Heat kernel estimates on Harnack manifolds and beyond (Q6118129)

The paper under review represents an excellent concise survey on the heat kernel, in particular of what concerns its large-scale behavior and its relation to global geometry. It starts handling existence questions, specially in the Riemannian context, but also considering non-Riemannian cases (unimodular Lie groups). After introducing Harnack manifolds, the author characterizes them in the beautiful Theorem 4.4, which is a byproduct of his own work, see [\textit{L. Saloff-Coste}, Int. Math. Res. Not. 1992, No. 2, 27--38 (1992; Zbl 0769.58054)], and that was also, independently, obtained by \textit{A. A. Grigor'yan} [Mat. Sb. 182, No. 1, 55--87 (1991; Zbl 0743.58031)]. The author's book [Aspects of Sobolev-type inequalities. Cambridge: Cambridge University Press (2002; Zbl 0991.35002)] presents a more leisured path to it. The conspicuous result has been substantially extended by \textit{K.-T. Sturm} [Osaka J. Math. 32, No. 2, 275--312 (1995; Zbl 0854.35015); J. Math. Pures Appl. (9) 75, No. 3, 273--297 (1996; Zbl 0854.35016)]. Altogether with Maria Gordina and Nate Eldredge the author established in the paper two significant conjectures, 4.5 and 4.6, relative to the case of compact Lie groups. The last part of the article concerns Non-Harnack manifolds, i.e., manifolds for which Theorem 4.4 does not hold. Here the approach is to consider several concrete manifolds that are built up piecewise out of Harnack manifolds with different characteristics, e.g., manifolds with finitely many non-parabolic Harnack ends and non-parabolic manifolds with finitely many Harnack ends. Further, it is considered the case of parabolic manifolds with finitely many Harnack ends. In all these cases the aim is to obtain two-sided heat kernel estimates. Finally, for the purpose of deriving heat kernel estimates, with Dirichlet boundary condition on domains obtained from a Harnack manifold, it is then specifically handled the problem of heat kernel estimates for manifolds with Neumann and Dirichlet boundary conditions. The nice Theorem 10.1, obtained by the author with \textit{P. Gyrya} [Neumann and Dirichlet heat kernels in inner uniform domains. Paris: Société Mathématique de France (SMF) (2011; Zbl 1222.58001)], exemplifies such procedure. The paper concludes mentioning some of the author's more recent results concerning mixed boundary conditions on manifolds and attachments along noncompact submanifolds, see [\textit{E. Dautenhahn} and \textit{L. Saloff-Coste}, ``Heat kernel estimates on manifolds with ends with mixed boundary condition'', Preprint, \url{arXiv:2108.05790}]. For the entire collection see [Zbl 07816360].