2025-11-14T22:46:11.583626

Asymptotic behavior of zeros of Bessel function derivatives

Baricz, Kumar, Ponnusamy
We derive two distinct asymptotic expansions for the zeros $j_{ν,k}^{(n)}$ of the $n$-th derivative of Bessel function $J_ν^{(n)}(x)$. The first is a McMahon-type expansion for the case when $k \to \infty$ with fixed $ν$, for which we also establish an explicit error bound. The second addresses the case when $ν\to \infty$ with fixed $k$ and it involves the zeros of Airy functions and their derivatives. These results extend and refine the classical work of Wong, Lang, and Olver on the zeros of Bessel functions. In the course of obtaining our main results, we also generalize several auxiliary results, which in turn provide a broader framework for the study of zeros of special functions.
academic

Asymptotic behavior of zeros of Bessel function derivatives

Basic Information

  • Paper ID: 2510.12353
  • Title: Asymptotic behavior of zeros of Bessel function derivatives
  • Authors: Árpád Baricz, Pranav Kumar, Saminathan Ponnusamy
  • Classification: math.CA (Classical Analysis and ODEs)
  • Publication Date: October 14, 2025
  • Paper Link: https://arxiv.org/abs/2510.12353v1

Abstract

This paper derives two distinct asymptotic expansions for the zeros jν,k(n)j_{\nu,k}^{(n)} of the n-th derivative Jν(n)(x)J_\nu^{(n)}(x) of the Bessel function. The first is a McMahon-type expansion as kk \to \infty with ν\nu fixed, establishing explicit error bounds. The second addresses the case ν\nu \to \infty with kk fixed, involving zeros of Airy functions and their derivatives. These results extend and improve classical work by Wong, Lang, and Olver on Bessel function zeros. In deriving the main results, several auxiliary results are generalized, providing a broader framework for studying zeros of special functions.

Research Background and Motivation

Importance of the Problem

Bessel function zeros play a crucial role in various problems in applied mathematics and mathematical physics, particularly in quantum mechanics, scattering theory, and wave propagation. Due to their widespread applicability, understanding the asymptotic behavior of Bessel functions and their zeros has received considerable attention.

Limitations of Existing Work

  1. Restrictions of classical McMahon expansion: Early work by McMahon on zeros of the Bessel function Jν(x)J_\nu(x) focused primarily on the function itself, not its derivatives.
  2. Work by Wong and Lang: Although they extended these ideas to obtain asymptotic expansions for zeros of Jν(x)J''_\nu(x), this was limited to second-order derivatives.
  3. Research gap: Despite significant progress in understanding properties of n-th order Bessel function derivative zeros, asymptotic expansions for these zeros have not been explored in the literature.

Research Motivation

This paper aims to fill this gap by presenting McMahon-type asymptotic expansions and analyzing the case ν\nu \to \infty with n fixed. The research employs standard methods introduced by McMahon to derive asymptotic expansions of Jν(n)(x)J_\nu^{(n)}(x) and establish asymptotic behavior of jν,k(n)j_{\nu,k}^{(n)} as kk \to \infty.

Core Contributions

  1. Establishment of McMahon-type asymptotic expansions: For zeros of n-th order derivatives of Bessel functions, asymptotic expansions as kk \to \infty with ν\nu fixed.
  2. Provision of explicit error bounds: Establishing definite error estimates for McMahon-type expansions, quantifying approximation accuracy.
  3. Derivation of large-parameter asymptotic expansions: Addressing the case ν\nu \to \infty with kk fixed, involving zeros of Airy functions and their derivatives.
  4. Generalization of auxiliary results: Generalizing several foundational theoretical results during the derivation of main results, providing a broader framework for studying special function zeros.
  5. Extension of classical theory: Extending classical work by Wong, Lang, and Olver from Bessel functions themselves to their arbitrary-order derivatives.

Detailed Methodology

Task Definition

Study the asymptotic behavior of positive zeros jν,k(n)j_{\nu,k}^{(n)} of the n-th derivative Jν(n)(x)J_\nu^{(n)}(x) of the Bessel function of the first kind Jν(x)J_\nu(x), where:

  • ν>0\nu > 0 is the order of the Bessel function
  • n1n \geq 1 is the order of the derivative
  • kk is the index of the zero

Core Theoretical Framework

1. Asymptotic Representation of Bessel Function Derivatives

Theorem 2.1: For nN0n \in \mathbb{N}_0 and large xx, the 2n-th derivative of the Bessel function can be expressed as:

12πxJν(2n)(x)=cos(xνπ2π4)τν(2n)(x)+sin(xνπ2π4)θν(2n)(x)\sqrt{\frac{1}{2}\pi x} J_\nu^{(2n)}(x) = \cos\left(x - \frac{\nu\pi}{2} - \frac{\pi}{4}\right)\tau_\nu^{(2n)}(x) + \sin\left(x - \frac{\nu\pi}{2} - \frac{\pi}{4}\right)\theta_\nu^{(2n)}(x)

where: τν(2n)(x)=m=0α2m,2nx2m,θν(2n)(x)=m=0α2m+1,2nx2m+1\tau_\nu^{(2n)}(x) = \sum_{m=0}^{\infty} \frac{\alpha_{2m,2n}}{x^{2m}}, \quad \theta_\nu^{(2n)}(x) = \sum_{m=0}^{\infty} \frac{\alpha_{2m+1,2n}}{x^{2m+1}}

