Effective equidistribution of Galois orbits for mildly regular test functions

Basic Information

• Paper ID: 2411.16294
• Title: Effective equidistribution of Galois orbits for mildly regular test functions
• Authors: Emanuel Carneiro, Mithun Kumar Das
• Classification: math.NT (Number Theory), math.CA (Classical Analysis)
• Publication Date: October 28, 2025 (arXiv v3)
• Paper Link: https://arxiv.org/abs/2411.16294

Abstract

This paper provides a detailed effective version of Bilu's equidistribution theorem for Galois orbits of sequences of small height points on N-dimensional algebraic tori. The authors identify the qualitative dependence of convergence rates on the regularity of test functions and develop a general Fourier analysis framework, extending earlier results by Petsche (2005) and D'Andrea, Narváez-Clauss, and Sombra (2017).

Research Background and Motivation

Research Problem

This paper studies the effective version of Bilu's equidistribution theorem, namely quantifying the convergence rate of uniform distribution on the unit polytorus $(S^1)^N$ for Galois orbits of sequences of points with small Weil height in $(\mathbb{Q}^×)^N$.

Problem Significance

1. Theoretical Importance: Bilu's theorem is a fundamental result in algebraic number theory, establishing connections between Galois orbits and equidistribution. Its effective version is crucial for understanding convergence mechanisms.
2. Application Value: Effective estimates have broad applications in Diophantine approximation, distribution of algebraic numbers, and related fields.
3. Analytical Significance: From an analytical perspective, there exists a vast space of function classes between continuous and Lipschitz continuous functions, requiring systematic investigation.

Limitations of Existing Methods

1. Dimensional Restrictions: The one-dimensional case (N=1) has been extensively studied (Petsche, Favre-Rivera-Letelier, Pritsker, etc.), but the multidimensional case remains less developed.
2. Regularity Requirements: Existing effective estimates typically require test functions to possess strong regularity properties (e.g., Lipschitz continuity).
3. Estimate Precision: D'Andrea et al.'s results yield convergence rate $hD(\xi)^{1/2}$, but impose stringent requirements on function regularity.

Research Motivation

To systematically bridge the gap between continuous and Lipschitz continuous functions by introducing Hölder continuity and fractional derivatives, thereby establishing optimal qualitative dependence relationships.

Core Contributions

1. Establishment of General Fourier Analysis Framework: Extends the settings of Petsche and D'Andrea et al., providing a broader theoretical framework.
2. Optimal Convergence Rates: For test functions with fractional regularity $\gamma \in (0, 1/2]$, proves convergence rate $O(hD(\xi)^\gamma$ and establishes that this exponent is qualitatively optimal.
3. Two Perspectives on Regularity:
   • Fourier Space Regularity (Theorem 2, Corollary 3): Characterized through integrability conditions on $\hat{F}$.
   • Angular Regularity (Theorem 5, Corollary 6): Characterized through modulus of continuity of the logarithmic radial part and Fourier regularity of the angular part.
4. Estimates Under Weak Regularity: For functions satisfying only $\hat{F} \in L^1$, provides effective estimates depending on tail functions (Theorems 4, 7).
5. Applications:
   • New bounds for multidimensional angular discrepancy (Appendix A, Theorem 10)
   • Completion of the proof of Bilu's equidistribution theorem (Appendix C)

Methodology Details

Task Definition

Input:

• Algebraic point $\xi = (\xi_1, \ldots, \xi_N) \in (\mathbb{Q}^×)^N$
• Test function $F: (\mathbb{C}^×)^N \to \mathbb{C}$

Output: Quantitative estimate $E(F, \xi) := \left|\int_{(\mathbb{C}^×)^N} F d\mu_S - \int_{(\mathbb{C}^×)^N} F d\mu_{(S^1)^N}\right|$ where $S$ is the Galois orbit of $\xi$, and $\mu_S$ is the corresponding discrete probability measure.

Objective: Establish bounds on $E(F, \xi)$ in terms of height $h(\xi)$ and generalized degree $D(\xi)$, depending on the regularity of $F$.

Core Technical Framework

1. Logarithmic Polar Coordinates

Via the transformation $(θ, s) \mapsto (e^{2πiθ_1+s_1}, \ldots, e^{2πiθ_N+s_N})$, identify $(\mathbb{C}^×)^N$ with $\mathbb{T}^N \times \mathbb{R}^N$, where $\mathbb{T} = \mathbb{R}/\mathbb{Z}$.

