Geometric inequalities for manifolds with Ricci curvature in the Kato class (Q2213342)

This quite interesting and well-written paper presents volume growth estimates for complete Riemannian manifolds satisfying a Sobolev inequality and a spectral-type condition on the negative part of its Ricci curvature. The paper is divided into five sections, the first two being introductory, where besides a general description and basic concepts in the first section, the author gathers and develops in the second section the analytical ingredients, Sobolev inequalities, Schrödinger operators, the Doob transform and Green kernel among others including an Agmon-type volume estimate, see the papers by \textit{S. Agmon} [in: Methods of functional analysis and theory of elliptic equations, Proc. Int. Meet. dedic. Mem. C. Miranda, Naples/Italy 1982, 19--52 (1983; Zbl 0595.58044); \textit{P. Li} and \textit{J. Wang} [J. Differ. Geom. 58, No. 3, 501--534 (2001; Zbl 1032.58016)], and \textit{P. Li} and \textit{J. Wang} [Ann. Sci. Éc. Norm. Supér. (4) 39, No. 6, 921--982 (2006; Zbl 1120.53018)], that play a fundamental role in further parts of the paper and particularly in the proof of its main result. The central aim, Theorem 1.4, corresponds to an extension to higher dimensions of the distinguished result by \textit{P. Castillon} [Comment. Math. Helv. 81, No. 2, 271--286 (2006; Zbl 1114.58025)]. This theorem, under more stringent conditions than Castillon's result, provides lower and upper estimates for volume growth, whereas the lower bound follows from well-known results. The upper one, the crucial point of the paper, involves extensive work. Among the consequences, one detaches that, under the hypothesis of the main result, one infers that the complete manifold is doubling, that is, the volume of balls centered at any point with a given radius is dominated, up to a constant, by the volume of balls with the same center with half radius and it satisfies the Poincaré inequality. The corresponding heat kernel satisfies the Li-Yau estimates as well [\textit{P. Li} and \textit{S. T. Yau}, Acta Math. 156, 154--201 (1986; Zbl 0611.58045)]. The author also proves in the second section a new interesting elliptic estimate, for the gradient of a positive harmonic function, by a variation of the De Giorgi-Nash-Moser iteration scheme (Subsection 2.8, Proposition 2.30). Further, the author also obtains eigenvalue estimates, heat kernels estimates, and Betti number estimates for closed manifolds for which the negative part of its Ricci curvature is bounded in the class, introduced by [\textit{T. Kato}, Isr. J. Math. 13, 135--148 (1973; Zbl 0246.35025)]. For a more recent account on Kato's class see, e.g., [\textit{C. Rose} and \textit{P. Stollmann}, ``Manifolds with Ricci curvature in the Kato class: heat kernel bounds and applications'', Preprint, \url{arXiv:1804.04094}]. In the third section of the paper, the author considers the case of closed manifolds. He obtains geometric and topological estimates based on the Kato constant of the Ricci curvature. For this purpose, the author, in a detailed and quite elaborated way, properly adapts methods introduced by \textit{Q. S. Zhang} and \textit{M. Zhu} [J. Funct. Anal. 275, No. 2, 478--515 (2018; Zbl 1407.35108)], to the paper's context. The fourth and fifth sections handle volume growth estimates. In the fourth section, the author treats global results, and in the fifth section local ones. Both sections forcibly are quite technical, where proper gradient estimates are aimed, step by step a formidable amount of analytical difficulties are overcome. Partial results, in both sections, are firstly obtained by assuming more stringent conditions (e.g., Subsections 4.3 and 5.2) that later are proved to be satisfied what leads to the conclusion of the proof of Theorem 1.4, whose crucial argument partially follows, mutatis mutandis, the reasoning of \textit{T. H. Colding} [Acta Math. 209, No. 2, 229--263 (2012; Zbl 1277.53066)]. After proving the main result, the author concludes with the proofs of Theorems 1.15 (localization at infinity) and 1.20 (localization in a geodesic ball).