Noncommutative geometry and Painlevé equations (Q500222)

The classical Painlevé equations are special 2-dimensional dynamical systems, and the authors claims to have rather practical goals for the article. The actual dynamical systems appears in simple but challenging problems in probability and mathematical physics. That is planar dimer with changing boundary. One of the main subjects of this article is to give the link to Artin-Style noncommutative geometry, and to prove the applications of this to probabilistic and dynamical applications. The results are stated throughout in the special situation of interest, with explicit examples, indicating the results in general. The various group actions in algebraic geometry is a necessary ingredient in noncommutative (algebraic) geometry. Let \(S\subset\mathbb P^N\) be a projective variety with homogeneous coordinate ring \(A=\mathbb C[x_0,\dots,x_N]/I\). Let \(\text{Coh}S\) denote the quotient category of finitely generated graded \(A\)-modules \(M\) by the finite dimensional modules. As this depends on a finite set of parameters, there is a (coarse) moduli space \(X\) which is a countable union of algebraic varieties. The general theory of such varieties are well developed, and this article gives a concrete description of \(X\) by giving generators and relations for the points \(M\in X\). The group \(\text{Pic}S\) of line bundles \(\mathcal L\) on \(S\) acts on \(X\) by \(\widetilde M\mapsto\mathcal L\otimes\widetilde M\), and when \(\mathcal L\) is topologically nontrivial, this permutes the components of \(X\). The integrable actions of abelian groups can be understood as invariants, one of the invariants of the dynamics being the cycle in \(S\) given by the support of \(\widetilde M=\mathcal M\). The starting point of the article is that \(A\) can be noncommutative, and the theory of Artin, Stafford, and Van den Bergh (see [\textit{J. T. Stafford} and \textit{M. Van den Bergh}, Bull. Am. Math. Soc., New Ser. 38, No. 2, 171--216 (2001; Zbl 1042.16016)]) can be applied to the study of graded algebras on a good category. Given a right \(A\)-module \(L\), then \(L\otimes_A M\in\text{Mod}A^\prime,\;A^\prime=\text{End}_A L\). If \(L\) is a deformation of a line bundle, then in general \(A^\prime\neq A\), and so there is a morphism of moduli spaces given by \(X\overset{L\otimes}\longrightarrow X^\prime\). This article considers the special situation in which \(S\) is a blowup of a surface \(S_0\) in \(p\in S_0\) and \(L\) is the exceptional divisor. When \(A\) is noncommutative, tensoring with \(L\) makes the point \(p\) move in \(S_0\) by an amount proportional to the degree of noncommutativity. Let \(\mathcal B\) denote the blowups of \(\mathbb P^2\) where the centres \(p_1,\dots,p_n\) of the blowup can move on a fixed cubic curve \(E\subset\mathbb P^2\). There is a \(\mathbb Z^n\)-action on theses that covers a \(\mathbb Z^n\)-action on \(E^n\) by translations. This deforms the dynamics. Also, there is no notion of support of a sheaf in this noncommutative geometry, meaning that the tensor-action above is not connected to support; the orbits are typically dense. After this introduction, the authors focus on the 3-generator Sklyanin algebra (the elliptic quantum \(\mathbb P^2\)) and discuss the dynamics of this from several angles, including an explicit linear algebra description of it which can be reformulated as addition on a moving Jacobian. The first non-trivial case considered gives the elliptic Painlevé equations of \textit{H. Sakai} [Commun. Math. Phys. 220, No. 1, 165--229 (2001; Zbl 1010.34083)] which gives all other Painlevé equations by degenerations and continuous limits. An explicit computation of this case shows that the system of isomorphisms between moduli spaces agrees with the corresponding system of isomorphisms between rational surfaces considered by Sakai. The semiclassical limit of the elliptic quantum \(\mathbb P^2\) degenerates to a Poisson structure on a commutative \(\mathbb P^2\), inducing a Poisson structure on suitable moduli spaces of sheaves. The authors show that these structures on moduli spaces carry over to the noncommutative setting. The article is built on blowups and Hecke modifications: A coherent sheaf \(\mathcal M\) on \(\mathbb P^2\) is an object in the category \(X\) above with \(A=\mathbb C[x_0,x_1,x_2]\), and the moduli of stable sheaves exists. Then the authors study \textit{birational} group actions. Thus it is sufficient to study open dense subsets of the moduli spaces of the form \(\mathcal M=\iota_\ast L\) where \(\iota:C\hookrightarrow S\) is an inclusion of a smooth curve of degree \(d\), and \(L\) is a line bundle on \(C\). Such sheaves are stable, and their moduli space is a fibration over the base \(B=\mathbb P^{d(d+3)/2}\setminus\{\text{singular curves}\}\) of non-singular curves \(C\) with fibre \(\text{Jac}_{\deg L}C.\) A curve \(C\subset\mathbb P^2\) meets a point \(p\) if and only if \(C=\text{Bl}\widetilde C\) with \(\widetilde C\subset S\). Line bundles \(\widetilde L\) on \(\widetilde C\) can be pushed forward to \(\mathbb P^2\) to give sheaves that surject to the structure sheaf \(\mathcal O_p\) of \(p\). If \(\widetilde{\mathcal M}\) is such a line bundle viewed as a sheaf on \(S\), then the sheaves \(\mathcal M=\text{Bl}_\ast\widetilde{\mathcal M}\) and \(\mathcal M^\prime=\text{Bl}_\ast\widetilde{\mathcal M}(-\mathcal E)\), \(\mathcal E\) the exceptional divisor of the blowup, fit in an exact sequence \(0\rightarrow\mathcal M^\prime\rightarrow\mathcal M\rightarrow\mathcal O_p\rightarrow 0\). When this is the case, \(\mathcal M\) and \(\mathcal M^\prime\) are \textit{Hecke modifications} of each other: Hecke modifications at \(p\) corresponds to twists by the exceptional divisor on the blowup with center \(p\). The noncommutative analogue of the correspondence between Hecke modifications and twists on the blowup give needs to describe the geometric constructions in module-theoretic terms. This leads to an explicit study of Poisson structures, via Van den Bergh's noncommutative blowup, the sheaves on quantum planes, the Weyl group action on parabolic sheaves. The first teorems in the article give expicit descriptions of the actions, and in the case \(d=3\) in particular, the \textit{elliptic Painlevé equation} appear. A nice contribution to noncommutative geometry that is based on geometric group actions. The article makes a lot of simplifications in the beginning, making the introductory sections straightforward. The last chapters are more demanding, but of great value.