A Long Time Ago in a Galaxy Far, Far Away: A Candidate z ∼ 12 Galaxy in Early JWST CEERS Imaging

We report the discovery of a candidate galaxy with a photo-z of z ∼ 12 in the first epoch of the James Webb Space Telescope (JWST) Cosmic Evolution Early Release Science Survey. Following conservative selection criteria, we identify a source with a robust z phot = 11.8−0.2+0.3 (1σ uncertainty) with m F200W = 27.3 and ≳7σ detections in five filters. The source is not detected at λ < 1.4 μm in deep imaging from both Hubble Space Telescope (HST) and JWST and has faint ∼3σ detections in JWST F150W and HST F160W, which signal a Lyα break near the red edge of both filters, implying z ∼ 12. This object (Maisie’s Galaxy) exhibits F115W − F200W > 1.9 mag (2σ lower limit) with a blue continuum slope, resulting in 99.6% of the photo-z probability distribution function favoring z > 11. All data-quality images show no artifacts at the candidate’s position, and independent analyses consistently find a strong preference for z > 11. Its colors are inconsistent with Galactic stars, and it is resolved (r h = 340 ± 14 pc). Maisie’s Galaxy has log M */M ⊙ ∼ 8.5 and is highly star-forming (log sSFR ∼ −8.2 yr−1), with a blue rest-UV color (β ∼ −2.5) indicating little dust, though not extremely low metallicity. While the presence of this source is in tension with most predictions, it agrees with empirical extrapolations assuming UV luminosity functions that smoothly decline with increasing redshift. Should follow-up spectroscopy validate this redshift, our universe was already aglow with galaxies less than 400 Myr after the Big Bang.


INTRODUCTION
stevenf@astro.as.utexas.edu The study of galaxy evolution is the ultimate human origin story -not just how did our species, planet or Solar System come to be, but this field seeks to answer how our Milky Way Galaxy came to be.One method to study our Galactic origins is to study the earliest build-ing blocks of the Milky Way by searching for and analyzing galaxies forming in the early Universe.The advent of the Wide Field Camera 3 (WFC3) on the Hubble Space Telescope (HST) pushed our cosmic horizons well into the epoch of reionization, the time when energetic photons (presumably from massive stars in early galaxies) ionized the gas in the intergalactic medium (IGM; e.g.Finkelstein 2016;Stark 2016;Robertson 2021, and references therein).These studies found that the z = 6-10 universe is teeming with galaxies, with thousands of galaxy candidates known, including spectroscopic confirmations out to z ∼ 11 (Oesch et al. 2016;Jiang et al. 2021).
One key focus in these studies has been the evolution of the cosmic star-formation rate density (SFRD).This quantity is well known to rise from the present day to the peak of cosmic star-formation at z ∼ 2-3, then decline again to early times (e.g.Madau & Dickinson 2014).As the aforementioned WFC3 studies pushed to higher redshifts, it became of interest to study whether the cosmic SFRD, which exhibited a smooth decline from z = 4-8 (e.g.Bouwens et al. 2015;Finkelstein et al. 2015), continued to decline smoothly to even higher redshifts.Results in the literature were mixed, with some studies finding evidence for an accelerated decline in the SFRD (e.g.Oesch et al. 2018;Bouwens et al. 2021), while others found that observations supported a continued smooth decline (e.g.Coe et al. 2013;McLeod et al. 2016;Finkelstein & Bagley 2022).Simulations do make predictions for the evolution of the SFRD, but these predictions span a wide range (e.g.Gnedin 2016;Dayal & Ferrara 2018;Tacchella et al. 2018;Yung et al. 2019;Behroozi et al. 2020).
Part of the difficulty of such studies is the near-heroic observational effort needed to study galaxies at z ∼ 10 with HST.These galaxies become more and more difficult to see with this 2.4m ultraviolet (UV)/optical/near-IR telescope, and at these high-redshifts they become single-band detections, leaving the z 11 universe opaque to our understanding.To avoid being dominated by spurious sources, studies employ a variety of vetting criteria to ensure robust samples of candidate galaxies (e.g.Bouwens et al. 2021;Bagley et al. 2022;Finkelstein et al. 2022), which makes it difficult to estimate the sample completeness and thereby to obtain a robust estimate of the SFRD.This all changes with the advent of the James Webb Space Telescope (JWST).The dramatic increase in lightgathering power coupled with the infrared sensitivity makes this telescope the ideal machine to push our cosmic horizons to the epoch of the first galaxies.As the first JWST images arrive it is natural to wonder what these early data tell us about the rise of star-formation in the early universe.If the SFRD really declines as steeply at z > 8 as has been proposed, few galaxies at z > 11 should be detectable in early JWST data.If the decline is instead more gradual one might expect to discover galaxies at z ∼ 12 or even higher.In just the first week since the data have been released exciting results already indicate significant star formation is occurring at z > 11 (e.g.Castellano et al. 2022;Naidu et al. 2022).
