2006MNRAS.365..555S

There is still no consensus as to what causes galactic discs to become warped. Successful models should account for the frequent occurrence of warps in quite isolated galaxies, their amplitude as well as the observed azimuthal and vertical distributions of the HI layer. Intergalactic accretion flows and intergalactic magnetic fields may bend the outer parts of spiral galaxies. In this paper we consider the viability of these non-gravitational torques to take the gas off the plane. We show that magnetically generated warps are clearly flawed because they would wrap up into a spiral in less than two or three galactic rotations. The inclusion of any magnetic diffusivity to dilute the wrapping effect causes the amplitude of the warp to damp. We also consider the observational consequences of the accretion of an intergalactic plane-parallel flow at infinity. We have computed the amplitude and warp asymmetry in the accretion model, for a disc embedded in a flattened dark matter halo, including self-consistently the contribution of the modes with azimuthal wavenumbers m= 0 and m= 1. Since the m= 0 component, giving a U-shaped profile, is not negligible compared to the m= 1 component, this model predicts quite asymmetric warps, maximum gas displacements on the two sides in the ratio 3 : 2 for the preferred Galactic parameters, and the presence of a fraction ∼3.5 per cent of U-shaped warps, at least. The azimuthal dependence of the moment transfer by the ram pressure would produce a strong asymmetry in the thickness of the HI layer and asymmetric density distributions in z, in conflict with observational data for the warp in our Galaxy and in external galaxies. The amount of accretion that is required to explain the Galactic warp would give gas scaleheights in the far outer disc that are too small. We conclude that accretion of a flow with no net angular momentum cannot be the main and only cause of warps.