2. Fourier Decomposition

For $F \in L^1(\mathbb{T}^N \times \mathbb{R}^N)$, the Fourier transform is: $\hat{F}(n, t) := \int_{\mathbb{T}^N} \int_{\mathbb{R}^N} F(θ, s) e^{-2πin·θ} e^{-2πit·s} ds dθ$

Via Fourier inversion formula, decompose $E(F, \xi)$ into two parts:

• $I_1$: Contribution from the radial part, depending on $h(\xi)$
• $I_2$: Contribution from the angular part, depending on $D(\xi)$

3. Refined Application of Siegel's Lemma (Lemma 9)

Key Lemma: For algebraic number $\xi$ and its Galois orbit $S = \{\xi_1, \ldots, \xi_d\}$, writing $\xi_j = |\xi_j|e^{2πiθ_j}$, we have $\left|\frac{1}{d}\sum_{j=1}^d e^{2πiθ_j}\right| \leq 2\sqrt{6}\left(\frac{h(\xi) + \frac{\log(2d)}{3d}}{1}\right)^{1/2}$

Proof Strategy:

1. Utilize the Bombieri-Vaaler version of Siegel's lemma to construct polynomial $Q(x)$ such that $\xi$ is at least an $m$-fold zero.
2. Apply Soundararajan's technique in proving the Erdős-Turán inequality to relate the angular distribution of roots to the logarithmic Mahler measure.
3. Optimize the parameter $λ = \sqrt{6}(h(\xi) + \frac{\log(2d)}{3d})^{1/2}$ to obtain the optimal bound.

This lemma is the key tool for improving upon results by Petsche and D'Andrea et al.

Proof Framework for Theorem 2

Estimating $I_1$ (Radial Contribution)

1. Regional Estimates: For $t \in \mathbb{R}^N \setminus \{0\}$, define $\Gamma_δ = \{z \in (\mathbb{C}^×)^N : \sum_{j=1}^N |\log|z_j|| \leq δ\}$
2. Key Inequality (3.6): $\left|\frac{1}{|S|}\sum_{α \in S} e^{2πin·θ(α)}(e^{2πit·s(α)} - 1)\right| \leq \min\{2\sqrt{8πh(\xi)\|t\|_∞}, 2\}$
3. Parameter Optimization: Choose weight function $G(x)$ and parameter $M = (8πh(\xi))^{-1}$, utilizing monotonicity of $G(x)/\sqrt{x}$ to obtain $|I_1| \leq \frac{2C_1(F,G)}{G((8πh(\xi))^{-1})}$

Estimating $I_2$ (Angular Contribution)

1. Galois Orbit Decomposition: For $n \neq 0$, let $S_n$ be the Galois orbit of $χ_n(\xi) = \xi_1^{n_1}\cdots\xi_N^{n_N}$.
2. Application of Lemma 9: Combined with the definition of generalized degree, obtain $\left|\frac{1}{|S|}\sum_{α \in S} e^{2πin·θ(α)}\right| \leq \min\{2\sqrt{6hD(\xi)}\|n\|_1, 1\}$
3. Parameter Optimization: Choose $H(x)$ and $M = (24hD(\xi))^{-1}$, yielding $|I_2| \leq \frac{C_2(F,H)}{H((24hD(\xi))^{-1})}$

Methodology for Theorem 5 (Angular Regularity Perspective)

Key Distinction: Rather than assuming complete Fourier transform integrability of $F$, assume:

1. Radial Regularity: Uniform modulus of continuity condition $|F(θ, s) - F(θ, 0)| \leq ω(|s|)$ where $ω$ is a non-decreasing concave function.
2. Angular Regularity: Only require that Fourier coefficients of $F_0(θ) = F(θ, 0)$ satisfy integrability conditions.

Estimating $I_1$: Utilize Jensen's inequality and modulus of continuity to directly obtain $|I_1| \leq ω(2h(\xi))$

Estimating $I_2$: Same as in Theorem 2.

Experimental Setup

This is a pure theoretical mathematics paper without computational experiments, but includes:

Verification of Qualitative Optimality (Section 5)

Construction 1: Optimality of Corollary 3

Test Function: $\hat{F}(0, t) = \frac{1}{(1+|t|^2)^{(N+γ)/2}\log(20+|t|^2)(\log\log(20+|t|^2))^2}$$\hat{F}(n, t) = 0, \quad n \neq 0$

Point Sequence: $\xi_k = (\xi_{k,1}, \ldots, \xi_{k,N})$, where $\xi_{k,j}$ is a root of $P_{k,j}(x) = x^{d_{k,j}} - d_{k,j}$, with $d_{k,j}$ prime satisfying $2^k < d_{k,1} < \cdots < d_{k,N} < 2^{k+1}$.