As another early probe of this epoch, here we report on a search for the highest redshift (z 12) galaxies in the first epoch of imaging from the Cosmic Evolution Early Release Science Survey (CEERS; Finkelstein et al. in prep).These data were among the first Cycle 1 science exposures taken, and were included in the first publicly released data on July 14. §2 describes the observations and data reduction, while §3 describes our photometry, photometric redshift measurement, and sample selection procedure.§4 presents our results, and we discuss these results in §5.Our conclusions are presented in §6.In this paper we assume the latest Planck flat ΛCDM cosmology with H 0 =67.36,Ω m =0.3153, and Ω Λ =0.6847 (Planck Collaboration et al. 2020).All magnitudes are in the absolute bolometric system (AB Oke & Gunn 1983).

CEERS Data
CEERS is one of 13 early release science surveys designed to obtain data covering all areas of astronomy early in Cycle 1. CEERS is based around a mosaic of 10 NIRCam (Rieke et al. 2005) pointings, with six obtaining NIRSpec (Jakobsen et al. 2022) in parallel, and four with MIRI (Rieke et al. 2015) in parallel (four of these pointings also include NIRCam wide-field slitless grism spectroscopy; Greene et al. 2016).Here we make use of the first four CEERS NIRCam pointings, obtained on 21 June 2022, known as CEERS1, CEERS2, CEERS3, and CEERS6.
In each pointing, data were obtained in the shortwavelength (SW) channel F115W, F150W, and F200W filters, and long-wavelength (LW) channel F277W, F356W, F410M, and F444W filters.The total exposure time for pixels observed in all three dithers was typically 2835 s per filter.The exception is F115W, which obtained double the exposure time to increase the depth on the filter covering the wavelength range below the Lyman-α break at z > 10.The full details on the readout and dither patterns will be available in the CEERS overview paper (Finkelstein et al. in prep).

Data Reduction
We performed a careful initial reduction of the NIR-Cam images in all four pointings, using version 1.6.2 of the JWST Calibration Pipeline1 with some custom modifications.We used the current (29 July 2022) set of NIRCam reference files2 , which includes in flight readnoise, superbias, distortion, and photometric flux calibration references.We note that the flats were created pre-flight.We describe our reduction steps below, and present more details in Bagley et al. (in prep).
Beginning with the raw data, we used Stage 1 of the pipeline with all default parameters to apply detectorlevel corrections, fit the ramps in each integration, and output countrate maps.We next subtracted the "wisp" features, stray light that is reflected off the secondary mirror supports, from detectors A3, B3 and B4 for filters F150W and F200W.For each image I, we scaled the corresponding wisp template3 W (available as of 8 July 2022) by the coefficient a that minimized Var(I − aW ), and subtracted the scaled template.We then performed a custom step to remove 1/f noise, which is correlated noise introduced in the images during the detector readout that presents as horizontal and vertical striping patterns (Schlawin et al. 2020).We applied the flat field to the countrate maps to ensure we were measuring the 1/f noise pattern on a flat image.We masked all bad pixels and source flux, using Photutils (Bradley et al. 2020) to detect sources and implementing a tiered approach to source masking.This approach convolves the image with progressively smaller kernels, identifying sources at each step.We use four tiers, with Gaussian kernels of σ = 25, 15, 5 and 2 pixels (on the original 0. 031/pixel and 0. 063/pixel scales for the SW and LW channels, respectively).These values were chosen after experimenting with several filter kernels to aggressively mask as much source flux as possible.First for each row and then each column, we measured a sigma-clipped median value and subtracted this value from the un-flat-fielded countrate map.This correction was performed amplifier-byamplifier in all filters except F444W, for which we did not perform any correction because the significant residual flat field structure present in the images dominated any 1/f pattern.
