A note on the moduli of stable sheaves on elliptic ruled surfaces.
Kōta Yoshioka · 2026
Read A note on the moduli of stable sheaves on elliptic ruled surfaces. by Kōta Yoshioka (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-26208.txt.
Introduction
Kota Yoshioka studies whether moduli spaces of stable sheaves on an elliptic ruled surface with a nef anticanonical bundle are actually nonempty. Such moduli questions ask whether the geometric objects being classified genuinely occur under the stated conditions, rather than forming an empty formal space. The note takes up that existence question in this ruled-surface setting, where the anticanonical bundle's nefness constrains which stable sheaves can occur.
Catalog record
| CREATOR | Kōta Yoshioka |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 17460 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.26208
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A note on the moduli of stable sheaves on elliptic ruled surfaces. Kōta Yoshioka Abstract 0. Introduction Let \pi:X\to C be a {\mathbb{P}}^{1} -bundle over a smooth projective curve C of genus g . Let C_{0} be a minimal section of \pi with (C_{0}^{2})=-e and f a fiber of \pi . Then K_{X}\equiv-2C_{0}-(2-2g+e)f . If ((K_{X}+f)\cdot H)<0 . then \operatorname{Ext}^{2}(E,E(-f))=0 for all \mu -semi-stable sheaves. Thus E is a prioritary sheaf in the sense of Walter. Then Walter [ 7 ] proved the irreducibility by using deformation theory of prioritary sheaves. In this note, we shall slightly generalize the irreducibility of the moduli spaces. Thus we shall prove the irreducibility of the moduli spaces under the assumption (K_{X}\cdot H)<0 . We note that (K_{X}\cdot H)<0 for any ample divisor H , if g=0,1 . Indeed -K_{X} is numerically equivalent to an effective divisor. If g=0 , then -K_{X}-f is also effective, and hence the irreducibility is a consequence of [ 7 ] . On the other hand if g=1 , then -K_{X}-f is numerically equivalent to an effective divisor if and only if e\geq 1 . So we are interested in the case where e=0,-1 . We also study the existence condition of the \mu -semistable sheaves in these cases. For a coherent sheaf on X , \operatorname{rk}E denotes the rank of E and we set \tau(E):=(\operatorname{rk}E,c_{1}(E),\chi(E)) . Let {\mathcal{M}}_{H}({\bf e})^{\mu ss} be the moduli stack of \mu -semi-stable sheaves E with \tau(E)={\bf e} . Then we have the following result. Theorem 0.1 (Proposition 2.7 , 2.9 ) . Let X be an elliptic ruled surface and H(x):=C_{0}+xf be an ample divisor on X . Then for {\bf e}=(r,\xi,\chi)\in{\mathbb{Z}}_{>0}\times\operatorname{NS}(X)\times{\mathbb{Z}} , with 0\leq(\xi\cdot f)<r , we have the following. (1) {\mathcal{M}}_{H(x)}(r,\xi,\chi)^{\mu ss} is irreducible. (2) {\mathcal{M}}_{H(x)}(r,\xi,\chi)^{\mu ss}\neq\emptyset if and only if r^{2}\Delta({\bf e})\geq r_{1}(r-r_{1})(x-\frac{e}{2}) , where r_{1}=(\xi\cdot f) . For the definition of \Delta({\bf e}) , see ( 1.2 ). (3) Assume that r>1 . Let E\in{\mathcal{M}}_{H(x)}(r,\xi,\chi)^{\mu ss} be a general member. Then E is a \mu -stable locally free sheaf such that E_{|\pi^{-1}(t)} is a rigid vector bundle for a general t\in C . Let X be an elliptic ruled surface and assume that -K_{X} is nef, that is, e=0,-1 . Then (K_{X}^{2})=0 and there is an adjacent chamber for {\bf e}=(r,\xi,\chi) . We denote the moduli stack by {\mathcal{M}}_{-K_{X}}({\bf e})^{\mu ss} . Corollary 0.2 . Let X be an elliptic ruled surface and assume that -K_{X} is nef, that is, e=0,-1 . Then {\mathcal{M}}_{-K_{X}}(r,\xi,\chi)^{\mu ss}\neq\emptyset if and only if \Delta({\bf e})>0 or r\mid(\xi\cdot f) and \Delta({\bf e})=0 . 