High-Redshift Galaxy Candidates at z > 6 as Revealed by JWST Observations of MACS0647

TL;DR

This study leverages JWST NIRCam imaging (7 filters) and NIRSpec spectroscopy, together with archival HST data, to assemble a catalog of $57$ high-$z$ galaxy candidates behind MACS0647, including $14$ spectroscopic confirmations up to $z=9.25$. The analysis combines photometric redshifts and SED fitting (via EAZY and BAGPIPES) with detailed emission-line spectroscopy (PRISM and G395H) to derive stellar masses ($10^8$–$10^9\,M_\odot$), metallicities ($\sim10$–$40\%\,Z_\odot$), and star-formation properties, while accounting for lensing magnification ($\mu$ up to about $11$). A prominent overdensity at $z\approx6.1$ is confirmed with nine members, suggesting a protocluster or structured region along the line of sight; the dataset also reveals Little Red Dots with AGN-like features and a highly magnified, extreme source EBG-1 at $z=9.25$ with a possible damped Ly$\alpha$ turnover. The work demonstrates JWST’s capability to map early galaxy assembly, constrain chemical enrichment, and probe the interplay between lensing and intrinsic properties in the epoch of reionization. Continued spectroscopic follow-up will refine redshift confirmations and illuminate the diversity of galaxies in the first billion years.

Abstract

We present a catalog of 57 high-redshift $z>6$ galaxy candidates, including 14 spectroscopic confirmations ($z = 6.10$ -- 9.25), 2 Little Red Dots ($z = 4.77$, 5.81), and 2 interlopers ($z = 3.23$, 3.72), based on \JWST\ NIRCam imaging (7 filters), NIRSpec spectroscopy (PRISM and G395H), and archival \HST\ imaging (17 filters) of the strong lensing galaxy cluster MACS0647. Our highest redshift confirmation ($z = 9.25$) is an Extremely Blue Galaxy (presented in~\citealt{Yanagisawa2024}), and here we identify a spectral turnover likely due to damped Lyman-$α$. We identify an overdensity of galaxies with spectroscopic redshifts $z = 6.1$, confirming the $z \sim 6$ overdensity identified in \HST\ images. In one of these galaxies, our high-resolution G395H spectroscopy reveals two spatially resolved components with a velocity difference of $\sim$90 km/s; if these components are gravitationally bound, this would imply a dynamical mass on the order of $\sim 10^8\ M_\odot$ given their projected separation. We present spectral line fluxes, widths, and derived physical properties, including stellar masses ($10^8 - 10^9 \ \mathrm{M}_{\odot}$) and metallicities ($10\% - 40\% \ \mathrm{Z}_{\odot}$) for our spectroscopic sample. We note half of our NIRSpec data was obtained with standard 3-slitlet nods and half was obtained with single slitlets yielding similar results, demonstrating the power to observe more sources on a densely packed NIRSpec MSA.

Figures (8)

• Figure 1: Left: Overview MACS0647 NIRCam color image based on DJA v7 reductions of observations in six wide filters. High-redshift candidates are labeled in white (v7 ID numbers), while galaxies with spectroscopic redshifts are indicated in cyan. Right:JWST NIRCam $4\arcsec \times 4\arcsec$ color images, each centered on a high-redshift candidate in the final sample.
• Figure 2: Example of a high-redshift galaxy (v4 3754 / v7 5191, $z = 7.464$) observed with NIRSpec PRISM in both Cycle 1 slitlet configurations. This target was observed in Obs 21 using single-slitlet placements and in Obs 23 using the standard 3-slitlet nod pattern (left). The PRISM spectrum (middle) shows multiple strong emission lines, and the corresponding SED fit is shown on the right.
• Figure 3: Four example members of the overdensity at $6.12 < z < 6.15$ identified with NIRSpec PRISM. For each galaxy we show Left: the NIRCam color image with slit orientation, Middle: the 2D and 1D PRISM spectra, and Right: the NIRCam photometry together with the SED model constrained by the spectrum. These sources, together with five additional galaxies in the same narrow redshift slice, indicate a prominent overdensity at $z \sim 6.1$ behind MACS0647.
• Figure 4: . Galaxy v4 3568 (v7 4922) at $z=9.25$ with NIRSpec PRISM, reported previously by Yanagisawa2024 and dubbed EBG-1 (Extremely Blue Galaxy). The source is modestly magnified by the foreground cluster, with $\mu = 2.72^{+0.09}_{-0.09}$. Left: NIRCam color images with slits overlaid. Middle: 2D and 1D spectra from DJA v3. Right: NIRCam photometry and SED fit with NIRSpec spectrum scaled to match.
• Figure 5: Candidate $z > 11$ galaxies (based on our NIRCam v4 photometry) shown to be interlopers at $z < 4$ with NIRSpec PRISM spectroscopy. In both cases, Balmer breaks are mistaken for Lyman breaks. Top: In the case of v4 3533 (v7 4870), strong emission lines [O iii] and H$\alpha$ conspire with the $z = 3.23$ Balmer break to mimic a blue SED ($\beta < -2$) with a Lyman break at $z \sim 11$. Bottom: Emission lines do not significantly affect the photometry of v4 1395 (v7 1658). A faint detection in F150W blueward of the break was missed in the v4 photometric catalog, then detected in the v7 catalog.