The coefficients satisfy the recurrence relations: α2m,2n=α2m,2n14m12α2m1,2n1\alpha_{2m,2n} = \alpha_{2m,2n-1} - \frac{4m-1}{2}\alpha_{2m-1,2n-1}α2m+1,2n=α2m1,2n14m+12α2m,2n1\alpha_{2m+1,2n} = -\alpha_{2m-1,2n-1} - \frac{4m+1}{2}\alpha_{2m,2n-1}

2. McMahon-type Asymptotic Expansion

Through inversion techniques, McMahon-type expansions for zeros are obtained:

Even-order derivatives (zeros of Jν(2m)J_\nu^{(2m)}): jν,k+1(2m)=α+α1,2mα0,2m1α+O(1α3)j_{\nu,k+1}^{(2m)} = \alpha + \frac{\alpha_{1,2m}}{\alpha_{0,2m}} \cdot \frac{1}{\alpha} + O\left(\frac{1}{\alpha^3}\right)

where α=kπ+νπ2π4\alpha = k\pi + \frac{\nu\pi}{2} - \frac{\pi}{4}

Odd-order derivatives (zeros of Jν(2m+1)J_\nu^{(2m+1)}): jν,k(2m+1)=βα1,2m+1α0,2m+11β+O(1β3)j_{\nu,k}^{(2m+1)} = \beta - \frac{\alpha_{1,2m+1}}{\alpha_{0,2m+1}} \cdot \frac{1}{\beta} + O\left(\frac{1}{\beta^3}\right)

where β=kπ+νπ23π4\beta = k\pi + \frac{\nu\pi}{2} - \frac{3\pi}{4}

3. Error Bound Estimation

Theorem 2.2: For error terms in the asymptotic expansion, bounded estimates are:

For 2n-th order derivatives, when ν2n+12\nu \geq -2n + \frac{1}{2}, nNn \in \mathbb{N} and xx is sufficiently large: δ2n(ν,x)4(ν+2n)214xexp{4(ν+2n)214x}|\delta_{2n}(\nu,x)| \leq \frac{4(\nu+2n)^2-1}{4x} \exp\left\{\frac{4(\nu+2n)^2-1}{4x}\right\}

4. Asymptotic Expansion for Large Parameters

When ν\nu \to \infty with kk fixed, utilizing Olver's uniform asymptotic expansion theory and Airy functions:

For even-order derivatives: jν,k(2n)=νak21/3ν1/3+21/3ak210ν1/3+O(ν1)j_{\nu,k}^{(2n)} = \nu - \frac{a_k}{2^{1/3}}\nu^{1/3} + \frac{2^{1/3}a_k^2}{10}\nu^{-1/3} + O(\nu^{-1})

For odd-order derivatives: jν,k(2n+1)=νak21/3ν1/3+21/3(ak)210ν1/3+O(ν1)j_{\nu,k}^{(2n+1)} = \nu - \frac{a'_k}{2^{1/3}}\nu^{1/3} + \frac{2^{1/3}(a'_k)^2}{10}\nu^{-1/3} + O(\nu^{-1})

where aka_k and aka'_k are the k-th negative zeros of the Airy function Ai(x)\text{Ai}(x) and Ai(x)\text{Ai}'(x), respectively.

Technical Innovations

  1. Unified recurrence framework: Establishing unified recurrence relations for coefficients of all derivative orders, avoiding individual calculations for each order.
  2. Improved error analysis: Through modification of Hethcote's method, obtaining more precise error bounds than classical results.
  3. Extended applicability: Extending classical theory from specific low-order derivatives to arbitrary-order derivatives.
  4. Application of uniform asymptotic theory: Skillfully applying Olver's uniform asymptotic expansion theory to high-order derivative cases.

Experimental Setup

Theoretical Verification Methods

As this is pure mathematical theoretical research, "experiments" primarily refer to theoretical verification and numerical validation:

  1. Mathematical induction verification: Using mathematical induction to verify correctness of recurrence relations
  2. Asymptotic analysis: Verifying validity of asymptotic expansions through limit analysis
  3. Error bound verification: Verifying tightness of error bounds through concrete calculations

Comparison Benchmarks

  • McMahon's original results (for zeros of Jν(x)J_\nu(x))
  • Wong and Lang's results (for zeros of Jν(x)J''_\nu(x))
  • Olver's uniform asymptotic expansion theory

Experimental Results

Main Theoretical Results

  1. Establishment of McMahon-type expansions: Successfully deriving McMahon-type asymptotic expansions for zeros of arbitrary-order Bessel function derivatives with explicit coefficient expressions.
  2. Improvement of error bounds: Compared to existing results, new error bounds are more precise, particularly for high-order derivatives.
  3. Completeness of large-parameter expansions: For the case ν\nu \to \infty, providing complete asymptotic series including leading terms and correction terms.

