Ω-Admissible Theory (Q4487592)

This paper is devoted to what is commonly called Arakelov theory for arithmetic surfaces. Namely, in 1974, \textit{S. Ju. Arakelov} [Math. USSR, Izv. 38 (1974), 1167--1180 (1976; Zbl 0355.14002)] had defined a real-valued intersection product for divisors on a regular arithmetic surface \(X\) depending on a fixed Hermitian metric \(\omega\) on the Riemann surface \(X(\mathbb{C})\). About ten years later, in his famous paper ``Calculus on arithmetic surfaces'' [Ann. Math. (2) 119, 387--424 (1984; Zbl 0559.14005)], \textit{G. Faltings} developed this arithmetic intersection theory much further, thereby demonstrating the existence of an amazing analogy with the classical complex-geometric theory. As to the fixed Hermitian metric on \(X(\mathbb{C})\), Faltings chose the one defined by the canonical volume form on \(X(\mathbb{C})\). However, from the moduli point of view, it has become evident that hyperbolic metrics on Riemann surfaces are more natural and convenient, at least so in the complex-geometric context. With a view to this very fact, the author of the article under review develops a more general ``\(\omega\)-admissible theory'', that is an intersection theory on arithmetic surfaces based upon an arbitrary fixed (and normalized) volume form \(\omega\) on \(X(\mathbb{C})\) à la Arakelov-Faltings. In Section 1, such a general admissible theory is explicitly constructed, partly along the line drawn by Arakelov and Faltings in their foregoing works, but also by using more recent results such as the theory of the Quillen metric and the arithmetic Riemann-Roch theorem of \textit{H. Grillet} and \textit{Ch. Soulé} [Invent. Math. 110, 473--543 (1992; Zbl 0777.14008)]. Moreover, the author shows that various \(\omega\)-admissible theories are closely related by the so-called mean value lemma. Section 2 turns to some applications of the author's general \(\omega\)-admissible theory. First it is shown that the new general approach really provides the classical Arakelov-Faltings theory when fixing the canonical Hermitian metric. The second application is obtained by taking \(\omega\) as the (normalized) hyperbolic Poincaré volume form \(\omega_{\text{hyp}}\) on \(X(\mathbb{C})\). The corresponding \(\omega_{\text{hyp}}\)-admissible theory transpires a new invariant, the so-called Arakelov-Poincaré volume on a compact Riemann surface, and one of the author's main results says that this invariant actually forms a Weil function on the compactified moduli space \(\overline M_g\) of stable compact Riemann surfaces of genus \(g\). In addition, the author derives an upper bound for the self-intersection of the classical Arakelov divisor in terms of Petersson norms of modular forms. In the final third section, the author deals with singular arithmetic surfaces and develops an \(\omega\)-admissible (arithmetic intersection) theory for this even more general case. His approach is motivated by the work ``An arithmetic Riemann-Roch theorem for singular arithmetic surfaces'' by \textit{W. Aitken} [Mem. Am. Math. Soc. 573 (1996; Zbl 0846.14008)] and yields, as an interesting by-product, another Riemann-Roch theorem for singular arithmetic surfaces. As for the latter result, the author introduces a suitable metric on the (Deligne) determinant of cohomology. This interesting and important paper ends with an appendix, in which Deligne pairings are calculated in terms of determinant line bundles, mainly with a view towards applications to Chow-Mumford stability and Kähler-Einstein metrics in the sequel.