betaFeedback

WorksMathematics

Bloch's conjecture for (anti-)symplectic automorphisms of K3 surfaces

Abstract

We prove Bloch's conjecture for (anti-)symplectic automorphisms of smooth projective complex K3 surfaces of Picard number two. Together with the recent result for abelian surfaces and the known K3 cases, this settles the remaining case of Bloch's conjecture for (anti-)symplectic automorphisms of smooth projective complex surfaces with trivial canonical bundle. More generally, we prove Bloch's conjecture for (anti-)symplectic derived autoequivalences of projective K3 surfaces whose Picard number is at least three, or is two with anisotropic Néron–Severi lattice. The key new result is the congruence property of the subgroup generated by reflective involutions in O(U⊕N)O(U\oplus N) for anisotropic even lattices NN of signature (1,1)(1,1). Combined with the local generation theorem and Chow-theoretic criterion of Li–Yu–Zhang, this gives the required reflective decompositions and their consequences for zero-cycles.

Introduction

Throughout, Chow groups are taken with rational coefficients unless otherwise stated. Let SS be a smooth projective complex K3 surface. The Bloch–Beilinson conjectures predict that the action of a correspondence on zero-cycles is constrained by its action on holomorphic forms. For automorphisms of SS, this leads to the following formulation of Bloch’s conjecture.

Conjecture 1.1. For any g1,g2∈Aut⁡(S)g_1, g_2 \in\operatorname{Aut}(S),

g1∗∣T(S)=g2∗∣T(S)⟹g1∗=g2∗on CH⁡0(S).g_1^*|_{T(S)} = g_2^*|_{T(S)} \Longrightarrow g_1^* = g_2^* \quad\text{on } \operatorname{CH}_0(S).

where T(S)T(S) is the transcendental lattice of SS.

Equivalently, every symplectic automorphism gg of SS satisfies

g∗=id⁡on CH⁡0(S).(1)g_* = \operatorname{id} \quad\text{on } \operatorname{CH}_0(S). \tag*{(1)}

Here gg is symplectic if g∗ωS=ωSg^*\omega_S = \omega_S for a nonzero ωS∈H0(S,ΩS2)\omega_S \in H^0(S,\Omega_S^2). For projective K3 surfaces, this is equivalent to g∗∣T(S)=id⁡g^*|_{T(S)} = \operatorname{id}. An automorphism is anti-symplectic if g∗ωS=−ωSg^*\omega_S = -\omega_S, equivalently g∗∣T(S)=−id⁡g^*|_{T(S)} = -\operatorname{id}; the corresponding prediction is g∗=−id⁡g_* = -\operatorname{id} on CH⁡0(S)hom\operatorname{CH}_0(S)_{\mathrm{hom}}.

Summary of known results

The following results establish the conjecture in several important cases.

Theorem 1.2. Let gg be a symplectic automorphism of SS. Then g∗=id⁡g_* = \operatorname{id} on CH⁡0(S)\operatorname{CH}_0(S) in each of the following cases:

(1) rk⁡2NS⁡(S)≥4\operatorname{rk}_2 \operatorname{NS}(S) \ge4 and rk⁡3NS⁡(S)≥3\operatorname{rk}_3 \operatorname{NS}(S) \ge3 [5];

(2) gg has finite order [13, 6];

(3) gg preserves an elliptic fibration on SS [3];

(4) ρ(S)≥3\rho(S) \ge3 [8 Theorem 1.3];

(5) ρ(S)=2\rho(S) = 2 and Aut⁡(S)≄Z\operatorname{Aut}(S) \not\simeq\mathbb{Z} [8 Section 7.2].

Here rk⁡p\operatorname{rk}_p is the maximal rank of a sublattice whose discriminant is prime to pp.

We briefly recall the methods underlying these results.

  1. Huybrechts relates the action of spherical twists on Chow groups to reflections in Mukai cohomology. Kneser’s generation theorem then gives the conjecture under the local lattice conditions in Theorem 1.2(1) [5].

  1. Voisin proves the conjecture for symplectic involutions using Prym varieties associated with curves on the quotient surface [13]. Combining Voisin’s theorem with his earlier results, Huybrechts proves the conjecture for symplectic automorphisms of arbitrary finite order [6 Theorem 0.1].

  1. Du–Liu use the relative Jacobian and the group structure of elliptic fibers to treat symplectic automorphisms preserving an elliptic fibration [3]. Their result also applies to automorphisms of infinite order.

  1. Building on Huybrechts’ approach, Li–Yu–Zhang study Bloch’s conjecture for derived autoequivalences. A Fourier–Mukai autoequivalence Φ:Db(S)→Db(S)\Phi: D^b(S) \to D^b(S) induces an action

ΦCH⁡:CH⁡0(S)hom⟶CH⁡0(S)hom.\Phi^{\operatorname{CH}} : \operatorname{CH}_0(S)_{\mathrm{hom}} \longrightarrow\operatorname{CH}_0(S)_{\mathrm{hom}}.

