A Furuta-like inequality for spin orbifolds and the minimal genus problem (Q5939834)

The authors generalize Furuta's 10/8ths inequality on smooth spin \(4\)-manifolds to spin \(4\)-orbifolds. This generalization admits many applications to smooth spin \(4\)-orbifolds. Recall that the famous 11/8ths conjecture states that for any closed connected oriented smooth spin \(4\)-manifold \(X\) the second Betti number \(b_2(X)\) and the signature \(\sigma(X)\) are related via the inequality \(b_2(X)\geq (11/8) \sigma(X)\). Although the conjecture has not been proven until now, Furuta has proven a weaker inequality: \[ b_2(X)\geq (10/8) \sigma(X) + 2 \] [\textit{M. Furuta}, Monopole equation and the \(\frac{11}{8}\)-conjecture, Math. Res. Lett. 8, No. 3, 279-291 (2001; Zbl 0984.57011)]. The author generalizes these results to orbifolds. He proves: Theorem. Let \(Y\) be a \(4\)-dimensional spin orbifold with a finite number of non-manifold singularities which are cones on lens spaces. Then, \[ \begin{aligned} B_2^+(Y) \geq \text{ind}(D) +1&\quad \text{if}\quad\text{ind}(D)>0 \text{ and }b_2^+(Y)>0;\\ b_2^-(Y)\geq |\text{ind}(D)|+1&\quad \text{if}\quad \text{ind}(D)<0 \text{ and } b_2^-(Y)<0, \end{aligned} \] where \(\text{ind}(D)\) is the (complex) index of the Dirac operator of the orbifold. This theorem admits applications to the theory of smooth \(4\)-manifolds. Theorem. Suppose \(\xi\) is a characteristic homology class in an indefinite, smooth, oriented, \(4\)-manifold \(X\) with nonpositive signature which is represented by an embedded \(2\)-sphere. Then \[ \begin{aligned} |\xi\cdot \xi -\sigma(X)|&\leq 8(b_2^-(X)-2) \quad \text{if}\quad\xi\cdot\xi\leq \sigma(X) \text{ and }b_2^-(X)>1;\\ |\xi\cdot \xi -\sigma(X)|&\leq 8(b_2^+(X)-2) \quad \text{if}\quad 0\leq \xi\cdot\xi \text{ and } b_2^+(X)>1;\\ |\xi\cdot \xi -\sigma(X)|&\leq 8(b_2^+(X)-1) \quad \text{if}\quad \sigma(X)\leq \xi\cdot\xi\leq 0. \end{aligned} \] Theorem. Suppose \(\xi\) is a characteristic homology class in a smooth, oriented, simply connected \(4\)-manifold \(X\) with nonpositive signature which is represented by an embedded oriented surface of genus \(g\). If \(\xi\cdot \xi\equiv \sigma(X)\bmod 16\), then \[ \begin{aligned} |\xi\cdot \xi -\sigma(X)|&\leq 8(g+b_2^-(X)-2) \quad \text{if}\quad \xi\cdot\xi\leq \sigma(X) \text{ and } g+b_2^-(X)>1;\\ |\xi\cdot \xi -\sigma(X)|&\leq 8(g+b_2^+(X)-2) \quad \text{if}\quad 0\leq \xi\cdot\xi \text{ and } g+b_2^+(X)>1;\\ |\xi\cdot \xi -\sigma(X)|&\leq 8(g+b_2^+(X)-1) \quad \text{if}\quad \sigma(X)\leq \xi\cdot\xi\leq 0 \text{ and }g+b_2^+(X)>1.\end{aligned} \] If \(X\) is a smooth \(4\)-manifold with positive definite intersection form of rank \(n\geq 2\), he shows that a smoothly embedded characteristic sphere \(\xi\) satisfies \(\xi\cdot \xi \leq 9n- 16\). For proving the results, the author develops a version of Seiberg-Witten theory which holds on spin orbifolds, including a nice presentation of spin structures, Dirac operators and and the Atiyah-Singer-index theorem on such spin orbifolds.