---
title: Big Heegner Point Euler System
url: https://www.emergentmind.com/topics/big-heegner-point-euler-system
type: topic
---

# Big Heegner Point Euler System

The Big Heegner Point Euler System is the $\Lambda$-adic system of Heegner cohomology classes attached to an ordinary Hida family and an imaginary quadratic field $K$, organized so that both the modular form and the anticyclotomic conductor vary compatibly. In Howard’s construction, the system lives in the Galois cohomology of the critical twist of the big ordinary Galois representation attached to the family, specializes to Heegner classes for arithmetic points of the Hida family, satisfies Hecke- and corestriction relations in the conductor variable, and lies in Greenberg Selmer groups [1202.6358]. Later work related these classes to two-variable anticyclotomic $p$-adic $L$-functions, extended the construction to quaternionic Shimura curves and to totally real settings, and analyzed exceptional zero and Kolyvagin-system phenomena [1410.6591; 1507.04260; 2507.04980; 2510.26332].

## 1. Hida-theoretic framework and critical twists

The ambient object is a Hida family of ordinary modular forms. In Howard’s formulation one starts from Hida’s big ordinary Hecke algebra and a branch $R$ through a fixed ordinary eigenform; the associated big Galois representation $T$ is free of rank two over $R$ and carries a continuous $G_{\mathbf Q}$-action [1202.6358]. In the notation used by Castella, one likewise has a finite flat local extension $I$ of the Iwasawa algebra $\Lambda = O_L[[1+p\mathbf Z_p]]$ and a Hida family
\[
\mathbf f=\sum_{n\ge 1}\mathbf a_n q^n \in I[[q]],
\]
together with a free rank-$2$ $I$-module $\mathbf T$ endowed with a continuous Galois action [1507.04260].

A defining feature is ordinarity at $p$. For each place $v\mid p$ there is an exact sequence
\[
0 \to F_v^+(T) \to T \to F_v^-(T) \to 0,
\]
with both $F_v^\pm(T)$ free of rank one; after passing to Castella’s notation, the ordinary filtration takes the form
\[
0\to \mathbf{T}^+ \to \mathbf{T} \to \mathbf{T}^- \to 0,
\]
with $\mathbf T^-$ unramified and Frobenius acting by $\mathbf a_p\in I^\times$ [1202.6358; 1507.04260]. The critical character $\vartheta$ or $\Theta$ is then used to define a critical, self-dual twist $T^\dagger$ or $\mathbf T^\dagger$, and Howard records a perfect alternating, $G_{\mathbf Q}$-invariant $R$-bilinear pairing
\[
T^{\dagger} \times T^{\dagger} \to R(1)
\]
for the twisted representation [1202.6358].

The imaginary quadratic field $K$ is required to satisfy a Heegner hypothesis. In Howard’s 2012 construction, this is the existence of an ideal $\mathfrak n\subset \mathcal O_K$ with $\mathcal O_K/\mathfrak n \simeq \mathbf Z/N\mathbf Z$; for the results in §3 of that paper, $N$ is also coprime to $\mathrm{disc}(K)$, implying that all prime divisors of $N$ are split in $K$ [1202.6358]. Howard explicitly remarks that there is no restriction on the behavior of $p$ in $K$; it may be split, ramified, or inert [1202.6358]. In Castella’s explicit reciprocity and exceptional-zero work, by contrast, $p$ is assumed to split in $K$ [1410.6591; 1507.04260].

Within this framework, “big” refers to $\Lambda$-adic variation. The Hecke algebra, the Galois representation, the Selmer groups, and the Heegner classes are defined over a coefficient ring finite flat over an Iwasawa algebra, so that arithmetic specialization recovers the objects attached to ordinary modular forms of varying weight and character [1303.1568].

