[Paper Review] Asymptotics of a Fredholm determinant involving the second Painlevé transcendent

Thomas Bothner, Alexander Its|arXiv (Cornell University)|Sep 24, 2012

Random Matrices and Applications21 references3 citations

TL;DR

This paper derives the large-s asymptotics of a Fredholm determinant involving the Hastings-McLeod solution of the second Painlevé equation, using the Riemann-Hilbert method. It establishes that the logarithm of the determinant behaves as $-\frac{2}{3}s^6 - xs^4 - \frac{3}{4}\ln s + \int_x^\infty (y-x)u^2(y)\,dy + 3\zeta'(-1) - \frac{1}{6}\ln 2 + O(s^{-1})$, capturing critical edge behavior in random matrix theory near a quadratic zero of the eigenvalue density.

ABSTRACT

We study the determinant $\det(I-K_{ extnormal{PII}})$ of an integrable Fredholm operator $K_{ extnormal{PII}}$ acting on the interval $(-s,s)$ whose kernel is constructed out of the $Ψ$-function associated with the Hastings-McLeod solution of the second Painlevé equation. This Fredholm determinant describes the critical behavior of the eigenvalue gap probabilities of a random Hermitian matrix chosen from the Unitary Ensemble in the bulk double scaling limit near a quadratic zero of the limiting mean eigenvalue density. Using the Riemann-Hilbert method, we evaluate the large $s$-asymptotics of $\det(I-K_{ extnormal{PII}})$.

Motivation & Objective

• To analyze the large-s asymptotic behavior of the Fredholm determinant $\det(I - K_{\textnormal{PII}})$ associated with the Hastings-McLeod solution of the second Painlevé equation.
• To describe the critical eigenvalue gap probability in the bulk double scaling limit near a quadratic zero of the limiting mean eigenvalue density in the Unitary Ensemble.
• To establish precise asymptotic expansions for the Fredholm determinant using the Riemann-Hilbert method and Painlevé transcendents.
• To connect the asymptotics to universal statistical behavior in random matrix theory near spectral edge singularities.

Proposed method

• The Fredholm determinant $\det(I - K_{\textnormal{PII}})$ is analyzed on the interval $(-s,s)$, with the kernel constructed from the $\Psi$-function of the second Painlevé equation's Hastings-McLeod solution.
• The Riemann-Hilbert problem for the $\Psi$-function is solved asymptotically using Deift-Zhou nonlinear steepest descent techniques.
• The asymptotic expansion of the determinant is derived by analyzing the trace of matrix-valued functions and their logarithmic derivatives via contour integration and residue expansions.
• Key components include the trace of products of matrix coefficients $\Theta_1, \Theta_2, \Theta_3$ and their contributions to the logarithmic derivative of the determinant.
• The method involves careful expansion of terms in inverse powers of $s$, retaining terms up to $O(s^{-3})$, and integrating the logarithmic derivative to recover the determinant's logarithm.
• The analysis relies on uniform convergence on compact subsets of the parameter space defined by $x \in \mathbb{R}$, ensuring robustness of the asymptotic formula.

Experimental results

Research questions

• RQ1 How does the Fredholm determinant $\det(I - K_{\textnormal{PII}})$ behave asymptotically as $s \to \infty$ in the critical bulk double scaling limit?
• RQ2 What is the precise asymptotic expansion of the eigenvalue gap probability near a quadratic zero of the mean eigenvalue density in the Unitary Ensemble?
• RQ3 How do Painlevé transcendents and their associated Riemann-Hilbert problems govern the universal behavior of gap probabilities at spectral edges?
• RQ4 What role does the Hastings-McLeod solution of the second Painlevé equation play in describing critical edge statistics in random matrix theory?
• RQ5 How can the Riemann-Hilbert method be used to derive logarithmic terms and constant terms in the asymptotic expansion of Fredholm determinants involving Painlevé functions?

Key findings

• The logarithm of the Fredholm determinant $\det(I - K_{\textnormal{PII}})$ admits the asymptotic expansion $\ln\det(I - K_{\textnormal{PII}}) = -\frac{2}{3}s^6 - xs^4 - \frac{3}{4}\ln s + \int_x^\infty (y-x)u^2(y)\,dy + 3\zeta'(-1) - \frac{1}{6}\ln 2 + O(s^{-1})$ as $s \to \infty$, uniformly on compact subsets of $x \in \mathbb{R}$.
• The leading-order term $-\frac{2}{3}s^6$ captures the dominant growth of the gap probability in the critical regime near a quadratic vanishing of the eigenvalue density.
• The logarithmic term $-\frac{3}{4}\ln s$ arises from the structure of the Riemann-Hilbert solution and contributes to the universal scaling behavior of the gap probability.
• The constant term $3\zeta'(-1) - \frac{1}{6}\ln 2$ is universal and independent of the specific potential $V$, reflecting deep connections to zeta-function regularization.
• The asymptotic formula includes a non-trivial integral term $\int_x^\infty (y-x)u^2(y)\,dy$, where $u(x)$ is the Hastings-McLeod solution of the second Painlevé equation, encoding the non-universal edge behavior.
• The error term $O(s^{-1})$ is uniform on compact subsets of the parameter space, ensuring the validity of the expansion for all fixed $x$ as $s \to \infty$.