Here CH⁡0(S)hom\operatorname{CH}_0(S)_{\mathrm{hom}} is the degree-zero subgroup. The autoequivalence is symplectic (resp. anti-symplectic) if its cohomological action on H2,0(S)H^{2,0}(S) is id⁡\operatorname{id} (resp. −id⁡-\operatorname{id}), equivalently if its restriction to T(S)T(S) is id⁡\operatorname{id} (resp. −id⁡-\operatorname{id}). Li–Yu–Zhang relate its action on zero-cycles to the induced isometry of the algebraic Mukai lattice

NS⁡~(S)=Z⊕NS⁡(S)⊕Z≃U⊕NS⁡(S).\widetilde{\operatorname{NS}}(S) = \mathbb{Z} \oplus\operatorname{NS}(S) \oplus\mathbb{Z} \simeq U \oplus\operatorname{NS}(S).

Their main criterion is the following.

Theorem 1.3 ([8 Theorem 1.2]). Let Φ\Phi be a symplectic (resp. anti-symplectic) derived autoequivalence of SS. If its action on NS⁡~(S)\widetilde{\operatorname{NS}}(S) is a product of reflective involutions, then

ΦCH⁡=id⁡(resp. −id⁡)on CH⁡0(S)hom.\Phi^{\operatorname{CH}} = \operatorname{id} \quad\text{(resp. } -\operatorname{id}\text{)} \quad\text{on } \operatorname{CH}_0(S)_{\mathrm{hom}}.

The connection with the stable orthogonal group is as follows. Every derived autoequivalence preserves the orientation of the positive two-planes in NS⁡(S)\operatorname{NS}(S) [7]. Since the Mukai lattice is unimodular, the actions on the discriminant groups of NS⁡(S)\operatorname{NS}(S) and its orthogonal complement T(S)T(S) agree under the gluing identification. Consequently,

Φ symplectic⟹ΦNS⁡~∈O~+(NS⁡~(S)),Φ anti-symplectic⟹−ΦNS⁡~∈O~+(NS⁡~(S)).(2)\begin{aligned} \Phi\ \text{symplectic} &\Longrightarrow\Phi^{\widetilde{\operatorname{NS}}} \in\widetilde{O}^{+}(\widetilde{\operatorname{NS}}(S)),\\ \Phi\ \text{anti-symplectic} &\Longrightarrow-\Phi^{\widetilde{\operatorname{NS}}} \in\widetilde{O}^{+}(\widetilde{\operatorname{NS}}(S)). \tag*{(2)} \end{aligned}

Here O~+\widetilde{O}^{+} consists of orientation-preserving isometries acting trivially on the discriminant group. For the second implication, −id⁡-\operatorname{id} preserves the orientation of positive two-planes. Equivalently, Φ[1]\Phi[1] is symplectic, and (Φ[1])CH=−ΦCH(\Phi[1])^{\mathrm{CH}}=-\Phi^{\mathrm{CH}}. Thus the anti-symplectic case reduces to the symplectic case by a shift.

For ρ(S)≥3\rho(S)\ge3, Li–Yu–Zhang obtain the required decomposition by proving that the reflective subgroup is a congruence subgroup; see Theorem 2.2. When ρ(S)=2\rho(S)=2, they also prove Conjecture 1.1 under the assumption Aut⁡(S)≇Z\operatorname{Aut}(S)\not\cong\mathbb{Z} [8 Section 7.2].

The remaining automorphism case has Aut⁡(S)≅Z\operatorname{Aut}(S)\cong\mathbb{Z}. Its Néron–Severi lattice is anisotropic; see Proposition 2.4. We establish the congruence property for every anisotropic even lattice of signature (1,1)(1,1), which also yields a result for derived autoequivalences.

Main result

Our result extends the reflective decomposition method to Picard number two with anisotropic Néron–Severi lattice.

Theorem 1.4. Let SS be a smooth projective complex K3 surface. Assume that either ρ(S)≥3\rho(S)\ge3, or ρ(S)=2\rho(S)=2 and NS⁡(S)\operatorname{NS}(S) contains no nonzero isotropic vector. Then every symplectic (resp. anti-symplectic) derived autoequivalence Φ\Phi of SS satisfies

ΦCH=id⁡(resp. −id⁡)on CH⁡0(S)hom.\Phi^{\mathrm{CH}}=\operatorname{id}\quad\text{(resp. }-\operatorname{id}\text{)}\quad\text{on }\operatorname{CH}_{0}(S)_{\mathrm{hom}}.

Together with the known cases, this proves the conjecture for (anti-)symplectic automorphisms.

Theorem 1.5. Conjecture 1.1 holds for every smooth projective complex K3 surface. Moreover, if gg is a symplectic (resp. anti-symplectic) automorphism of SS, then

g∗=id⁡(resp. −id⁡)on CH⁡0(S)hom.g_{*}=\operatorname{id}\quad\text{(resp. }-\operatorname{id}\text{)}\quad\text{on }\operatorname{CH}_{0}(S)_{\mathrm{hom}}.

Remark 1.6. Theorem 1.5 also holds with integral coefficients. Indeed, Roitman’s theorem [10] implies that the integral degree-zero Chow group of a complex K3 surface is torsion-free, so the rational equalities lift to integral ones.

Remark 1.7. For abelian surfaces, Chen–Li–Zhang [2 Theorem A] prove the corresponding statement for all (anti-)symplectic derived autoequivalences on the Albanese kernel. Together with Theorem 1.5, this establishes Bloch’s conjecture for (anti-)symplectic automorphisms of every smooth projective complex surface with trivial canonical bundle.

