Marked points of families of hyperbolic automorphisms of smooth complex projective varieties
Yugang Zhang
TL;DR
This work develops a two-pronged framework to quantify stability for marked points in flat families of smooth projective varieties with fiberwise hyperbolic automorphisms of positive entropy. It constructs both geometric geometric canonical heights (via relative divisors D^± and their λ-Scaling) and analytic Green currents (with continuous potentials) and proves their coincidence when the base is a curve, yielding a unified stability theory. The paper establishes a weak Northcott property and a function-field Kawaguchi–Silverman result for non-isotrivial families, and it develops a relative divisor toolkit (including the augmented base locus) alongside Chow- and Hilbert-scheme parameter spaces to study the sparsity and density of stable marked points. These results link dynamical complexity, arithmetic growth, and geometric positivity, providing a mechanism to transfer dynamical stability information to function-field height considerations and to draw KS-type conclusions in this setting.
Abstract
Let $π: X\to Λ$ be a flat family of smooth complex projective varieties parameterized by a smooth quasi-projective variety $Λ$, and let $f: X\to X$ be a family of automorphisms with positive topological entropy. Suppose $σ: Λ\to X$ is a marked point, i.e., it is a rational section of $π$. We propose two methods to measure the stability, normality, or periodicity of the family given by $t \mapsto f_t^n(σ(t))$. First, from an algebraic perspective, we construct geometric canonical height functions that have desirable properties. Second, from an analytic viewpoint, we construct a positive closed $(1,1)$-current with continuous local potential. When $Λ$ is a curve, we demonstrate that these two constructions actually coincide, providing a unified approach to understanding the dynamical behavior of the family. As an application of the algebraic method, we prove a special case of the Kawaguchi-Silverman conjecture over complex function fields.
