Early results from GLASS-JWST. III: Galaxy candidates at z$\sim$9-15

TL;DR

GLASS-JWST presents the first search for galaxy candidates at $z>9$ using deep seven-band NIRCam imaging in a Frontier Fields flanking field. The study combines color-color Lyman-break selections with cross-validated photometric redshifts from two codes (EAzY and zphot) and employs simulations to quantify completeness and contamination, finding roughly 70% completeness at the bright end. It identifies six $z>9$ candidates, including two bright systems at $z\simeq 10.6$ and $z\simeq 12.2$ with $M_{UV}\lesssim -21$, broadly consistent with a non-evolving UV luminosity function within uncertainties, though the bright $z>10$ detections are surprising given the small survey volume. The results demonstrate JWST's capability to probe the high-redshift frontier and provide compelling targets for Cycle-2 spectroscopic follow-up.

Abstract

We present the results of a first search for galaxy candidates at z$\sim$9--15 on deep seven-bands NIRCam imaging acquired as part of the GLASS-JWST Early Release Science Program on a flanking field of the Frontier Fields cluster A2744. Candidates are selected via two different renditions of the Lyman-break technique, isolating objects at z$\sim$9-11, and z$\sim$9-15, respectively, supplemented by photometric redshifts obtained with two independent codes. We find five color-selected candidates at z$>$9, plus one additional candidate with photometric redshift z$_{phot}\geq$9. In particular, we identify two bright candidates at $M_{UV}\simeq -21$ that are unambiguously placed at $z\simeq 10.6$ and $z\simeq 12.2$, respectively. The total number of galaxies discovered at $z>9$ is in line with the predictions of a non-evolving LF. The two bright ones at $z>10$ are unexpected given the survey volume, although cosmic variance and small number statistics limits general conclusions. This first search demonstrates the unique power of JWST to discover galaxies at the high redshift frontier. The candidates are ideal targets for spectroscopic follow-up in Cycle$-2$.

Figures (4)

• Figure 1: Color selection diagrams (left panels), and redshift distribution of selected galaxies (right panels), for LBGs at z$\sim$9-11 (top, Selection #1) and z$\sim$9-15 (bottom, Selection #2) exploiting the NIRCam filters of the GLASS-JWST-ERS parallel. The catalog comprises objects at $z=0-15$ over an area of 0.12 sq. deg. Black stars show the position of brown dwarf models from Marley2021 normalized at 26.0$\leq$F444W$\leq$28.0 that meet our detection criteria for high-redshift galaxies. All fluxes have been perturbed with realistic noise properties.
• Figure 2: Observed color selection diagrams for LBGs at z$\sim$9-11 (left, Selection #1) and z$\sim$9-15 (right, Selection #2) in GLASS-JWST. Grey points show all objects with SNR(F444W)$>$8 in the GLASS catalog. Green circles indicate the color-selected candidates. The additional candidate selected on the basis of photometric redshift is shown as a purple empty circle. The z$\sim$9-11 LBGs from the Selection #1 diagram are shown as dark orange crosses in the Selection #2 one. Upper limits are indicated by arrows. All error-bars and upper limits are at 1$\sigma$.
• Figure 3: The two high-quality bright high-redshift candidates from the GLASS-JWST NIRCAM field taken in parallel to NIRISS. Photometry and best-fit SEDs at the best-fit redshift are given in the main quadrant. Redshift probability distributions P(z) from zphot (grey) and EAzY (red) are shown in the inset. Upper limits are reported at the $2\sigma$ level, including a conservative estimate of the error budget, especially in the bluest bands (M22). Thumbnails, from left to right, show the objects in the F090W, F115W, F150W, F200W, F277W, F356W and F444W bands.
• Figure 4: As in Figure \ref{['highz1']}, for fainter high-redshift candidates from GLASS-JWST. These candidates need further verification owing to the non-negligible probability of a low-redshift solution.