MOND prediction of a new giant shell in the elliptical galaxy NGC3923

Context. Stellar shells, which form axially symmetric systems of arcs in some elliptical galaxies, are most likely remnants of radial minor mergers. They are observed up a radius of $\sim$100 kpc. The stars in them oscillate in radial orbits. The radius of a shell depends on the free-fall time at the position of the shell and on the time since the merger. We previously verified the consistency of shell radii in the elliptical galaxy NGC 3923 with its most probable MOND potential. Our results implied that an as~yet undiscovered shell exists at the outskirts of the galaxy. Aims. We here extend our study by assuming more general models for the gravitational potential to verify the prediction of the new shell and to estimate its position. Methods. We tested the consistency of the shell radial distribution observed in NGC 3923 with a wide variety of MOND potentials of the galaxy. The potentials differed in the mass-to-light ratio and in distance to the galaxy. We considered different MOND interpolation functions, values of the acceleration constant $a_0$, and density profiles of the galaxy. We verified the functionality of our code on a Newtonian self-consistent simulation of the formation of a shell galaxy. Results. Our method reliably predicts that exactly one new outermost shell exists at a galactocentric radius of about 1900$^{\prime\prime}$ ($\sim$210 kpc) on the southwestern side of the galaxy. Its estimated surface brightness is about 28 mag arcsec$^{-2}$ in $B$ - a value accessible by current instruments. This prediction enables a rare test of MOND in an elliptical down to an acceleration of $a_0/10$. The predictive power of our method is verified by reconstructing the position of the largest known shell from the distribution of the remaining shells.