Isogenies of CM Elliptic Curves
Edgar Assing, Yingkun Li, Tian Wang, Jiacheng Xia · 2025
Read Isogenies of CM Elliptic Curves by Edgar Assing, Yingkun Li, Tian Wang, Jiacheng Xia (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-2503-05685.txt.
Introduction
Edgar Assing, Yingkun Li, Tian Wang, and Jiacheng Xia present a 2025 mathematics paper on maps between special elliptic curves. Curves with complex multiplication, called CM curves, have extra symmetry. An isogeny is a structured map between two curves, and the authors ask how often two such curves become linked after reduction modulo a prime.
They prove a polynomial lower bound: for any fixed degree m, the count of primes where the reduced curves are m-isogenous grows at least polynomially in m. The proof leans on higher Green functions and the celebrated Gross-Zagier circle of theorems. One trick approximates the coefficients of an incoherent Eisenstein series by coherent ones. A side result bounds the size of any modular form from finitely many of its coefficients.
Catalog record
| CREATOR | Edgar Assing, Yingkun Li, Tian Wang, Jiacheng Xia |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 51 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-05685
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Isogenies of CM Elliptic Curves Edgar Assing, Yingkun Li, Tian Wang, Jiacheng Xia Abstract . Given two CM elliptic curves over a number field and a natural number m m , we establish a polynomial lower bound (in terms of m m ) for the number of rational primes p p such that the reductions of these elliptic curves modulo a prime above p p are m m -isogenous. The proof relies on higher Green functions and theorems of Gross-Zagier and Gross-Kohnen-Zagier. A crucial observation is that the Fourier coefficients of incoherent Eisenstein series can be approximated by those of coherent Eisenstein series of increasing level. Another key ingredient is an explicit upper bound for the Petersson norm of an arbitrary elliptic modular form in terms of finitely many of its Fourier coefficients at the cusp infinity, which is a result of independent interest. 2010 Mathematics Subject Classification Primary 11F30, 11G15, 14K02; Secondary 11G05 1. Introduction In the study of the arithmetic of elliptic curves, it is of great interest to understand isogenies of elliptic curves in analogy to understanding their torsion points. Let E_{1} and E_{2} be two elliptic curves defined over a number field K . Let \mathfrak{p} be a prime ideal of the ring of integers \mathcal{O}_{K} such that both E_{1} and E_{2} have good reductions at \mathfrak{p} . Let E_{1,\mathfrak{p}} and E_{2,\mathfrak{p}} denote these two reductions, respectively. If there is a cyclic m -isogeny between E_{1} and E_{2} , then it is clear that there is also a cyclic m -isogeny between E_{1,\mathfrak{p}} and E_{2,\mathfrak{p}} . Conversely, one can ask for a fixed positive integer m , if E_{1,\mathfrak{p}} and E_{2,\mathfrak{p}} are m -isogenous for sufficiently many primes \mathfrak{p} , does it imply that E_{1} and E_{2} are also m -isogenous? For a rational prime \ell , such a local-global principle for non-trivial \ell -torsion points of the Mordell-Weil group of an elliptic curve was investigated by Katz [ Kat81 ] . Subsequently, the local-global principle for \ell -isogenies of elliptic curves has been extensively studied in works such as [ Ann14 , BC14 , Sut12 , Vog20 ] . In general, addressing this local-global principle is closely related to the broader question of understanding the image of the mod \ell Galois representation attached to elliptic curves. There is a significant distinction between cyclic isogenies in CM and non-CM cases. For instance, for any two non-CM elliptic curves over a number field, there is at most one number that can appear as degrees of cyclic isogenies between them. In contrast, there are infinitely many different degrees of cyclic isogenies between any two rationally isogenous CM elliptic curves. In terms of the local-global principle for cyclic m -isogenies, we want to understand how many primes are “sufficiently many” in the CM case, which leads us to the following natural question. Suppose E_{1} and E_{2} are two CM elliptic curves over K which are not m -isogenous. For a fixed positive integer m , how many prime ideals in \mathcal{O}_{K} are there such that the reductions E_{1,\mathfrak{p}} and E_{2,\mathfrak{p}} at \mathfrak{p} become m -isogenous? Note that, without fixing the degree m , similar questions