Chow Rings of Hurwitz Spaces with Marked Ramification
Emily Clader, Zhengning Hu, Hannah Larson, Amy Q. Li, Rose Lopez · 2025
Read Chow Rings of Hurwitz Spaces with Marked Ramification by Emily Clader, Zhengning Hu, Hannah Larson, Amy Q. Li, Rose Lopez (2025) 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-2505-04903.txt.
Introduction
Emily Clader and co-authors open a study of Chow rings on Hurwitz spaces, shapes that collect curves with a map of set degree to a line. Their 2025 paper starts with the degree three case, the first whose ring was not already known. Chow rings grade how sub-shapes sit inside the whole, a core tool in algebraic geometry.
For degree three, the second-level Chow group is spanned by the classes of border strata of matching depth, they prove. The route views border pieces as glued images of smaller Hurwitz spaces with one marked special fiber. Those smaller spaces, at degrees two and three, turn out to have trivial Chow rings, which clears the path.
Catalog record
| CREATOR | Emily Clader, Zhengning Hu, Hannah Larson, Amy Q. Li, Rose Lopez |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 113 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-04903
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Chow Rings of Hurwitz Spaces with Marked Ramification Emily Clader, Zhengning Hu, Hannah Larson, Amy Q. Li, Rose Lopez Abstract . The Hurwitz space ℋ ¯ k , g \overline{\mathscr{H}}_{k,g} is a compactification of the space of smooth genus- g g curves with a simply-branched degree- k k map to ℙ 1 \mathbb{P}^{1} . In this paper, we initiate a study of the Chow rings of these spaces, proving in particular that when k = 3 k=3 (which is the first case in which the Chow ring is not already known), the codimension-2 Chow group is generated by the fundamental classes of codimension-2 boundary strata. The key tool is to realize the codimension-1 boundary strata of ℋ ¯ 3 , g \overline{\mathscr{H}}_{3,g} as the images of gluing maps whose domains are products of Hurwitz spaces ℋ k ′ , g ′ ( μ ) \mathscr{H}_{k^{\prime},g^{\prime}}(\mu) with a single marked fiber of prescribed (not necessarily simple) ramification profile μ \mu , and to prove that the spaces ℋ k ′ , g ′ ( μ ) \mathscr{H}_{k^{\prime},g^{\prime}}(\mu) with k ′ = 2 , 3 k^{\prime}=2,3 have trivial Chow ring. 2020 Mathematics Subject Classification 14H10, 14H30, 14C15, 14C17 1. Introduction The moduli space \mathcal{M}_{g} of smooth genus- g curves admits a compactification \overline{\mathcal{M}}_{g} parameterizing stable curves. Since stable curves are built by gluing together smooth curves at marked points, the boundary of \overline{\mathcal{M}}_{g} admits a stratification by topological type where each stratum is a finite group quotient of a product of moduli spaces \mathcal{M}_{g^{\prime},n^{\prime}} . In [ CL24 ] , the authors use this stratification together with a study of the Chow rings of \mathcal{M}_{g^{\prime},n^{\prime}} to obtain generators for the Chow rings of \overline{\mathcal{M}}_{g} for g\leq 7 . The goal of the present paper is to initiate a similar strategy for Hurwitz spaces and their compactification by moduli of admissible covers. Let \mathscr{H}_{k,g} be the moduli space of simply-branched, degree- k , genus- g covers of smooth genus-zero curves. By the Riemann–Hurwitz formula, such covers are branched over b:=2g-2+2k distinct points on the target. Here, we study the admissible covers compactification \overline{\mathscr{H}}_{k,g}\supseteq\mathscr{H}_{k,g} , which is the quotient of the moduli space of admissible covers by the S_{b} -action permuting the branch points. Thus, in our case, the branch divisor defines a morphism \overline{\mathscr{H}}_{k,g}\to\overline{\mathcal{M}}_{0,b}/S_{b} . When k=2 , this morphism induces an isomorphism on coarse moduli spaces, and so an isomorphism on Chow rings with rational coefficients: (1) A^{*}(\overline{\mathscr{H}}_{2,g})=A^{*}(\overline{\mathcal{M}}_{0,b}/S_{b})=A^{*}(\overline{\mathcal{M}}_{0,b})^{S_{b}}. The boundary of \overline{\mathscr{H}}_{k,g} admits a stratification by the topological type and branching behavior of the source curve. As a first question towards understanding the intersection theory of \overline{\mathscr{H}}_{k,g} , we may ask: When do the fundamental classes of closures of boundary strata generate the Chow groups A^{i}(\overline{\mathscr{H}}_{k,g}) ? For k=2 , ( 1 ) implies that the fundamental classes of the closures of boundary strata generate the Chow groups of \overline{\mathscr{H}}_{2,g} for all codimensions i , because the analogous statement is true on \overline{\mathcal{M}}_{0,b} by [ Kee92 ] . For i=1 , the Picard Rank Conjecture predicts that A^{1}(\overline{\mathscr{H}}_{k,g}) is generated by the fundamental classes of boundary divisors (that is, the closures of codimension- 1 boundary strata). The conjecture is known