Insights from probability distribution functions of intensity maps

TL;DR

This paper introduces the voxel intensity distribution (VID) as a new statistical method for analyzing intensity-mapping surveys, capturing non-Gaussian information beyond the power spectrum to improve astrophysical constraints.

Contribution

The paper proposes the VID as a novel statistic for intensity maps, demonstrating its effectiveness in extracting additional information and constraining the CO luminosity function at high redshift.

Findings

VID provides substantial constraints beyond power spectrum analysis.

Future surveys could constrain CO luminosity function to about 10%.

VID remains effective even with contamination and interloper effects.

Abstract

In the next few years, intensity-mapping surveys that target lines such as CO, Ly$\alpha$, and CII stand to provide powerful probes of high-redshift astrophysics. However, these line emissions are highly non-Gaussian, and so the typical power-spectrum methods used to study these maps will leave out a significant amount of information. We propose a new statistic, the probability distribution of voxel intensities, which can access this extra information. Using a model of a CO intensity map at $z\sim3$ as an example, we demonstrate that this voxel intensity distribution (VID) provides substantial constraining power beyond what is obtainable from the power spectrum alone. We find that a future survey similar to the planned COMAP Full experiment could constrain the CO luminosity function to order $\sim10\%$. We also explore the effects of contamination from continuum emission, interloper lines, and gravitational lensing on our constraints and find that the VID statistic retains significant constraining power even in pessimistic scenarios.