stabilization on ideal class groups in potential cyclic towers
Jianing Li · 2025
Read stabilization on ideal class groups in potential cyclic towers by Jianing Li (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-2505-15224.txt.
Introduction
Jianing Li studies towers of number fields where the arithmetic eventually stops changing. In Iwasawa theory, Fukuda showed that class groups along a cyclic p-tower stabilize. Li widens the stage to what he calls potential cyclic p-towers, where the Galois group mixes a p-part with an arbitrary second group, as in radical towers built by stacking p-th roots. He extends Fukuda's stabilization result to this broader family of towers. The paper also carries Iwasawa's class number formula over from the classic setting to potential Zp-extensions.
Catalog record
| CREATOR | Jianing Li |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 19 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-15224
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
stabilization on ideal class groups in potential cyclic towers Jianing Li Abstract . Let p p be a prime and let F F be a number field. Consider a Galois extension K / F K/F with Galois group H ⋊ Δ H\rtimes\Delta where H ≅ ℤ p H\cong{\mathbb{Z}}_{p} or ℤ / p d ℤ {\mathbb{Z}}/p^{d}{\mathbb{Z}} , and Δ \Delta is an arbitrary Galois group. The subfields fixed by H p n ⋊ Δ H^{p^{n}}\rtimes\Delta ( n = 0 , 1 , ⋯ ) (n=0,1,\cdots) form a tower which we call it a potential cyclic p p -tower in this paper. A radical p p -tower is a typical example, say ℚ ⊂ ℚ ( a p ) ⊂ ℚ ( a p 2 ) ⊂ ⋯ {\mathbb{Q}}\subset{\mathbb{Q}}(\sqrt[p]{a})\subset{\mathbb{Q}}(\sqrt[p^{2}]{a})\subset\cdots where a ∈ ℤ a\in{\mathbb{Z}} . We extend the stabilization result of Fukuda in Iwasawa theory on p p -class groups in cyclic p p -towers to potential cyclic p p -towers. We also extend Iwasawa’s class number formula in ℤ p {\mathbb{Z}}_{p} -extensions to potential ℤ p {\mathbb{Z}}_{p} -extensions. 1. Introduction Let p be a prime number and let F be a number field. Let F_{\infty}/F be a {\mathbb{Z}}_{p} -extension with n -th layer F_{n} . The classical Iwasawa theory studies the p -class groups of F_{n} when n varies. The stationary result in Iwasawa theory proved by Fukuda [ Fuk94 ] states that if the p -class groups of F_{n_{0}} and F_{n_{0}+m} are isomorphic for m=1 , then the same holds for each m\geq 1 ; here n_{0} is some integer such that F_{\infty}/F_{n_{0}} is totally ramified at every ramified prime. Using a similar argument, a stabilization result can be proved in a finite cyclic p -tower; see [ LOXZ22 , Proposition 3.2] . It is used there to explicitly compute certain p -class groups of fields in radical towers, like {\mathbb{Q}}(\mu_{p^{2}},\sqrt[p^{n}]{a}) with n varies, which are not Galois extensions in general. It motivates the author to think whether there are stationary results for p -class groups in non-cyclic radical towers, say {\mathbb{Q}}\subset{\mathbb{Q}}(\sqrt[p]{a})\subset{\mathbb{Q}}(\sqrt[p^{2}]{a})\subset\cdots , where a is an integer. In this paper they will be called potential cyclic p -towers, and we prove a stationary result in such towers and obtain an ”Iwasawa like” class number formula for such a tower. Definition 1.1 (Potential cyclic p -tower) . Let p be a prime number. Let F be a number field. Let K/F be a Galois extension and let G={\mathrm{Gal}}(K/F) . Assume that G=H\rtimes\Delta is a semidirect product of two subgroups ( H being normal), and that H\cong{\mathbb{Z}}/p^{d}{\mathbb{Z}} or H\cong{\mathbb{Z}}_{p} , while \Delta may be an arbitrary (finite or infinite) Galois group. Let F_{n} be the intermediate field of K/F fixed by the group H^{p^{n}}\rtimes\Delta , so F_{0}=F . We then get a tower F_{0}\subset F_{1}\subset\cdots\subset F_{d} ( d=\infty in the case H={\mathbb{Z}}_{p} ) with [F_{n+1}:F_{n}]=p for each n , and we say this tower is a potential cyclic p-tower (in the case H\cong{\mathbb{Z}}_{p} , we may call it a potential {\mathbb{Z}}_{p} -tower). Example 1.2 . (1) {\mathbb{Z}}_{p} -extensions and {\mathbb{Z}}/p^{d}{\mathbb{Z}} -extensions clearly give potential cyclic p -towers. (2) Suppose a\in F such that x^{p}-a is an irreducible polynomial. Then the radical tower F_{n}=F(\sqrt[p^{n}]{a}) ( n=0,1,\cdots,\infty ) is a potential cyclic p-tower, with K=F_{\infty}(\mu_{p^{\infty}}) . A particular interesting example is {\mathbb{Q}}\subset{\mathbb{Q}}(\sqrt[p]{p})\subset{\mathbb{Q}}(\sqrt[p^{2}]{p})\subset\cdots\subset{\mathbb{Q}}(\sqrt[p^{\infty}]{p}) , in which every subextension is not Galois. To obtain results on class groups, we make the following hypothesis on ramifications: RamHyp: Let K/F and \{F_{n}\} be as in a potential cyclic p -tower. If v is a prime of F which is ramified in K/F , we assume that the inertia group I_{v}(K/F) of v is conjugate to a subgroup of G of the form H\rtimes\Delta_{v} , where \Delta_{v} is a subgroup of \Delta . Moreover, assume that there exists a prime v which is totally ramified in K/F , in other words I_{v}(K/F)=G . Example 1.3 . (1) In a {\mathbb{Z}}_{p} -extension F_{\infty}/F , RamHyp holds for F_{\infty}/F_{n_{0}} when n_{0} is sufficiently large. (2) Regarding Example 1.2 (2), we consider the 2 -dimensional p -adic tower F_{n,m}=F(\sqrt[p^{n}]{a},\mu_{p^{m}}) and let K=F_{\infty,\infty} . Then one can show that the potential cyclic tower arising from K/F_{N,N} satisfies RamHyp when N is sufficiently large. If E is any number field, let A_{E} denote the p -class group (i.e. p -Sylow subgroup of the class group) of E . Our first main result is the following: Theorem 1.1 . Let \{F_{n}\} be a potential cyclic p -tower satisfying RamHyp . If A_{F_{1}}\cong A_{F_{0}} , then A_{F_{n}}\cong A_{F_{0}} for n\geq 1 . More generally, for each k\geq 1 , if A_{F_{1}}/p^{k}A_{F_{1}}\cong A_{F_{0}}/p^{k}A_{F_{0}} , then A_{F_{n}}/p^{k}A_{F_{n}}\cong A_{F_{0}}/p^{k}A_{F_{0}} for n\geq 1 . By RamHyp , the norm map A_{F_{n}}\to A_{F_{n-1}} is always surjective for each n , so the condition A_{F_{1}}\cong A_{F_{0}} is equivalent to |A_{F_{1}}|=|A_{F_{0}}| . For a finite abelian group A , recall that its p^{i} -rank (i\geq 1) is defined as {\mathrm{rank}}_{p^{i}}A:=\mathrm{dim}_{{\mathbb{F}}_{p}}(p^{i-1}A)/(p^{i}A) . A direct consequence is the following: Corollary 1.4 . If |A_{F_{1}}|=|A_{F_{0}}| , then |A_{F_{n}}|=|A_{F_{0}}| . More generally, for each k\geq 1 , if {\mathrm{rank}}_{p^{i}}A_{F_{1}}={\mathrm{rank}}_{p^{i}}A_{F_{0}} holds for every i\leq k , then the same equalities hold for A_{F_{n}} and A_{F_{0}} for each n\geq 1 . Fix an integer m\geq 0 , then the tower F_{n}={\mathbb{Q}}(\mu_{p^{m}},\sqrt[p^{n}]{p}) ( n=0,1,\cdots ) satisfies (see [ LOXZ22 , Lemma 4.1] ) the condition in Theorem 1.1 with K={\mathbb{Q}}(\mu_{p^{\infty}},\sqrt[p^{\infty}]{p}) . If p is a regular prime, it is easy to show that A_{F_{n}} is trivial for each n and each m . Let p=37 , the smallest irregular prime with A_{{\mathbb{Q}}(\mu_{p})}={\mathbb{Z}}/p{\mathbb{Z}} ; then it is computed by McCallum and Sharifi that A_{{\mathbb{Q}}(\sqrt[p]{p})}=0 and A_{{\mathbb{Q}}(\mu_{p},\sqrt[p]{p})}={\mathbb{Z}}/p{\mathbb{Z}} ; see [ MS03 ] . It follows from Theorem 1.1 we indeed have A_{{\mathbb{Q}}(\mu_{p},\sqrt[p^{n}]{p})}={\mathbb{Z}}/p{\mathbb{Z}} and A_{{\mathbb{Q}}(\sqrt[p^{n}]{p})}=0 for each n\geq 1 . Other explicit examples can also be found in [ LOXZ22 ] and whose proofs can be simplified using Theorem 1.1 . Our second main result is a generalization of Iwasawa’s classical class number formula (see [ Iwa58 ] ) in a {\mathbb{Z}}_{p} -extension. Theorem 1.2 . Let \{F_{n}\} be a potential cyclic p -tower satisfying RamHyp . Let e_{n}\geq 0 such that p^{e_{n}}=|A_{F_{n}}| . Then there exit non-negative integers \mu,\lambda,\nu such that e_{n}=\mu p^{n}+\lambda n+\nu when n is sufficiently large. This has been proved by Lei [ Lei17 ] under the following three conditions (1): \Delta\cong{\mathbb{Z}}_{p} ; (2): there is only one prime {\mathfrak{p}} of F lying above p ; (3): {\mathfrak{p}} is totally ramified in K/F . We mention a related result of Caputo and Nuccio [ CN23 ] . Let k be an imaginary quadratic field, let K/k be the anti-cyclotomic {\mathbb{Z}}_{p} -extension and suppose p\neq 2 . Then {\mathrm{Gal}}(K/{\mathbb{Q}})\cong{\mathbb{Z}}_{p}\rtimes{\mathbb{Z}}/2{\mathbb{Z}}:=H\rtimes\Delta . The subgroups H^{p^{n}}\rtimes\Delta correspond to a potential {\mathbb{Z}}_{p} -tower {\mathbb{Q}}=F_{0}\subset F_{1}\subset\cdots F_{\infty} , which are called ”fake {\mathbb{Z}}_{p} -extensions of dihedral type” in [ CN23 ] . They prove an ”Iwasawa like” class number formula but the \mu,\nu -invariants are only in {\mathbb{Z}}[1/2] (see [ CN23 , Theorem 4.6] ). We remark that Theorem 1.2 can not directly apply to this situation since the RamHyp does not hold in general. Acknowledgment. The author thanks Yuan Ren and Lim Meng-Fai for helpful discussions on this paper. The