On $β=6$ Tracy-Widom distribution and the second Calogero-Painlevé system

The Calogero-Painlevé systems were introduced in 2001 by K. Takasaki as a natural generalization of the classical Painlevé equations to the case of the several Painlevé ``particles’’ coupled via the Calogero type interactions. In 2014, I. Rumanov discovered the remarkable fact that a particular case of the second Calogero-Painlevé II equation describes the Tracy-Widom distribution function for the general beta-ensembles with even values of the parameter beta. Most recently, in 2017 work of M. Bertola, M. Cafasso, and V. Rubtsov, it was proven that all Calogero-Painlevé systems are Lax integrable, and hence their solutions admit a Riemann-Hilbert representation. This important observation has opened the door to rigorous, based on the Deift-Zhou nonlinear steepest descent method, asymptotic analysis of the Calogero-Painlevé equations. This in turn yields the possibility of rigorous evaluation of the asymptotic behavior of the Tracy-Widom distributions for the values of beta beyond the classical $β=1, 2, 4.$ In this work we shall start an asymptotic analysis of the Calogero-Painlevé system with a special focus on the Calogero-Painlevé system corresponding to $β= 6$ Tracy-Widom distribution function. The principle technical challenge is the implementation of the nonlinear steepest descent approach beyond the $2\times 2$ matrix dimension of the corresponding Riemann-Hilbert problem; in our case, it is $6\times 6$.

💡 Research Summary

This paper presents a rigorous asymptotic analysis of the Tracy-Widom distribution function for the general beta-ensemble in the specific case of β = 6. The classical values β = 1, 2, 4 are well-understood through orthogonal polynomial methods and 2x2 matrix Riemann-Hilbert problems (RHPs). For general β, and specifically for β=6, a different approach is required.

The study leverages a profound connection discovered by Rumanov: for even values of β, the Tracy-Widom distribution can be described by solutions of particular “Calogero-Painlevé systems.” These systems, introduced by Takasaki, are natural multi-particle generalizations of the classical Painlevé equations, where particles interact via Calogero-type inverse-square potentials. For β = 6, the relevant object is the second Calogero-Painlevé II system for three particles (Equation 1.17). The logarithm of the distribution function F_6(t) itself can be expressed as an integral involving the Hamiltonian of this system (Equation 1.18).

The central problem thus becomes a “connection problem” for this Calogero-Painlevé system: finding the precise asymptotic behavior of its solution as t → +∞ and t → -∞, which corresponds to the right and left tails of F_6(t), respectively. Recent work by Bertola, Cafasso, and Rubtsov proved that all Calogero-Painlevé systems are Lax integrable, meaning their solutions admit a representation as the solution to a matrix Riemann-Hilbert problem. For the β=6 case, this RHP is of 6x6 matrix dimension.

The main technical achievement of this paper is the successful implementation of the Deift-Zhou nonlinear steepest descent method to analyze this higher-dimensional (6x6) RHP. The analysis is carried out separately for the two asymptotic regimes:

• As t → +∞: The authors perform a series of canonical transformations on the RHP. They construct a “global parametrix” that approximates the solution away from singular points (λ = ±i, 0). Near these singular points, they build precise “local parametrices” using special functions (parabolic cylinder functions and modified Bessel functions). Applying a small-norm theorem, they prove that the leading asymptotic terms of the Calogero-Painlevé solution are given by these parametrices, leading to the asymptotic description in Equation (1.24).
• As t → -∞: A similar but distinct steepest descent analysis is performed for t → -∞. The contour structure and singular points change (e.g., to λ = ±√2, 0). The local parametrices in this regime involve Airy functions and confluent hypergeometric functions. This analysis yields the asymptotic behavior of the system’s solution as t → -∞, as shown in Equation (1.20).

The asymptotic results for the Calogero-Painlevé solution, particularly as t → -∞, provide via the Rumanov formalism the necessary “connection formulae” to derive the asymptotic expansion for log F_6(t) in that limit. This offers a rigorous pathway towards verifying the detailed conjectured asymptotics for the Tracy-Widom β=6 distribution proposed by Borot, Eynard, Majumdar, and Nadal (Equation 1.21).

In summary, this work pioneers the application of the nonlinear steepest descent method to a higher-dimensional RHP arising from an integrable system, establishing a powerful framework for the rigorous asymptotic study of Tracy-Widom distributions beyond the classical β values. It marks a significant step in bridging random matrix theory, integrable systems, and asymptotic analysis.