A peculiar galaxy appears at redshift 11: properties of a moderate redshift interloper

TL;DR

This study re-evaluates a galaxy candidate initially thought to be at very high redshift, revealing it is actually a moderate redshift galaxy at z=2.082 through spectroscopic observations, highlighting challenges in photometric redshift determination.

Contribution

The paper provides spectroscopic confirmation that a candidate high-redshift galaxy is actually a lower redshift interloper, emphasizing the importance of spectroscopic follow-up for accurate redshift estimation.

Findings

Spectroscopic data shows the galaxy is at z=2.082, not z=11.

Line emission contributes minimally to the broadband flux, not affecting dropout selection.

Deep optical imaging is required to distinguish between high-z and mid-z solutions photometrically.

Abstract

Laporte et al. (2011) reported a very high redshift galaxy candidate: a lensed J-band dropout (A2667-J1). J1 has a photometric redshift of z=9.6-12, the probability density function for which permits no low or intermediate z solution. We here report new spectroscopic observations of this galaxy with VLT/XShooter, which show clear [OIII]5007AA, Ly-alpha, H-alpha, and H-beta emission and place the galaxy firmly at z=2.082. The oxygen lines contribute only ~25% to the H-band flux, and do not significantly affect the dropout selection of J1. After correcting the broadband fluxes for line emission, we identify two roughly equally plausible natures for A2667-J1: either it is young heavily reddened starburst, or a maximally old system with a very pronounced 4000AA break, upon which a minor secondary burst of star formation is superimposed. Fits show that to make a 3 sigma detection of this object in the B-band (V-band), imaging of depth AB=30.2 (29.5) would be required - despite the relatively bright NIR magnitude, we would need optical data of equivalent depth to the Hubble Ultra Deep Field to rule out the mid-z solution on purely photometric grounds. Assuming that this stellar population can be scaled to the NIR magnitudes of recent HST/WFC3 IR-selected galaxies, we conclude that infeasibly deep optical data AB~32 would be required for the same level of security. There is a population of galaxies at z~2 with continuum colours alone that mimic those of our z=7-12 candidates.