The key input is the following statement. We write W(L)\mathcal{W}(L) for the subgroup generated by reflective involutions.

Theorem 1.8. Let NN be an even lattice of signature (1,ρ−1)(1,\rho-1), where ρ≥2\rho\ge2. If NN contains no nonzero isotropic vector, then W(U⊕N)\mathcal{W}(U \oplus N) is a congruence subgroup of O(U⊕N)O(U \oplus N).

The case ρ≥3\rho\ge3 follows from [8 Corollary 6.11]. For ρ=2\rho= 2, we adapt their argument to prove Theorem 1.8 in three steps.

  1. Anisotropy identifies the rational spin group with Res⁡K/QSL2\operatorname{Res}_{K/\mathbb{Q}}\mathrm{SL}_2 for a real quadratic field KK. The corresponding arithmetic lattice is irreducible in SL2(R)2\mathrm{SL}_2(\mathbb{R})^2.

  1. The inverse image of W±\mathcal{W}^{\pm} contains an infinite normal subgroup. Margulis’s normal subgroup theorem gives finite index, and Serre’s theorem gives the congruence property.

  1. Strong approximation for vectors of fixed norm corrects the local signs of spin lifts, giving the congruence property in the orthogonal group.

The local generation theorem of Li–Yu–Zhang then gives the reflective decomposition, and Theorem 1.3 gives the action on zero-cycles. The isotropic example in § 3 explains why the full reflective decomposition need not hold for every lattice of rank two.

Acknowledgements

The author thanks Ruxuan Zhang for helpful discussions, and Zheng Yuan and Zhihong Wang for their kind help. This work was supported by NSFC grant 12425105 and the Shanghai Pilot Program for Basic Research (No. 21TQ001).

History and AI disclosure

Our approach to Bloch’s conjecture was to decompose complicated derived autoequivalences into simpler ones whose actions on zero-cycles are understood. However, carrying out this decomposition directly in the derived category of a K3 surface runs into difficulties at low Picard number. In fact, Example 2.3 gives an obstruction to such a decomposition in general. To address these difficulties, we pursued two approaches over the past two years:

A) Lift the problem to a higher-dimensional hyper-Kähler variety, for instance a variety of K3[n]K3^{[n]}-type or O’Grady’s ten-dimensional exceptional hyper-Kähler type. The advantage of this approach is that the increased rank of the relevant lattice ensures that the desired decomposition exists. The difficulty is that the corresponding instances of Bloch’s conjecture are not known for all the reflections occurring in such a decomposition.

B) Restrict the decomposition problem to the actions induced by automorphisms, rather than arbitrary autoequivalences. These actions are represented by certain diagonal matrices, so the aim is to decompose only these matrices into reflective involutions. The difficulty is that we also knew of counterexamples to the desired decomposition for diagonal matrices in general.

The turning point came from an experiment rather than a planned step in the proof. While testing Seed Pro 2.1, I asked whether the reflection group and the diagonal group could have a nontrivial intersection. The model first proposed a counterexample already known to us. I then asked whether it could find one under the additional assumption that the Néron–Severi lattice contains no nonzero vector of square 0, 2, or −2-2. For a projective K3 surface of Picard number two, this condition is equivalent to Aut⁡(S)≅Z\operatorname{Aut}(S) \cong\mathbb{Z}. In response, it drew my attention to Serre’s congruence subgroup theorem.

That suggestion became the starting point of this paper. Building on it, one can easily use our previous approach in [8] to prove that, when the Néron–Severi lattice has rank two and is anisotropic, the associated reflective group is a congruence subgroup. I record this exchange both to acknowledge that contribution and to preserve an unexpected part of the history of the work. LLM were also used to refine and improve the exposition of the manuscript.

Preliminaries

Reflective involutions and Eichler transformations

Let U=Ze⊕ZfU = \mathbb{Z}e \oplus\mathbb{Z}f with e2=f2=0e^2 = f^2 = 0 and (e,f)=1(e,f) = 1. For an even nondegenerate lattice LL, put q(x)=(x,x)/2q(x) = (x,x)/2 and write L∨L^\vee for its dual lattice and AL=L∨/LA_L = L^\vee/L for its discriminant group. Set

O~(L)=ker⁡(O(L)⟶O(AL)).\widetilde{O}(L) = \ker\left(O(L) \longrightarrow O(A_L)\right).

For a lattice of signature (2,n)(2,n), let O+(L)O^+(L) denote the isometries preserving an orientation of the positive two-planes, and put O~+(L)=O~(L)∩O+(L)\widetilde{O}^+(L) = \widetilde{O}(L) \cap O^+(L). We recall the reflective involutions of [8 Definition 3.3].

Definition 2.1. Let W(L)W(L) be the subgroup of O(L)O(L) generated by the following integral isometries:

  1. sr(x)=x−2(x,r)r/(r,r)s_r(x) = x - 2(x,r)r/(r,r) for r2=2r^2 = 2 or r2=−2r^2 = -2;

  2. involutions jj acting trivially on ALA_L for which there is a primitive isotropic u∈Lu \in L with j(u)=−uj(u) = -u and j∣u⊥/Zu=id⁡j|_{u^\perp/\mathbb{Z}u} = \operatorname{id}.

