Non-Archimedean Equidistribution and Group Actions on Trees

Basic Information

• Paper ID: 1607.00153
• Title: Equidistribution non-archimédienne et actions de groupes sur les arbres (Non-Archimedean Equidistribution and Group Actions on Trees)
• Authors: Anne Broise-Alamichel, Jouni Parkkonen, Frédéric Paulin
• Classification: math.NT (Number Theory)
• Publication Date: July 1, 2016
• Paper Link: https://arxiv.org/abs/1607.00153

Abstract

This paper establishes equidistribution results for elements of function fields over finite fields and their quadratic irrationals in complete local fields. These results are derived from equidistribution theorems for common perpendiculars in tree quotient spaces, which are proved using ergodic properties of discrete geodesic flows. The research establishes profound connections between arithmetic problems and geometric dynamical systems.

Research Background and Motivation

Problems to be Addressed

This paper studies Diophantine approximation problems in non-Archimedean local fields, specifically:

1. How elements of function fields K over finite fields are equidistributed in their completions Kv
2. The equidistribution laws of quadratic irrationals in these complete fields
3. How integral representations of quadratic forms are distributed

Importance of the Problems

1. Classical Analogy: This is the function field analogue of the classical Mertens theorem on equidistribution of rationals in the reals. Mertens' theorem states: $\frac{\pi^2}{6}s^{-2}\sum_{p,q\in\mathbb{Z}, (p,q)=1, |q|\leq s}\Delta_{\frac{p}{q}} \stackrel{*}{\rightharpoonup} \text{Haar}_{\mathbb{R}}$
2. Theoretical Value: Reveals profound connections between equidistribution phenomena in number theory and ergodic theory of geodesic flows in geometric dynamical systems
3. Methodological Innovation: First systematic application of ergodic theory on Bruhat-Tits trees to Diophantine approximation problems in function fields

Limitations of Existing Methods

• Classical equidistribution theory for real and p-adic fields is well-developed
• However, systematic theory over function fields is limited, particularly for arbitrary finite index subgroups
• Lack of a unified geometric framework for handling different types of equidistribution problems

Research Motivation

Through the geometric structure of Bruhat-Tits trees and automorphic group actions, establish correspondence between arithmetic equidistribution and equidistribution of common perpendiculars on trees, thereby utilizing the powerful tools of ergodic theory to solve arithmetic problems.

Core Contributions

1. Arithmetic Equidistribution Theorems: Provides four equidistribution theorems (Theorems 1.1-1.4) for orbits of finite index subgroups of the modular group PGL₂(Rv)
2. Geometric Equidistribution Theorem: Proves simultaneous equidistribution theorem for common perpendiculars in tree quotient spaces (Theorem 2.1), the core technical tool
3. Methodological Innovation: Establishes systematic derivation method from geometric equidistribution on trees to arithmetic equidistribution
4. Error Term Estimates: For geometrically finite lattices, provides exponentially decaying error terms O(q^(-κ))
5. Generality of Results: Applies to arbitrary finite index subgroups, not limited to congruence subgroups

Detailed Methods

Task Definition

Main Task: Study the following equidistribution problems

Task 1 (Equidistribution of Rational Elements):

• Input: Function field K over finite field Fq, valuation v, subgroup G ⊂ GL₂(Rv)
• Output: Equidistribution law of G·∞ in P¹(Kv)
• Constraint: As the absolute value of denominators tends to infinity

Task 2 (Equidistribution of Quadratic Irrationals):

• Input: Quadratic irrational α₀ ∈ Kv, complexity function h(α) = 1/|α - α^σ|v
• Output: Equidistribution of orbit G·α₀ according to complexity h
• Constraint: h(α) ≤ s, as s → +∞

Model Architecture

1. Geometric Framework

Bruhat-Tits Tree X:

• Vertex set VX: Homology classes of Ov-lattices in Kv × Kv
• (q+1)-regular tree structure
• Boundary ∂∞X ≅ P¹(Kv) = Kv ∪ {∞}

Geodesic Space:

• $\hat{GX}$: Bartels-Lück space consisting of generalized geodesics
• Geodesic ℓ: ℝ → X satisfying:
  • ℓ(0) ∈ VX
  • Isometric on some closed interval
  • Constant on each connected component of the complement
• Discrete geodesic flow: (t,ℓ) ↦ {s ↦ ℓ(s+t)}, t ∈ ℤ

