First eigenvalue of the \(p\)-Laplacian under integral curvature condition (Q2402702): Difference between revisions

The \(p\)-Laplacian on a compact Riemannian manifold \((M^n,g)\), for \(1<p<\infty\), is given by \[ \Delta_p(f) = \operatorname{div} (|\nabla f|^{p-2} \nabla f). \] Of course, \(\Delta_2\) coincides with the Laplace-Beltrami operator on \((M,g)\). \textit{A.-M. Matei} [Nonlinear Anal., Theory Methods Appl., Ser. A, Theory Methods 39, No. 8, 1051--1068 (2000; Zbl 0948.35090)] extended several eigenvalue estimates from \(\Delta_2\) to \(\Delta_p\) for any \(p\). As usual, these theorems assume \(\text{Ric}\geq (n-1)K\) for some real number \(K\). The article under review extends the results by Matei to a situation where one only has an integral bound for the part of the Ricci curvature which lies below \((n-1)K\). Roughly speaking, the assumption \(\text{Ric}\geq (n-1)K\) is replaced by an upper bound for \[ \|\text{Ric}_{-}^K\|_{q,R} = \sup_{x\in M} \left( \int_{B(x,R)} \left(\text{Ric}_{-}^K\right)^q \; d\text{vol} \right)^{1/q}, \] for an adequate \(q\), where \(\text{Ric}_{-}^K(x)= \max\{0,(n-1)K-\rho(x)\}\) and \(\rho(x)\) denotes the smallest eigenvalue for the Ricci tensor. Clearly, \(\|\text{Ric}_{-}^K\|_{q,R}=0\) if and only if \(\text{Ric}\geq (n-1)K\). The results generalized are the ones known as Cheng estimate, Lichnerowicz estimate, Lichnerowicz-Obata estimate, and Faber-Krahn estimate. The article may also be seen as an extension from \(\Delta_2\) to \(\Delta_p\) of previous results by the second author and \textit{P. Petersen} on generalizations of geometrical results under integral curvature bounds, for instance [Geom. Funct. Anal. 7, No. 6, 1031--1045 (1997; Zbl 0910.53029)].