Optimal cobordisms between torus knots (Q2396640)

An optimal cobordism between two torus knots \(K_0\) and \(K_1\) is a smooth cobordism from \(K_0\) to \(K_1\) of genus \(|g(K_0)-g(K_1)|\); this is the smallest possible genus allowed by the Milnor conjecture. These cobordisms are related to deformations of curve singularities, and more generally to singular curves in the complex plane; indeed, an algebraic (or even symplectic) cobordism between two torus knots is known to be optimal [\textit{P. B. Kronheimer} and \textit{T. S. Mrowka}, Topology 32, No. 4, 773--826 (1993; Zbl 0799.57007)]. This is where the motivation of the paper lies, and the main results (Theorems 1 and 2) assert that, when one of the two torus knots is of the form \(T(2,n)\) and the other one has braid index 3 or 4, then there exists an optimal smooth cobordism between them if and only if there is an algebraic cobordism between them. (Equivalent numerical conditions are given, too.) Corollary 3 measures the failure of optimality for torus knots whose sum of braid indices is at most 6; this failure is expressed in terms of two concordance invariants coming from Heegaard Floer homology, namely \(\tau\) [\textit{P. Ozsváth} and \textit{Z. Szabó}, Geom. Topol. 7, 615--639 (2003; Zbl 1037.57027)] and \(\upsilon\) [\textit{P. S. Ozsváth}, \textit{A. I. Stipsicz}, and \textit{Z. Szabó}, Adv. Math. 315, 366--426 (2017; Zbl 1383.57020)].