The redshift distribution of submillimetre galaxies at different wavelengths

TL;DR

This study uses simulations to show that the varied redshift distributions of submillimetre galaxies at different wavelengths can be explained by a single parent distribution influenced by observational factors, challenging previous claims of intrinsic differences.

Contribution

The paper demonstrates through simulations that the different observed redshift distributions of SMGs at various wavelengths are consistent with a single underlying population, considering observational biases.

Findings

Redshift distributions at different wavelengths can be derived from one parent distribution.

Wavelength selection, survey depth, and resolution explain observed differences.

Additional populations or biases are needed to explain the 450 um distribution.

Abstract

Using simulations we demonstrate that some of the published redshift distributions of submillimetre galaxies (SMGs) at different wavelengths, which were previously reported to be statistically different, are consistent with a parent distribution of the same population of galaxies. The redshift distributions which peak at z_med=2.9, 2.6, 2.2, 2.2, and 2.0 for galaxies selected at 2 and 1.1 mm, and 870, 850, and 450 um respectively, can be derived from a single parent redshift distribution, in contrast with previous studies. The differences can be explained through wavelength selection, depth of the surveys, and to a lesser degree, angular resolution. The main differences are attributed to the temperature of the spectral energy distributions, as shorter-wavelength maps select a hotter population of galaxies. Using the same parent distribution and taking into account lensing bias we can also reproduce the redshift distribution of 1.4 mm-selected ultra-bright galaxies, which peaks at z_med=3.4. However, the redshift distribution of 450 um-selected galaxies in the deepest surveys, which peaks at z_med=1.4, cannot be reproduced from the same parent population with just these selection effects. In order to explain this distribution we have to add another population of galaxies or include different selection biases.