After processing the cleaned countrate maps through Stage 2 of the pipeline, we performed an astrometric calibration using an edited version of the TweakReg step of the pipeline.The TweakReg step detects sources in each input image, identifies their counterparts in the refer-ence catalog, and calculates a transformation to correct the image WCS.We ran Source Extractor (Bertin & Arnouts 1996) on each individual image to replace the internal TweakReg source identification, finding that Source Extractor did a better job of identifying and deblending real sources.In lieu of using the default options that allow for alignment to Gaia DR2, we used a reference catalog derived from a HST F160W 0. 03/pixel mosaic 4 in the EGS field with astrometry tied to Gaia-EDR3 (see Koekemoer et al. 2011, for details).We first determined relative offsets between images of the same detector, allowing for shifts in x and y.The RMS of this relative astrometry is ∼3-6 mas.We then aligned all images to our HST F160W reference catalog, allowing for shifts in x and y, rotations and, in the LW images only, a scaling to account for any additional distortion (though we note that the scaling factor is ∼ 1 × 10 −5 ).The RMS of this absolute alignment is ∼ 12 − 15mas, and the alignment between NIRCam images in different filters has an RMS of 5 − 10mas.
We fit and removed a single value in MJy/sr from each calibrated detector image separately before coadding the images onto a common output grid.For each image, we additionally calculate a scaling factor for the readnoise variance array (VAR RNOISE) from the backgroundsubtracted sky pixels, again avoiding source flux and bad pixels.We apply this scaling factor to the variance arrays so that they include the robustly-measured sky variance.The coadding was performed using the drizzle algorithm with an inverse variance map weighting (Fruchter & Hook 2002;Casertano et al. 2000) via the Resample step in the pipeline.The output mosaics have pixels scales of 0.03 /pixel.The usable total area covered by these observations, calculated from the number of pixels with low effective error-map values in all of the F115W, F150W, F200W, F277W, and the detection image (see below) is 34.5 arcmin 2 .
We note that our data reduction represents a preliminary version, with several aspects that will be improved with the release of updated NIRCam reference files.We also have not removed the features known as "snowballs" from the mosaics at this time.However, we have carefully inspected all input exposures to ensure that the fluxes in all filters at the positions of galaxies of interest are unaffected by snowballs (see Section 4.2).The full details of our photometric analysis will be presented in Finkelstein et al. (2022d, in prep); here we briefly summarize our procedures (many of which are similar to Finkelstein et al. 2022).The data products from our modified data reduction pipeline come in the form of multi-extension "i2d" files.We first estimate and subtract any residual background using a custom Python-based algorithm.This routine iteratively convolves the image with Gaussian kernels of progressively smaller sizes (with σ =25, 15, 5 and 2 pixels), then uses Photutils to mask pixels identified with sources in four iterations to mask progressively smaller sources, dilating the masks in between iterations (by 33, 25, 21 and 19 pixels), then measuring the background after masking with photutils.Background2D.This final background image-construction step used the BkgZoom-Interpolator algorithm to construct a smooth background based on the robust sigma-clipped means measured within boxes of 10x10 pixels, and median-filtered over 5x5 adjacent boxes.
The i2d file was split into separate extensions, subtracting this background from the SCI extension.Empirical PSFs were made by stacking stars, and the F115W, F150W, F200W, and F277W images were PSFmatched to the F356W image using pypher.Photometry was computed on the PSF-matched images using Source Extractor (hereafter SE; Bertin & Arnouts 1996) v2.25.0 in two image mode, with an inversevariance weighted combination of the PSF-matched F277W and F356W images as the detection image, with photometry measured on all seven bands.
Colors were measured in small Kron apertures with a Kron factor of 0.8 and a Kron minimum radius of 1.1 pixels; this is smaller than previous studies, which we found necessary to keep the elliptical aperture close to the significant isophotes of small, faint galaxies.An aperture correction was derived in the F356W catalog as the ratio between the flux measured in the default Kron aperture (with PHOT AUTOPARAMS 2.5, 3.5) to that in our small Kron aperture.This correction was applied to all fluxes and uncertainties.We use the CEERS simulated imaging 5 to test the accuracy of this procedure, finding that after this aperture correction, total fluxes were underestimated by ∼10-15%, rising to 22% in F444W (understandable due to the larger point-spread function [PSF] in F444W as the photometric apertures were defined on F356W).We apply these simulationbased corrections (comparable to similar corrections applied in HST studies; Finkelstein et al. e.g. 2022) to all 5 Simulated Data Release 3; ceers.github.io/sdr3.htmlfluxes and uncertainties to complete our total flux measurements.All fluxes and uncertainties were corrected for Galactic attenuation assuming a field-averaged E(B-V)=0.006and a Cardelli et al. (1989) Milky Way attenuation curve.We also measure fluxes in a range of circular apertures; as these are used for detection significance tests, we do not correct them to total fluxes (though they are still corrected for Galactic attenuation).
