Orthogonal polynomials in the spherical ensemble with two insertions
Sung-Soo Byun, Peter J. Forrester, Arno B.J. Kuijlaars, Sampad Lahiry · 2025
Read Orthogonal polynomials in the spherical ensemble with two insertions by Sung-Soo Byun, Peter J. Forrester, Arno B.J. Kuijlaars, Sampad Lahiry (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-15732.txt.
Introduction
Sung-Soo Byun, Peter Forrester, Arno Kuijlaars, and Sampad Lahiry study special polynomials tied to a random gas of charges in this 2025 paper. The polynomials are orthogonal over the whole flat plane under a weight built from two inserted charges. Their behavior at large degree mirrors how the charges arrange themselves.
The team derives exact large-scale formulas for the polynomials, their sizes, and where their zeros gather. The results cover the early phase of the charge system, when the charged region stays in one piece. Their method rewrites the flat-plane problem as one along a contour, then applies a steepest-descent analysis borrowed from wave theory. The work sharpens tools used across random matrix theory.
Catalog record
| CREATOR | Sung-Soo Byun, Peter J. Forrester, Arno B.J. Kuijlaars, Sampad Lahiry |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 84 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-15732
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Orthogonal polynomials in the spherical ensemble with two insertions Sung-Soo Byun, Peter J. Forrester, Arno B.J. Kuijlaars, Sampad Lahiry Abstract We consider asymptotics of planar orthogonal polynomials P n , N P_{n,N} (where deg P n , N = n \mathrm{deg}P_{n,N}=n ) with respect to the weight | z − w | 2 N Q 1 ( 1 + | z | 2 ) N ( 1 + Q 0 + Q 1 ) + 1 , ( Q 0 , Q 1 > 0 ) \frac{|z-w|^{2NQ_{1}}}{(1+|z|^{2})^{N(1+Q_{0}+Q_{1})+1}},\quad(Q_{0},Q_{1}>0) in the whole complex plane. With n , N → ∞ n,N\rightarrow\infty and N − n N-n fixed, we obtain the strong asymptotics of the polynomials, asymptotics for the weighted L 2 L^{2} norm and the limiting zero counting measure. These results apply to the pre-critical phase of the underlying two-dimensional Coulomb gas system, when the support of the equilibrium measure is simply connected. Our method relies on specifying the mother body of the two-dimensional potential problem. It relies too on the fact that the planar orthogonality can be rewritten as a non-Hermitian contour orthogonality. This allows us to perform the Deift-Zhou steepest descent analysis of the associated 2 × 2 2\times 2 Riemann-Hilbert problem. 1 Introduction and statement of results 1.1 Underlying Coulomb gas, random matrix ensemble and relation to planar orthogonal polynomials Let \{\mathbf{r}_{j}\}_{j=1}^{N} be N points on the surface of the sphere of radius \frac{1}{2} embedded in \mathbb{R}^{3} and centred at the origin. Let ||\mathbf{x}|| denote the usual Euclidean length in \mathbb{R}^{3} . Suppose these N unit points are in fact charges obeying Poisson’s equation as defined on the surface of the sphere. This means that a point at \mathbf{x} and a point at \mathbf{x}^{\prime} interact via the pair potential -\log||\mathbf{x}-\mathbf{x}^{\prime}|| ; see e.g. [ 29 , Eq. (15.105)] . In addition, let there be two fixed charges of (macroscopic) strengths NQ_{0} and NQ_{1} at positions \mathbf{R}_{0} and \mathbf{R}_{1} , with