A combinatorial proof of simplified Kollár's theorem on Pell's equations (Q2192154)

Consider two algebraic varieties \(U\) and \(V\). One of the fundamental problems between logic and algebraic geometry is isomorphism problem: Does there exist a checking isomorphism algorithm? It was found that, in particular, for topological varieties such algorithm does not exist. The paper [\textit{A. Ya. Kanel'-Belov} and \textit{A. A. Chilikov}, Math. Notes 106, No. 2, 299--302 (2019; Zbl 1442.14185); translation from Mat. Zametki 106, No. 2, 307--310 (2019)] treats embedding problem. The proof plays with Pell equation and results provided by \textit{J. Kollár} [Acta Math. Hung. 160, No. 2, 478--518 (2020; Zbl 07200111)]. In 1826 \textit{N. H. Abel} [J. Reine Angew. Math. 1, 185--221 (1826; Zbl 02751787)] started the study of the polynomial Pell equation \(x^2-g(u)y^2=1\). Its solvability in polynomials \(x(u), y(u)\) depends on a certain torsion point on the Jacobian of the hyperelliptic curve \(v^2=g(u)\). In his paper Kollár study the affine surfaces defined by the Pell equations in \(3\)-space with coordinates \(x, y,u\), and aim to describe all affine lines on it. These are polynomial solutions of the equation \(x(t)^2-g(u(t))y(t)^2=1\). His results are rather complete when the degree of \(g\) is even but the odd degree cases are left completely open. For even degrees he also describe all curves on these Pell surfaces that have only \(1\) place at infinity. This proof is very deep and highly non-trivial and far from been elementary. In the refereed paper author provide elementary proof of some partial case, needed for logarithmical insolvability for embedding problem. Consider Pell equation \(X^2-g(t)*Y^2 = 1\) in the ring of polynomials \(F[t]\). Then the set of solutions is an abelian group \(G\) respect to operation \((X,Y)*(X',Y') = (XX'+gYY', XY'+X'Y)\). It was found that \(G\simeq\mathbb{Z}_2 \oplus \mathbb{Z}^r\), \(r\le 1\). When \(g(t) = (u(t))^2 -1\) we have \(G\simeq\mathbb{Z}_2 \oplus\mathbb{Z}\) and the solution \((u,1)\) is fundamental. The authors also prove the Kollár theorem supposing that non-trivial solution exist. Existence problem, resolved by J. Kollár, they did not handle. J. Kollár's paper inspired an project on the conference \url{https://www.turgor.ru/lktg/2019/} Efremov, Zveryk and Khomich were high school students at that time and the refereed paper is an output of it.