We derive flux uncertainties directly from the data, following Finkelstein et al. (2022), based on previous methodology outlined in Papovich et al. (2016).We fit for the noise as a function of aperture size by measuring the fluxes at ∼5×10 3 randomly-placed positions in 15 circular apertures with diameters ranging from 1 -100 pixels, fitting a polynomial function to the standard deviation in aperture fluxes as a function of the number of pixels in each aperture.We then use this function to calculate the photometric uncertainties for each object for a given aperture area.These values were scaled by the ratio of the error image value at the central position of a given source to the median error value of the whole map.All aperture and Galactic attenuation corrections were applied to these uncertainties.Finally, around each source in our catalog, we calculate a "local" noise estimate, as the standard deviation in flux values from the 200 closest of these previously placed random apertures.

Photometric Redshifts
We use the EAZY (Brammer et al. 2008) software package to estimate photometric redshifts for all sources in our photometric catalog.eazy fits non-negative linear combinations of user-supplied templates to derive probability distribution functions (PDFs) for the redshift, based on the quality of fit of the various template combinations to the observed photometry for a given source.The template set we use includes the "tweak fsps QSF 12 v3" set of 12 FSPS (Conroy & Gunn 2010) templates recommended by the eazy documentation.To this we add a set of six additional templates spanning bluer colors than the FSPS models, as Larson et al. (2022, in prep) found that these improve the accuracy of photometric redshift fits for the expected blue colors of z > 9 galaxies.
The new templates were created using stellar population models created with BPASS (Eldridge & Stanway 2009).To generate bluer rest-UV colors than the initial set of FSPS templates, we selected BPASS templates with low metallicities (5% solar), young stellar populations (log stellar ages of 6, 6.5, and 7 Myr), and were inclusive of binary stars.We added an additional set of these models inclusive of nebular emission lines derived with CLOUDY (Ferland et al. 1998) using a high ionization parameter (log U = −2), a similarly low metallicity, and with nebular continuum emission.Larson et al. (in prep) validated the efficacy of including these models by testing the recovery of photometric redshifts from a mock catalog derived by a semi-analytic model (Yung et al. 2022), finding that the inclusion of these six additional templates significantly improved the photometric redshift estimates for blue high-redshift galaxies.
We do not use a luminosity prior (e.g., a flat prior is assumed) as the epoch in question is completely unexplored, and we include a systematic error of 5% of the observed flux values.Our fiducial eazy run uses our total fluxes derived from our Kron-aperture measured colors, and we use the measured fluxes and errors even in the case of non-detections (as opposed to using upper limits).We also perform two ancillary runs which we use for later vetting.One uses fluxes measured in 0.3 circular apertures (to cover the possibility that a Kron ellipse was drawn inaccurately, which happens in the presence of bright neighbors).A second run had a maximum redshift of z = 7 to allow the exploration of secondary redshift solutions.

Sample Selection
To select our sample of candidate very high redshift galaxies, we follow previous work done by our team (Finkelstein et al. 2015;Rojas-Ruiz et al. 2020;Finkelstein et al. 2022;Bagley et al. 2022).We make use of photometric signal-to-noise criteria, to ensure robust photometric detections (to minimize the chance of a spurious signal), and ensure robust non-detections below the Lyman-α break.We add to these several criteria based off of the full eazy redshift PDF (denoted P(z)).We note that the criteria imposed here are fairly conservative -we wish to identify the most robust highestredshift candidates.Future work will explore how to relax some of these criteria to improve sample completeness and push to z < 12, without introducing unacceptable levels of contamination.
To derive an initial sample of z ≥ 12 galaxies, we first impose all following requirements: • Signal-to-noise (SNR) in both F200W and F277W > 6 in conservatively small 0.2 (6.7-pixel) diameter apertures for these measurements, using both the fiducial (global) and local noise estimates.We note that enforcing a SNR cut in F200W effectively limits this sample to z 15, though by enforcing detections in both a short and long-wavelength detector, we limit inclusion of detector-specific spurious sources.
• Error map values < 1000 (indicating coverage by the majority of exposures) in F115W, F150W, F200W, F277W and the detection image.
• Initial more inclusive photometric redshift cuts of P(z > 8) ≥ 0.9, z best > 8.5, χ 2 EAZY < 20 (to reject poor eazy fits), and that the ∆z = 1 integer redshift bin (z sample ) with the largest integrated P(z) to be at z sample ≥ 9.
