Strong Heegaard diagrams and strong L-spaces (Q728312): Difference between revisions

In Heegaard Floer homology theory, L-spaces are the simplest \(3\)-manifolds. Their Heegaard Floer homology groups are free abelian groups with rank equal to the order of the first homology group. The purpose of the present paper is to study a subfamily of L-spaces called strong L-spaces. Given a closed oriented \(3\)-manifold \(Y\), a Heegaard diagram \(H\) of \(Y\) yields a chain complex \(\widehat{CF}(H)\) which is freely generated by a certain finite set. The minimum size \(M(Y)\) of this generating set ranging over all Heegaard diagrams is a measure of the topological complexity of \(Y\). Then the inequality \(M(Y)\geq\mathrm{rank }\widehat{\mathrm{HF}}(Y)\geq \det Y\) holds. A strong L-space is a closed oriented \(3\)-manifold \(Y\) with \(M(Y)=\det Y\). This notion was first introduced by \textit{A. Levine} and \textit{S. Lewallen} [Math. Res. Lett. 19, No. 6, 1237--1244 (2012; Zbl 1283.57019)]. Any strong L-space admits a strong Heegaard diagram which gives the above generating set of size equal to its determinant. The most typical L-spaces are lens spaces, which are always strong L-spaces. Also, the double branched covers of non-split alternating links are strong L-spaces. However, the notion of strong L-spaces is much stronger than that of L-spaces. For example, the only strong L-space with determinant one is the \(3\)-sphere. This implies that the Poincaré homology \(3\)-sphere is an L-space but not a strong L-space. The authors ask whether any strong L-space is the double branched cover of an alternating link or not. This question is still open, but it is solved affirmatively for genus two strong L-spaces and strong L-spaces with determinant at most \(8\). In the former case, strong L-spaces of Heegaard genus two are completely determined as the double branched covers of certain links. In the latter case, strong L-spaces are the connected sums of lens spaces. More generally, it is shown that there exist only finitely many rational homology spheres \(Y\) with bounded \(M(Y)\). As a corollary, there exist only finitely many strong L-spaces with bounded determinant. This is in contrast to L-spaces. The arguments are topological and graph theoretic. The last part of the paper discusses Floer simple knots contained in strong L-spaces, and formal L-spaces.