Integration of twisted Dirac brackets

From MaRDI portal



Abstract: The correspondence between Poisson structures and symplectic groupoids, analogous to the one of Lie algebras and Lie groups, plays an important role in Poisson geometry; it offers, in particular, a unifying framework for the study of hamiltonian and Poisson actions. In this paper, we extend this correspondence to the context of Dirac structures twisted by a closed 3-form. More generally, given a Lie groupoid G over a manifold M, we show that multiplicative 2-forms on G relatively closed with respect to a closed 3-form phi on M correspond to maps from the Lie algebroid of G into the cotangent bundle T∗M of M, satisfying an algebraic condition and a differential condition with respect to the phi-twisted Courant bracket. This correspondence describes, as a special case, the global objects associated to twisted Dirac structures. As applications, we relate our results to equivariant cohomology and foliation theory, and we give a new description of quasi-hamiltonian spaces and group-valued momentum maps.


Dirac structures were introduced by \textit{T. J. Courant} [Trans. Am. Math. Soc. 319, No. 2, 631--661 (1990; Zbl 0850.70212), and with \textit{A. Weinstein} [Trav. Cours 27, 39--49 (1988; Zbl 0698.58020)] as a generalization of Poisson structures, presymplectic forms and regular foliations. Courant's main motivation was to study constrained mechanical systems. Connection of Poisson structures to topological sigma-models led to the definition of Poisson structures twisted by a closed \(3\)-form and more generally twisted Dirac structures [\textit{P. Severa} and \textit{A. Weinstein}, Prog. Theor. Phys., Suppl. 144, 145--154 (2001; Zbl 1029.53090)]. In [\textit{M. Crainic} and \textit{R. L. Fernandes}, Ann. Math. (2) 157, No. 2, 575--620 (2003; Zbl 1037.22003)] it was shown that many Lie algebroids are integrable, i.e.~there is a Lie groupoid such that the given Lie algebroid is the Lie algebroid to this Lie groupoid. This result by Crainic and Fernandes should be viewed as a generalization of Lie's third theorem stating that an abstract Lie algebra is always the Lie algebra to a Lie group. The integration of such a Lie algebroid is not only an enourmous conceptual progress, but it rapidly turned out that this integration is very useful for solving analytical problems on non-compact, complete manifolds and on foliated manifolds. In a similar way, it was shown in [\textit{A. Coste, P. Dazord} and \textit{A. Weinstein}, Publ. Dép. Math., Nouv. Sér., Univ. Claude Bernard, Lyon 2/A, 1--62 (1987; Zbl 0668.58017), \textit{M. Crainic} and \textit{R. L. Fernandes}, J. Differ. Geom. 66, No. 1, 71--137 (2004; Zbl 1066.53131)] that the infinitesimal object ``integrable (twisted) Poisson structure could be integrated to a (twisted) symplectic groupoid. In view of this, it is natural to ask for an integration of the more general twisted Dirac structures. This is the subject of the article under review. A twisted Dirac structure gives rise to a Lie algebroid structure. This Lie algebroid structure can be integrated to a Lie groupoid. The main task is now to determine the additional structure on the Lie groupoid that arises by integrating the twisted Dirac structure. A Lie groupoid with such an additional structure is called \textit{\(\phi\)-twisted presymplectic groupoid}. To give some more details: Let \(G\) be a \(2n\)-dimensional Lie groupoid over an \(n\)-dimensional manifold \(M\) with target map \(t\) and source map \(s\). We assume that \(\phi\) is a closed \(3\)-form on \(M\). Then \((G,\omega,\phi)\) is called a \textit{\(\phi\)-twisted presymplectic groupoid} if \(\omega\) is a multiplicative \(2\)-form on \(G\) such that \[ d\omega = s^*\phi -t^*\phi \] and \[ \text{ ker\;}(\omega_x)\cap \text{ ker}(ds)_x \cap \text{ ker}(dt)_x=\{0\} \] for all \(x\in M\). The main result of the article is that modulo integrability issues, there is a \(\phi\)-twisted presymplectic groupoid for any \(\phi\)-twisted Dirac structure, and that \(\phi\)-twisted Dirac structure arise infinitesimally from \(\phi\)-twisted presymplectic groupoids. This correspondence is bijective if we only consider \(s\)-simply connected Lie groupoids. The article starts with a well written introduction to the subjects, which provides motivation, the historical background and suitable references. Section~2 gives definitions and basic properties of the main objects of the article such as ``twisted Dirac structures, ``groupoids and ``multiplicative \(2\)-forms. At the end of this section, the main results are properly stated. The following sections are devoted to the proof. After having proved several technical preliminaries about multiplicative \(2\)-forms in section 3, the authors describe the passage from twisted presymplectic groupoids to twisted Dirac structures (section 4), and the inverse procedure in section~5. Section~6 is mainly devoted to examples. The example of Cartan-Dirac structures on Lie groups leads to section~7, in which the associated Alekseev-Malkin-Meinrenken groupoid is studied. Finally, section 8 studies multiplicative \(2\)-forms on foliation groupoids, and the connections to the spectral sequence of this foliation.



Cites work


Cited in
(84)








This page was built for publication: Integration of twisted Dirac brackets

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1880386)