## 2. Construction from CM points and Kummer maps

Howard’s construction begins with CM points on the tower of modular curves $X_s$ attached to $\Gamma_0(N)\cap \Gamma_1(p^s)$. For $c\ge 1$ prime to $N$, one considers the order $\mathcal O_c$ of conductor $c$, the ring class field $H_c$, and CM elliptic curves
\[
E_{c,s}(\mathbf C)=\mathbf C/\mathcal O_{cp^s}
\]
equipped with level structures. This produces Heegner points
\[
h_{c,s}=(E_{c,s},\mathfrak n_{c,s},\pi_{c,s})\in X_s(\mathbf C)
\]
and, after ordinary and weight projection, classes $y_{c,s}$ with Galois transformation law
\[
y_{c,s}^\sigma=\vartheta(\sigma)y_{c,s}
\]
for $\sigma\in \mathrm{Gal}(\overline{\mathbf Q}/H_{cp^s})$ [1202.6358].

Howard then applies a twisted Kummer map. For $s\ge 1$,
\[
\mathrm{Kum}_s: H^0(H_c, J_s(L_{c,s})^{\mathrm{ord}}\otimes \zeta_s)\to H^1(\mathcal O_c,\mathrm{Ta}^{\mathrm{ord}}(J_s)\otimes \zeta_s)
\]
sends a point to a cocycle built from compatible $p^n$-division points. The resulting classes
\[
\mathfrak c_{c,s}\in H^1(\mathcal O_c,\mathrm{Ta}^{\mathrm{ord}}(J_s)\otimes \zeta_s)
\]
are made compatible in $s$ by the degeneracy maps and the $U_p$-distribution relation, and their inverse limit yields Howard’s big Heegner point
\[
\mathfrak X_c\in H^1(\mathcal O_c,T^\dagger)
\]
[1202.6358].

In the anticyclotomic direction, Castella formulates the compatible family as
\[
\mathfrak Z_\infty=\{\mathfrak Z_n\}_{n\ge 0}\in H^1_{\mathrm{Iw}}(K_\infty,\mathbf T^\dagger):=\varprojlim_n H^1(K_n,\mathbf T^\dagger),
\]
where $K_\infty/K$ is the anticyclotomic $\mathbf Z_p$-extension and $\Gamma_\infty=\mathrm{Gal}(K_\infty/K)\simeq \mathbf Z_p$ [1507.04260]. The finite-level classes satisfy a Greenberg Selmer condition, and for each prime $\mathfrak q\mid p$ the localization is constrained by
\[
\mathrm{loc}_{\mathfrak q}(\mathfrak Z_\infty)\in \ker\!\left(H^1_{\mathrm{Iw}}(K_{\infty,\mathfrak q},\mathbf T^\dagger)\to H^1_{\mathrm{Iw}}(K_{\infty,\mathfrak q},\mathbf T^{-\,\dagger})\right)
\]
[1507.04260]. Complex conjugation acts by a sign:
\[
\mathfrak Z_\infty^*=w\cdot \mathfrak Z_\infty,\qquad w\in\{\pm 1\}
\]
[1507.04260].

The construction has a direct interpolation property. For any arithmetic prime of the Hida branch, the specialization of the big Heegner system produces the Heegner-point Euler system for the specialized ordinary modular form; in weight two and trivial character it recovers the Kummer images of classical Heegner points on the modular abelian variety [1202.6358].

## 3. Euler relations, local conditions, and specialization

The designation “Euler system” is justified by precise corestriction relations. At $p$, Howard proves the $U_p$-distribution relation
\[
\alpha_s(\mathfrak c_{c,s+1}) = U_p\cdot \mathfrak c_{c,s},
\]
which induces the corresponding relation for the big classes [1202.6358]. For primes $\ell\nmid Np$ inert in $K$, Howard proves
\[
\mathrm{Cor}_{H_{c\ell}/H_c}(\mathcal Z_{c\ell}) = T_\ell\cdot \mathcal Z_c
\]
and a local Frobenius compatibility asserting that $\mathcal Z_{c\ell}$ and $\mathrm{Fr}_v(\mathcal Z_c)$ have the same image in local cohomology [1202.6358].