2. Lattices and Quotient Spaces

Nagao Lattice Γ = PGL₂(Rv):

• Discrete subgroup in Aut(X)
• Volume of quotient graph Γ\X: $\|\text{vol}_{\Gamma\\X}\| = \sum_{[x]\in\Gamma\backslash VX}\frac{1}{|\Gamma_x|}\Delta_{[x]}$
• For K = Fq(Y), v = v∞: $\|\text{vol}_{\Gamma\\X}\| = \frac{2}{(q-1)(q^2-1)}$

3. Subtrees and Measures

Key Subtrees:

• D₋, D₊: Appropriate non-empty subtrees of X
• Family (γD±)γ∈Γ/ΓD± locally finite in X

Natural Measures:

• $\partial^1_∓D^±$: Space of geodesic rays entering/leaving D±
• $\tilde{\sigma}^∓_{D^±}$: Natural Borel measure on it
• Aut(X)-invariant probability measure on ray subspaces with fixed starting point

4. Common Perpendiculars

When D₋ and γD₊ are disjoint:

• λγ = d(D₋, γD₊): Length of common perpendicular
• α⁻γ: 0, λγ → X: Parameterization starting from D₋
• α⁺γ: -λγ, 0 → X: Parameterization reaching D₊

Technical Innovation Points

1. Geometric-Arithmetic Correspondence

Key Observation: Establish the following correspondence

2. Ergodic Theory Tools

Mixing Properties:

• Utilize Bowen-Margulis measure on complete geodesic space GX
• Geodesic flow squared has mixing property
• For geometrically finite lattices: exponential mixing with decay rate O(q^(-κ))

Key Lemma: Image of Bowen-Margulis measure under origin map is a multiple of vol_Γ\X

3. Derivation Strategy

Proof Strategy for Theorem 1.1:

1. Take D₋ = D₊ = H∞ (horoball centered at ∞)
2. Center of γD₊ is γ∞ = a/c, (a,c) = 1
3. Common perpendicular length: -2v(c) = 2ln|c|v/ln qv
4. Apply Theorem 2.1 via continuous map θ: ∂¹₊D₋ → ∂∞X
5. Use θ*(σ̃⁺D₋) = HaarKv

Proof Strategy for Theorem 1.2:

1. Take D₋ = H∞, D₊ = ]α₀, α₀^σ[
2. Common perpendicular length relates to h(α)
3. Action of γ maps D₊ to ]γα₀, (γα₀)^σ[
4. Equidistribution to Haar measure

Experimental Setup

Theoretical Framework Parameters

Function Field Setup:

• Fq: Finite field with q elements
• K: Function field of geometrically irreducible smooth projective curve C of genus g over Fq
• v: (Normalized discrete) valuation on K
• qv: Order of residue field

Typical Example (Simplified Case):

• K = Fq(Y): Rational function field
• v = v∞: Degree valuation, v∞(P/Q) = deg Q - deg P
• g = 0, qv = q, Rv = FqY

Choice of Lattices

Standard Case: Γ = PGL₂(Rv) (Nagao lattice)

• Geometrically finite
• Quotient graph Γ\X: Cuspidal ray glued at origin of an edge
• Known volume formula

General Case: G is finite index subgroup of GL₂(Rv)

• Need not be congruence subgroup
• Requires computing index GL₂(Rv) : G

Complexity Functions

Theorem 1.1: |y|v (absolute value of denominator)

Theorem 1.2: h(α) = 1/|α - α^σ|v

Theorem 1.3: hβ(α) = max{|α, β, β^σ, α^σ|v, |α^σ, β, β^σ, α|v} where a,b,c,d = (c-a)(d-b)/(c-b)(d-a) is the cross-ratio

Theorem 1.4: Norm |n(x - yβ)|v

Experimental Results

Main Theorem Statements

Theorem 1.1 (Equidistribution of Rational Elements): For any finite index subgroup G of GL₂(Rv), as s → +∞: $\frac{(q_v^2-1)(q_v+1)\zeta_K(-1)[GL_2(R_v):G]}{q_v^3 q^{g-1}[GL_2(R_v)_{(1,0)}:G_{(1,0)}]}s^{-2}\sum_{(x,y)\in G_{(1,0)}, |y|_v\leq s}\Delta_{\frac{x}{y}} \stackrel{*}{\rightharpoonup} \text{Haar}_{K_v}$

Coefficient Interpretation:

