Consequences of feedback from early supernovae for disk assembly

In this letter we examine the role of the first supernovae in proto-galaxies, their role in feedback and the consequences for disk assembly. Extending the picture proposed by Dekel & Silk (1986), we argue that energetic supernovae winds can expel baryons from all proto-galaxies with varying degrees of efficiency. The fraction of baryons retained and hence available to assemble into the baryonic disk is therefore, a function of the central velocity dispersion of the halo. Such a coupling of the baryonic component to the dark halo leads to the following interesting consequence, a prediction for a weak scaling of the zero-point of the Tully-Fisher relation or alternatively, the mass-to-light ratio with the central velocity dispersion of the halo. On applying to the case of the Milky Way halo, this feedback mechanism implies: (i) that the Milky Way halo lost approximately 10% of its original gas content; (ii) a range in the inferred redshift of formation $z_f$, and the local baryon fraction $f_b$ for the Milky Way that depends on the initial spin parameter of the halo. We find that for a low spin halo - $z_f<1$, $f_b\sim 2$; for a median spin halo - $z_f\sim 1-2.5$, $f_b\sim 5$; and for a high spin halo

• $z_f\sim 4-8$, $f_b\sim 20$. The observationally determined ages for the oldest disk stars in the Milky Way seem to rule out a low value for the spin parameter. Given the shape of the spin distribution of halos obtained in N-body simulations, while a high value of the spin parameter is not very probable, it is interesting to note that if this is indeed the case for the Milky Way halo, then feedback processes can cause the local baryon fraction to differ significantly from the universal value.