In the Hida-family formulation, the conductor relation is commonly written as
\[
\mathrm{cor}_{K(m\ell)/K(m)}\big(\mathfrak z_{m\ell}\big)=P_\ell\big(\mathrm{Frob}_\ell^{-1}\big)\cdot \mathfrak z_m,
\]
where after specialization at an arithmetic point $\nu$ of weight $k_\nu$ the Hecke polynomial is
\[
P_\ell(X):=1-\mathbf a_\ell X+\varepsilon(\ell)\,\ell^{k_\nu-1}X^2
\]
[1507.04260]. At primes dividing $N$, there are analogous relations involving $T_\ell$ and Atkin–Lehner involutions [1507.04260].

These classes are built to satisfy Greenberg local conditions. Howard defines the strict Greenberg local condition by unramifiedness away from $p$ and by the kernel of the map to $H^1(L_v,F_v^-(M))$ at $v\mid p$; the Selmer group is then
\[
\mathrm{Sel}_{\mathrm{Gr}}(L,M)=\ker\!\left(H^1(L,M)\to \prod_v H^1(L_v,M)/H^1_{\mathrm{Gr}}(L_v,M)\right)
\]
[1202.6358]. Proposition 2.4.5 in Howard shows that the big Heegner classes satisfy these local conditions [1202.6358].

A central feature of Howard’s theory is nontrivial specialization. Theorem 3.1.1 and Corollary 3.1.2 show that for any arithmetic prime $\mathfrak p$, and for all $s\gg 0$, the specialization $e_\chi\cdot \mathcal Z_{p^s}$ is nontrivial in $\mathrm{Sel}_{\mathrm{Gr}}(H_{p^s},V_{\mathfrak p})$ [1202.6358]. Howard explicitly states that this extends the Cornut–Vatsal nonvanishing from weight $2$ and trivial character to all ordinary modular forms in the Hida family [1202.6358].

The same local and norm compatibilities are the input for Kolyvagin descent. Büyükboduk develops Kolyvagin’s descent for Howard’s system, interpolates and controls the Tamagawa factors at bad primes, and constructs a big Kolyvagin system
\[
\boldsymbol{\kappa}\in \overline{KS}\bigl(T\otimes \Lambda^{\mathrm{ac}},F_{\mathrm{Gr}}\bigr)
\]
with initial class
\[
\kappa_1=\mathfrak z_\infty\in \widetilde H^1_{f,\mathrm{Iw}}(K_\infty,T)
\]
under the hypotheses H.Tam and H.stz [1303.1568].

## 4. Reciprocity laws and analytic interpolation

Castella proves an explicit reciprocity law relating Howard’s big Heegner points to a two-variable anticyclotomic $p$-adic $L$-function $\mathscr L_{p,\xi}(\mathbf f)$ that interpolates the Bertolini–Darmon–Prasanna $p$-adic Rankin $L$-series in Hida families [1410.6591]. On the Galois side, the key input is a Perrin–Riou-style anticyclotomic regulator
\[
\mathrm{tw}_{-1}\mathcal L_{\omega_{\mathbf f}}^\Gamma:
H^1_{\mathrm{Iw}}(K_{\infty,\mathfrak p},{}^+T(1))\to \widetilde I[[\Gamma]]
\]
whose specializations interpolate Bloch–Kato logarithms or dual exponentials, together with explicit local factors [1410.6591]. The reciprocity law itself states
\[
\mathrm{tw}_{-1}\mathcal L_{\omega_{\mathbf f}}^\Gamma\bigl(\mathrm{res}_{\mathfrak p}(\mathfrak Z_{c,\infty}^{\xi^{-1}})\bigr)
=
\mathscr L_{p,\xi}(\mathbf f)\cdot \sigma_{-1,\mathfrak p}
\]
in $\widetilde I[[\Gamma]]$ [1410.6591].

The same paper also identifies higher-weight specializations of big Heegner points with classical generalized Heegner cycles. For arithmetic $\nu$ of weight $2r>2$, one has
\[
(\mathfrak Z_{c,\infty})_\nu\cdot c^{r-1}=\mathbf z_{f_\nu,c,\alpha}
\]
in $\mathrm{Sel}_{\mathrm{Gr}}(H_{cp^\infty},T_{f_\nu}(r))$, and at finite level
\[
(\mathfrak Z_{c,0})_\nu=
\bigl(1-p^{r-1}\mathbf a_p(\nu)^{-1}\bigr)^2\cdot
\frac{\Phi^{\acute{e}t}_{f_\nu,H_c}(\Delta^{\mathrm{heeg}}_{r,c})}
{u_c(2c\sqrt{-D_K})^{r-1}}
\]
[1410.6591]. This is the precise comparison between the Greenberg-Selmer specialization of a big Heegner point and the étale Abel–Jacobi image of a classical Heegner cycle.

