The formation of galactic haloes in the neutrino-adiabatic theory

Abstract

Recent evidence of a finite rest mass for the neutrino has led to a revival of the idea that these particles may provide the unseen but dynamically indicated matter on cosmological scales of galactic halo and larger size. In the early Universe, the free streaming of collisionless neutrinos is seen to damp out perturbations less than of the order of tens of Mpc long^1–5. A numerical simulation of the clustering of collisionless particles in a universe with this type of perturbation spectrum⁶ has led to the formulation of a cell structure with large voids. Many observations^7,8 support the existence of such a structure on scales of tens of Mpc. However, collapse of perturbations of this size leads to velocities of the order of 1,000 km s⁻¹, and as the neutrinos presumably cannot cool, this seems to be too large for their inclusion in galactic haloes, which have much smaller velocity dispersions. Thus it has been suggested that neutrinos cannot cluster in galactic haloes without some means to ‘save’ smaller-scale perturbations^3,4,9. Furthermore, if neutrinos of mass ∼30 eV form galactic haloes, the phase space density must be very close to the primordial one¹⁰—and phase mixing is often associated with collapse. Although the average value of the velocity dispersion may be reduced if the perturbations are anisotropic¹¹, a disparity still remains. I now present a numerical simulation which shows that if the initial perturbations are sufficiently anisotropic, the collapse of very large perturbations of collisionless particles leads to a condensation of particles with low velocity dispersion and high phase-space density, which may easily fragment.