SIMBAD references

We study how well the complex gas velocity fields induced by massive spiral arms are modelled by the hydrodynamical simulations that we used recently to constrain the dark matter fraction in nearby spiral galaxies. More specifically, we explore the dependence of the positions and amplitudes of features in the gas flow on the temperature of the interstellar medium (assumed to behave as a one-component isothermal fluid), the non-axisymmetric disc contribution to the galactic potential, the pattern speed Ω_p, and finally the numerical resolution of the simulation. We argue that, after constraining the pattern speed reasonably well by matching the simulations to the observed spiral arm morphology, the amplitude of the non-axisymmetric perturbation (the disc fraction) is left as the primary parameter determining the gas dynamics. However, owing to the sensitivity of the positions of the shocks to modelling parameters, one has to be cautious when quantitatively comparing the simulations to observations. In particular, we show that a global least-squares analysis is not the optimal method for distinguishing different models, as it tends to slightly favour low disc fraction models. Nevertheless, we conclude that, given observational data of reasonably high spatial resolution and an accurate shock-resolving hydro-code, this method tightly constrains the dark matter content within spiral galaxies. We further argue that, even if the perturbations induced by spiral arms are weaker than those of strong bars, they are better suited for this kind of analysis because the spiral arms extend to larger radii where effects like inflows due to numerical viscosity and morphological dependence on gas sound speed are less of a concern than they are in the centres of discs.