On a problem à la Kummer-Vandiver for function fields (Q1932375): Difference between revisions

From MaRDI portal
Importer (talk | contribs)
Changed an Item
Normalize DOI.
 
(One intermediate revision by one other user not shown)
Property / DOI
 
Property / DOI: 10.1016/j.jnt.2012.02.008 / rank
Normal rank
 
Property / cites work
 
Property / cites work: Log-algebraicity of twisted \(A\)-harmonic series and special values of \(L\)-series in characteristic \(p\) / rank
 
Normal rank
Property / cites work
 
Property / cites work: On \(L\)-functions of cyclotomic function fields / rank
 
Normal rank
Property / cites work
 
Property / cites work: Irregular primes to 163 million / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4714153 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Class-groups of function fields / rank
 
Normal rank
Property / cites work
 
Property / cites work: Kummer's theory for function fields / rank
 
Normal rank
Property / cites work
 
Property / cites work: A Herbrand-Ribet theorem for function fields / rank
 
Normal rank
Property / DOI
 
Property / DOI: 10.1016/J.JNT.2012.02.008 / rank
 
Normal rank

Latest revision as of 13:36, 16 December 2024

scientific article
Language Label Description Also known as
English
On a problem à la Kummer-Vandiver for function fields
scientific article

    Statements

    On a problem à la Kummer-Vandiver for function fields (English)
    0 references
    0 references
    0 references
    18 January 2013
    0 references
    Let \(p\) be a prime number and let \(F\) be the maximal real subfield of the \(p\)-th cyclotomic field \({\mathbb Q}(\mu_p).\) The Kummer-Vandiver conjecture predicts that \(p\) does not divide the class number of \(F\). A function field analogue of the classical Kummer-Vandiver conjecture was formulated in~[Invent. Math. 188, No. 2, 253--275 (2012; Zbl 1278.11102)] by the second named author. The purpose of this paper under review is to construct a counterexample to the function field version of the Kummer-Vandiver statement. To state the main result, let us recall the setting and statements of this analogue. Let \(A = {\mathbb F}_q[T]\) be the polynomial ring and let \(k = {\mathbb F}_q(T)\) be the rational function field in variable \(T\) over the finite field \({\mathbb F}_q\) of \(q = p^r\) elements. Let \(C\) be the Carlitz module over \(A.\) Let \(P\in A\) be a monic irreducible element and let \(K = k(C[P])\) be the extension field by adjoining \(k\) the \(P\)-torsion points \(C[P]\) of \(C.\) Then \(K/k\) is a Galois extension and there is a canonical isomorphism \(\omega_P : \mathrm{Gal}(K/k) \to \left(A/P\right)^{\times}\) which can be viewed as the mod \(P\) Teichmüller character in the function field setting. Let \(R\) be the integral closure of \(A\) in \(K\) and put \(Y_P = \mathrm{Spec} R.\) Consider the flat cohomology group \(H_P := H^1\left(Y_{P, fl}, C[P]^D\right)\) where \(C[P]^D\) denotes the Cartier dual of \(C[P].\) Then the Galois group \(\mathrm{Gal}(K/k)\) acts on the \(A/P\)-vector space \(H_P\) and it decomposes as \(H_P = \bigoplus_{n=1}^{q^{\deg P -1}} H_P\left(\omega_P^{n}\right).\) As explained in the paper of Taelman [op. cit.], the function field version of the Kummer-Vandiver question is equivalent to asking whether or not \(H_P(\omega_P^{n-1}) = 0\) if \(n\) is not divisible by \(q-1\). Let \(A_{+}\) denote the set of monic elements in \(A\) and let \(i : A_{+} \to {\mathbb Z}/p{\mathbb Z}\) be the function given by \(i\left(T^n + \alpha_1 T + \cdots + \alpha_n\right) = \mathrm{Tr}_{{\mathbb F}_q/{\mathbb F}_p}(\alpha_1) \in {\mathbb Z}/p{\mathbb Z}\). For a positive integer \(n\), we put \(\beta(n) = \sum_{m\geq 0}\sum_{a\in A_m} a^n\) (the Bernoulli-Goss polynomial) and \(\gamma(n) = \sum_{m\geq 0}\sum_{a\in A_m} i(a) a^n\) where \(A_m\) denotes the subset of \(A_+\) of monic elements of degree \(m\). The main results (Theorem~2 and Theorem~3) of this paper can be stated as follows. Let \(P\in A\) be monic irreducible polynomial of degree \(d\) and such that \(i(P) \neq 0. \) Let \(n\) be an integer satisfying the following conditions: (1) if \(n\) is not divisible by \(q-1\) then \(\beta(n)\) is divisible by \(P\) for odd \(p\) or for \(p =2\) the special value \(L(1, K/k, \omega_P^n)\) of the \(L\)-function associated to \(K/k\) and \(\omega_P^n\) is divisible by 4; and (2) \(\gamma(n)\) is divisible by \(P\). Let \(Q(T) = P(T^p - T)\) and let \(N = n(q^{pd}-1)/(q^d -1)\) for odd \(p\) otherwise \(N = n(q^d +1)\) for \(p=2\). Then \(Q\) is irreducible and \(H^1\left(Y_{Q, fl}, C[Q]^D\right)\left( \omega_Q^{-N-1} \right) \neq \{0\}.\) As an application of their main results, the authors provide explicit counterexamples as a negative answer to the function field analogue of the Kummer-Vandiver question. For instance, for \(q=3, P(T)= T^3 - t^2 + 1\) and \(n = 13\) the authors show that \(H^1\left(Y_{Q, fl}, C[Q]^D\right)\left( \omega_Q^{-N-1} \right) \neq \{0\}\) with \(N = 9841.\) Another counterexample for \(q = 4\) is also given. In the last section of the paper, the authors provide a heuristic argument to show that for odd \(q\) and sufficiently large \(X\), one could expect that there are at least \(c X^{1/p}(\log X)^{-1}\) (for some constant \(c>0\)) prime elements of residue cardinality at most \(X\) that are counterexamples to the function field analogue of the Kummer-Vandiver question.
    0 references
    Kummer-Vandiver conjecture
    0 references
    Carlitz module
    0 references
    cyclotomic function field
    0 references
    cyclotomic character
    0 references

    Identifiers