Researcher profile

Aron Wennman

Aron Wennman contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author
6works
0followers
6topics
3close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

6 published item(s)

preprint2020arXiv

Planar orthogonal polynomials and boundary universality in the random normal matrix model

We show that the planar normalized orthogonal polynomials $P_{m,n}(z)$ of degree $n$ with respect to an exponentially varying planar measure $\mathrm{e}^{-2mQ}\mathrm{dA}$ enjoy an asymptotic expansion \[ P_{m,n}(z)\sim m^{\frac{1}{4}}\sqrt{ϕ_τ'(z)}[ϕ_τ(z)]^n \mathrm{e}^{m\mathcal{Q}_τ(z)}\left(\mathcal{B}_{τ, 0}(z) +m^{-1}\mathcal{B}_{τ, 1}(z)+m^{-2} \mathcal{B}_{τ,2}(z)+\ldots\right), \] as $n,m\to\infty$ while the ratio $τ=\frac{n}{m}$ is fixed. Here $\mathcal{S}_τ$ denotes the droplet, the boundary of which is assumed to be a smooth simple closed curve, and $ϕ_τ$ is a conformal mapping from the complement $\mathcal{S}_τ^c$ to the exterior disk $\Bbb{D}_\mathrm{e}$. The functions $\mathcal{Q}_τ$ and $\mathcal{B}_{τ, j}$ are bounded holomorphic functions which may be expressed in terms of $Q$ and $\mathcal{S}_τ$. We apply these results to obtain boundary universality in the random normal matrix model for smooth droplets, i.e., that the limiting rescaled process is the random process with correlation kernel \[ \mathrm{k}(ξ,η)= \mathrm{e}^{ξ\barη\,-\frac12(\lvertξ\rvert^2+\lvert η\rvert^2)} \,\mathrm{erf}\,(ξ+\barη). \] A key ingredient in the proof of the asymptotic expansion of the orthogonal polynomials is the construction of an orthogonal foliation -- a smooth flow of closed curves near $\partial\mathcal{S}_τ$, on each of which $P_{m,n}$ is appropriately orthogonal to lower order polynomials. To compute the coefficient functions, we develop an algorithm which determines the coefficients $\mathcal{B}_{τ, j}$ successively in terms of inhomogeneous Toeplitz kernel conditions. These inhomogeneous Toeplitz kernel conditions may be understood in terms of scalar Riemann-Hilbert problems.

preprint2020arXiv

Riemann-Hilbert hierarchies for hard edge planar orthogonal polynomials

We obtain a full asymptotic expansion for orthogonal polynomials with respect to weighted area measure on a Jordan domain $\mathscr{D}$ with real-analytic boundary. The weight is fixed and assumed to be real-analytically smooth and strictly positive, and for any given precision $\varkappa$, the expansion holds with an $\mathrm{O}(N^{-\varkappa-1})$ error in $N$-dependent neighborhoods of the exterior region as the degree $N$ tends to infinity. The main ingredient is the derivation and analysis of Riemann-Hilbert hierarchies - sequences of scalar Riemann-Hilbert problems - which allows us to express all higher order correction terms in closed form. In fact, the expansion may be understood as a Neumann series involving an explicit operator. The expansion theorem leads to a semiclassical asymptotic expansion of the corresponding hard edge probability wave function in terms of distributions supported on $\partial\mathscr{D}$.

preprint2020arXiv

The forbidden region for random zeros: appearance of quadrature domains

Our main discovery is a surprising interplay between quadrature domains on the one hand, and the zero process of the Gaussian Entire Function (GEF) on the other. Specifically, consider the GEF conditioned on the rare hole event that there are no zeros in a given large Jordan domain. We show that in the natural scaling limit, a quadrature domain enclosing the hole emerges as a forbidden region, where the zero density vanishes. Moreover, we give a description of those holes for which the forbidden region is a disk. The connecting link between random zeros and potential theory is supplied by a constrained extremal problem for the Zeitouni-Zelditch functional. To solve this problem, we recast it in terms of a seemingly novel obstacle problem, where the solution is forced to be harmonic inside the hole.

preprint2019arXiv

Off-spectral analysis of Bergman kernels

The asymptotic analysis of Bergman kernels with respect to exponentially varying measures near emergent interfaces has attracted recent attention. Such interfaces typically occur when the associated limiting Bergman density function vanishes on a portion of the plane, the off-spectral region. This type of behaviour is observed when the metric is negatively curved somewhere, or when we study partial Bergman kernels in the context of positively curved metrics. In this work, we cover these two situations in a unified way, for exponentially varying planar measures on the complex plane. We obtain uniform asymptotic expansions of root functions, which are essentially normalized partial Bergman kernels at an off-spectral point, valid in the entire off-spectral component and protruding into the spectrum as well, which allows us to show error function transition behaviour of the original kernel along the interface. In contrast, previous work on asymptotic expansions of Bergman kernels is typically local, and valid only in the bulk region of the spectrum.

preprint2016arXiv

Discrepancy densities for planar and hyperbolic Zero Packing

We study the problem of geometric zero packing, recently introduced by Hedenmalm. There are two natural densities associated to this problem: the discrepancy density $ρ_{\mathbb{H}}$, given by $$ ρ_{\mathbb{H}} = \liminf_{r\to 1^-} \inf_{f} \frac{\int_{\mathbb{D}(0,r)} \left((1-\lvert z\rvert^2) \lvert f(z)\rvert-1\right)^2 \frac{dA(z)}{1-\lvert z\rvert^2}} {\int_{\mathbb{D}(0,r)} \frac{dA(z)}{1-\lvert z\rvert^2}} $$ which measures the discrepancy in optimal approximation of $(1-\lvert z\rvert^2)^{-1}$ with the modulus of polynomials $f$, and it's relative, the tight discrepancy density $ρ_{\mathbb{H}}^*$, which will trivially satisfy $ρ_{\mathbb{H}}\leqρ_{\mathbb{H}}^*$. These densities have deep connections to the boundary behaviour of conformal mappings with $k$-quasiconformal extensions, which can be seen from the Hedenmalm's result that the universal asymptotic variance $Σ^2$ is related to $ρ_{\mathbb{H}}^*$ by $Σ^2=1-ρ_{\mathbb{H}}^*$. Here we prove that in fact $ρ_{\mathbb{H}}=ρ_{\mathbb{H}}^*$, resolving a conjecture by Hedenmalm in the positive. The natural planar analogues $ρ_{\mathbb{C}}$ and $ρ_{\mathbb{C}}^*$ to these densities make contact with work of Abrikosov on Bose-Einstein condensates. As a second result we prove that also $ρ_{\mathbb{C}}=ρ_{\mathbb{C}}^*$. The methods are based on Ameur, Hedenmalm and Makarov's Hörmander-type $\bar\partial$-estimates with polynomial growth control. As a consequence we obtain sufficiency results on the degrees of approximately optimal polynomials.

preprint2014arXiv

Carleman-Sobolev classes for small exponents

This paper is devoted to the study of a generalization of Sobolev spaces for small $L^{p}$ exponents, i.e. $0<p<1$. We consider spaces defined as abstract completions of certain classes of smooth functions with respect to weighted quasi-norms, simultaneously inspired by Carleman classes and classical Sobolev spaces. If the class is restricted with a growth condition on the supremum norms of the derivatives, we prove that there exists a condition which guarantees that the resulting space can be embedded into $C^{\infty}(\mathbb{R})$. This is sharp up to some regularity conditions on the weight sequence. We also show that the growth condition is necessary, in the sense that if we drop it entirely we can naturally embed $L^p$ into the resulting completion.