The group W(L)W(L) is normal in O(L)O(L). For L=U⊕NL = U \oplus N, it contains the integral Eichler transformations Te(v)T_e(v) and Tf(v)T_f(v) for v∈Nv \in N; see [8 Section 6.4]. We use the convention

Te(v)e=e,Te(v)f=f+v−v22e,Te(v)n=n−(v,n)e(n∈N),T_e(v)e = e,\qquad T_e(v)f = f + v - \frac{v^2}{2}e,\qquad T_e(v)n = n - (v,n)e\quad(n \in N),

and define Tf(v)T_f(v) by interchanging ee and ff.

Congruence subgroups and reflective decompositions

For an integral lattice MM, write

O(M,m)=ker⁡(O(M)⟶Aut⁡(M/mM)),SO(M,m)=SO(M)∩O(M,m).O(M,m) = \ker\left(O(M) \longrightarrow\operatorname{Aut}(M/mM)\right),\qquad SO(M,m) = SO(M) \cap O(M,m).

A subgroup of O(M)O(M) is a congruence subgroup if it contains O(M,m)O(M,m) for some integer m>0m>0. We use W±(M)W^{\pm}(M) in the sense of [8 Section 6.4]: it is generated by pairs srsss_r s_s with r2=s2=±2r^2=s^2=\pm2 and even products of Huybrechts involutions. For M=U⊕NM=U\oplus N, these are JTe(v)JT_e(v) and their conjugates by reflections in vectors of square ±2\pm2, where J=(−id⁡U)⊕id⁡NJ=(-\operatorname{id}_U)\oplus\operatorname{id}_N; see [8 Definition 6.6]. These Huybrechts involutions are reflective, so W±(M)⊆W(M)W^{\pm}(M)\subseteq W(M).

The following is a consequence of [8 Lemma 6.1 and Theorem 6.9].

Theorem 2.2. Let NN be an even lattice of signature (1,ρ−1)(1,\rho-1) with ρ≥2\rho\geq2, and put L=U⊕NL=U\oplus N. If W±(L(−1))W^{\pm}(L(-1)) is a congruence subgroup of SO(L(−1))SO(L(-1)), then

W(L)∩O+(L)=O~+(L).W(L)\cap O^{+}(L)=\widetilde{O}^{+}(L).

Proof. Every generator of W(L)W(L) acts trivially on ALA_L. Conversely, let g∈O~+(L)g\in\widetilde{O}^{+}(L). Multiplying by se−fs_{e-f} if necessary, we may assume that det⁡g=1\det g=1. By [8 Lemma 6.1], gg has trivial rational spinor norm on L(−1)L(-1). The congruence hypothesis and [8 Theorem 6.9(1′) and (3′)] therefore give g∈W±(L(−1))⊆W(L)g\in W^{\pm}(L(-1))\subseteq W(L). Here changing the sign of the form preserves reflective involutions.

The congruence hypothesis holds when ρ≥3\rho\geq3 [8 Corollary 6.11], but the resulting equality can fail in Picard number two, as the following example shows.

Example 2.3. Let SS be a projective complex K3 surface with NS⁡(S)≃U(9)\operatorname{NS}(S)\simeq U(9), and put L=U⊕U(9)L=U\oplus U(9). In an isotropic basis (e,f,a,b)(e,f,a,b) with (e,f)=1(e,f)=1 and (a,b)=9(a,b)=9, the isometry

h=(−800−27010−27003−8030010)h=\begin{pmatrix} -8 & 0 & 0 & -27 \\ 0 & 10 & -27 & 0 \\ 0 & 3 & -8 & 0 \\ 3 & 0 & 0 & 10 \end{pmatrix}

belongs to O~+(L)\widetilde{O}^{+}(L) but not to W(L)W(L). The first assertion follows by direct verification. For the second, Appendix A constructs a homomorphism Ψ:H→Z\Psi:H\to\mathbb{Z}, where HH contains W(L)W(L) and hh, with

Ψ∣W(L)=0,Ψ(h)=−3.\left.\Psi\right|_{W(L)}=0,\qquad\Psi(h)=-3.

Thus W(L)∩O+(L)W(L)\cap O^{+}(L) has infinite index in O~+(L)\widetilde{O}^{+}(L). Here N=U(9)N=U(9) is isotropic, so this example lies outside Theorem 1.8.

For anisotropic NN of rank two, Proposition 3.1 establishes the congruence hypothesis directly.

K3 surfaces of Picard number two

Let SS be a smooth projective complex K3 surface with ρ(S)=2\rho(S)=2, and put N=NS⁡(S)N=\operatorname{NS}(S). The following criterion is [8 Lemma 7.6].

Proposition 2.4. The group Aut⁡(S)\operatorname{Aut}(S) is infinite cyclic if and only if NN contains no nonzero vector of square 00, 22, or −2-2.

This identifies the remaining automorphism case in the introduction. For the proof of Theorem 1.5, we use the following consequence of [4 Proposition 3.2].

Proposition 2.5. Let gg be an infinite-order symplectic automorphism of SS, and put A=g∗∣NA = g^{*}|_{N}. Then NN contains no nonzero isotropic vector or (−2)(-2)-vector, and

A∈SO+(N)∖{id},(A−id)N∨⊂N.A \in\mathrm{SO}^{+}(N) \setminus\{\mathrm{id}\}, \qquad(A-\mathrm{id})N^{\vee} \subset N.

