Recent studies have shown that dark matter with a superfluid phase in which phonons mediate a long-distance force gives rise to the phenomenologically well-established regularities of Modified Newtonian Dynamics (MOND). Superfluid dark matter, therefore, has emerged as a promising explanation for astrophysical observations by combining the benefits of both particle dark matter and MOND, or its relativistic completions, respectively. We here investigate whether superfluid dark matter can reproduce the observed Milky Way rotation curve for $ R < 25\,\rm{kpc}$ and are able to answer this question in the affirmative. Our analysis demonstrates that superfluid dark matter fits the data well with parameters in reasonable ranges. The most notable difference between superfluid dark matter and MOND is that superfluid dark matter requires about $ 20\% $ less total baryonic mass (with a suitable interpolation function). Our analysis further allows us to estimate the radius of the Milky Way's superfluid core and the total mass of dark matter in both the superfluid and the normal phase.

In superfluid dark matter the exchange of phonons can create an additional force that has an effect similar to Modified Newtonian Dynamics (MOND). To test whether this hypothesis is compatible with observation, we study a set of strong gravitational lenses from the SLACS survey and check whether the measurements can be explained by a superfluid in the central region of galaxies. Concretely, we try to simultaneously fit each lens's Einstein radius and velocity dispersion with a spherically symmetric density profile of a fluid that has both a normal and a superfluid component. We demonstrate that we can successfully fit all galaxies except one, and that the fits have reasonable stellar mass-to-light-ratios. We conclude that strong gravitational lensing does not pose a challenge for the idea that superfluid dark matter mimics modified gravity.

Modified Newtonian Dynamics has one free parameter and requires an interpolation function to recover the normal Newtonian limit. We here show that this interpolation function is unnecessary in a recently proposed covariant completion of Erik Verlinde's emergent gravity, and that Verlinde's approach moreover fixes the function's one free parameter. The so-derived correlation between the observed acceleration (inferred from rotation curves) and the gravitational acceleration due to merely the baryonic matter fits well with data. We then argue that the redshift-dependence of galactic rotation curves could offer a way to tell apart different versions of modified gravity from particle dark matter.