An effective function field version of Schmidt's subspace theorem for projective varieties, with arbitrary families of homogeneous polynomials (Q2167513)

The relation between Diophantine approximation and Nevanlinna theory are based on the works of \textit{C. F. Osgood} [Indian J. Math. 23, 1--15 (1981; Zbl 0528.30021); J. Number Theory 21, 347--389 (1985; Zbl 0575.10032)] and \textit{P. Vojta} [Lect. Notes Math. 1239, Diophantine approximations and value distribution theory. Berlin, etc: Springer (1987; Zbl 0609.14011)], where the later provided a correspondence between the results. For instead, the subspace theorem in Diophantine approximation corresponds to Cartan's second main theorem in Nevanlinna theory, and the effective versions of the Schmidt subspace theorem over function fields correspond to the second main theorem for moving targets. In the paper under review, by giving a new lower bound estimate for the Chow weight of a projective variety, the author proved an effective Schmidt subspace theorem for projective varieties on the rational function field with arbitrary families of higher degree homogeneous polynomials. The main result of the paper gives a generalization and also an improvement of author's previous work [J. Geom. Anal. 32, No. 2, Paper No. 52, 29 p. (2022; Zbl 1487.32088)], based on the methods in [Zbl 0609.14011; \textit{M. Ru} and \textit{J. T. Y. Wang}, Int. Math. Res. Not. 2012, No. 3, 651--684 (2012; Zbl 1267.11083); \textit{J. T. Y. Wang}, Math. Z. 246, No. 4, 811--844 (2004; Zbl 1051.11041)], which is stated below. Theorem (1.1): Let \(K\) be the function field of a non-singular projective variety \(V\) defined over an algebraically closed field of characteristic zero and let \(S\) be a finite set of prime divisors of \(V\). Let \(X\) be a smooth projective subvariety of \(\mathbb{P}^N\) defined over \(K\) of dimension \(n\) and degree \(\Delta_X\). Let \(Q_1, \dots, Q_q\) be \(q\) homogeneous polynomials of degree \(d_i\) in \(K[X_0, \dots, X_N]\). Let \(\delta_X\) be the distributive constant of \(\{Q_i\}_{i=1}^q\) with respect to \(X\). Then for every \(\varepsilon > 0\), there exist an effectively computable finite union \(W_\varepsilon \) of proper algebraic subsets of \(\mathbb{P}^N(K)\) not containing \(X\). and effectively computable constants \(C_\varepsilon, C'_\varepsilon\) such that for every \(x \in X \backslash W_\varepsilon\) we have either \[h(x) \leq C_\varepsilon\] or \[\sum_{i=1}^{q} \sum_{\mathfrak{p} \in S} \frac{\lambda_{\mathfrak{p}, Q_i} (x)}{d_i} \leq (\delta_X (n+1) + \varepsilon) h(x) + C'_\varepsilon\] The algebraic subsets in \(W_\varepsilon \) and the constants \(C_\varepsilon, C'_\varepsilon\) depend on \(\varepsilon, N, q, K, S, X\) and the \(Q_i\). Furthermore, the degrees of the algebraic subsets in \(W_\varepsilon \) can be bounded above by \[2 (2n +1) d^{n+1} \Delta_X \left( \binom{d + N}{N} + q+1 \right) \delta_X \varepsilon^{-1} +d, \] where \(d= \operatorname{lcm}(d_1, \dots, d_q)\).