• ζK(-1): Dedekind zeta function at -1
• Exponential terms: Correction for subgroup index
• s^(-2): Same decay rate as Mertens theorem

Theorem 1.2 (Equidistribution of Quadratic Irrationals I): For any finite index subgroup G and quadratic irrational α₀: $\frac{(q_v+1)^2\zeta_K(-1)m_0[GL_2(R_v):G]}{2q_v^2(q-1)|v(\text{tr }g_0)|}s^{-1}\sum_{\alpha\in G\cdot\alpha_0, h(\alpha)\leq s}\Delta_\alpha \stackrel{*}{\rightharpoonup} \text{Haar}_{K_v}$

where:

• g₀ ∈ G fixes α₀ and v(tr g₀) ≠ 0
• m₀ = Gα₀ : g₀^ℤ
• Decay rate s^(-1)

Theorem 1.3 (Equidistribution of Quadratic Irrationals II): On Kv - {β, β^σ}: $\frac{(q_v+1)^2\zeta_K(-1)m_0[GL_2(R_v):G]}{2q_v^2(q-1)|\beta-\beta^\sigma|_v|v(\text{tr }g_0)|}s^{-1}\sum_{\alpha\in G\cdot\alpha_0, h_\beta(\alpha)\leq s}\Delta_\alpha \stackrel{*}{\rightharpoonup} \frac{d\text{Haar}_{K_v}(z)}{|z-\beta|_v|z-\beta^\sigma|_v}$

Innovation Points:

• Limiting measure is no longer Haar measure
• But absolutely continuous with respect to Haar measure
• Density function 1/(|z-β|v|z-β^σ|v)
• PGL₂(Rv)-invariant at β

Theorem 1.4 (Representation of Quadratic Forms): For ideal I and quadratic irrational β: $\frac{(q_v^2-1)(q_v+1)\zeta_K(-1)N(I)\prod_{p|I}(1+\frac{1}{N(p)})}{q_v^3(q-1)^2q^{g-1}}s^{-1}\sum_{\substack{(x,y)\in R_v\times I, xR_v+yR_v=R_v \\ |n(x-y\beta)|_v\leq s}}\Delta_{\frac{x}{y}} \stackrel{*}{\rightharpoonup} \frac{d\text{Haar}_{K_v}(z)}{|z-\beta|_v|z-\beta^\sigma|_v}$

Geometric Theorem (Core Tool)

Theorem 2.1 (Simultaneous Equidistribution of Common Perpendiculars): As t → +∞, on $\hat{GX}\times\hat{GX}$: $\frac{(q^2-1)(q+1)}{2q^2}\|\text{vol}_{\Gamma\\X}\|q^{-t}\sum_{\substack{\gamma\in\Gamma/\Gamma_{D^+} \\ 0<\lambda_\gamma\leq t}}\Delta_{\alpha_\gamma^-}\otimes\Delta_{\gamma^{-1}\alpha_\gamma^-} \stackrel{*}{\rightharpoonup} \tilde{\sigma}_{D^-}^+\otimes\tilde{\sigma}_{D^+}^-$

Key Features:

• Simultaneous equidistribution (two components)
• Exponential decay q^(-t)
• Coefficients involve tree regularity and quotient space volume

Error Terms

Geometrically Finite Case: All theorems evaluated on locally constant compactly supported functions have error terms O(s^(-κ)) or O(q^(-κ)), κ > 0

Proof Dependence:

• Exponential mixing property of geodesic flow
• Key result announced by Kwon 2016 on effective mixing in Bruhat-Tits trees

Concrete Numerical Example

For K = Fq(Y), v = v∞, Γ = PGL₂(FqY):

• Volume: $\|\text{vol}_{\Gamma\\X}\| = \frac{2}{(q-1)(q^2-1)}$
• Genus: g = 0
• Residue field order: qv = q

Substituting into Theorem 1.1, the coefficient simplifies to: $\frac{(q^2-1)(q+1)\zeta_{F_q(Y)}(-1)}{q^3(q-1)[GL_2(F_q[Y])_{(1,0)}:G_{(1,0)}]}$

Related Work

Classical Number Theory

1. Mertens Theorem (1874):
   • Equidistribution of rationals in reals
   • Direct inspiration for this paper
   • Coefficient π²/6 relates to Riemann zeta function
2. Farey Sequence Theory:
   • Studies reduced fractions with bounded denominators
   • Related to fundamental domain of modular group
   • This paper generalizes to function fields

p-adic Number Theory

3. Parkkonen-Paulin 2014:
   • Cross-ratios and generalized Mertens formulas
   • Provides geometric methods
   • This paper builds on these ideas
4. Similar Results in Qp:
   • Analogous theorems for Qp included in paper 1
   • This paper focuses on function field case

