Narrow band selected high redshift galaxy candidates contaminated by lower redshift O[III] ultrastrong emitter line galaxies

TL;DR

This study investigates how ultra-strong O[III] emission lines at redshift ~1.1 can contaminate high-redshift galaxy surveys, highlighting the need for better characterization of O[III] luminosity to improve sample purity.

Contribution

It demonstrates that O[III] ultrastrong emitters can mimic high-redshift galaxy signals, emphasizing the importance of accounting for this contamination in surveys.

Findings

O[III] lines identified in three low-redshift interlopers

Ultrastrong line emitters at z~1.1 can contaminate high-z galaxy samples

Suggests characterizing O[III] luminosity function at z~3+

Abstract

Context. Lyman Break Galaxies (LBG) and Narrow Band (NB) surveys have been successful at detecting large samples of high-redshift galaxies. Both methods are subject to contamination from low-redshift interlopers. Aims. In this paper, our aim is to investigate the nature of low-redshift interlopers in NB Lyman-$\alpha$ emitters (LAE) searches. Methods. From previous HAWK-I NB imaging at z $\sim$7.7 we identify three objects that would have been selected as high-redshift LAEs had our optical data been one magnitude shallower (but still one to two magnitudes fainter than the near infrared data). We follow-up these objects in spectroscopy with XSHOOTER at the VLT. Results. Despite low quality data due to bad weather conditions, for each of the three objects we identify one, and only one emission line, in the spectra of the objects, that we identify as the O[III]5007A line. This result combined to spectral energy density fitting and tests based on line ratios of several populations of galaxies we infer that the 3 objects are ultrastrong line emitters at redshifts $\sim$1.1. Conclusions. From this work and the literature we remark that the O[III] line appears to be a common source of contamination in high-redshift LBG and LAE samples and we suggest that efforts be put to characterize with high accuracy the O[III] luminosity function out to redshift $\sim$3 or higher.