Here SO+(N)\mathrm{SO}^{+}(N) denotes the determinant-one isometries preserving a chosen component of the positive cone.

Proof. The assertions concerning NN and A∈SO+(N)A \in\mathrm{SO}^{+}(N) are [4 Proposition 3.2]. Since g∗∣T(S)=idg^{*}|_{T(S)} = \mathrm{id}, the unimodularity of H2(S,Z)H^{2}(S,\mathbb{Z}) implies that AA acts trivially on ANA_N, which is the remaining assertion. □

Normal subgroup and congruence subgroup theorems

We use the following form of Margulis’s normal subgroup theorem; see [9].

Theorem 2.6. Let Γ\Gamma be an irreducible lattice in SL2(R)×SL2(R)\mathrm{SL}_{2}(\mathbb{R}) \times\mathrm{SL}_{2}(\mathbb{R}). Every infinite normal subgroup of Γ\Gamma has finite index.

We also use Serre’s congruence subgroup theorem for real quadratic fields [11 p. 499, Corollary 3].

Theorem 2.7. Let FF be a real quadratic field with ring of integers OF\mathcal{O}_{F}. Every finite-index subgroup of SL2(OF)\mathrm{SL}_{2}(\mathcal{O}_{F}) contains

ker⁡(SL2(OF)⟶SL2(OF/qOF))\operatorname{ker}\left(\mathrm{SL}_{2}(\mathcal{O}_{F}) \longrightarrow\mathrm{SL}_{2}(\mathcal{O}_{F}/q\mathcal{O}_{F})\right)

for some integer $q > 0.

A congruence subgroup theorem in signature (2,2)

Key result

We now establish the congruence property in the remaining case of an anisotropic lattice of rank two.

Proposition 3.1. Let NN be an anisotropic even lattice of signature (1,1)(1,1), and put L=U⊕NL = U \oplus N. Then W±(L)W^{\pm}(L) is a congruence subgroup of SO(L)\mathrm{SO}(L). In particular, W(L)W(L) is a congruence subgroup of O(L)\mathrm{O}(L).

Proof. Set V=L⊗QV = L \otimes\mathbb{Q} and D=−det⁡ND = -\det N. Since NN is anisotropic, D>0D > 0 is not a square. As VV contains a hyperbolic plane,

G:=Spin⁡(V)≃Res⁡K/QSL2,K=Q(D).G := \operatorname{Spin}(V) \simeq\operatorname{Res}_{K/\mathbb{Q}} \mathrm{SL}_{2}, \qquad K = \mathbb{Q}(\sqrt{D}).

Let π:G→SO(V)\pi: G \to\mathrm{SO}(V) be the spin covering, and put

Γ=π−1(SO(L))∩G(Q),W^=π−1(W±(L))∩G(Q).\Gamma= \pi^{-1}(\mathrm{SO}(L)) \cap G(\mathbb{Q}), \qquad\widehat{W} = \pi^{-1}(W^{\pm}(L)) \cap G(\mathbb{Q}).

Since KK is a real quadratic field, Γ\Gamma is an irreducible arithmetic lattice in SL⁡2(R)\operatorname{SL}_{2}(\mathbb{R}). The group W^\widehat{W} contains the normal subgroup of Γ\Gamma generated by lifts of all products susvs_{u}s_{v} with u2=v2=−2u^{2}=v^{2}=-2; conjugation by Γ\Gamma permutes these products and their lifts. This subgroup is infinite: for h∈Nh\in N with q(h)>0q(h)>0, take

u=e−f,v=e−(1+q(h))f+h;(u,v)=−2−q(h)<−2.u=e-f,\qquad v=e-(1+q(h))f+h;\qquad(u,v)=-2-q(h)<-2.

Then susvs_{u}s_{v} has infinite order. Theorem 2.6 gives [Γ:W^]<∞[\Gamma:\widehat{W}]<\infty. Since Γ\Gamma is commensurable with SL⁡2(OK)\operatorname{SL}_{2}(\mathcal{O}_{K}), Theorem 2.7 implies that W^\widehat{W} is congruence.

It remains to show that W±(L)W^{\pm}(L) contains a principal congruence subgroup of SO⁡(L)\operatorname{SO}(L). Since W^\widehat{W} is congruence, choose a finite set of primes TT and open subgroups Cp⊂G(Qp)C_{p}\subset G(\mathbb{Q}_{p}) such that

{t∈Γ:tp∈Cp for all p∈T}⊆W^.\{t\in\Gamma:t_{p}\in C_{p}\text{ for all }p\in T\}\subseteq\widehat{W}.

Since SO⁡(L)\operatorname{SO}(L) is finitely generated, its rational spinor norm has finite image, and each nontrivial value is detected at a finite place. By local continuity, we may therefore choose mm such that every g∈SO⁡(L,m)g\in\operatorname{SO}(L,m) admits a rational lift t∈Γt\in\Gamma. Since π(Cp)\pi(C_{p}) is an open neighbourhood of the identity in SO⁡(V)(Qp)\operatorname{SO}(V)(\mathbb{Q}_{p}) and π−1(π(Cp))={±1}Cp\pi^{-1}(\pi(C_{p}))=\{\pm1\}C_{p}, increasing mm further ensures that

tp∈εpCp,εp∈{±1},p∈T.t_{p}\in\varepsilon_{p}C_{p},\qquad\varepsilon_{p}\in\{\pm1\},\qquad p\in T.