1. Basic results on moduli spaces 1.1. Notation. In this note, instead of using equivariant cohomology of suitable quot-schemes as in [ 9 ] , we use moduli stacks of coherent sheaves. Let {\mathcal{M}}({\bf e}) be the stack of coherent sheaves E with \tau(E)={\bf e} . Let {\mathcal{M}}_{H}({\bf e})^{\mu ss} be the open substack of {\mathcal{M}}({\bf e}) consisting of \mu -semi-stable sheaves and {\mathcal{M}}_{H}({\bf e})^{\mu s} the open substack consisting of \mu -stable sheaves. Let {\mathcal{M}}_{H}({\bf e})^{ss} (resp. {\mathcal{M}}_{H}({\bf e})^{s} ) be the substack of {\mathcal{M}}_{H}({\bf e})^{\mu ss} consisting of semi-stable sheaves (resp. stable sheaves). Let {\mathcal{M}}_{H}({\bf e})^{\mu s}_{0} be the open substack of {\mathcal{M}}_{H}({\bf e})^{\mu s} consisting of \mu -stable locally free sheaves. Let \overline{M}_{H}({\bf e}) be the moduli space of S -equivalence classes of semi-stable sheaves E with \tau(E)={\bf e} and M_{H}({\bf e}) the open subscheme consisting of stable sheaves. Let M_{H}({\bf e})_{0}^{\mu s} be the moduli space of \mu -stable locally free sheaves. For a coherent sheaf E with \operatorname{rk}E=r\neq 0 , we set (1.1) \begin{split}\mu(E):=&\frac{c_{1}(E)}{r},\\ \Delta(E):=&\frac{2rc_{2}(E)-(r-1)(c_{1}(E)^{2})}{2r^{2}}\\ =&\frac{-2r\chi(E)+2r^{2}\chi({\mathcal{O}}_{X})-r(c_{1}(E)\cdot K_{X})+(c_{1}(E)^{2})}{2r^{2}}.\end{split} For {\bf e}=(r,\xi,\chi) , we also set (1.2) \mu({\bf e}):=\frac{\xi}{r},\;\Delta({\bf e}):=\frac{-2r\chi+2r^{2}\chi({\mathcal{O}}_{X})-r(\xi\cdot K_{X})+(\xi^{2})}{2r^{2}}. For coherent sheaves E,F with \operatorname{rk}E,\operatorname{rk}F>0 , Riemann-Roch theorem implies (1.3) \chi(E,F)=\operatorname{rk}E\operatorname{rk}F\left(\frac{((\mu(F)-\mu(E))^{2})-((\mu(F)-\mu(E))\cdot K_{X})}{2}+\chi({\mathcal{O}}_{X})-\Delta(E)-\Delta(F)\right). 1.2. Stack of filtrations For {\bf e}_{i}:=(r_{i},\xi_{i},\chi_{i}) (1\leq i\leq s) , let {\mathcal{F}}_{H}({\bf e}_{1},{\bf e}_{2},...,{\bf e}_{s}) be the stack of filtrations (1.4) 0\subset F_{1}\subset F_{2}\subset\cdots\subset F_{s}=E such that E_{i}:=F_{i}/F_{i-1}\in{\mathcal{M}}_{H}({\bf e}_{i})^{\mu ss} (1\leq i\leq s) and \operatorname{Ext}^{2}(E_{j},E_{i})=0 ( i<j ). Proposition 1.1 . \dim{\mathcal{F}}_{H}({\bf e}_{1},{\bf e}_{2},...,{\bf e}_{s})=2r^{2}\Delta({\bf e})-r^{2}\chi({\mathcal{O}}_{X})+\sum_{i<j}\chi({\bf e}_{i},{\bf e}_{j}) . Proof. By the proof of [ 2 , Lem. 5.2] and Proposition 1.5 , we see that (1.5) \begin{split}\dim{\mathcal{F}}_{H}({\bf e}_{1},{\bf e}_{2},...,{\bf e}_{s})=&\sum_{i}\dim{\mathcal{M}}_{H}({\bf e}_{i})^{\mu ss}-\sum_{i>j}\chi({\bf e}_{i},{\bf e}_{j})\\ =&-\chi({\bf e},{\bf e})+\sum_{i<j}\chi({\bf e}_{i},{\bf e}_{j}).\end{split} ∎ Remark 1.2 . (i) Assume that \frac{(\xi_{1}\cdot H)}{r_{1}}>\frac{(\xi_{2}\cdot H)}{r_{2}}>\cdots>\frac{(\xi_{s}\cdot H)}{r_{s}}. Then ( 1.4 ) is the Harder-Narasimhan filtration of E . In this case, {\mathcal{F}}_{H}({\bf e}_{1},{\bf e}_{2},...,{\bf e}_{s})\to{\mathcal{M}}({\bf e}) is an immersion. (ii) Assume that \frac{(\xi_{1}\cdot H)}{r_{1}}=\frac{(\xi_{2}\cdot H)}{r_{2}}=\cdots=\frac{(\xi_{s}\cdot H)}{r_{s}}. Then ( 1.4 ) is a Jordan-Hölder filtration of E . In this case, the image of {\mathcal{F}}_{H}({\bf e}_{1},{\bf e}_{2},...,{\bf e}_{s})\to{\mathcal{M}}({\bf e}) parameterizes properly \mu -semi-stable sheaves having Jordan-Hölder filtrations ( 1.4 ). 