(X)_{>2}\). Moreover, the author shows that under the same hypothesis on \(X\) and \(L\) is ample, if \(S\) is a spherical sheaf (i. e. \(\Hom(S,S)=\mathrm{Ext}^{2}(S,S)=\mathbb{C}\) and \(\mathrm{Ext}^{1}(S,S)=0\)), then \(S\) is \(\mu-\)stable and locally free. If it has rank at most \(\sqrt{d}\), then it is \(\sigma-\)stable for every \(\sigma\in U(X)_{>2}\). The assumption on the rank is optimal. As an application of these results, the author shows that if \(X\) and \(Y\) are two projective \(K3\) surfaces of Picard number 1 and \(\Phi:D(Y)\longrightarrow D(X)\) is an equivalence such that \(\Phi_{*}(U(Y))=U(X)\), then there is an isomorphism \(f:X\longrightarrow Y\), a line bundle \(M\in \mathrm{Pic}(X)\) and \(n\in\mathbb{Z}\) such that \(\Phi(\cdot)=M\otimes f_{*}(\cdot)[n]\). This means, in particular, that \(D(X)\) and \(U(X)\) together determine the \(K3\) surface \(X\). We know that this is not true for the only \(D(X)\): \textit{P. Stellari} in [Geom. Dedicata 108, 1--14 (2004; Zbl 1072.14043)] shows that if \(\Phi:D(Y)\longrightarrow D(X)\) is an equivalence of projective \(K3\) surfaces where \(X\) has Picard number 1, then \(Y\) is a moduli space of stable sheaves on \(X\) which, in general, is not isomorphic to \(X\). The author shows that \(D(X)\) and \(U(X)\) do not determine \(X\), if it is an abelian surface: if \(X\) and \(Y\) are abelian surfaces and \(\Phi:D(Y)\longrightarrow D(X)\) is an equivalence, then we always have \(\Phi_{*}(U(Y))=U(X)\).
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references