Prym-Brill-Noether theory for ramified double covers
Andrei Bud · 2024
Read Prym-Brill-Noether theory for ramified double covers by Andrei Bud (2024) 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-2411-00716.txt.
Description
Abstract We initiate the study of Prym-Brill-Noether theory for ramified double covers, extending several key results from classical Prym-Brill-Noether theory to this new framework. In particular, we improve Kanev’s results on the dimension of pointed Prym-Brill-Noether loci for ramified double covers.
Catalog record
| CREATOR | Andrei Bud |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 86 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2411-00716
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Prym-Brill-Noether theory for ramified double covers Andrei Bud Abstract We initiate the study of Prym-Brill-Noether theory for ramified double covers, extending several key results from classical Prym-Brill-Noether theory to this new framework. In particular, we improve Kanev’s results on the dimension of pointed Prym-Brill-Noether loci for ramified double covers. Additionally, we compute the dimension of twisted Prym-Brill-Noether loci with vanishing conditions at points, thus extending the results of Tarasca. Furthermore, we compute the class of the twisted Prym-Brill-Noether loci inside (a translation of) the Prym variety, thus extending the results of de Concini and Pragacz to ramified double covers. Finally, we prove that a generic Du Val curve is Prym-Brill-Noether general. 1 Introduction Starting with the fundamental work of Mumford and Beauville, see [ Mum74 ] and [ Bea77b ] , the moduli space of Prym curves \mathcal{R}_{g}\coloneqq\left\{[C,\eta]\ |\ [C]\in\mathcal{M}_{g},\ \eta^{\otimes 2}\cong\mathcal{O}_{C}\ \mathrm{and}\ \eta\neq\mathcal{O}_{C}\right\} became an important object of study in Algebraic Geometry. A key motivation for this study is the Prym map \mathcal{P}_{g}\colon\mathcal{R}_{g}\rightarrow\mathcal{A}_{g-1} which links the geometry of curves to the geometry of principally polarized Abelian varieties. In recent years, several variations of this moduli space of Prym curves have been studied, including pointed Prym curves, see [ Tar23 ] , and ramified Prym curves, see [ LCKV23 ] , [ DLC23 ] and [ Bud24a ] . Thus it is natural to consider the moduli spaces of ramified Prym curves \mathcal{R}_{g,2k}\coloneqq\left\{[C,\eta,B]\ |\ [C]\in\mathcal{M}_{g},\ \eta\in\textrm{Pic}^{k}(C)\ \textrm{and}\ B\ \textrm{is a reduced divisor in}\ |\eta^{\otimes 2}|\right\} and study their geometry. The equivalence between such tuples [C,\eta,B] and double covers f\colon\widetilde{C}\rightarrow C with branch divisor B will be particularly useful in this study. In many instances, we will think of the moduli space \mathcal{R}_{g,2k} as parametrizing double covers ramified at 2k points instead. Moreover, the moduli spaces \mathcal{R}_{g,2k} appear naturally as boundary loci in the compactification \overline{\mathcal{R}}_{g} of \mathcal{R}_{g} , see [ FL10 ] and [ BCF04 ] . As such, Brill-Noether questions on \mathcal{R}_{g} can be reduced via degeneration to questions on these moduli of ramified Prym curves. One advantage of this is that the elements in \mathcal{R}_{g,2} , \mathcal{R}_{g,4} and \mathcal{R}_{g,6} are better behaved from the perspective of Brill-Noether theory than their counterpart in \mathcal{R}_{g} , see [ Bud24a ] and [ DLC23 ] . Motivated by the many applications of Brill-Noether Theory in understanding the birational geometry of \mathcal{M}_{g,n} (cf. [ HM82 ] , [ Har84 ] , [ FJP20 ] and the references therein), as well as the role of Prym-Brill-Noether Theory in studying the geometry of \mathcal{R}_{g} (cf. [ FV16 ] and [ Bud24b ] ), our aim is to understand the geometry of Prym-Brill-Noether loci of a double cover f\colon\widetilde{C}\rightarrow C corresponding to a generic element [C,\eta,B] of \mathcal{R}_{g,2k} . Given such a cover f\colon\widetilde{C}\rightarrow C , the Prym-Brill-Noether locus, for some r\geq 0 , is defined as the closed set V^{r}(f)=\overline{\left\{L\in\textrm{Pic}^{2g-2}(\widetilde{C})\ |\ \textrm{Nm}_{f}(L)=\omega_{C},\ \ h^{0}(\widetilde{C},L)=r+1\ \right\}}. Similarly, the twisted Prym-Brill-Noether locus is defined as V^{r}_{\eta}(f)\coloneqq\left\{L\in\mathrm{Pic}^{2g-2+k}(\widetilde{C})\ |\ \textrm{Nm}_{f}(L)=\omega_{C}\otimes\eta,\ \textrm{and}\ h^{0}(\widetilde{C},L)\geq r+1\right\}. The goal of this paper is to study the geometry of Prym-Brill-Noether loci. Specifically, we will compute their dimension as well as their class inside the singular cohomology. Moreover, once we have the characterization of these loci as degeneracy loci of expected dimension, we immediately obtain a description of their singular locus. Our results improve on the existing literature in multiple ways. The dimension of these Prym-Brill-Noether loci was estimated from bellow in [ Kan96 ] ; while the class of these loci inside the singular cohomology was computed in [ DCP95 ] only in the unramified case [f\colon\widetilde{C}\rightarrow C]\in\mathcal{R}_{g} . 