• Objects at z sample > 10 must have SNR ≤ 2.0 in F115W, while objects at z sample > 13 must have a SNR ≤ 2.0 in both F115W and F150W (in both the global and local noise in 0.2 -diameter apertures).These redshifts correspond to the wavelength of the Lyman-α break leaving a given dropout filter.
• F200W magnitude < 29, to focus on well-detected objects regardless of formal SNR.
After this initial set of selection criteria, we examined the resulting objects.We inspected their spectral-energy distributions (SEDs), image stamps, and P(z) plots.We noticed several low-confidence sources which could be identified with further automated cuts.We thus implemented this additional set of selection criteria: • We additionally implement all of the above detection significance criteria, both in the detection and dropout bands, in a 0.3 -diameter aperture to account for situations where faint flux was visible slightly off-center of the source barycenter.
• We require the χ 2 from an additional eazy run with a maximum redshift of seven to have a significantly worse fit than our fiducial run via < 1 for objects with z sample ≤ 13 to reduce the incidence of red low-redshift interlopers.This is similar to the color cuts simulated by Hainline et al. (2020) and implemented by Castellano et al. (2022).For objects with z sample = 13-18 the Lyα break falls in the F200W filter, thus we require F277W -F444W < 1.
• To account for situations where the Kron aperture could be affected by nearby bright sources, we also require P(z > 8) ≥ 0.5 from an independent eazy run performed with colors measured in 0.3 circular apertures.
As our focus here is on the highest-redshift sources, we limit our analysis to objects with z sample ≥ 12.Running Note-TBigBang is the time elapsed from the Big Bang to the photometric redshift for our assumed cosmology.MUV and β were computed from the Prospector models, using the same techniques as in Tacchella et al. (2022).The physical properties listed below the horizontal line were derived with Prospector.
the above selection process on all four fields, we find a single galaxy candidate which satisfies all of the above criteria.We perform an initial visual inspection of this candidate, inspecting 1.5 image stamps in all filters, and 5 image cutouts in F200W and the detection image, and find that this object appears astrophysical in origin, and is not an artifact.

A Robust Galaxy Candidate at z ≈ 12
This source, CEERSJ141946.36+525632.8,hereafter known as "Maisie's Galaxy"6 was detected in the CEERS2 field.Promisingly, it was first identified in the earliest (v0.02) internal CEERS reduction in this field, being the first z 12 candidate viewed on 18 July, 2022.In each subsequent reduction, this source continued to satisfy all selection criteria, becoming progressively more robust as the data became cleaner.Several CEERS team members viewed this potential z 12 candidate on 22 July 2022, and agreed on the robustness of this source.The photometric redshift of this source with NIRCam photometry alone is z = 12.0 +0.2 −0.4 .
Given the much greater sensitivity of NIRCam, we do not expect significant changes to our interpretation based on the inclusion of HST images.These images were not included in our fiducial SE analysis as they had not yet been pixel aligned given the short time since NIRCam data acquisition.However, upon inspection we find a hint of a positive signal at the position of the source in the F160W image.Indeed, while the source is not in in the published Stefanon et al. (2017) and Skelton et al. (2014) catalogs, there is a 3.5σ detection at a separation of 0.15 in the Finkelstein et al. (2022) catalog (this object has SNR < 2 in all other HST filters in this catalog).
Using SE we perform forced photometry at this position on the CANDELS (e.g., Grogin et al. 2011;Koekemoer et al. 2011) 30 mas images (using the updated images provided by the CEERS team which have registered the astrometry to Gaia) in the ACS F606W and F814W, and WFC3 F125W, F140W and F160W bands.Following Finkelstein et al. (2022), we apply an additional flux correction of 1.2 to all it HST bands to account for missing wings of the PSF in the larger Kron aperture.We find SNR <2 in all bands except WFC3 F160W, which has a ∼4σ detection, with a flux consistent with the weak F150W detection for this source.Including this photometry (listed in Table 3) in our photometric redshift fit slightly changes our photometric redshift estimate to z = 11.8 +0.2 −0.3 , which (for our assumed cosmology) corresponds to an age of the Universe of 373 +16 −8 Myr.We use these results inclusive of HST photometry as our fiducial values.The properties of this galaxy are summarized in Table 1, and we list its photometry in Tables 2 and 3.
