We investigate the ten dimensional origin of six dimensional F 4 variant supergravity with supersymmetric de Sitter background. We address first the issue of spontaneous compactification, showing that it consists of a warped compactification on a four sphere of a variant massive type IIA supergravity. Moreover we illustrate how the known D4-D8 brane solution, whose near horizon geometry yields AdS 6 ⊗ S 4 , is accordingly modified to a system including Euclidean branes. Finally, we discuss the relation between this latter solution and the D4-D8 brane system, showing how it represents a generalisation of the DW/Cosmology correspondence.
IntroductionAmong the supergravity theories with supersymmetric AdS vacua, D = 6 N = 2, supergravity based on the exceptional supergroup F 4 [1] is somehow peculiar. Indeed, F 4 appears to be the only supergroup admitting two real sections 1 whose bosonic generators span respectively the algebra SO(2, 5) ⊗ SU(2) and SO(1, 6) ⊗ SU(2). This reflects into the existence of two version of F 4 supergravity: the standard one 2 , F 4 (1, 5), with supersymmetric AdS 6 background [1,4,5], and a variant version, F * 4 (1, 5), with supersymmetric dS 6 background [6]. F * 4 (1, 5) is a "variant" theory in the sense discussed by Hull [7]. Variant type II supergravities were introduced [7] considering T-duality transformations involving timelike circles; consequently lower dimensional variant supergravities naturally arise e.g. from compactifications on non-Euclidean tori. Hence, variant supergravities can occur in non-Lorentzian signatures. Quite generally they also have ghosts, and in lower dimensions they may have non-compact R-symmetry groups. F * 4 (1, 5) supergravity has Lorentzian signature, nevertheless is a variant theory since its vector fields are ghosts. Remarkably the R-symmetry group is compact, since it is SU(2) for both real sections. This fact turns out to be quite relevant in the understanding of its ten dimensional origin.It is in fact well known [8] that F 4 (1, 5) supergravity can be obtained from a consistent Kaluza-Klein compactification of massive IIA m (1, 9) [9] on a four-sphere. More precisely not on the whole sphere, rather on an hemisphereS 4 viewed as a foliation of three-spheres S 3 , whose rigid deformations parametrise SU(2). This observation strongly suggests that F * 4 (1, 5) must come from a similar compactification where at least the foliating S 3 is a genuine compact three-sphere.In Section 1 we will see that this is actually the case. Modifying the ansatz in [8] we show that F * 4 (1, 5) can be obtained from a compactification of 3 IIA * m (5, 5) on a timelike foursphereS 4 . The signature (5, 5) is rather peculiar: it is in fact the only signature, together with (1, 9), for which we can impose (pseudo-)Majorana and Weyl conditions at the same time. Moreover, apart from the space-time signature, the action of IIA m (5, 5) coincides with the action of IIA m (1, 9). The same happens for IIA * m (5, 5) and IIA * m (1, 9), since the action of both theor...