Higgs bundles and \((A,B,A)\)-branes (Q406662): Difference between revisions

\textit{M. F. Atiyah} in [Q. J. Math., Oxf. II. Ser. 17, 367--386 (1966; Zbl 0146.19101)] advocated the philosophy that real and quaternionic geometric objects are to be thought of as complex objects, invariant under an involution. Hence a (smooth, geometrically connected, projective) real algebraic curve is seen as a pair, consisting of a Riemann surface \(\Sigma\) and an anti-holomorphic involution \(f\), encoding the action of the absolute Galois group \(\text{Gal}(\mathbb{C}/\mathbb{R})\simeq \mathbb{Z}/2\mathbb{Z}\). Let \(G\) be a connected reductive complex Lie group and \(G^\rho\) a compact real form, corresponding to an anti-linear involution \(\rho\) on \(\text{Lie } G\). Given a holomorphic \(G\)-bundle \(P\to \Sigma\) with a \(G^\rho\)-reduction, one can construct a fibrewise anti-linear involution on \(f^\ast P\). This induces an involution on the (coarse) moduli space of (polystable) bundles. Conceivably, the geometry and topology of the fixed loci -- which are often physicists' branes -- can be understood by incorporating the involutions into the gauge theory and studying the minima of the Yang--Mills functional. This would be a natural extension of the seminal work of \textit{M. F. Atiyah} and \textit{R. Bott} [Philos. Trans. R. Soc. Lond., Ser. A 308, 523--615 (1983; Zbl 0509.14014)] to the real (or quaternionic) context. A first step in this direction -- the study of stable vector bundles -- was made by \textit{I. Biswas} et al. [Math. Ann. 347, No. 1, 201--233 (2010; Zbl 1195.14048)]. Naturally, one should next proceed with \(G\)-bundles and ``decorated bundles'', e.g., Higgs bundles or bundles with connection. The present article is concerned with the non-abelian Hodge theory of the pair \((\Sigma, f)\), for \(\Sigma\) of genus \(g\geq 2\). The strategy is to consider Hitchin's moduli space of solutions to the self-duality equations [\textit{N. J. Hitchin}, Proc. Lond. Math. Soc. (3) 55, 59--126 (1987; Zbl 0634.53045)] and induce a real structure on it. This space is a hyperkaehler orbifold, and, as usual, we denote its Dolbeault (Higgs) complex structure by \(I\), and its de Rham (flat) complex structure by \(J\). In \S 2, the authors introduce an action of \(\mathbb{Z}/2\mathbb{Z}\) on the fundamental group of \(\Sigma\) as follows. A choice of basepoint \(x_0\in\Sigma\) and a path \(\gamma\) joining \(x_0\) to \(f(x_0)\) determine an order-2 automorphism \(\widehat{f}\) of \(\pi:= \pi_1(\Sigma,x_0)\), defined up to an inner automorphism -- this is the natural map \(f_\ast: \pi_1(\Sigma, x_0)\to\pi_1(\Sigma, f(x_0))\) followed by conjugation with \(\gamma\). By composition, \(\widehat{f}\) induces an involution on the set of reductive representations \(\text{Rep}^{+}(\pi, G)= \text{Hom}^{+}(\pi_1, G)/G\). By Proposition 8, \(\widehat{f}\) preserves the locus of simple (regular) representations and is holomorphic and anti-symplectic (for the Goldman symplectic structure) there. The smooth points of the fixed locus, \(\mathcal{L}_{G}\), constitute a smooth complex Lagrangian submanifold of \(\text{Rep}^{+}(\pi, G)\), by Proposition 10. The involutions \(f\) and \(\rho\) induce an involution on the moduli space of solutions to Hitchin's equations. This involution is \(I\)-anti-holomorphic and \(J\)-holomorphic (Proposition 12), and covers the naturally induced anti-holomorphic involution of the Hitchin base. Consequently, by Theorem 14, \(\mathcal{L}_{G}\) is an \((A,B,A)\)-brane (in the sense of Kapustin and Witten) on the moduli space of (\(K_\Sigma\)-valued) Higgs \(G\)-bundles. By Theorem 17, the Hitchin map makes \(\mathcal{L}_{G}\) a real integrable system. The spectral data and the connected components of the generic fibre (for \(\mathrm{GL}_n\) and \(\mathrm{SL}_n\)) are studied in \S\S 7, 8. Some relations to the gauge-theoretic approach of \textit{A. Kapustin} and \textit{E. Witten} to the geometric Langlands program [Commun. Number Theory Phys. 1, No. 1, 1--236 (2007; Zbl 1128.22013)] and \textit{S. Gukov}'s work on 3-manifolds [in: New trends in mathematical physics. Selected contributions of the XVth international congress on mathematical physics, Rio de Janeiro, Brazil, August 5--11, 2006. Dordrecht: Springer. 313--343 (2009; Zbl 1175.81170)] are indicated in \S 12 and \S 11. These topics are discussed in greater detail in the authors' work ``Real structures on moduli spaces of Higgs bundles'' [\url{arXiv:1309.1195}, to appear in Adv. Theor. Math. Phys].