Figure 1 shows cutouts of this candidate galaxy in the NIRCam bands, while Figure 2 shows two color composites.Figure 3 shows the observed spectral energy distribution of our candidate with photometric redshift fits.The confidence of this source as a robust very highredshift galaxy is easy to see from all three of these figures.The Lyman-α break color, here F115W-F200W, is >1.9 mag (2σ lower limit), completely eliminating any known low-redshift interloper (the F150W-F200W color is 1.3 mag; still strong, though smaller in amplitude due to the Lyα break being present at the very red edge of F150W).Such a model would need to have an extremely red color to match our F115W-F200W >1.9 mag break, but then have a very blue color.While lower-redshift passive or dusty galaxies can mimic highredshift Lyman-α breaks, the observed >1.9 mag break is much larger than known populations of low-redshift galaxies.Such galaxies would also be fairly red redward of the break.3 ) and a significant detection in the bluest detection band (F200W in this case).Very faint flux is visible (at ∼2.8σ significance when measured in a 0.2 diameter aperture) in F150W, which drives the redshift to z ∼ 12.The wide wavelength range of NIRCam allows this source to be well-detected in multiple filters, and in the imaging alone it is clear this source exhibits a blue spectral shape.Bottom) Same ordering as the top, for sky-uncertainty maps constructed from the variance of the readout noise, all using a linear scale from 0.33 to 3× the robustly-measured sky standard deviation in each band.The patchiness of the uncertainties is due to loss of exposure time when cosmic-rays are detected and rejected in the multiple readouts or in outlier rejection when combining the dithered exposures (the 2× larger original pixel scale of the long-wavelength channels results in larger patches than the short-wavelength channels).The uncertainty arrays show no excess in rejected pixels near the candidate galaxy.
Though differential geometry could accommodate UV spectral slopes as blue as β ∼ − 1, this object has β ∼ −2.5 (see §5.1; Casey et al. 2014).The significant detection in four broadband filters also rules out low-redshift extreme emission line galaxies.We show as the orange curve in Figure 3 eazy's best-fitting lowredshift model, which is ruled out at high confidence.Based on the non-detection in F115W and strong detection in F200W, the implied redshift is z > 11.This is confirmed by the eazy fit, shown as the blue line, which prefers z ∼ 11.8.

Fidelity of Candidate
Figure 1 shows 1.8 cutout images of this source at multiple wavelengths.This source shows the expected pattern for a high-redshift galaxy, with no significant flux in multiple dropout bands, with robust flux in redder bands.The very sharp break between F115W and F200W is consistent with a redshift of z > 11 (the faint detection in F150W pushes the redshift solution to z ∼ 12 as opposed to higher redshift).The advantage of JWST is clear here, as this source is well-detected in all five NIRCam filters redward of the break.This multi-band detection essentially eliminates the possibility of a spurious nature.Of note is that while persistence from previous observations affected several HST programs (see discussion in ters first, thus any flux from persistence would be most apparent in F115W. To further rule out a spurious nature, the science, error, and data-quality images were visually inspected at the position(s) of the best candidate(s).This is to ensure that the detected sources in the co-added images are not just chance super-positions of regions that were Left) The circles denote our fiducial photometry, with green and red denoting HST/ACS+WFC3 and NIRCam instruments, respectively.This SED exhibits the hallmark shape of a high-redshift galaxy, with several non-detections in blue filters, followed by significant detections with a blue spectral slope.The arrows denote 1σ upper limits.The F115W-F200W break color is >1.9 mag (2σ), which is sufficient to rule out all low-redshift solutions.The blue curve shows the best-fitting eazy model at z = 11.8.The orange curve shows the result if we force eazy to find a solution at z < 7.This model is unable to match the amplitude of the break as well as the slope redward of the break, and is correspondingly ruled out at high confidence affected by cosmic rays or other artifacts.In the case of the z ∼ 12 candidate, the source is visible in all of the individual F200W, F277W, and F356W exposures, and overlaps with a cosmic ray in only a few images.Even in those cases, the cosmic rays that are masked in the dataquality array are of the typical size that is cleanly rejected in the jump-detection step of the pipeline.There were no overlaps with the larger "snowball" chargedparticle events.
As an additional check, we measured photometry at the position of this source on our images without our post-processing residual background subtraction step, to ensure any systematic effects at this source position did not affect our results.The images already have a pedestal background subtracted in the pipeline, so the relative colors should be secure when measured in this way.We found that this set of photometry was consistent with our fiducial photometry, and eazy returns P(z > 11) = 0.995.Therefore it is unlikely that our sky subtraction routine negatively affected our result.

