Fornax dwarf spheroidal in MOND: its formation and the survival of its globular clusters

Abstract

The Fornax dwarf spheroidal galaxy has five massive globular clusters (GCs). They are often used for testing different dark matter and modified gravity theories, because it is difficult to reconcile their old stellar ages with the short time they need to settle in the center of the galaxy due to dynamical friction. Using high resolution $N$-body simulations with the Phantom of Ramses code, we investigate whether the GCs of Fornax can be reconciled with the modified Newtonian dynamics (MOND), namely its QUMOND formulation. Observational data interpreted in MOND indicate that Fornax is a tidal dwarf galaxy formed at redshift $z=0.9$ in a flyby of the Milky Way (MW) and Andromeda galaxies, and that its GCs were initially massive star clusters in the disk of the MW. This helps us to set up and interpret the simulations. In the simulations, a point-mass GC orbits Fornax, and they both orbit the MW. When we ran multiple simulations with varying initial conditions for the GC, we found a 20% probability of Fornax being observed with five unsunk GCs. The unsunk GCs have the observed radial distribution. Moreover, we found: 1) In MOND, Fornax has an orbit around the MW such that the pericenters coincide with the observed peaks in the star formation history of Fornax; 2) The simulations reproduce the observed ``diffuse stellar halo'' of Fornax; 3) The simulations predict that Fornax has a stellar stream, which could be detectable in the existing data. 4) An extra simulation shows that if Fornax was initially a rotating disky tidal dwarf galaxy, the gravitational influence of the MW would be able to transform it into a nonrotating spheroidal. 5) Sometimes Phantom of Ramses does not conserve angular momentum. This makes the GC sink too fast if it is simulated as an $N$-body object.

Figures (14)

• Figure 1: Separation of Fornax and the MW as a function of the lookback time. We indicate the time of the MW-M 31 flyby banik18 that we assume ejected the material forming Fornax. We also indicate our assumed moment of the formation of Fornax, which is the starting point of our simulations, just as the times of the two recent observed peaks of star formation in Fornax rusakov21. The orbit of Fornax was not tuned to have apocenters at the observed starbursts peaks.
• Figure 2: Internal and external gravitational accelerations in Fornax. The black curve indicates the profile of internal acceleration of Fornax calculated as if it were isolated. The horizontal band indicates the range of the external acceleration experienced by Fornax as it orbits the MW.
• Figure 3: Spatially resolved star formation history of Fornax. The data come from deboer12. The color indicates the star formation rate in the given radial and age bin. Each radial bin contains an almost equal number of stars. The edges of the radial bins are marked by the short red lines. The vertical dotted line marks the effective radius of the galaxy, the horizontal full line the assumed time of the MW-M 31 flyby banik18, when the material of Fornax left the MW in our scenario.
• Figure 4: Orbital decay of the GCs of Fornax, depending on their mass and initial trajectory. The vertical axis denotes the distance between the GG and the center of Fornax. For the light GCs, the two dotted horizontal lines indicate the range of the projected distances of the four lightest GCs of the real Fornax. For the massive GC, the horizontal dotted line indicates the projected distance of the most massive GC of the real Fornax.
• Figure 5: Comparison of the simulated and observed line-of-sight velocity dispersion profiles of Fornax. The unreliable observational data points are indicated by the fainter color.