In the definite quaternionic setting, Longo and Vigni recast the geometric construction into big theta elements. Starting from Heegner points $P_{p^n,m}$ on definite Shimura sets satisfying the $U_p$-distribution relations, they define compatible elements
\[
\Theta_\infty^{\mathrm{alg}}(f)\in I[[\Gamma]],\qquad
L^{\mathrm{alg}}(f/K):=\Theta_\infty^{\mathrm{alg}}(f)\cdot \Theta_\infty^{\mathrm{alg}}(f)^*
\]
and prove that for arithmetic $\kappa$ of trivial nebentypus,
\[
\kappa\bigl(L^{\mathrm{alg}}(f/K)\bigr)=A_\kappa^2\cdot L^{\mathrm{an}}(f_\kappa/K),
\]
where $L^{\mathrm{an}}(f_\kappa/K)$ is the anticyclotomic $p$-adic $L$-function constructed from Chida–Hsieh theta elements [1412.7071]. In that setting, the geometric Euler system is encoded by the compatible Heegner-point system and its $\Gamma$-adic theta elements rather than by $H^1$-classes.

A quaternionic explicit reciprocity law of the same general shape is established by Longo, Occhipinti, and Vigni. For the big Heegner point
\[
\mathbf z_c=\mathrm{Cor}_{H_{cp^\infty}/K}(\mathfrak X_{cp^\infty})\in H^1_{\mathrm{Iw}}(\Gamma_c,T^\dagger),
\]
the big algebraic $p$-adic $L$-function defined through a Perrin–Riou logarithm satisfies
\[
\mathcal L_p^{\mathrm{alg}}
=
\frac{1}{\sqrt{-D_K}\,\phi_p(-1)}\cdot \mathcal L_p^{\mathrm{an}}
\qquad \text{in } I[[\Gamma_c]]
\]
[2510.04306]. This is the quaternionic analogue of Castella’s explicit reciprocity law.

## 5. Exceptional zero phenomena and derived Heegner classes

A major refinement of the theory concerns exceptional arithmetic points, especially weight-$2$ specializations with multiplicative reduction at $p$. Castella studies the semistable non-crystalline case, where $V_{f_\kappa}|_{G_{\mathbf Q_p}}$ is semistable but not crystalline, so that $D_{\mathrm{st}}(V_{f_\kappa})$ has nontrivial monodromy $N\neq 0$ and the usual interpolation factor at $p$ may vanish [1507.04260].

In this setting Castella first extends the Bertolini–Darmon–Prasanna $p$-adic Gross–Zagier formula to the $p$-new semistable non-crystalline case. For a $p$-new $f\in S_k(\Gamma_0(Np))$ of weight $k=r+2\ge 2$, and a Hecke character $\chi$ of conductor $c$ and infinity type $(r+1-j,1+j)$ with $0\le j\le r$, the formula is
\[
\frac{L_p(f)(\chi)}{\Omega_p^{\,r-2j}}
=
\Bigl(1-a_p(f)\,\chi^{-1}(\overline{\mathfrak p})\Bigr)\cdot
\left(\frac{c^{-j}}{j!}\sum_{[\mathfrak a]\in\mathrm{Pic}(\mathcal O_c)}
\chi^{-1}(\mathfrak a)\,\mathrm N(\mathfrak a)\cdot
\mathrm{AJ}_F(\Delta_{\varphi_{\mathfrak a}\varphi_0})(\omega_f\wedge \omega_A^j\eta_A^{r-j})\right)
\]
[1507.04260]. After specialization to the norm character $\mathbf N_K$, this becomes
\[
L_p(f)(\mathbf N_K)=\bigl(1-a_p(f)p^{-1}\bigr)\cdot
\langle \log_{V_f}(\mathrm{loc}_p(\kappa_f)),\omega_f\rangle
\]
[1507.04260].

