Spectral geometry and asymptotically conic convergence

6Citations
Citations of this article
5Readers
Mendeley users who have this article in their library.

Abstract

We define a new conic metric collapse, asymptotically conic convergence,1 in which a family of smooth Riemannian metrics degenerates to have an isolated conic singularity. For a conic metric (Mo, go) and an asymptotically conic or "scattering" metric (Z, gz), we construct a new non-standard blowup, the resolution blowup. in which the conic singularity in Mo is resolved by Z. This blowup induces a smooth family of metrics {gε} on the compact resolution space M. (M, g ε) is said to converge asymptotically conically to (M o, go) as ε → 0. Let Δε and Δ0 be geometric Laplacians on (M, g0) and (Mo, go), respectively. Our first result is convergence of the spectrum of Δε to the spectrum of Δ0 as ε → 0. Note that this result implies spectral convergence for the k-form Laplacian under certain geometric hypotheses. This theorem is proven using rescating arguments, standard elliptic techniques and the b-calculus of [33]. Our second result is technical: we construct a parameter (ε) dependent heat operator calculus which contains, and hence describes precisely, the heat kernel for Δ as ε → 0. The consequences of this result include the existence of a polyhomogeneous asymptotic expansion for Hε as ε → 0, with uniform convergence down to t = 0. To prove this result, we construct heat spaces as manifolds with corners using both standard and non-standard blowups on which we construct suitable heat operator calculi. A parametrix construction modeled after Melrose's heat kernel construction [33] and a maximum principle argument complete the proof.

Cite

CITATION STYLE

APA

Rowlett, J. (2008). Spectral geometry and asymptotically conic convergence. Communications in Analysis and Geometry, 16(4), 735–798. https://doi.org/10.4310/CAG.2008.v16.n4.a2

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free