have already been addressed in previous works. For instance, an upper bound of such primes with bounded norm was established in [ CHW24 , Won22 ] for non-isogenous non-CM elliptic curves using analytic tools such as effective Chebotarev density theorem and effective joint Sato-Tate distribution. For non-isogenous CM elliptic curves, an asymptotic formula can be easily derived using Deuring’s criterion (see Remark 1.3 ). For both CM and non-CM elliptic curves, Charles [ Cha18 ] applied Arakelov intersection theory to prove the existence of infinitely many primes \mathfrak{p} for which E_{1,\mathfrak{p}} and E_{2,\mathfrak{p}} are isogenous. Nonetheless, as far as we know, there is no such results in the literature when we fix the degree of cyclic isogenies. In this note, we establish an polynomial lower bound, in terms of m , for the number of primes where the reductions of two CM elliptic curves become m -isogenous. For i=1,2 , let E_{i} be fixed elliptic curves with CM by orders \mathcal{O}_{D_{i}}:=\mathbb{Z}[(D_{i}+\sqrt{D_{i}})/2] in imaginary quadratic fields K_{i}=\mathbb{Q}(\sqrt{D_{i}}) . They are defined over the ring class fields H_{D_{i}} for \mathcal{O}_{D_{i}} . The composite H:=H_{D_{1}}H_{D_{2}} is an abelian extension of K:=K_{1}K_{2} , which is a biquadratic field and a quadratic CM extension of the real quadratic field F=\mathbb{Q}(\sqrt{D}),D:=D_{1}D_{2} when D is not a rational square. Let p be a rational prime inert in K_{i} . Then it splits completely in H_{D_{i}} , hence also in H . For a prime \mathfrak{p} in H above such p , we will denote E_{i,\mathfrak{p}} the supersingular elliptic curve defined over \mathcal{O}_{H}/\mathfrak{p}\cong\mathbb{F}_{p^{2}} obtained from E_{i} by reduction modulo \mathfrak{p} . 1 1 1 It is an exercise to show that this rational prime p splits into two primes of the same inertia degree in K . Let \mathfrak{p}_{K} be any prime of K above p . Then {\mathrm{Nm}}(\mathfrak{p}_{K})=p^{2} and it is obvious that \mathfrak{p}_{K} splits completely in H_{i}K for i=1,2 . As H=H_{1}H_{2}=(H_{1}K)(H_{2}K) , \mathfrak{p}_{K} also splits completely in H . So for a prime \mathfrak{p} in H lying above p (hence above \mathfrak{p}_{K} ), we have {\mathrm{Nm}}(\mathfrak{p})=p^{2} . Then they are isogenous over \overline{\mathbb{F}}_{p} . For m\in\mathbb{N} , denote E_{1,\mathfrak{p}}\sim_{m}E_{2,\mathfrak{p}} if there is a cyclic isogeny of degree m between E_{1,\mathfrak{p}} and E_{2,\mathfrak{p}} . In this note, we will investigate the number of p such that E_{1,\mathfrak{p}}\sim_{m}E_{2,\mathfrak{p}} as a function of m , i.e., the size of the set (1.1) \pi(m):=\pi_{E_{1},E_{2}}(m):=\{p\text{ prime}\colon p\nmid m,\,E_{1,\mathfrak{p}}\sim_{m}E_{2,\mathfrak{p}}\text{ for a prime }\mathfrak{p}\text{ over }p\}. This is a finite set with elements p\in\pi(m) bounded by Dm^{2}/4 (see Remark 2.8 ). Furthermore, only primes p that are non-split in K_{i} could appear in \pi(m) , otherwise, E_{i,\mathfrak{p}} would be ordinary and have different endomorphism rings. In [ Li21 ] , one of us showed that \pi(m) is non-empty for any m . This is improved in the following theorem. Theorem 1.1 . Suppose D_{1},D_{2} are fundamental discriminants and co-prime. Then (1.2) \#\pi(m)\gg_{E_{1},E_{2},\epsilon}m^{1/5-\epsilon} for any \epsilon>0 . Remark 1.2 . The condition on D_{i} is technical, and can be removed with more work. Furthermore, Theorem 1.1 is effective as the proof uses equidistribution of Hecke orbits [ COU01 ] . In principle, one can make the constants explicit and obtain a quantitative dependence on the discriminants D_{1} and D_{2} . Remark 1.3 . We provide a rough estimate for an expected upper bound of \#\pi(m) . Assume the CM elliptic curves E_{1} and E_{2} are defined over \mathbb{Q} . Counting primes in \pi(m) is almost equivalent to counting the number of distinct prime divisors of \varphi_{m}(j_{E_{1}},j_{E_{2}}) , where \varphi_{m}(X,Y)\in{\mathbb{Z}}[X,Y] is the m -th modular polynomial and j(E_{i}) is the j -invariant of E_{i} (see also ( 1.4 )). By a result on the heights of modular polynomials [ Coh84 ] , the number of prime divisors of \varphi_{m}(j(E_{1}),j(E_{2})) is bounded above by O\left(\log|\varphi_{m}(j(E_{1}),j(E_{2}))|\right)=O_{E_{1},E_{2}}(m\log m\log\log m) . If E_{1} and E_{2} are defined over a general number field K , the O -term also depends on K . Additionally, we note that applying Deuring’s criterion alone does not provide meaningful information about the distribution of primes in \pi(m)