It suffices to find c∈W^c\in\widehat{W} with cp∈εpCpc_{p}\in\varepsilon_{p}C_{p} for every p∈T:thenp\in T: thentc^{-1}\in\widehat{W}$ by the choice of the CpC_{p}, and hence g=π(t)∈W±(L)g=\pi(t)\in W^{\pm}(L).

To construct cc, put u=e−fu=e-f, so that q(u)=−1q(u)=-1. The quadric q(v)=−1q(v)=-1 has the rational point uu, and dim⁡V=4\dim V=4 with VRV_{\mathbb{R}} isotropic. Strong approximation away from the real place [1 Theorem 1.2], with integrality imposed outside TT, therefore gives v∈Lv\in L sufficiently close to εpu\varepsilon_{p}u for every p∈Tp\in T. In the Clifford algebra, set c=−uvc=-uv. Then

π(c)=susv∈W±(L),c∈W^.\pi(c)=s_{u}s_{v}\in W^{\pm}(L),\qquad c\in\widehat{W}.

Since −u(εpu)=εp-u(\varepsilon_{p}u)=\varepsilon_{p}, choosing vv sufficiently close ensures that cp∈εpCpc_{p}\in\varepsilon_{p}C_{p} for all p∈Tp\in T, as required. Thus SO⁡(L,m)⊆W±(L)\operatorname{SO}(L,m)\subseteq W^{\pm}(L). Taking 3∣m3\mid m also gives O⁡(L,m)⊆SO⁡(L,m)⊆W(L)\operatorname{O}(L,m)\subseteq\operatorname{SO}(L,m)\subseteq W(L). □\square

Theorem 1.8 follows from Proposition 3.1 when ρ=2\rho=2. For ρ≥3\rho\geq3, [8 Corollary 6.11] identifies a subgroup of W(U⊕N)W(U\oplus N) with O⁡~(U⊕N)\widetilde{\operatorname{O}}(U\oplus N), which contains a principal congruence subgroup.

Proof of Theorem 1.4

By (2) and the shift argument following it, we may assume that Φ\Phi is symplectic. The case ρ(S)≥3\rho(S)\geq3 is [8 Theorem 1.3]. For ρ(S)=2\rho(S)=2, put L=NS⁡~(S)≃U⊕NS⁡(S)L=\widetilde{\operatorname{NS}}(S)\simeq U\oplus\operatorname{NS}(S). Since NS⁡(S)(−1)\operatorname{NS}(S)(-1) is again anisotropic of signature (1,1)(1,1), Proposition 3.1 applied to L(−1)≃U⊕NS⁡(S)(−1)L(-1)\simeq U\oplus\operatorname{NS}(S)(-1) shows that W±(L(−1))W^{\pm}(L(-1)) is a congruence subgroup. Theorem 2.2 gives

ΦNS⁡~∈O⁡~+(L)=W(L)∩O⁡+(L).\Phi^{\widetilde{\operatorname{NS}}}\in\widetilde{\operatorname{O}}^{+}(L)=W(L)\cap\operatorname{O}^{+}(L).

Theorem 1.3 proves the assertion.

Proof of Theorem 1.5

The cases ρ(S)≥3\rho(S) \ge3 and ρ(S)=2\rho(S) = 2 with Aut⁡(S)≇Z\operatorname{Aut}(S) \not\cong\mathbb{Z} are covered by [8 Theorem 1.3 and Section 7.2]. In Picard number one, Aut⁡(S)\operatorname{Aut}(S) is either trivial or Z/2Z\mathbb{Z}/2\mathbb{Z}. The remaining case has ρ(S)=2\rho(S) = 2 and Aut⁡(S)≅Z\operatorname{Aut}(S) \cong\mathbb{Z}. Then NS⁡(S)\operatorname{NS}(S) is anisotropic by Proposition 2.4, and Theorem 1.4 applies to the derived pullback.

Appendix A. A character of an integral orthogonal group

We establish the obstruction asserted in Example 2.3. Realize L=U⊕U(9)L = U \oplus U(9) as the order

R={(xz−9wy):x,y,z,w∈Z},q(X)=det⁡X.R = \left\{ \begin{pmatrix} x & z \\ -9w & y \end{pmatrix} : x,y,z,w \in\mathbb{Z} \right\}, \qquad q(X) = \det X.

Put D=diag⁡(1,−1)D = \operatorname{diag}(1,-1) and Γ∗=R×=Γ0(9)⋊⟨D⟩\Gamma^* = R^\times= \Gamma_0(9) \rtimes\langle D \rangle.

Lemma A.1. There is a homomorphism χ:Γ∗→Z\chi: \Gamma^* \to\mathbb{Z} such that

χ(−I)=χ(D)=0,χ((−21−94))=−1.\chi(-I) = \chi(D) = 0, \qquad\chi\left(\begin{pmatrix} -2 & 1 \\ -9 & 4 \end{pmatrix}\right) = -1.