Q_{0},Q_{1}>0 . Up to a multiplicative constant, the corresponding Boltzmann factor has the functional form ||\mathbf{R}_{0}-\mathbf{R}_{1}||^{\beta N^{2}Q_{0}Q_{1}}\prod_{l=1}^{N}||\mathbf{R}_{0}-\mathbf{r}_{l}||^{\beta NQ_{0}}||\mathbf{R}_{1}-\mathbf{r}_{l}||^{\beta NQ_{1}}\prod_{1\leq j<k\leq N}||\mathbf{r}_{k}-\mathbf{r}_{j}||^{\beta}, (1.1) where \beta>0 is the inverse temperature. By rotational invariance of the sphere, it is always possible to choose \mathbf{R}_{0} to be at the south pole (0,0,-\frac{1}{2}) , which we henceforth assume. From the viewpoint of charge neutrality in the Coulomb gas picture, it has been noted in [ 14 ] that associated with fixed charges at positions \mathbf{R}_{0} and \mathbf{R}_{1} are spherical caps of area \pi Q_{0}/(1+Q_{0}+Q_{1}) and \pi Q_{1}/(1+Q_{0}+Q_{1}) , respectively. As already evident from potential theoretic viewpoints [ 22 , 10 , 20 , 43 , 21 ] the Coulomb gas exists in two phases depending on there being no overlap of the spherical caps (referred to as the post-critical phase ), or the spherical caps overlapping (referred to as the pre-critical phase ). With (\theta,\phi) the polar angle, azimuthal angle pair corresponding to the point \mathbf{R}_{1} , we have from [ 14 , Eq. (1.7)] that the condition for no overlap, and thus the post-critical phase is \cos^{2}(\theta/2)>{1\over(\gamma_{1}+\gamma_{2}+2)^{2}}\Big(\sqrt{(\gamma_{1}+1)(\gamma_{1}+\gamma_{2}+1)}+\sqrt{\gamma_{2}+1}\Big)^{2}\Big|_{\begin{subarray}{c}\gamma_{1}=-1+Q_{0}/Q_{1}\\ \gamma_{2}=1/Q_{1}\end{subarray}}. (1.2) Conversely, if this inequality is in the other direction, the Coulomb gas system is in the pre-critical phase. Suppose now that the points on the sphere are stereographically projected to the complex plane, viewed to be tangent to the north pole; this construction is illustrated e.g. in [ 29 , Figure 15.2] . With \mathbf{r} , \mathbf{r}^{\prime} points on the sphere and z , z^{\prime} the corresponding points on the plane, we then have that [ 29 , equivalent to Eq. (15.126)] ||\mathbf{r}-\mathbf{r}^{\prime}||={|z-z^{\prime}|\over(1+|z|^{2})^{1/2}(1+|z^{\prime}|^{2})^{1/2}}. (1.3) We have too the relation between the differential on the surface of the sphere dS say, and the flat measure on the complex plane viewed as \mathbb{R}^{2} [ 29 , Eq. (15.127)] dS={1\over(1+|z|^{2})^{2}}\,dx\,dy. (1.4) These two facts together give that the stereographically projected form of ( 1.1 ) is equal to {1\over(1+|w|^{2})^{\beta Q_{0}Q_{1}N^{2}}}\prod_{l=1}^{N}{|w-z_{l}|^{\beta Q_{1}N}\over(1+|z_{l}|^{2})^{\beta(1+Q_{0}+Q_{1})N/2+2-\beta/2}}\prod_{1\leq j<k\leq N}|z_{k}-z_{j}|^{\beta}, (1.5) where w is the stereographic projection onto the complex plane of \mathbf{R}_{1} . It is convenient to rotate the sphere about the veritical axis so that w>0 . The equation ( 1.2 ) can be rewritten in terms of w . Thus the requirement that the spherical caps associated with two charges overlap restricts Q_{0},Q_{1},w to be such that [ 14 , Eq. (2.74)] w>w_{\rm cri}:=\Big(2Q_{0}Q_{1}+Q_{0}+Q_{1}+2\sqrt{Q_{0}Q_{1}(1+Q_{0})(1+Q_{1})}\Big)^{-1/2}. (1.6) Now set \beta=2,\quad Q_{1}N=r,\quad Q_{0}N=K (1.7) in ( 1.5 ). The factor in ( 1.5 ) independent of w is then \prod_{l=1}^{N}{1\over(1+|z_{l}|^{2})^{K+r+N+1}}\prod_{1\leq j<k\leq N}|z_{k}-z_{j}|^{2}. (1.8) As revised in [ 14 , §3.1] this