Group Actions on Trees

5. Serre "Arbres, amalgames, SL₂" 1983:
   • Foundational tree theory
   • Bass-Serre theory
   • Basic tools for this paper
6. Paulin 2004:
   • Geometrically finite automorphism groups
   • Diophantine approximation in trees
   • Defines geometric finiteness
7. Weil 1970:
   • Modular group analogue in characteristic p
   • Early study of Nagao lattices

Ergodic Theory

8. Bowen-Margulis Measure:
   • Entropy-maximizing measure in negatively curved spaces
   • Discrete version on trees
   • Core tool in this paper's proofs
9. Kwon 2016:
   • Effective mixing in Bruhat-Tits trees
   • Provides exponential decay estimates
   • Basis for error terms in this paper

Advantages of This Paper

• Unified Framework: Incorporates multiple equidistribution problems into unified geometric framework
• Generality: Applies to arbitrary finite index subgroups, not limited to congruence subgroups
• Quantitative Results: Provides explicit coefficients and error terms
• Methodological Innovation: Systematically applies ergodic theory to arithmetic problems

Conclusions and Discussion

Main Conclusions

1. Arithmetic-Geometric Correspondence: Successfully establishes precise correspondence between equidistribution problems in function fields and equidistribution of common perpendiculars on trees
2. Effectiveness of Ergodic Methods: Demonstrates that ergodic theory of discrete geodesic flows is powerful tool for studying non-Archimedean equidistribution
3. Quantitative Control: All equidistribution results have explicit constants and (for geometrically finite lattices) error terms
4. Broad Applicability: Methods apply to:
   • Arbitrary function fields (any genus)
   • Arbitrary finite index subgroups
   • Multiple complexity functions

Limitations

1. Characteristic Restriction: Theorems 1.2-1.4 require characteristic ≠ 2 (quadratic extensions needed)
2. Geometric Finiteness: Error terms only proved for geometrically finite lattices; general lattices have only asymptotic results
3. Discrete Valuation: Methods depend on discrete valuations, not directly applicable to more general valuations
4. Missing Complete Proofs: This paper is announcement-style note; complete proofs in preparation 1
5. Computational Complexity: Computing constants involves:
   • Dedekind zeta function values
   • Stabilizer subgroup indices
   • May be difficult to compute explicitly

Future Directions

Directions explicitly indicated in paper:

1. Complete Monograph: 1 Broise-Alamichel, Parkkonen, Paulin will include:
   • Complete proofs
   • More general versions
   • Analogous results for Qp
   • Further applications
2. Effective Mixing: Depends on Kwon 2016 announced results; awaiting complete proofs

Potential research directions:

3. Higher Dimensional Generalizations:
   • Actions of GLn(Rv)
   • Higher dimensional Bruhat-Tits buildings
   • More complex algebraic groups
4. Other Complexity Functions:
   • Explore other geometrically meaningful complexity functions
   • Correspond to different subtree configurations
5. Arithmetic Applications:
   • Refined results in Diophantine approximation
   • Function field version of continued fraction theory
   • Connections with coding theory
6. Dynamical Systems:
   • Deeper ergodic properties
   • Other conserved quantities
   • Connections with symbolic dynamics

In-Depth Evaluation

Strengths

1. Methodological Innovation

• Interdisciplinary Fusion: Skillfully combines number theory, algebraic geometry, ergodic theory, and geometric group theory
• Unified Perspective: Reduces seemingly different equidistribution problems to unified geometric framework
• Technical Depth: Requires deep understanding of multiple fields to construct this theory

2. Completeness of Results

• Multi-level: From geometric theorem (Theorem 2.1) to multiple arithmetic applications (Theorems 1.1-1.4)
• Quantitative Precision: Not only gives asymptotic equidistribution but explicit constants
• Error Control: Provides exponentially decaying error terms for geometrically finite lattices

3. Theoretical Value

• New Tools: Provides new research tools for Diophantine approximation in function fields
• Deep Connections: Reveals profound links between number theory and dynamical systems
• Generalizability: Framework has potential for generalization to other settings

4. Writing Quality

• Clear Structure: English abstract followed by detailed French exposition
• Clear Motivation: Naturally motivated from Mertens theorem
• Technical Accuracy: Rigorous and precise mathematical exposition

