Speaker
Description
We present a kinematic and mass-modelling analysis of the nearby late-type galaxy NGC 45 using high-quality HI observations from the IMAGINE survey, complemented by a stellar mass profile derived from Spitzer IRAC 1 (3.6 μm) imaging. From the well-resolved HI cube, we construct a rotation curve and decompose it into stellar, gaseous, and dark matter (DM) components. Mass models are fitted using two widely adopted halo parametrisations, the Navarro–Frenk–White (NFW) profile and the pseudo-isothermal (ISO) profile, exploring their parameter spaces with a Bayesian MCMC framework. Both halo models reproduce the observed rotation curve within uncertainties, and the present data do not allow a clear distinction between a cuspy (NFW) or cored (ISO) DM distribution.
The main aim of this work is to quantify how observational limitations impact the recovered galaxy dynamics and mass. By systematically degrading the spatial resolution and sensitivity of the HI cube, we isolate the effects of beam smearing and low S/N on the kinematic and mass-modelling outputs. With only approximately 3 beams across the disc, spatial resolution loss drives, systematic biases: rotation velocities are underestimated by ~15%, the inner rise is flattened by ~34%, velocity dispersions are artificially inflated by ~40%, and the gas surface density is reduced by ~21%, producing correspondingly biased DM halo parameters. Reducing the sensitivity from S/N ≈ 34 to ≈ 5 produces a different failure mode: the inner rise is flattened by ~53%, the outer rotation suppressed by ~20%, and the velocity dispersion severely underestimated (~82%), leading to 20–40% reductions in halo mass and artificially compact DM profiles. When both degradations are applied simultaneously, the biases compound, yielding large distortions in the kinematics and mass models and virial masses suppressed by up to ~75%.
These tests demonstrate that observational quality, critically governs the reliability of rotation curve decompositions.
| Stream | Science or Engineering |
|---|