Result: Prove that there exists constant $c > 0$ such that $E(F, \xi_k) \geq c \frac{hD(\xi_k)^γ}{|\log hD(\xi_k)|(\log|\log hD(\xi_k)|)^2}$

This demonstrates that the exponent $γ$ in Corollary 3 is qualitatively optimal (up to logarithmic factors).

Construction 2: Optimality of Corollary 6

Test Function: $F(θ, s) = |s|^γ$

Point Sequence: Same as above.

Result: Direct computation yields $E(F, \xi_k) \asymp hD(\xi_k)^γ$.

Experimental Results

Main Theoretical Results

Theorem 2 (Fourier Space Regularity)

For $F \in \mathcal{A}$ and weight functions $G, H$ satisfying specific monotonicity conditions, $E(F, \xi) \leq \frac{2C_1(F,G)}{G((8πh(\xi))^{-1})} + \frac{C_2(F,H)}{H((24hD(\xi))^{-1})}$

where $C_1(F,G) = \sum_{n \in \mathbb{Z}^N} \int_{\mathbb{R}^N} |\hat{F}(n,t)|G(\|t\|_∞)dt$$C_2(F,H) = \sum_{n \in \mathbb{Z}^N \setminus \{0\}} |\hat{F}_0(n)|H(\|n\|_1)$

Corollary 3 (Fractional Derivative Case)

For $0 < γ \leq 1/2$, if $C(F) = \sum_{n \in \mathbb{Z}^N} \int_{\mathbb{R}^N} |\hat{F}(n,t)|(2(8π)^γ\|t\|_∞^γ + 24^γ\|n\|_1^γ)dt < \infty$ then $E(F, \xi) \leq C(F)hD(\xi)^γ$

and the exponent $γ$ is qualitatively optimal.

Theorem 5 (Angular Regularity)

If $F$ satisfies uniform modulus of continuity $|F(θ,s) - F(θ,0)| \leq ω(|s|)$, then $E(F, \xi) \leq ω(2h(\xi)) + \frac{C_2(F,H)}{H((24hD(\xi))^{-1})}$

Corollary 6 (Hölder Continuity Case)

For $0 < γ \leq 1/2$, if $L_γ(F) = \sup_{s \neq 0} \frac{|F(θ,s) - F(θ,0)|}{|s|^γ} < \infty$, then $E(F, \xi) \leq C(F)hD(\xi)^γ$ where $C(F) = 2^γL_γ(F) + 24^γ\sum_{n}|\hat{F}_0(n)|\|n\|_1^γ$.

Comparison with Prior Work

Improvements:

1. Achieves the same $1/2$ exponent with weaker regularity requirements.
2. Proves that $1/2$ is the qualitatively optimal exponent.
3. Provides a unified Fourier analysis framework.

Application Results

Theorem 10 (Multidimensional Angular Discrepancy)

For $hD(\xi) \leq e^{-1}$, $\Delta(\xi) \leq \left(9\left(\frac{3}{2}\right)^N + 14N\right)hD(\xi)^{1/3}|\log hD(\xi)|^{2(N-1)/3}$

This generalizes the one-dimensional result of Langevin-Mignotte to the multidimensional setting.

Related Work

Main Research Directions

1. Effective Estimates in One Dimension:
   • Petsche (2005): Fourier analysis and Siegel's lemma
   • Favre-Rivera-Letelier (2006), Pritsker (2011): Potential theory methods
   • Baker-Masser (2023): Latest improvements, removing degree dependence
2. Multidimensional Generalizations:
   • D'Andrea, Narváez-Clauss, Sombra (2017): Multidimensional Lipschitz case
   • Burgos Gil et al. (2019): Generalizations to toric varieties
3. Related Equidistribution Theory:
   • Bilu (1997): Original equidistribution theorem
   • Szpiro-Ullmo-Zhang (1997): Equidistribution on abelian varieties
   • Numerous extensions to different heights and locations

Positioning of This Paper

Strengths:

1. Systematically bridges the gap between continuous and Lipschitz functions.
2. Establishes optimality and provides qualitative bounds.
3. Provides a unified Fourier analysis framework.

Complementarity:

• Complements the potential theory methods of Baker-Masser et al. (different function classes)
• When N=1, both frameworks have respective advantages

Conclusions and Discussion

Main Conclusions

1. Optimal Convergence Rates: For test functions with fractional regularity $γ \in (0, 1/2]$, the convergence rate of Bilu's equidistribution is $O(hD(\xi)^γ)$, and this exponent is qualitatively optimal.
2. $1/2$ as Natural Limit: Regardless of how high the regularity of the test function, the natural limit of this method is $hD(\xi)^{1/2}$.
3. Estimates Under Weak Regularity: Even assuming only $\hat{F} \in L^1$, effective estimates depending on tail functions can be provided.
4. Unified Framework: Establishes a unified Fourier analysis framework for handling test functions with different regularities.

