“An accurate measurement of the baryonic Tully-Fisher relation with heavily gas-dominated ALFALFA galaxies” by E. Papastergis et. al.
Calibrated the baryonic Tully-Fisher relation on a sample of 97 galaxies observed by the ALFALFA 21cm survey. The BTFR relates rotational velocities to the baryonic mass in a galaxy.
To reduce measurement errors, only gas-dominated galaxies, oriented edge-on, were selected. There is a large systematic uncertainty in measuring galactic stellar mass, so by choosing gas-dominated galaxies, they reduce that contribution to the error in the mass estimate. See Equation 15 on page 7 for the quantification of this. By selecting edge-on galaxies, the rotational velocity can be measured with the width of the 21cm line, after it has been shifted to the rest-frame. See Figure 2 on page 8 for the raw relation.
This has been partitioned into subsets, those with high- and low-kurtosis line profiles. See Figure 5 on page 10 for types of line profiles. High-kurtosis profiles are “peaked”, which are heavily influenced by the rising portion of the rotation curve. Low-kurtosis profiles, on the other hand, are “double-horned”, which is indicative of the flat outer rotation curves. See Figure 6 on page 11 for the observed relation, with the data partitioned based on kurtosis. In the double-horned case, the slope is consistent with MOND predictions.
They compare predictions of ΛCDM and MOND on page 14 in Figures 9 and 10. Both appear to be in relative agreement. Note that there may be some added error in the MOND predictions, as the MOND predictions are based on the velocities in the flat region of the rotation curves, while the data are based on the 21cm line, which relies on the entire galaxy.