The impact of the supersonic baryon-dark matter velocity difference on the z~20 21cm background

TL;DR

This paper investigates how the supersonic relative velocity between baryons and dark matter during the cosmic Dark Ages affects early structure formation, heating, and the 21cm background, revealing modest impacts but notable acoustic features.

Contribution

It provides the first self-consistent simulations of gas pressure and baryon-dark matter velocity, quantifying their effects on early universe structure and 21cm signals.

Findings

Supersonic velocities suppress small-scale structure formation.

Velocity differences source shocks and impact gas accretion onto minihalos.

The 21cm fluctuations from velocity effects are limited to about 10% of total fluctuations.

Abstract

Recently, Tseliakhovich and Hirata (2010) showed that during the cosmic Dark Ages the baryons were typically moving supersonically with respect to the dark matter with a spatially variable Mach number. Such supersonic motion may source shocks that heat the Universe. This motion may also suppress star formation in the first halos. Even a small amount of coupling of the 21cm signal to this motion has the potential to vastly enhance the 21cm brightness temperature fluctuations at 15<z<40 as well as to imprint acoustic oscillations in this signal. We present estimates for the size of this coupling, which we calibrate with a suite of cosmological simulations. Our simulations, discussed in detail in a companion paper, are initialized to self-consistently account for gas pressure and the dark matter-baryon relative velocity, v_bc (in contrast to prior simulations). We find that the supersonic velocity difference dramatically suppresses structure formation at 10-100 comoving kpc scales, it sources shocks throughout the Universe, and it impacts the accretion of gas onto the first star-forming minihalos (even for halo masses as large as ~10^7 Msun). However, we find that the v_bc-sourced temperature fluctuations can contribute only as much as ~10% of the fluctuations in the 21cm signal. We do find that v_bc could source an O(1) component in the power spectrum of the 21cm signal via the X-ray (but not ultraviolet) backgrounds produced once the first stars formed. In a scenario in which ~10^6 Msun minihalos reheated the Universe via their X-ray backgrounds, we find that the pre-reionization 21cm signal would be larger than previously anticipated and exhibit significant acoustic features. We show that structure formation shocks are unable to heat the Universe sufficiently to erase a strong 21cm absorption trough at z ~ 20 that is found in most models of the sky-averaged 21cm intensity.