Witten's conjecture for many four-manifolds of simple type (Q2344500)

In the 1980s Donaldson demonstrated that solutions to the Yang-Mills equations encoded subtle information about the topology of smooth \(4\)-manifolds. The combined work of many people advanced the state of knowledge of these equations and their solutions. One relevant result here is the Kronheimer-Mrowka structure theorem. This result demonstrated that the Donaldson invariants could be combined into a generating function that was specified by numerical invariants associated to a finite number of basic classes [\textit{P. B. Kronheimer} and \textit{T. S. Mrowka}, J. Differ. Geom. 41, No. 3, 573--734 (1995; Zbl 0842.57022)]. Seiberg and Witten used ideas from quantum field theory to give a physical explanation of the structure theorem, and introduced a new system of equations (the Seiberg-Witten equations) encoding similar information to the Yang-Mills equations [\textit{E. Witten}, Math. Res. Lett. 1, No.6, 769--796 (1994; Zbl 0867.57029)]. Physical arguments let to a formula relating the Seiberg-Witten invariants and Donaldson invariants. This formula is known as Witten's conjecture. Shortly after these new invariants were introduced, \textit{V. Y. Pidstrigach} and \textit{A. N. Tyurin} proposed a system of equations (the SO\((3)\)-monopole equations) containing the Yang-Mills and Seiberg-Witten equations as special cases [Russ. Acad. Sci., Izv., Math. 40, No.2, 267--351 (1993); translation from Izv. Ross. Akad. Nauk, Ser. Mat. 56, No.2, 279--371 (1992; Zbl 0796.14024)]. These equations point the way to a mathematical proof of Witten's conjecture. The way, however, appears to be fairly long with a number of necessary technical details. Paul Feehan and Thomas Leness have been making steady progress along this way in a series of at least ten significant papers [\textit{P. M. N. Feehan} et al., Math. Res. Lett. 6, No. 2, 169--182 (1999; Zbl 0967.57027)], [\textit{P. M. N. Feehan} and \textit{T. G. Leness}, ``Donaldson invariants and wall-crossing formulas. I: Continuity of gluing and obstruction maps'', \url{arXiv:math/9812060}; ``PU(2) monopoles. III: Existence of gluing and obstruction maps'', \url{arXiv:math/9907107}; ``A general SO(3)-monopole cobordism formula relating Donaldson and Seiberg-Witten invariants'', Mem. Amer. Math. Soc., to appear; \url{arXiv:math/0203047}; Witten's conjecture for four-manifolds of simple type'', \url{arXiv:math/0609530v1}; Topology Appl. 88, No. 1--2, 111--145 (1998; Zbl 0931.58012); J. Differ. Geom. 49, No. 2, 265--410 (1998; Zbl 0998.57057); J. Reine Angew. Math. 538, 57--133 (2001; Zbl 0983.57024); J. Reine Angew. Math. 538, 135--212 (2001; Zbl 0983.57025); in: Geometry and topology of manifolds. Papers from the conference held at McMaster University, Hamilton, ON, USA, May 14--18, 2004. Providence, RI: American Mathematical Society (AMS). 97--118 (2005; Zbl 1094.58003)] since 1994. Ignoring technical details, the space of solutions to the the SO\((3)\)-monopole equations is a cobordism connecting the spaces related to the Seiberg-Witten equations to the spaces related to the Donaldson invariants together with some standard parts depending on the underlying manifold. Under the hypothesis that this cobordism fits together as expected, Feehan and Leness proved the monopole cobordism formula. (They are still working on establishing this hypothesis.) This is a formula relating the Donaldson and Seiberg-Witten invariants. This formula contains some polynomials in characteristic classes with universal coefficients that are independent of the underlying manifold. It takes a rather different form from Witten's conjecture. In this paper, Feehan and Leness consider the important special case of manifolds of simple-type. Conjecturally, all standard manifolds will have simple-type. They identify the coefficients of all of the terms in the monopole cobordism formula that appear for manifolds of simple-type. To do so, they apply the formula to a sufficiently large family of smooth \(4\)-manifolds for which one can compute the Donaldson invariants and the Seiberg-Witten invariants. They further show that the monopole cobordism formula implies Witten's conjecture, at least for abundant, Seiberg-Witten simple type manifolds as well as other Seiberg-Witten simple-type manifolds satisfying a certain geography restriction. The paper begins with a nice overview of Donaldson theory and Seiberg-Witten theory. It is yet another piece to the eventual proof of the Witten conjecture relating the Donaldson and Seiberg-Witten invariants.