Weaknesses

1. Completeness Issues

• Missing Proofs: As announcement-style note, complete proofs require reference to preparation monograph
• Verifiability: Readers cannot immediately verify all details
• Dependencies: Some results depend on unpublished work (Kwon's effective mixing)

2. Technical Threshold

• Multi-field Knowledge: Requires reader background in multiple areas
• Geometric Intuition: Tree geometry less intuitive than hyperbolic geometry
• Symbol Density: Abundant notation may hinder readability

3. Application Scope

• Function Field Restriction: Primarily for function fields; essential differences from number fields
• Characteristic Restriction: Some results require characteristic ≠ 2
• Geometric Finiteness: Strongest results require geometric finiteness assumption

4. Computational Practicality

• Complex Constants: Involves zeta function values, exponents; actual computation may be difficult
• Asymptotic Nature: Limited understanding of behavior for finite s
• Numerical Verification: Lacks concrete numerical examples verifying theory

Impact

1. Contribution to Field

• Pioneering: First systematic application of tree ergodic theory to function field equidistribution
• Methodological: Provides paradigm for studying similar problems
• Theoretical Deepening: Enhances understanding of connections between non-Archimedean geometry and number theory

2. Practical Value

• Theoretical Tools: Provides new tools for studying function field Diophantine properties
• Cryptography: Function fields have cryptographic applications; equidistribution results potentially relevant
• Coding Theory: Connections with algebraic geometry codes worth exploring

3. Reproducibility

• Theoretical Nature: As pure mathematics theory, focuses on proofs rather than implementation
• Monograph Dependence: Complete reproduction requires publication of monograph 1
• Technical Difficulty: Requires high-level mathematical background to understand and apply

4. Subsequent Research

• Citation Potential: Provides foundational tools for function field geometry and number theory; high citation potential
• Research Directions: Opens multiple directions worthy of deep investigation
• Teaching Value: Exemplifies interdisciplinary research in modern mathematics

Applicable Scenarios

1. Direct Applications

• Function Field Diophantine Approximation: Study rational approximation problems in function fields
• Quadratic Form Theory: Representation problems for quadratic forms over function fields
• Modular Forms: Connections with function field modular forms

2. Method Borrowing

• Other Algebraic Groups: Generalization to GLn or other reductive groups
• Higher Dimensional Buildings: Application to higher dimensional Bruhat-Tits buildings
• Other Local Fields: Adaptation to other types of local fields

3. Theoretical Research

• Ergodic Theory: Case study of ergodic theory applications
• Geometric Group Theory: Example of lattice actions on trees
• Algebraic Number Theory: Study of function field arithmetic properties

4. Teaching Purposes

• Advanced Courses: Suitable as advanced topic in number theory, geometric group theory, or ergodic theory
• Interdisciplinary Examples: Demonstrates how different mathematical branches interact
• Research Training: Provides example of modern mathematical research for graduate students

References

Key references cited in the paper:

1 Broise-Alamichel, Parkkonen, Paulin. Equidistribution and counting under equilibrium states in negatively curved spaces and graphs of groups. Monograph in preparation.

• Complete version of this paper with all proofs

2 Kwon, S. Effective mixing and counting in Bruhat-Tits trees. Preprint arXiv:1506.04306, 2015.

• Provides key exponential mixing results

3 Parkkonen, J. and Paulin, F. On the arithmetic of crossratios and generalised Mertens' formulas. Ann. Fac. Scien. Toulouse 23 (2014) 967-1022.

• Prior work on geometric methods

4 Paulin, F. Groupes géométriquement finis d'automorphismes d'arbres et approximation diophantienne dans les arbres. Manuscripta Math. 113 (2004) 1-23.

• Geometric finiteness definition and properties

5 Serre, J.-P. Arbres, amalgames, SL₂. 3ème éd., Astérisque 46, 1983.

• Classical reference for tree theory

6 Weil, A. On the analogue of the modular group in characteristic p. Springer, 1970.

• Early study of function field modular groups

Overall Assessment: This is a high-level number theory research paper that skillfully applies ergodic theory to equidistribution problems in function fields. Although as an announcement-style note it lacks complete proofs, its innovative methodology, profound mathematical insights, and broad application prospects make it an important contribution to the field. For researchers working in function field number theory, geometric group theory, or ergodic theory, this paper provides valuable ideas and tools.