Limitations

1. Restriction of $1/2$ Exponent: The method itself limits the exponent to at most $1/2$, preventing improvement even for highly regular functions (e.g., $C^\infty$).
2. Logarithmic Factors: Optimality proofs contain logarithmic factor gaps; whether these can be completely removed remains unclear.
3. Dependence on Generalized Degree: Estimates depend on $D(\xi)$ rather than only $h(\xi)$. Whether this can be removed in the multidimensional case, as Baker-Masser achieved in one dimension, remains open.
4. Special Function Classes: Discontinuous functions like characteristic functions require special treatment (e.g., Appendix A).

Future Directions

1. Removing Logarithmic Factors: Can logarithmic terms be completely removed from optimality results?
2. Removing Degree Dependence: Can the multidimensional case remove dependence on $D(\xi)$ like Baker-Masser's one-dimensional result?
3. Other Function Spaces: Explore other regularity concepts (e.g., Besov spaces, Sobolev spaces).
4. Computational Methods: Develop algorithms for computing the equidistribution error of specific algebraic points.
5. Generalizations to Other Varieties: Extend methods to more general algebraic varieties.

In-Depth Evaluation

Strengths

1. Theoretical Depth:
   • Sophisticated Fourier analysis techniques
   • Innovative application of Siegel's lemma (Lemma 9)
   • Complete optimality proofs
2. Systematicity:
   • Two complementary perspectives on regularity
   • Complete spectrum from strong to weak regularity
   • Unified theoretical framework
3. Technical Innovation:
   • Combines Soundararajan's Erdős-Turán inequality proof technique
   • Refined parameter optimization strategies
   • Clever domain partitioning methods
4. Completeness:
   • Detailed proofs
   • Rich appendices (angular discrepancy, auxiliary lemmas, Bilu theorem proof)
   • Clear literature review
5. Writing Quality:
   • Clear structure and rigorous logic
   • Consistent notation system
   • Sufficient motivation

Weaknesses

1. Method Limitations: Essential restriction of $1/2$ exponent prevents handling higher regularity cases.
2. Application Scope: Primarily theoretical results, lacking numerical computation examples for specific algebraic numbers.
3. Technical Complexity: Complex proof techniques may limit further generalizations.
4. Logarithmic Factors: Logarithmic factor gaps in optimality results remain unresolved.

Impact

1. Theoretical Contribution:
   • Perfects the effective version theory of Bilu's equidistribution theorem
   • Establishes standard framework for the field
   • Resolves questions left by Petsche and D'Andrea et al.
2. Methodological Value:
   • Exemplifies Fourier analysis applications in algebraic number theory
   • Techniques may inspire research on other equidistribution problems
3. Practical Value:
   • Angular discrepancy estimates have applications in Diophantine approximation
   • Provides theoretical guarantees for numerical computation
4. Reproducibility:
   • Complete theoretical proofs are verifiable
   • Optimality constructions are explicit and checkable

Applicable Scenarios

1. Theoretical Research:
   • Distribution theory of algebraic numbers
   • Height theory
   • Equidistribution problems
2. Related Fields:
   • Diophantine approximation
   • Transcendental number theory
   • Equidistribution in dynamical systems
3. Method Borrowing:
   • Problems requiring refined Fourier analysis
   • Problems involving Galois orbits of algebraic numbers
   • Quantitative studies of height and distribution relationships

References

The paper cites 31 important references, key ones including:

1. Bilu (1997): Original equidistribution theorem
2. Petsche (2005): One-dimensional effective version
3. D'Andrea, Narváez-Clauss, Sombra (2017): Multidimensional Lipschitz case
4. Bombieri-Vaaler (1983, 1987): Siegel's lemma
5. Baker-Masser (2023): Latest one-dimensional improvements
6. Soundararajan (2019): Erdős-Turán inequality proof technique

Overall Assessment: This is a high-quality theoretical paper at the intersection of number theory and analysis, systematically addressing the dependence on test function regularity in the effective version of Bilu's equidistribution theorem. It establishes a unified Fourier analysis framework and proves optimality of results. The paper demonstrates sophisticated techniques, complete proofs, and makes important contributions to the field. Main limitations are the inherent $1/2$ exponent restriction of the method and its primarily theoretical nature, but these do not diminish its status as a landmark work in this direction.