One always has to be careful of chi2 values in rotation curve fits, as most formal procedures give overly optimistic (small) error bars. This seems to be well handled here, but there does not appear to be any consideration of distance uncertainties, with H0 = 75 km/s/Mpc assumed. Slight deviations from the assumed distance can matter, and ESO 116-G12 shows the classic signature of a fit which would be greatly improved if the distance were increased a few percent. At any rate, there should be a term in the error budget for the uncertainty in the acceleration V2/R stemming from the uncertainty in the distance. Without it, chi2 is overestimated.
NGC 1090 is potentially a more serious problem. In this case, the problem is caused mostly by the points in the region where the rotation curve turns over. These appear to be the only points in the sample where the Halpha and HI data disagree (Fig. 12); the velocity in this region might arguably be lower and less certain. The turn-over region is not well sampled, a line connecting the rising points to the high turnover points would appear to need to inflect upwards a bit, something no plausible model of any flavor is going to do. Given this, I'd have a hard time objecting to any of the models presented for this galaxy. Even the steep cusp of a Moore halo is not particularly objectionable.
Return to Literature page