Stellar Screening
Low-mass stars and brown-dwarfs can have colors that mimic high-redshift galaxies in broadband filters (e.g Yan et al. 2003;Ryan et al. 2005;Caballero et al. 2008;Wilkins et al. 2014) in the absence of longer wavelength observations (λ obs 2 µm).We explore this possibility following the methodology in Finkelstein et al. (2022).In brief, we derive a grid of models for the colors of lowmass stars and brown dwarfs (spectral types of M4-T8) in the NIRCam filters, by integrating the IRTF SpEX brown dwarf templates (Burgasser 2014).As these spectra end at 2.5µm, we use the tabulated 2MASS photometry to link each SpeX model with Spitzer/IRAC pho- tometry from Patten et al. (2006).As the differences in filter transmission are negligible, we assume we can map IRAC 3.6µm onto F356W and 4.5µm onto F444W, however this assumption will need to be revisited with future spectroscopic observations of brown dwarfs with JWST at λ 2.5 µm.We estimate the best brown dwarf template would be an L7.5-dwarf, and such a source would have blue near-infrared color of F115W−F200W = 0.9 mag.This is strongly ruled out by our observation of F115W−F200W>1.9 mag (2σ lower limit).Additionally, our size analysis in §5.2 shows that this object is inconsistent with a point source.

Photometric Accuracy
While our fiducial photometric measurements were derived in as robust a manner as possible, different software packages require different parameters and assumptions, which could lead to unknown systematic biases.We thus independently derive NIRCam photometry from our images with two independent software packages.The first method is Photutils from Python's astropy package (Bradley et al. 2020).Source detection was performed on a combined F277W and F356W image and the resulting segmentation image passed to the Photutils SourceCatalog routine, which carried out aperture-matched photometry on the backgroundsubtracted, PSF-matched images in each filter.
The second method is a custom photometry package, where photometry is measured in circular apertures with radii ranging from 0.10 to 0.35 , applying aperture corrections for point-like sources (<0.1 mag for r>0.25 ), and after locally (30 box) aligning the images (Pérez-González et al. 2008).Sky noise measurements in a 6 × 6 box around the source take into account correlated noise and are used to quote 5σ upper limits for nondetections.Photometric differences for each band are smaller than 0.1 mag for apertures between 0.2 and 0.35 , 0.3-0.6 mag fainter for smaller radii, indicating that the source is (slightly) resolved.This method was applied to the non-PSF-matched imaging.
Comparing results between our fiducial SE photometry and these independent methods, we find that both the Lyman break and rest-UV colors show extremely high consistency.The upper limits in F115W are similar to our fiducial values.The F115W−F200W Lyman-α break color is >2.3 mag (2σ) from the custom method, and >1.7 mag (2σ) for Photutils (compared to >1.9 mag [2σ] for our fiducial SE photometry).Comparing colors, our measured F200W−F444W color of −0.6 ±0.4 is highly consistent with the measurement from the custom method of −0.7±0.3.The Photutils measurement is even bluer (−1.0 ± 0.4) due to a 0.4 mag fainter F444W measurement.We conclude that while differences in photometric packages and associated assumptions (in particular aperture corrections) can affect the photometry at the ∼10-30% level in most cases, this does not affect the validity of our candidate as these independent methods find a consistently strong Lyman-α break followed by a blue spectral slope, fully consistent with our interpretation of a z ∼ 12 galaxy.

Photometric Redshift Accuracy
Similar to photometry, different photometric redshift packages can also impart biases on results.While we have used a well-tested fiducial package in eazy, and implemented a new set of templates customized for very high-redshift galaxies, it is prudent to explore whether other packages would find different photometric redshift results.As we discuss below, we have run the Prospector (Johnson et al. 2021), Bagpipes (Carnall et al. 2018), Cigale (Burgarella et al. 2005;Noll et al. 2009;Boquien et al. 2019) and Dense Basis (Iyer & Gawiser 2017;Iyer et al. 2019) SED-fitting codes on our fiducial photometry.While for the stellar population properties discussed below, we use our eazy-derived P(z) as a redshift prior, we also performed an independent run with the redshift as a free parameter.Figure 3 shows our fiducial eazy P(z) along with the redshift PDFs from these independent runs.These five results show remarkable consistency, all preferring z > 12 with no significant low-redshift solutions.All four codes find results simular to our fiducial eazy run.Prospector finds z = 11.8 +0.2 −0.3 , Bagpipes finds z = 11.6 +0.2 −0.2 , Cigale finds z = 11.8 +0.4 −0.4 , and Dense Basis finds 11.9 +0.4  −0.4 .Combining the posteriors of all four photometric redshift estimates provides a redshift PDF in agreement with our fiducial eazy results with a median redshift of 11.74, and a 97.5% confidence that z > 11.0.We conclude that systematic biases due to choices in photometric redshift analyses are not affecting our results.Our fiducial result uses that from eazy as it used templates trained on observations, while the full grids spanned by the other four codes may include unphysical parameter combinations.