Specific Numerical Results

For zeros of third and fourth order derivatives: jν,k(3)=νak21/3ν1/3+21/3ak210ν1/3+ak3700ν1+j_{\nu,k}^{(3)} = \nu - \frac{a_k}{2^{1/3}}\nu^{1/3} + \frac{2^{1/3}a_k^2}{10}\nu^{-1/3} + \frac{a_k^3}{700}\nu^{-1} + \cdots

jν,k(4)=νak21/3ν1/3+21/3(ak)210ν1/3+(ak)3700ν1+j_{\nu,k}^{(4)} = \nu - \frac{a'_k}{2^{1/3}}\nu^{1/3} + \frac{2^{1/3}(a'_k)^2}{10}\nu^{-1/3} + \frac{(a'_k)^3}{700}\nu^{-1} + \cdots

Specific Forms of Error Bounds

For McMahon-type expansions, when conditions are satisfied:

  • Even-order derivatives: jν,k+1(2n)(k+ν214)π0.26(4(ν+2n)21)π(k+ν20.314)|j_{\nu,k+1}^{(2n)} - (k + \frac{\nu}{2} - \frac{1}{4})\pi| \leq \frac{0.26(4(\nu+2n)^2-1)}{\pi(k + \frac{\nu}{2} - 0.314)}
  • Odd-order derivatives: jν,k(2n+1)(k+ν214)π0.26(4(ν+2n+1)21)π(k+ν20.314)|j_{\nu,k}^{(2n+1)} - (k + \frac{\nu}{2} - \frac{1}{4})\pi| \leq \frac{0.26(4(\nu+2n+1)^2-1)}{\pi(k + \frac{\nu}{2} - 0.314)}

Historical Development

  1. McMahon (1895): Establishing classical asymptotic expansions for Bessel function zeros
  2. Watson (1944): Perfecting theoretical foundations with more rigorous proofs
  3. Olver (1954): Developing uniform asymptotic expansion theory
  4. Wong & Lang (1990-1991): Extending theory to second-order derivatives of Bessel functions
  5. Baricz et al. (2018): Proving reality and positivity of n-th order derivative zeros

Position and Contribution of This Paper

This paper's contribution in this developmental trajectory is to completely generalize theory from specific low-order derivative cases to arbitrary-order derivatives, providing more precise error analysis. This represents an important theoretical refinement in the field.

Conclusions and Discussion

Main Conclusions

  1. Completeness: Establishing complete asymptotic theory for zeros of arbitrary-order Bessel function derivatives
  2. Precision: Providing more precise error bounds than existing results
  3. Uniformity: Providing a unified theoretical framework applicable to all derivative orders

Theoretical Significance

  1. Filling gaps: Completing the final piece of asymptotic theory for Bessel function derivative zeros
  2. Methodological innovation: Providing general methods for handling high-order derivative asymptotic problems
  3. Application potential: Providing theoretical tools for related physical and engineering problems

Limitations

  1. Applicability conditions: Some results require satisfaction of specific parameter range conditions
  2. Computational complexity: Calculation of high-order terms may become quite complex
  3. Numerical verification: Theoretical results require more numerical verification to confirm practical utility

Future Directions

  1. Extension to modified Bessel functions: Applying similar methods to modified Bessel functions and their derivatives
  2. Other special functions: Extending to Wright functions, hyper-Bessel functions, etc.
  3. Numerical algorithms: Developing efficient zero-finding algorithms based on theoretical results

In-depth Evaluation

Strengths

  1. Theoretical completeness: Providing complete theoretical solution to the problem
  2. Methodological rigor: Using mature mathematical analysis methods with rigorous proofs
  3. Generality of results: Not only solving specific problems but providing general theoretical framework
  4. Error analysis: Providing explicit error bounds, enhancing practical utility of results

Weaknesses

  1. Lack of numerical verification: As pure theoretical research, lacking sufficient numerical experiments
  2. Insufficient application examples: No provision of specific application scenarios and examples
  3. Computational complexity: For very high-order derivatives, computation may become impractical

Impact Assessment

  1. Academic impact: Perfecting an important branch of special function theory
  2. Practical value: Providing theoretical foundation for applications requiring Bessel function derivative zeros
  3. Methodological value: Techniques provided may be applicable to similar problems

Applicable Scenarios

  1. Mathematical physics: Analytical solutions to wave equations, heat conduction, etc.
  2. Engineering applications: Signal processing, vibration analysis, and related fields
  3. Numerical computation: Theoretical foundation for high-precision zero-finding algorithms

References

The paper cites key literature in the field, including:

  • McMahon's pioneering work (1895)
  • Watson's classical textbook (1944)
  • Olver's asymptotic theory (1954, 1962, 1974)
  • Wong & Lang's extension work (1990, 1991)
  • Recent related research (Baricz et al., 2018; Dimitrov & Lun, 2025)

These citations reflect historical continuity and theoretical completeness of the research.