The exceptional-zero theorem concerns Howard’s big Heegner points specialized at a split multiplicative point $f$ with $a_p(f)=1$. Writing
\[
\mathcal Z_{p,f,n}:=\mathrm{loc}_p\bigl(\nu_f(\mathfrak Z_n)\bigr)\in H^1(K_{n,p},{}^+V_f),
\]
Castella proves
\[
\mathcal Z_{p,f,0}=0,
\]
and constructs a derived class $\mathcal Z_{p,f,0}'\in H^1(K_p,V_f)$ such that
\[
\mathcal Z_{p,f,0}'=\mathscr L_p(f,K)\cdot \mathrm{loc}_p(\kappa_f)
\]
[1507.04260]. Here
\[
\mathscr L_p(f,K):=\mathscr L_p(f)-\mathscr L_p(\chi_K)
=\mathscr L_p(f)-\frac{\log_p(\varpi_p)}{\mathrm{ord}_p(\varpi_p)}
\]
is the difference between the Mazur–Tate–Teitelbaum $L$-invariant of $f$ and the Ferrero–Greenberg/Gross–Koblitz $L$-invariant of $\chi_K$ [1507.04260]. Pairing with $\omega_f$ yields
\[
(1-p^{-1})\cdot \langle \log_{V_f}(\mathcal Z_{p,f,0}'),\omega_f\rangle
=
\mathscr L_p(f,K)\cdot L_p(f)(\mathbf N_K)
\]
[1507.04260].

This exceptional-zero identity is the rank-one Heegner/Euler-system analogue of the Mazur–Tate–Teitelbaum phenomenon. Castella’s proof compares two evaluations of a two-variable anticyclotomic $p$-adic $L$-function along the line $(\nu,\phi)=(\nu_f,1)$, using a big logarithm map with explicit Euler factors and an improved Coleman-style map that removes the vanishing factor at the exceptional point [1507.04260].

## 6. Kolyvagin systems, quaternionic variants, and higher-rank reinterpretations

The Iwasawa-theoretic role of the Big Heegner Point Euler System is to provide a source of Kolyvagin systems and characteristic-ideal divisibilities. Büyükboduk’s descent on Howard’s Euler system yields a big Kolyvagin system over $T\otimes \Lambda^{\mathrm{ac}}$ and proves the one-sided divisibility
\[
\mathrm{char}\bigl(\widetilde H^2_{f,\mathrm{Iw}}(K_\infty,T)_{\mathrm{tors}}\bigr)
\mid
\mathrm{char}\bigl(\widetilde H^1_{f,\mathrm{Iw}}(K_\infty,T)/R_\infty\cdot \mathfrak z_\infty\bigr)^2
\]
under the stated hypotheses [1303.1568]. A key innovation there is the interpolation of Tamagawa factors through the Tamagawa element
\[
\boldsymbol\tau=\mathrm{ord}_v(\boldsymbol\partial(1))\in R,
\]
which controls bad-prime local conditions uniformly across the Hida family [1303.1568].

Quaternionic generalizations replace modular curves by Shimura curves and relax the classical Heegner hypothesis. Zerman constructs a modified universal Kolyvagin system starting from the big Heegner point Euler system of Longo–Vigni on towers of quaternionic Shimura curves, under a generalized Heegner hypothesis in which primes dividing $N^+$ split in $K$ and primes dividing $N^-$ are inert, with $N^-$ a square-free product of an even number of primes [2507.04980]. The resulting modified universal Kolyvagin system
\[
\kappa^{ac}\in \overline{\mathfrak K}(\mathbb T^{ac},\mathfrak F_{Gr},\mathcal P',\{\chi_{n,\ell}\})
\]
leads to the divisibility
\[
\mathrm{Char}_{\mathcal R^{ac}}\bigl(H^1_{\mathfrak F_{Gr}}(K,(\mathbb T^{ac})^\vee)_{\mathrm{tors}}\bigr)
\supseteq
\mathrm{Char}_{\mathcal R^{ac}}\bigl(H^1_{\mathfrak F_{Gr}}(K,\mathbb T^{ac})/\mathcal R^{ac}\cdot \kappa^{ac}(1)\bigr)^2
\]
[2507.04980].