to hold for k\leq 5 [ DP15 ] and k>g-1 [ Mul23 ] . However, little is known about the higher-codimension Chow groups. Here, we study the first case of interest, which is codimension i=2 for covers of degree k=3 . Theorem 1.1 . The Chow group A^{2}(\overline{\mathscr{H}}_{3,g}) is generated by the fundamental classes of closures of boundary strata of codimension 2 . It was previously shown in [ CL22 ] that A^{*}(\mathscr{H}_{3,g})=\mathbb{Q} , so by excision, any codimension- 2 class on \overline{\mathscr{H}}_{3,g} is the pushforward of a codimension- 1 class supported on the boundary. Thus, the main work in proving Theorem 1.1 is to show that the codimension- 1 locally closed boundary strata have trivial Chow groups in codimension 1 . In fact, we will show that each codimension- 1 boundary stratum of \overline{\mathscr{H}}_{3,g} has trivial Chow ring. For degree- 3 covers, each of these codimension- 1 boundary strata is a finite group quotient of a product of Hurwitz spaces with marked ramification , which, for partitions \mu of k , are the moduli spaces \mathscr{H}_{k^{\prime},g^{\prime}}(\mu) of degree- k^{\prime} covers C\to\mathbb{P}^{1} that have ramification profile \mu over a marked special fiber and are simply branched elsewhere. These spaces play a role analogous to \mathcal{M}_{g^{\prime},1} when studying the codimension- 1 boundary strata on \overline{\mathcal{M}}_{g} that occur as the images of gluing maps \mathcal{M}_{g^{\prime},1}\times\mathcal{M}_{g-g^{\prime},1}\to\overline{\mathcal{M}}_{g} . In Section 2.3 , we explicitly describe the codimension- 1 boundary strata of \overline{\mathscr{H}}_{3,g} : they are all finite group quotients of products of \mathscr{H}_{k^{\prime},g^{\prime}}(\mu) with k^{\prime}\leq 3 and g^{\prime}\leq g . Finite, proper maps induce surjections on Chow groups with rational coefficients, so it suffices to show triviality of the Chow rings of such products. In general, the Chow ring of a product of spaces need not be generated by the pullbacks of classes from either factor (that is, there is no Künneth formula in Chow). Nevertheless, some particularly nice spaces X satisfy the so-called Chow–Künneth generation Property (CKgP) , meaning that for any Y , the tensor product map A^{*}(X)\otimes A^{*}(Y)\to A^{*}(X\times Y) is surjective. Our main theorem is the following. Theorem 1.2 . For each partition \mu\in\{(3),(2,1),(1,1,1)\} and g\geq 0 , we have A^{*}(\mathscr{H}_{3,g}(\mu))=\mathbb{Q} and \mathscr{H}_{3,g}(\mu) has the Chow–Künneth generation Property. In Lemma 4.1 , we give a simple argument that \mathscr{H}_{2,g}(\mu) has trivial Chow ring and the CKgP for \mu=(2) and g\geq 1 , and for \mu=(1,1) and g\geq 0 . Combining this with Theorem 1.2 and the above discussion, we deduce Theorem 1.1 . 1.1. Strategy for Theorem 1.2 Our proof of Theorem 1.2 builds off of ideas used in [ CL22 ] to compute the Chow ring of \mathscr{H}_{3,g} , which we briefly recall. For consistency with later notation, we replace \mathscr{H}_{3,g} and \mathscr{H}_{3,g}(\mu) with slightly different spaces \mathcal{H}_{3,g} and \mathcal{H}_{3,g}(\mu) ; explicitly, whereas \mathscr{H}_{3,g} and \mathscr{H}_{3,g}(\mu) are quotients of corresponding parameterized Hurwitz spaces by an action of \mathrm{PGL}_{2} , the spaces \mathcal{H}_{3,g} and \mathcal{H}_{3,g}(\mu) are quotients by \mathrm{SL}_{2} . There are maps \mathcal{H}_{3,g}\rightarrow\mathscr{H}_{3,g}\;\;\;\text{ and }\;\;\;\mathcal{H}_{3,g}(\mu)\rightarrow\mathscr{H}_{3,g}(\mu) that are \mu_{2} -gerbes and hence induce isomorphisms on rational Chow rings. Therefore, for our calculations of rational Chow rings, the distinction between the two spaces is irrelevant; the technical reason why we prefer to work with the latter is explained in Section 4.2 below. A geometric point of \mathcal{H}_{3,g} is given by a degree- 3 cover \alpha:C\to\mathbb{P}^{1} . There is a naturally associated rank- 2 vector bundle E_{\alpha}:=(\alpha_{*}\mathcal{O}_{C}/\mathcal{O}_{\mathbb{P}^{1}})^{\vee} on \mathbb{P}^{1} , sometimes called the Tschirnhausen bundle . The degree of E_{\alpha} is g+2 , and there is a natural embedding C\hookrightarrow\mathbb{P}E^{\vee}_{\alpha} via which C can be viewed as the vanishing of a section f\in H^{0}(\mathbb{P}E_{\alpha}^{\vee},\gamma^{*}\det(E_{\alpha}^{\vee})\otimes\mathcal{O}_{\mathbb{P}E_{\alpha}^{\vee}}(3)), where \gamma:\mathbb{P}E_{\alpha}^{\vee}\rightarrow\mathbb{P}^{1} denotes the bundle projection. From this perspective, we have \alpha=\gamma|_{C} , and therefore the data of the degree- 3 cover \alpha:C\rightarrow\mathbb{P}^{1} can be recovered from the data of the bundle E_{\alpha} and the section f . The space of such sections forms a vector bundle \mathcal{U} over an open substack \mathcal{B} of the moduli space \mathcal{B}_{2,g+2} of rank-2, degree- (g+2) vector bundles on \mathbb{P}^{1} -bundles that