JWST GLIMPSE: Ultra-deep Study of High-Redshift Galaxies

• JWST GLIMPSE is an ultra-deep, gravitationally-lensed survey designed to detect ultrafaint high-redshift galaxies driving cosmic reionization.
• The program utilizes over 120 hours of JWST/NIRCam imaging across seven broadband and two medium-wide filters, leveraging Abell S1063 to reach intrinsic depths beyond AB ≈ 33 mag.
• Its analysis encompasses UV luminosity function measurements, resolved star-forming region studies, and constraints on early galaxy and AGN formation.

The JWST GLIMPSE (Gravitational Lensing Infrared Multi‐filter Probe of the Epoch of Reionization and Survival of the Earliest galaxies) program constitutes the deepest systematic investigation into faint, high-redshift galaxy populations and the mechanisms underpinning cosmic reionization. Employing ultra-deep JWST/NIRCam imaging across seven broad and two medium-wide filters and leveraging the gravitational magnification afforded by the galaxy cluster Abell S1063, GLIMPSE probes intrinsic depths beyond AB ≈ 33 mag in a compact field, reaching effective source-plane areas of ≈4.4 arcmin² at $z \sim 6$. An initial catalogue includes ∼540 galaxy candidates at $6 < z < 16$, extending UV luminosity function measurements to $M_{\mathrm{UV}}$ = $-$12, thereby enabling stringent constraints on the abundance and demographics of ultrafaint galaxies that likely dominated ionizing photon production in the reionization epoch.

1. Survey Architecture and Observing Strategy

GLIMPSE's field centers on the Hubble Frontier Fields cluster Abell S1063 ($z = 0.348$), featuring dual NIRCam footprints: Module B focusing on the cluster core to maximize lensing magnification, and Module A targeting a modestly lensed blank region. The program accumulates over 120 hours of JWST time distributed among seven broadband (F090W, F115W, F150W, F200W, F277W, F356W, F444W) and two medium-band (F410M, F480M) filters. Depths, determined in a 0.2″ diameter aperture, range from AB = 30.68 to 30.87 for broad bands and AB = 29.24–30.07 for the medium bands.

The observing strategy incorporates the MEDIUM8 readout, six large-scale INTRAMODULEBOX dithers, a four-point subpixel dither pattern, and careful avoidance of detector artifacts, optimizing both depth and spatial sampling. The corresponding table summarizes each filter's exposure time and measured depth:

Filter	Exposure (h)	5σ Depth (AB)
F090W	39.1	30.85
F115W	39.1	30.87
F150W	22.3	30.77
F200W	19.5	30.82
F277W	22.3	30.82
F356W	19.5	30.77
F410M	16.7	30.07
F444W	39.1	30.68
F480M	22.3	29.24

2. Gravitational Lensing Model and Source-Plane Mapping

Abell S1063 is characterized by a relaxed mass distribution comprising two dominant dark-matter halos and $\sim$300 galaxy-scale halos. Extensive VLT/MUSE spectroscopy provides over 70 multiply imaged sources with reliable redshifts, producing highly-constrained lensing models within the Frontier Fields sample.

Lensing properties are computed under the thin-lens approximation, with local magnification given by

$$\mu(\theta) = \frac{1}{\left[1 - \kappa(\theta)\right]^2 - \gamma(\theta)^2}$$

where $\kappa$ is the dimensionless surface mass density and $\gamma$ the shear. Surface brightness is conserved; thus, a small image-plane area $d\Omega_{\text{image}}$ is related to source-plane area $$d\Omega_{\text{source}}(\theta) = d\Omega_{\text{image}} / \mu(\theta)$$. Integration over the NIRCam mosaic using $z\simeq6$ $\kappa$ and $\gamma$ maps yields an effective source-plane coverage $A_{\text{source}}(z\approx6) \simeq 4.4$ arcmin².

Magnification factors reach $\mu \sim 5$–$10$ near critical curves, exceeding $\mu > 100$ in localized regions, corresponding to intrinsic point-source depths beyond AB$\gtrsim$33.