1.1 Moduli spaces of ramified curves and Brill-Noether conditions Starting with a ramified double cover f\colon\widetilde{C}\rightarrow C , generic in the moduli space \mathcal{R}_{g,2k} , it is natural to study the geometric properties of both curves and of the map itself. A study of the geometric properties of \widetilde{C} already appeared in the literature, with satisfying results when the number of ramification points is low (i.e. 2,4 or 6 ). This was motivated by the study of the birational geometry of \mathcal{R}_{g,2k} in both low genera, see [ DLC23 ] , [ LCKV23 ] , and high genus cases, see [ Bud24a ] . In these cases, it is known that \widetilde{C} is Brill-Noether general, see [ Bud24a ] and [ DLC23 ] . This property is crucial for finding effective divisors on \mathcal{R}_{g,2k} : under good numerical assumptions, the locus of covers f\colon\widetilde{C}\rightarrow C with \widetilde{C} not Brill-Noether general is an effective divisor in \overline{\mathcal{R}}_{g,2k} and its class can be explicitly computed. One of the objectives of this paper is to study the behaviour of the curve \widetilde{C} with respect to Gieseker-Petri conditions. In Theorem 3.3 and Theorem 3.6 we obtain results about the Gieseker-Petri generality of \widetilde{C} , when the double cover f has 2 or 4 ramification points: Theorem 1.1 . Let f\colon\widetilde{C}\rightarrow C be a generic double cover in \mathcal{R}_{g,2} or \mathcal{R}_{g,4} . Then the curve \widetilde{C} is Gieseker-Petri general. These results are obtained by deforming the double cover f\colon\widetilde{C}\rightarrow C to the boundary of \mathcal{R}_{g,2k} . The compactified space \overline{\mathcal{R}}_{g,2k} was described in [ Bud24a , Section 2] in terms of line bundles on the base, and can be alternatively described in terms of admissible double covers, see [ ACV03 ] . Since Gieseker-Petri conditions describe the tangent space for both twisted Prym-Brill-Noether loci, as well as their pointed counterparts, we obtain several results about the dimension and the singular locus of these spaces in Section 5 . 1.2 Prym-Brill-Noether loci for ramified double covers Similar to the case of \mathcal{R}_{g} , there is a natural way to construct Abelian varieties out of the datum of a ramified double cover, thereby relating the geometry of pointed curves to the geometry of Abelian varieties. Moreover, when k=1 , these Abelian varieties can be equipped with a principal polarization, giving a map \mathcal{P}_{g,2}\colon\mathcal{R}_{g,2}\rightarrow\mathcal{A}_{g}. that allows the study of principally polarized Abelian varieties via the geometry of curves. The image of this map sits as an intermediary between Jacobian varieties and Prym varieties, i.e. \overline{J}_{g}\subset\overline{\textrm{Im}(\mathcal{P}_{g,2})}\subset\overline{P}_{g+1}\coloneqq\overline{\textrm{Im}(\mathcal{P}_{g+1})}. This further motivates the study of \mathcal{R}_{g,2} from the perspective of the Schottky problem, which aims to identify explicit geometric, algebraic or analytic conditions that distinguish Jacobians and Prym varieties from generic Abelian varieties in \mathcal{A}_{g} . While the literature on the Schottky problem for Jacobians and Prym varieties is extensive, see [ AM67 ] , [ Mum74 ] , [ Bea77a ] among others, little is known about Abelian varieties in the image of \mathcal{P}_{g,2} . For a generic Prym variety, the singularities of its theta divisor are described through the Prym-Brill-Noether Theory of its corresponding Prym curve in \mathcal{R}_{g} . This suggests a similar phenomenon for Abelian varieties in the image of \mathcal{P}_{g,2} , further motivating the study of Prym-Brill-Noether conditions for ramified Prym curves. When f\colon\widetilde{C}\rightarrow C is a double cover in \mathcal{R}_{g,2k} , Kanev proved that V^{r}(f)=\overline{\left\{L\in\textrm{Pic}^{2g-2}(\widetilde{C})\ |\ \textrm{Nm}_{f}(L)=\omega_{C},\ \ h^{0}(\widetilde{C},L)=r+1\ \right\}}. has dimension at least g-1+k-k(r+1)-\frac{r(r+1)}{2} , see [ Kan96 ] . When k=1 , we also prove the reverse inequality for the dimension and obtain: Theorem 1.2 . Let r\geq 0 and f\colon\widetilde{C}\rightarrow C a generic double cover in \mathcal{R}_{g,2} .