functional form, up to proportionality, is the eigenvalue probability density function for the particular random matrix ensemble SrUE (N,K+r) . Here the (non-Hermitian) ensemble SrUE (N,s) is formed from random matrices (G^{\dagger}G)^{-1/2}X where G is an (N+s)\times N complex standard Gaussian matrix and X is an N\times N complex standard Gaussian matrix; the case s=0 corresponds to the complex spherical ensemble [ 13 , §2.5] . This allows ( 1.5 ) with the substitution ( 1.7 ) to be identified as proportional to {1\over(1+|w|^{2})^{2rK}}\Big\langle\prod_{l=1}^{N}|w-z_{l}|^{2r}\Big\rangle_{{\rm SrUE}_{(N,K+r)}}. (1.9) In words the second factor is the ensemble average with respect to {\rm SrUE}_{(N,K+r)} of the 2r -th power of the absolute value of the characteristic polynomial \prod_{l=1}^{N}|w-z_{l}| . A result from [ 30 ] gave that this ensemble average is, assuming K\geq r , equal to the different ensemble average \Big\langle\prod_{l=1}^{r}(|w|^{2}+t_{l})^{N}\Big\rangle_{{\rm JUE}_{r,(0,K-r)}}. (1.10) Here {\rm JUE}_{r,(0,K-r)} denotes the (Hermitian) Jacobi unitary ensemble of r eigenvalues \{t_{l}\} supported on the interval (0,1) and distributed according to the probability density function proportional to \prod_{l=1}^{r}(1-t_{l})^{K-r}\prod_{1\leq j<k\leq n}|t_{k}-t_{j}|^{2}; for more on this see [ 29 , Ch. 3] . Since the roles of N and r in the average in ( 1.9 ) and that of ( 1.10 ) are reversed their equality is referred to as a duality identity. In [ 14 ] this was used as one tool to study the large N form of the configuration integral associated with the \beta=2 case of the Boltzmann factor ( 1.1 ) or equivalently ( 1.5 ), in both the pre- and post-critical phases. A salient feature of ( 1.1 ) and ( 1.5 ) with \beta=2 is that they both specify determinantal point processes. Consider for definiteness ( 1.5 ). Excluding the first factor, this can be written \prod_{l=1}^{N}e^{-NV(z_{l})}\prod_{1\leq j<k\leq N}|z_{k}-z_{j}|^{2}, (1.11) where V(z)=\Big(1+\frac{1}{N}+Q_{0}+Q_{1}\Big)\log(1+|z|^{2})-2Q_{1}\log|z-w|. (1.12) Introduce monic orthogonal polynomials \{P_{k,N}(z)\}_{k=0}^{N-1} of degree k such that \int_{\mathbb{C}}P_{l,N}(z)\overline{P_{k,N}(z)}e^{-NV(z)}\,dA(z)=h_{k,N}\,\delta_{k,l},\quad(l,k=0,1\dots,N-1), (1.13) where dA(z)=dx\,dy . One notes that V(z) and thus the polynomials P_{k,N}(z) depend on w , although this dependance is suppressed in the notation. Standard theory (see e.g. [ 29 , §15.3] ) gives that \int_{\mathbb{C}}dA(z_{1})\cdots\int_{\mathbb{C}}dA(z_{N})\,\prod_{l=1}^{N}e^{-NV(z_{l})}\prod_{1\leq j<k\leq N}|z_{k}-z_{j}|^{2}=N!\prod_{l=0}^{N-1}h_{l,N}, (1.14) and \rho_{(k)}(z_{1},\dots,z_{k})=\det[K_{N}(z_{i},z_{j})]_{i,j=1}^{k}, (1.15) with K_{N}(w,z)=e^{-\frac{N}{2}\left(V(w)+V(z)\right)}\sum_{l=0}^{N-1}\frac{P_{l,N}(w)\overline{P_{l,N}(z)}}{h_{l,N}}. (1.16) In ( 1.15 ), \rho_{(k)} denotes the k -point correlation function, defined as a suitable normalisation times ( 1.11 ) integrated over z_{k+1},\dots,z_{N} ; see e.g. [ 29 , §5.1.1] . Of particular interest is the large N form of the partition function ( 1.14 ), which will be dependent on the phase (post or pre-critical), as well as the correlation function ( 1.16 ), which in fact can be proved to exhibit universal properties [ 1 , 2 , 34 ] in the microscopic scaling regime. The study of [ 34 ]