3. Scientific Objectives

The GLIMPSE program pursues several principal goals:

• UV Luminosity Function Measurement ($z \sim 6$–16): By extending sensitivity to $M_{\mathrm{UV}} \gtrsim -12$, GLIMPSE enables quantification of the faint end—regimes presumed to dominate ionizing-photon production responsible for cosmic reionization.
• Galaxy Formation and Reionization Model Constraints: Four magnitudes fainter than prior blank-field surveys, GLIMPSE supplies the necessary datapoints to test and refine galaxy evolution models.
• Resolved Studies of Star-Forming Regions: Physical spatial resolutions $\lesssim 10$ pc (after de-lensing) permit analyses of star-forming clumps and proto–globular cluster candidates within early galaxies.
• Identification and Characterization of High-$z$ AGN and Magnified Stars: Faint AGNs, as well as individual luminous stars magnified by $\mu\gtrsim10^3$, are detectable within the GLIMPSE field.

4. Analysis Methodologies and First Results

Candidate galaxies are selected via Lyman-break methodology and photometric redshift estimation, employing spectral energy distribution fitting (SED-fitting) with the EAZY code. The initial sample comprises ∼540 galaxy candidates with $6$M_{\mathrm{UV}}$ from $-20$ up to $-12$.

Volume densities are inferred using completeness and magnification corrections, enabling measurement of the UV luminosity function $\phi(M)$ for $6

$$\phi(M) = 0.4\,\ln 10\,\phi^*\,10^{-0.4(M - M^*)(\alpha + 1)}\,\exp\left[-10^{-0.4(M - M^*)}\right]$$

Preliminary fits yield faint-end slopes $\alpha\approx-2.1$ to $-2.3$ at $z\sim6$–8, with no strong evidence for turnover down to $M_{\mathrm{UV}} \approx -13$. This suggests that ultrafaint galaxies contribute substantially to the budget of reionizing photons.

5. Ancillary Data Integration and Public Release

GLIMPSE leverages extensive supplementary datasets:

• HST Frontier Fields: ACS (F435W, F606W, F814W) and WFC3/IR (F105W, F125W, F140W, F160W) with $\sim$29 AB depth.
• BUFFALO Imaging: ACS and WFC3/IR at $\sim$27.5 AB.
• FLASHLIGHTS: Long-pass optical images F200LP, F350LP.
• VLT/MUSE: Integral-field spectroscopy ($>$1000 h), securing $\sim$200 redshifts of lensed sources.
• JWST/NIRSpec: Medium-resolution prism and $R \sim 1000$ spectroscopy (DD 9223).

The first public data release (DR1) includes:

• PSF-mosaicked NIRCam full-depth images at 20 mas pixel$^{-1}$ (short wavelength) and 40 mas pixel$^{-1}$ (long wavelength).
• Co-aligned HST ACS+WFC3 mosaics drizzled to 40 mas pixel$^{-1}$.
• Photometric catalogs with aperture fluxes ($D=0.1$–1.2″), empirical uncertainties, and quality flags.
• Best-available Zitrin-analytic lensing models with $\kappa$, $\gamma$, and $\mu$ maps for $z=2$–10.

These assets support comprehensive analyses of faint galaxy populations, internal structures, AGN searches, and stellar population studies enabled by lensing.

6. Implications for Early Universe Studies

GLIMPSE achieves apparent depths of AB$\approx$30.9 and intrinsic depths beyond AB$\approx$33, exceeding prior surveys by probing galaxy demography at magnitudes and epochs inaccessible without gravitational lensing. This enables the most stringent constraints yet on galaxy populations regulating cosmic reionization, and provides a laboratory for spatially resolved investigations of star-forming regions, potential first black hole seeds, and individual stars at cosmological distances.

A plausible implication is that lack of a turnover in the UV luminosity function down to $M_{\mathrm{UV}} \approx -13$ suggests ultrafaint galaxies are major contributors to reionizing photon production. The multiwavelength approach and public release dramatically enhance community access and research flexibility, underpinning future work on reionization, population III star search efforts, and early galaxy assembly.