A further extension to totally real fields is developed by Dong and Wang. Under the weak Heegner hypothesis for a CM extension $K/F$, they construct big Heegner points in the definite setting and big Heegner classes in the indefinite setting for Hida families of Hilbert modular forms, with the same characteristic Hecke and corestriction compatibilities
\[
U_p(\mathfrak P_c)=\mathrm{cor}_{H_{cp}/H_c}(\mathfrak P_{cp}),\qquad
T_v(\mathfrak P_c)=\mathrm{cor}_{H_{cv}/H_c}(\mathfrak P_{cv})
\]
and
\[
U_p(\kappa_c)=\mathrm{cor}_{H_{cp}/H_c}(\kappa_{cp}),\qquad
T_v(\kappa_c)=\mathrm{cor}_{H_{cv}/H_c}(\kappa_{cv})
\]
[2510.26332]. In the definite case they also define a totally real analogue of Longo–Vigni’s two-variable $p$-adic $L$-function
\[
L_p(f/K)=\theta_\infty\cdot \theta_\infty^*\in R_\infty
\]
[2510.26332].

At a more conceptual level, Kataoka and Sano reinterpret Heegner points within the formalism of higher-rank Euler systems. For $E/\mathbf Q$ and an imaginary quadratic field $K$ satisfying the Heegner hypothesis, they set $M=h^1(E/K)(1)$, so that the basic rank is $r=2$, and define a $\Lambda$-adic Heegner element
\[
z_{\mathrm{Hg}}^\Lambda \in Q(\Lambda)\otimes_\Lambda \bigwedge\nolimits^2 H^1(\mathcal O_K,S,T_p(E)\otimes \Lambda)
\]
as the determinantal image of the big Heegner point $y_\infty\otimes y_\infty$ [2203.08342]. Their Theorem 5.18 shows that Perrin-Riou’s Heegner point main conjecture is equivalent to the rank-two Iwasawa main conjecture
\[
\mathrm{char}_\Lambda\bigl(\bigwedge\nolimits^2 H^1/\Lambda\cdot z_{\mathrm{Hg}}^\Lambda\bigr)
=
\mathrm{char}_\Lambda\bigl(H^2\bigr)
\]
[2203.08342]. This does not alter Howard’s original construction, but it places the big Heegner point in a determinant-theoretic framework adapted to higher-rank Euler systems.

In the anticyclotomic elliptic-curve setting, the broader Heegner-point Euler-system program culminates in the proof of Perrin-Riou’s Heegner point main conjecture by Castella, Çiperiani, Skinner, and Sprung. Their distinguished class
\[
\kappa_\infty\in \mathrm{Sel}(K,\mathbf T)
\]
for the anticyclotomic big representation $\mathbf T$ satisfies
\[
\mathrm{Char}_\Lambda(X_{\mathrm{tors}})
=
\mathrm{Char}_\Lambda(\mathrm{Sel}(K,\mathbf T)/\Lambda\cdot \kappa_\infty)^2,
\]
and, when $p$ splits in $K$, this is identified with the square of the Bertolini–Darmon–Prasanna $p$-adic $L$-function via an explicit reciprocity law [1908.09512]. Although this is not Howard’s Hida-family construction, it exhibits the same structural principle: a compatible anticyclotomic Heegner system controls Selmer groups through Euler-system and reciprocity mechanisms.

Across these variants, the Big Heegner Point Euler System remains the same type of object: a compatible family of Heegner classes over a large coefficient ring, subject to Hecke-theoretic norm relations and ordinary local conditions, and designed to interpolate arithmetic information across weight, conductor, and anticyclotomic direction. Its later refinements show that trivial zeros, quaternionic settings, and higher-rank determinant formalisms do not replace the original construction; they extend its range and clarify its role in anticyclotomic Iwasawa theory.

Source: https://www.emergentmind.com/topics/big-heegner-point-euler-system