A Geometric Finiteness Theory for Essential Surfaces in Knot Exteriors

Abstract

We develop a relative geometric finiteness theory for essential surfaces in knot exteriors. Let $\gamma$ be a unit-thickness representative of a knot type $K$, with $\mathrm{Len}(\gamma )\le \Lambda$, and let $F\subset E(\gamma )$ be a properly embedded essential surface with $\mathrm{Area}(F)\le \Delta$ and relative thickness at least $\tau$, defined using positive reach and controlled boundary collars. We prove that every bounded-geometry slice contains only finitely many pair-isotopy classes. We construct explicitly bounded canonical layered codes on a fixed ambient lattice and show that, at resolution $\epsilon \le c\min \{1,\tau \}$ with sufficiently fine angular quantization, equality of codes implies ambient pair-isotopy. Thus the topology of each bounded slice is recoverable from finite geometric data. For a fixed exterior, these classes form finite visible subcomplexes of the essential-surface complex; the subcomplexes are monotone, exhaust the full complex, and carry isometric actions levelwise and meridian-preserving $C^{1,1}$ actions with controlled reindexing. Positive-reach compactness also yields attainment results for fixed-exterior and compactified visibility problems. Finally, the peripheral geometry gives a writhe window for connected surfaces with nonempty non-meridional boundary: \[ |r|\leq C_{\mathrm{BS}}Λ^{4/3}+w(Δ,τ). \] This produces slope invisibility gaps and a linear joint-area lower bound for a Seifert surface and cabling annulus of a torus knot. The framework is triangulation-free and complementary to normal-surface, branched-surface, sutured-manifold, and Heegaard-theoretic methods; it does not assert finiteness without geometric bounds.

102 pages, 21 figures, 2 tables