Torsion zero-cycles on a product of canonical lifts of elliptic curves (Q702906)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Torsion zero-cycles on a product of canonical lifts of elliptic curves
scientific article

    Statements

    Torsion zero-cycles on a product of canonical lifts of elliptic curves (English)
    0 references
    0 references
    0 references
    19 January 2005
    0 references
    Let \(\text{CH}_0(X)\) be the Chow group of zero-cycles of a projective geometrically connected surface \(X\) over a finite extension \(K/\mathbb{Q}_p\) of a \(p\)-adic field \(\mathbb{Q}_p\), and \(T(X)\) be the kernel of the Albanese map. \textit{W. Raskind} and \textit{M. Spiess} [Compos. Math. 121, 1--33 (2000; Zbl 0985.14003)] conjectured that \(F(X)=T(X)/T(X)_{\text{div}}\) is a finite group, where \(T(X)_{\text{div}}\) is the maximal divisible subgroup of \(T(X)\). This conjecture was proved by Colliot-Thélène for \(X\) such that the Albanese map of the scalar extension of \(X\) to an algebraic closure of \(K\) is injective. In this case \(T(X)_{\text{div}}\) is trivial. Raskind and Spiess proved this conjecture for another class of \(X\) including products of two elliptic curves with good ordinary reduction. It is also conjectured that if \(X\) has good reduction, then for any prime \(l\neq p\), the \(l\)-primary torsion subgroup of \(T(X)\) is isomorphic to that of \(T(Y)\), where \(Y\) is the special fiber of \(X\). Colliot-Thélène and Raskind proved that this is so if \(H^2(Y, \mathcal{O}_Y)=0\). In the case where \(X\) is a product of two elliptic curves with good reduction this conjecture was verified by \textit{M. Spiess} [K-Theory 17, 363--383 (1999; Zbl 0952.14008)]. As for \(p\)-primary torsion subgroup of \(F(X)\), \textit{W. Raskind} and \textit{M. Spiess} [loc. cit.] have constructed an counter-example where \(F(X)/pF(X)\) is a group of order \(p\), but \(T(Y)\) is trivial. In the paper under review the author proves that the \(p\)-primary torsion subgroup of \(F(X)\) can be arbitrary large even if one fixes the special fiber \(Y\). Namely, consider the fiber product \(\mathcal{X}_0\) of the Serre-Tate canonical lifts of two ordinary elliptic curves \(E_1\) and \(E_2\) over a finite field \(F\) of characteristic \(p\geq 5\). The generic fiber \(X_0\) of \(\mathcal{X}_0\) is an abelian surface over an unramified finite extension \(K_0\) of \(\mathbb{Q}_p\). Let \(K_n\) be the extension of \(K_0\) generated by the \(p^n\)-torsion points of the formal Brauer group of \(\mathcal{X}_0\), and \(X_n =X_0 \times_{\text{Spec}\;K_0} \text{Spec} \;K_n\). Then \(X_n\) has good reduction and \(Y =E_1 \times_{ \text{Spec} \;F} E_2\) is the special fiber of \(X_n\) for every \(n\). The main result of the paper asserts that there exists an integer \(N>0\) such that, for any \(n\geq N\), the order of the \(p\)-primary torsion subgroup of \(F(X_n)\) is greater than \(p^{n-N}\). The proof is based on the Bloch formula \(\text{CH}_0(V) \simeq H^2(V, K_2(V))\) for any smooth surface \(V\) over a field, \(K\)-cohomology and syntomic cohomology, the main result of \textit{M. Spiess} [loc. cit.], and the results of the earlier author's paper [Duke Math. J. 102, 103--119 (2000; Zbl 1114.14017)].
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    abelian surface over local field
    0 references
    Chow group
    0 references
    syntomic cohomology
    0 references