Asymmetries in random motions of neutral hydrogen gas in spiral galaxies

Abstract

Context. The velocity dispersion ellipsoid of gas in galactic discs is usually assumed to be isotropic. Under this approximation, no projection effect occurs in the random motions of gas, as traced by the line-of-sight velocity dispersion. However, it has been recently shown that random motions of the neutral hydrogen gas of the Triangulum galaxy (M 33) exhibit a bisymmetric perturbation which is aligned with the minor axis of the galaxy, suggesting a projection effect. Aims. To investigate if perturbations in the velocity dispersion of nearby discs are comparable to those of M 33, the sample is extended to 32 galaxies from The H I Nearby Galaxy Survey (THINGS) and the Westerbork H I Survey of Spiral and Irregular Galaxies (WHISP). Methods. We studied velocity asymmetries in the disc planes by performing Fourier transforms of high-resolution H I velocity dispersion maps corrected for beam-smearing effects, and we measured the amplitudes and phase angles of the Fourier harmonics. Results. In all velocity dispersion maps, we find strong perturbations of first, second, and fourth orders. The strongest asymmetry is the bisymmetry, which is predominantly associated with the presence of spiral arms. The first order asymmetry is generally orientated close to the disc major axis, and the second and fourth order asymmetries are preferentially orientated along intermediate directions between the major and minor axes of the discs. These results are evidence that strong projection effects shape the H I velocity dispersion maps. The most likely source of systematic orientations is the anisotropy of velocities, through the projection of streaming motions that are stronger along one of the planar directions in the discs. Moreover, systematic phase angles of asymmetries in the H I velocity dispersion could arise from tilted velocity ellipsoids, that is when the velocities are correlated. We expect a larger incidence of correlation between the radial and tangential velocities of H I gas with |ρRθ|∼0.6, which could be tested against the kinematics of the youngest stellar populations of the Milky Way. Conclusions. H I velocity dispersions cannot be considered devoid of projection effects. The systematic orientations of asymmetries can be explained by the projection of unresolved streaming motions mainly arising from spiral arms. Our methodology is a powerful tool to constrain the dominant direction of streaming motions and thus the shape of the velocity ellipsoid of H I gas, which is de facto anisotropic at the angular scales probed by the observations. The next step is to study the shape of the velocity ellipsoids of molecular and ionised gas and their link with galaxy mass and/or morphology, in addition to extending the sample size.