Proof. Identify Γ0(9)\SL⁡2(Z)\Gamma_0(9) \backslash\operatorname{SL}_2(\mathbb{Z}) with P1(Z/9Z)\mathbb{P}^1(\mathbb{Z}/9\mathbb{Z}) via the projective class of the bottom row. Use the representatives

∞=[0:1],t=[1:t](t∈Z/9Z),c1=[3:1],c2=[3:2].\infty= [0:1], \qquad t = [1:t]\quad(t \in\mathbb{Z}/9\mathbb{Z}), \qquad c_1 = [3:1], \qquad c_2 = [3:2].

For S=(0−110)S = \begin{pmatrix} 0 & -1 \\ 1 & 0 \end{pmatrix} and T=(1101)T = \begin{pmatrix} 1 & 1 \\ 0 & 1 \end{pmatrix}, assign weight zero to every SS-edge in the Schreier graph and weight w(v)w(v) to v→vTv \to vT, where

w(3)=w(c1)=−1,w(5)=w(c2)=1,w(v)=0 otherwise.w(3) = w(c_1) = -1, \qquad w(5) = w(c_2) = 1, \qquad w(v) = 0\text{ otherwise}.

Reverse edges have opposite weights. At each vertex, the relations S2=(ST)3=1S^2 = (ST)^3 = 1 in PSL⁡2(Z)\operatorname{PSL}_2(\mathbb{Z}) have total weight zero. Hence path weights define a homomorphism χ0:Γ0(9)→Z\chi_0 : \Gamma_0(9) \to\mathbb{Z} with χ0(−I)=0\chi_0(-I) = 0. The weights also satisfy w(v)=−w(vDT−1)w(v) = -w(vDT^{-1}), so DSD=−SDSD = -S and DTD=T−1DTD = T^{-1} imply that χ0\chi_0 is invariant under conjugation by DD. Thus χ(γDϵ)=χ0(γ)\chi(\gamma D^\epsilon) = \chi_0(\gamma) defines the required extension. Finally,

S−1T−5S−1T−2S−1T−1=(−21−94)S^{-1}T^{-5}S^{-1}T^{-2}S^{-1}T^{-1} = \begin{pmatrix} -2 & 1 \\ -9 & 4 \end{pmatrix}

has total weight −1.□-1. \square

For P,Q∈Γ∗P,Q \in\Gamma^* with det⁡P=det⁡Q\det P = \det Q, set FP,Q(X)=PXQ−1F_{P,Q}(X) = PXQ^{-1} and τ(X)=−adj⁡(X)\tau(X) = -\operatorname{adj}(X). Let H⊂O(L)H \subset O(L) be generated by all such FP,QF_{P,Q} and by τ\tau. The identity τFP,Qτ=FQ,P\tau F_{P,Q}\tau= F_{Q,P} and uniqueness of (P,Q)(P,Q) up to simultaneous sign show that

Ψ:H⟶Z,Ψ(FP,Qτϵ)=χ(P)+χ(Q)(3)\Psi: H \longrightarrow\mathbb{Z}, \qquad\Psi(F_{P,Q}\tau^\epsilon) = \chi(P) + \chi(Q) \tag*{(3)}

is a homomorphism.

Proposition A.2. Every reflective involution belongs to HH, and W(L)⊂ker⁡ΨW(L) \subset\ker\Psi.

Proof. Put ω=(01/330)\omega= \begin{pmatrix} 0 & 1/3 \\ 3 & 0 \end{pmatrix}. The normalizer calculation for an Eichler order gives

{P∈GL⁡2(Q):PRP−1=R, det⁡P=±1}=R×⊔ωR×.\{P \in\operatorname{GL}_{2}(\mathbb{Q}) : PRP^{-1} = R,\ \det P = \pm1\} = R^{\times} \sqcup\omega R^{\times}.

see [12 Section 23.1, equation (23.1.4)]. Indeed, away from 3 the determinant condition forces the local scalar factor to be a unit; at 3 the additional normalizer coset is represented by ω\omega.

For a root r∈Rr \in R, put δ=det⁡r=±1\delta= \det r = \pm1. Then sr=Fr,δr−1τ∈Hs_r = F_{r,\delta r^{-1}}\tau\in H. Now let jj be an isotropic reflective involution with j(u)=−uj(u)=-u. By definition, its −1-1 eigenspace is nondegenerate and intersects u⊥u^{\perp} in Qu\mathbb{Q}u, hence has dimension two. Thus j∈SO⁡(LQ)j \in\operatorname{SO}(L_{\mathbb{Q}}). The left–right description of the special orthogonal group of the determinant form gives

j(X)=PXQ−1,P,Q∈GL⁡2(Q),det⁡P=det⁡Q.j(X) = PXQ^{-1}, \qquad P,Q \in\operatorname{GL}_{2}(\mathbb{Q}), \qquad\det P = \det Q.

