On the field of definition of \(p\)-torsion points on elliptic curves over the rationals (Q368631)

Let \(d\) be an integer \(\geq 1\). Let us denote by \(S(d)\) the set of primes \(p\) for which there exists a number field \(K\) of degree \(\leq d\) over \(\mathbb Q\) and an elliptic curve \(E/K\), such that \(p\) divides the order of the torsion subgroup \(E(K)_{\text{tors}}\). In 1978 \textit{B. Mazur} [Invent. Math. 44, 129--162 (1978; Zbl 0386.14009)] has proved that \(S(1)=\big\{ 2,3,5,7\big\}\). \textit{S. Kamienny} and \textit{B. Mazur} [Astérisque 228, 81--98 (1995; Zbl 0846.14012)] have established in 1995 that \(S(2)=\big\{ 2,3,5,7,11,13\big\}\). \textit{L.~Merel} [Invent. Math. 124, No. 1--3, 437--449 (1996; Zbl 0936.11037)] has proved that \(S(d)\) is finite, and if \(p\in S(d)\) then \(p\leq d^{3d^2}\). This bound has been improved by Oesterlé, replacing \(d^{3d^2}\) by \(\bigl(3^{{d\over 2}}+1\bigr)^2\). \textit{P. Parent} [J.. Théor. Nombres Bordx. 15, No. 3, 831--838 (2003; Zbl 1072.11037)] has proved the equality \(S(3)=S(2)\). Recently, \textit{M. Derickx} et al. [``Torsion points on elliptic curves over number fields of small degree'' (in preparation, private communication)] have shown that \(S(4)=S(3)\cup \big\{ 17\big\}\), \(S(5)=S(4)\cup \big\{ 19\big\}\), and that \(S(6)\) is contained in \(S(5)\cup \big\{ 37,73\big\}\). In this paper, the author investigates the case of elliptic curves \(E/K\) that arise from elliptic curves \(E/\mathbb Q\). Let \(S_{\mathbb Q}(d)\) be the set of primes \(p\) for which there exists a number field \(K\) of degree \(\leq d\) over \(\mathbb Q\) and an elliptic curve \(E/\mathbb Q\), such that \(p\) divides the order of the torsion subgroup \(E(K)_{\text{tors}}\). The set \(S_{\mathbb Q}(d)\) in contained in \(S(d)\) and \(S_{\mathbb Q}(1)=S(1)\). The author gives bounds for the primes in \(S_{\mathbb Q}(d)\). In particular, he shows that if \(p\in S_{\mathbb Q}(d)\) is a prime number \(\geq 11\), distinct from 13 and 37, then \(p\leq 2d+1\). Moreover, he determines \(S_{\mathbb Q}(d)\) for all \(d\leq 42\). For instance, for \(12\leq d\leq 20\), one has \(S_{\mathbb Q}(d)=\big\{ 2,3,5,7,11,13,17,19,37\big\}.\) He also gives a conjectural formula for all \(d\geq 1\). Furthermore, if Serre's uniformity problem is answered positively, his conjectural formula is valid for all sufficiently large \(d\). Under further assumptions on the non-cuspidal points on modular curves that parametrize the \(j\)-invariants associated to Cartan subgroups, his formula is valid for all \(d\geq 1\).