1.3. Walls and chambers Let \operatorname{Amp}(X) be the ample cone of X . Let {\mathcal{F}}({\bf e}) be the set of subsheaves F\subset E which satisfy (1) \tau(E)={\bf e} , (2) 0<\operatorname{rk}F<\operatorname{rk}E , (3) \Delta(F),\Delta(E/F)\geq 0 and (4) \mu(E)-\mu(F)\neq 0 and there is an element H\in\operatorname{Amp}(X) with ((\mu(E)-\mu(F))\cdot H)=0 . Definition 1.3 . For an element 0\subset F\subset E of {\mathcal{F}}({\bf e}) , we define a wall W^{F} by W^{F}:=\{H\in\operatorname{Amp}(X)|((\mu(E)-\mu(F))\cdot H)=0\}. \cup_{F}W^{F} is locally finite. We shall call chamber a connected component of \operatorname{Amp}(X)\setminus\cup_{F}W^{F} . Lemma 1.4 . Assume that (K_{X}\cdot H)<0 . Let E and F be \mu -semi-stable sheaves of positive rank satisfying \frac{(c_{1}(E)\cdot H)}{\operatorname{rk}E}=\frac{(c_{1}(F)\cdot H)}{\operatorname{rk}F} . Then \operatorname{Ext}^{2}(E,F)=0 . Proof. Since (-K_{X}\cdot H)>0 , we have \frac{(c_{1}(E(K_{X}))\cdot H)}{\operatorname{rk}E}<\frac{(c_{1}(F)\cdot H)}{\operatorname{rk}F} . Hence \operatorname{Ext}^{2}(E,F)=\operatorname{Hom}(F,E(K_{X}))^{\vee}=0. ∎ Proposition 1.5 . (1) Let {\mathcal{M}} be an irreducible component of {\mathcal{M}}_{H}({\bf e})^{\mu ss} . Then \dim{\mathcal{M}}\geq 2r^{2}\Delta({\bf e})-r^{2}\chi({\mathcal{O}}_{X}). (2) Assume that (K_{X}\cdot H)<0 . Then {\mathcal{M}}_{H}({\bf e})^{\mu ss} is smooth of \dim{\mathcal{M}}_{H}({\bf e})^{\mu ss}=2r^{2}\Delta({\bf e})-r^{2}\chi({\mathcal{O}}_{X}). Proof. By the deformation theory of coherent sheaves, \dim{\mathcal{M}}\geq\dim\operatorname{Ext}^{1}(E,E)-\dim\operatorname{Ext}^{2}(E,E)-\dim\operatorname{Hom}(E,E)=-\chi(E,E) for E\in{\mathcal{M}} . Hence we get the first claim. By (K_{X}\cdot H)<0 , \operatorname{Ext}^{2}(E,E)=0 for all E\in{\mathcal{M}}_{H}({\bf e})^{\mu ss} . Hence {\mathcal{M}}_{H}({\bf e})^{\mu ss} is smooth. ∎ Theorem 1.6 (Beauville [ 1 ] , Markman [ 3 ] ) . Let X be a smooth projective surface and H an ample divisor on X such that (K_{X}\cdot H)<0 . If \gcd(r,\xi,\chi)=1 , then M_{H}(r,\xi,\chi) is a smooth, projective and irreducible. Proof. By (K_{X}\cdot H)<0 , \operatorname{Ext}^{2}(E,F)=\operatorname{Hom}(F,E(K_{X}))^{\vee}=0 for all E,F\in M_{H}(r,\xi,\chi) . Hence [ 3 , Thm. 8] holds. Then the proof of [ 3 , Cor. 10] implies the claim. ∎ Corollary 1.7 . Let X be a smooth projective surface and H an ample divisor on X such that (K_{X}\cdot H)<0 . Then M_{H}(r,\xi,\chi)_{0}^{\mu s} is irreducible. Proof. For the proof, we shall apply Theorem 1.6 to the moduli space on a blow-up of X . By using a result of Nakashima [ 4 , Thm. 1] on the relations of moduli spaces on X and the blow-up, we shall prove the claim. Let p:\widetilde{X}\to X be the blow-up of X at a point x and C_{1} the exceptional divisor. For E\in M_{H}(r,\xi,\chi)_{0}^{\mu s} , let E^{\prime} be a locally free sheaf on \widetilde{X} fitting in an exact sequence (1.6) 0\to p^{*}(E)\to E^{\prime}\to{\mathcal{O}}_{C_{1}}(-1)\to 0. Then E^{\prime} is a \mu -stable locally free sheaf with respect to H_{n}:=np^{*}(H)-C_{1} for sufficiently large n [ 4 , Prop. 2.5] . We note that c_{1}(E^{\prime})=\pi^{*}(\xi)+C_{1} and \chi(E^{\prime})=\chi(E)=\chi . Since c_{1}(E^{\prime}) is primitive and