Since j2=1j^{2}=1, we have P2=Q2=cIP^{2}=Q^{2}=cI for some c∈Q×c \in\mathbb{Q}^{\times}. Since Pu=−uQPu=-uQ, the rational line im⁡u\operatorname{im}u is PP-invariant, whence c=λ2c=\lambda^{2} with λ∈Q×\lambda\in\mathbb{Q}^{\times}. Neither matrix is scalar: together with P2=Q2=cIP^{2}=Q^{2}=cI and det⁡P=det⁡Q\det P=\det Q, this would imply j=±1j=\pm1. Rescaling both by λ−1\lambda^{-1} gives P2=Q2=IP^{2}=Q^{2}=I and det⁡P=det⁡Q=−1\det P=\det Q=-1. The left and right multiplier orders of RR are both RR. Applying this to PRQ−1=RPRQ^{-1}=R gives PRP−1=QRQ−1=RPRP^{-1}=QRQ^{-1}=R. Thus P,QP,Q belong to R×⊔ωR×R^{\times}\sqcup\omega R^{\times}. Since PRQ−1=RPRQ^{-1}=R, we have PQ−1∈R×PQ^{-1}\in R^{\times}, so they belong to the same coset. If both lie in ωR×\omega R^{\times}, then jj exchanges the two summands of

AL=⟨a/9⟩⊕⟨b/9⟩≃(Z/9Z)2.A_L = \langle a/9\rangle\oplus\langle b/9\rangle\simeq(\mathbb{Z}/9\mathbb{Z})^{2}.

left–right multiplication by units acts diagonally on ALA_L, whereas X↦ωXω−1X \mapsto\omega X\omega^{-1} sends (z,w)(z,w) to (−w,−z)(-w,-z). This contradicts the trivial discriminant action of jj. Consequently P,Q∈R×P,Q \in R^{\times}, and j∈Hj \in H. Since Z\mathbb{Z} is torsion-free, Ψ\Psi vanishes on every involution in HH, and consequently on W(L)W(L).

The isometry in Example 2.3 is h=FB,Ih=F_{B,I}, where

B=(−21−94)3=(−83−2710)∈Γ0(9).B=\begin{pmatrix}-2 & 1 \\ -9 & 4\end{pmatrix}^{3}=\begin{pmatrix}-8 & 3 \\ -27 & 10\end{pmatrix}\in\Gamma_{0}(9).

Left multiplication by B∈SL⁡2(R)B\in\operatorname{SL}_{2}(\mathbb{R}) preserves the orientation of positive two-planes, and the third and fourth columns of h−Ih-I are divisible by 9. Thus h∈O~+(L)h\in\widetilde{O}^{+}(L). Since Ψ(h)=χ(B)=−3\Psi(h)=\chi(B)=-3, we have h∉W(L)h\notin W(L), and its powers represent infinitely many distinct cosets of W(L)∩O(L)W(L)\cap O(L) in O~+(L)\widetilde{O}^{+}(L).

References

References

  1. [1]Yang Cao, Zhizhong Huang, and Runlin Zhang, Effective equidistribution, arithmetic purity of strong approximation, and geometric sieve for affine quadrics, 2025, arXiv:2509.23320v1.arxiv.org/abs/2509.23320
  2. [2]Zaiyuan Chen, Zhiyuan Li, and Ruxuan Zhang, Bloch’s conjecture for equivalences between twisted abelian surfaces and applications, 2026, arXiv:2606.22323v2.arxiv.org/abs/2606.22323
  3. [3]Jiabin Du and Wenfei Liu, On symplectic automorphisms of elliptic surfaces acting on CH₀, 2022, arXiv:2103.03052v2.DOI
  4. [4]Kenji Hashimoto, JongHae Keum, and Kwangwoo Lee, K3 surfaces with Picard number 2, Salem polynomials and Pell equation, J. Pure Appl. Algebra 224 (2020), no. 1, 432–443.arxiv.org/abs/1711.02822
  5. [5]Daniel Huybrechts, Chow groups and derived categories of K3 surfaces, Current Developments in Algebraic Geometry, Math. Sci. Res. Inst. Publ., vol. 59, Cambridge University Press, Cambridge, 2012, pp. 177–195.
  6. [6]———, Symplectic automorphisms of K3 surfaces of arbitrary finite order, Math. Res. Lett. 19 (2012), no. 4, 947–951.DOI
  7. [7]Daniel Huybrechts, Emanuele Macrì, and Paolo Stellari, Derived equivalences of K3 surfaces and orientation, Duke Math. J. 149 (2009), no. 3, 461–507.DOI
  8. [8]Zhiyuan Li, Xun Yu, and Ruxuan Zhang, Bloch’s conjecture for (anti-)autoequivalences on K3 surfaces, 2024, arXiv:2305.10078v2.arxiv.org/abs/2305.10078
  9. [9]G. A. Margulis, Discrete subgroups of semisimple lie groups, Ergebnisse der Mathematik und ihrer Grenzgebiete (3), vol. 17, Springer-Verlag, Berlin, 1991.DOI
  10. [10]A. A. Roitman, The torsion of the group of 0-cycles modulo rational equivalence, Ann. of Math. (2) 111 (1980), no. 3, 553–569.
  11. [11]Jean-Pierre Serre, Le problème des groupes de congruence pour SL₂, Ann. of Math. (2) 92 (1970), no. 3, 489–527.
  12. [12]John Voight, Quaternion algebras, Graduate Texts in Mathematics, vol. 288, Springer, Cham, 2021.DOI
  13. [13]Claire Voisin, Symplectic involutions of K3 surfaces act trivially on CH₀, Doc. Math. 17 (2012), 851–860.

Paper details

vlnc:W_k6qerh8t
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 1

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents