Cosmological peculiar velocities in general relativity?

Abstract

Cosmological peculiar velocities have traditionally been studied within the framework of Newtonian theory. Around the turn of the century, a few quasi-Newtonian analyses appeared in the literature, but led to equations and results identical to those of the purely Newtonian approach [1]. More recently, a series of studies introduced a relativistic treatment of the peculiar-velocity problem, criticising the quasi-Newtonian approach as effectively Newtonian in nature [2]. These works also reported a linear growth-rate of $v \propto t$ for peculiar velocities, in contrast to the slower Newtonian/quasi-Newtonian scaling of $v \propto t^{1/3}$. In a manuscript uploaded to the archives a few days ago [3], the authors defended their earlier quasi-Newtonian work and criticised the more recent relativistic treatments. However, the limitations of the quasi-Newtonian approach are not a new concern, but they have been noted at least since [4]. There, it was clearly stated that the quasi-Newtonian approximation leads to Newtonian-like equations and results, and readers were cautioned against applying it to large-scale cosmological studies. Given that one of the authors of [3] was also a coauthor of [4], the self-contradiction is evident. The relativistic analyses have, in fact, confirmed the concerns of [4], clarified the underlying issues and shown how they can be resolved. Motivated by [3], we present a critical comparison of the two approaches and in the process identify several internal inconsistencies in that manuscript.