Instanton and the depth of taut foliations (Q2053660)

A sutured 3-manifold \((M, \gamma)\) is a compact orientable 3-manifold \(M\) with some extra data \(\gamma\), called sutures, on its boundary. Sutured 3-manifolds were introduced by \textit{D. Gabai} [J. Differ. Geom. 18, 445--503 (1983; Zbl 0533.57013)] to study two-dimensional transversely oriented foliations of 3-manifolds; the sutures record, qualitatively, how a foliation is allowed to intersect the boundary. Taut sutured manifolds are those sutured manifolds (besides \(S^2 \times S^1\) and \(S^2 \times I\)) that admit taut foliations; i.e. a foliation with a compact 1-manifold intersecting every leaf transversely. Balanced sutured 3-manifolds were introduced by \textit{A. Juhász} [Algebr. Geom. Topol. 6, 1429--1457 (2006; Zbl 1129.57039)] to define sutured (Heegaard) Floer homology, a generalisation of Heegaard Floer homology. An example of a taut balanced sutured 3-manifold to keep in mind is the following: start with the complement of a regular neighbourhood of a knot in \(S^3\) and cut along a minimal genus Seifert surface \(R\) for the knot. \textit{D. Gabai} [Comment. Math. Helv. 61, 519--555 (1986; Zbl 0621.57003)] observed that a knot in \(S^3\) is fibered if and only if the preceding sutured 3-manifold has a sutured manifold hierarchy of length \(0\); i.e. it is a product sutured 3-manifold \((R \times I, \partial R \times I)\). Juhász proved that given a taut balanced sutured 3-manifold \((M, \gamma)\) with vanishing second homology \(H_2(M)\), and the rank of the sutured Floer homology less than \(2^{k+1}\), \((M, \gamma)\) admits a sutured manifold hierarchy of length at most \(2k\). In particular, setting \(k=0\), sutured Floer homology detects product sutured 3-manifolds. \textit{A. Juhász} [Geom. Topol. 12, No. 1, 299--350 (2008; Zbl 1167.57005), Question 9.14] also asked whether \((M, \gamma)\) admits a taut foliation of depth less than \(2k\)? Motivated by Juhász's question, \textit{P. Kronheimer} and \textit{T. Mrowka} conjectured the following in the instanton context [J. Differ. Geom. 84, No. 2, 301--364 (2010; Zbl 1208.57008), Conjecture 7.27]: ``Let \(K \subset S^3\) be a knot, and consider the irreducible homomorphisms \(\rho \colon \pi_1(S^3 \setminus K) \rightarrow \mathrm{SU}(2)\) which map a chosen meridian \(m\) to the element \(i \in \mathrm{SU}(2)\). Suppose that these homomorphisms are non-degenerate and that the number of conjugacy classes of such homomorphisms is less than \(2^ {k+1}\). Then the knot complement \(S^3 \setminus N^\circ(K)\) admits a foliation of depth at most \(2k\), transverse to the torus boundary.'' The paper under review gives a partial answer to the above question of Juhász and the conjecture of Kronheimer and Mrowka, with the bound \(2k\) replaced by \(2^{k+6}\). The strategy of the proof is as follows: Following Juhász, one can cut along i) a disjoint union of product annuli, and then ii) a well-groomed surface, and reduce the rank of the sutured Floer homology by a factor of \(\frac{1}{2}\). By the work of \textit{D. Gabai} [Zbl 0533.57013, J. Differ. Geom. 26, 479--536 (1987; Zbl 0639.57008)], decomposing along a well-groomed surface increases the depth of the associated taut foliation by at most \(1\). However, the annuli in i) need not be well-groomed in general. Instead, the author shows that decomposing along the annuli in i) can be replaced by a sequence of decompositions (with a modified terminal manifold obtained by attaching product pieces along a subset of the sutures, but crucially with the same sutured Floer homology) where each decomposing surface is either well-groomed or has boundary contained in an annular neighbourhood of the sutures; this will suffice to bound the depth of the taut foliation obtained by gluing back along the decomposed surface. Moreover, the length of the substituted sequence of decompositions can be suitably bounded from above. An induction on \(k\) completes the proof. The proof works for sutured instanton Floer homology, and sutured monopole Floer homology as well. Some of the statements in the paper rely on recent joint work of the author with \textit{S. Ghosh} [``Decomposing sutured monopole and instanton Floer homologies'', Preprint, \url{arXiv:1910.10842}].