As one final test, we explore the impact of our inclusion of the six additional blue templates in our photometric redshift analysis.Refitting our photometry with the standard templates we find z = 11.50 +0.33  −0.33 , consistent within 1σ of our fiducial result.However, standard templates produce a significantly worse fit (χ 2 ∼ 34 versus ∼15 for the fiducial fit).The standard best-fitting template has a F200W-F277W color of −0.04 mag, while our fiducial template has F200W-F277W=−0.35;both can be compared to the observed color of F200W-F277W=−0.39.It is clear that our inclusion of bluer templates are better able to match the colors of ultra high-redshift galaxies such as the one we present here.

Contamination Estimation
To determine the likelihood that our selection criteria would produce a low-redshift contaminant we imposed our same selection criteria cuts on the simulated catalogs used for all the mock CEERS observations.We note that there are zero z > 10 sources in this catalog so recovery of any source using these selection criteria would indicate contamination of our high-redshift sample.More information about the simulation used can be found in Yung et al. (2022) and Somerville et al. (2021).We use the perturbed fluxes as described in Larson et al. (2022, in prep) which use the same method as determined by Bagley et al. (2022, in prep) where they modeled the noise in simulated JWST images to have a Voigt profile distribution.We used the 1σ-depth in each filter for our errors and ran the whole catalog through eazy.As our catalog-level fluxes do not have aperture-specific fluxes we cannot impose criteria based on those fluxes.We apply the following selection criteria to the simulated catalog: SNR in both F200W and F277W > 6, P(z > 8) ≥ 0.9, z best > 8.5, χ 2 EAZY < 20, SNR ≤ 1.5 in F606W & F814W & F115W & F150W, F200W magnitude < 29, F200W-F444W color < 1, and χ 2 Low−z − χ 2 f iducial > 4. Finally, matching the values to those of Maisie's Galaxy, which exhibits SNR > 10 in both F200W and F277W and P(z > 11) ≥ 0.99, we find zero sources that meet our criteria.This provides further evidence that Maisie's Galaxy has a high-redshift nature.

Physical Properties
The five photometric detections afforded by NIRCam allow us the unprecedented opportunity to study the physical properties of a galaxy potentially only ∼400 Myr after the Big Bang.Our fiducial stellar population modeling is done with the Prospector Bayesian SED fitting code (Johnson et al. 2021).We follow the same procedures as in Tacchella et al. (2022) and we refer the reader there for more details.Briefly, we model the SED with a 13-parameter model that includes redshift (where here prior is set to the posterior of Eazy, unline in §4.2.3 where it was free), stellar mass, stellar and gas-phase metallicities, dust attenuation (twocomponent dust model including birth-cloud dust attenuating young stars (< 10 Myr) and nebular emission, a diffuse component for the whole galaxy with a flexible attenuation law; 3 parameters), and an ionization parameter for the nebular emission.We adopt a flexible SFH prescription with six time bins (the first two lookback time bins are spaced at 0 − 5 Myr and 5 − 10 Myr, while the other four are log-spaced out to z = 20; five free parameters) and with the bursty-continuity prior.Furthermore, we assume the MIST stellar models (Choi et al. 2017) and a Chabrier (2003) IMF.
To explore how robust these properties are, we perform an independent fit with the Bayesian Bagpipes (Carnall et al. 2018), Cigale (Burgarella et al. 2005;Noll et al. 2009;Boquien et al. 2019) and Dense Basis (Iyer & Gawiser 2017;Iyer et al. 2019) SED-fitting codes.For Bagpipes we assumed a simple exponential star formation history with a Chabrier IMF, a Calzetti dust attenuation law and included nebular emission with an ionization parameter of 10 −3 , with Bruzual & Charlot (2003) stellar population models.For Cigale, we assume a delayed star formation history after checking that adding a burst does not significantly modify the results.Bruzual & Charlot (2003) models with a Chabrier IMF and was used.Dense Basis was run using the flexible non-parametric SFH model and priors described in Iyer et al. (2019), assuming a Calzetti dust law and a Chabrier IMF.Metallicites in all four codes were allowed   1), Bagpipes (black), Cigale (green dotted) and Dense Basis (blue dotted).Right) Posterior distributions of the key stellar population properties from all four codes.The panels show stellar mass, mass weighted age, dust attenuation and SFR averaged over the last 10 Myrs.The vertical dotted lines indicate the mean of the posteriors.Posteriors of attenuation are consistent between all four codes.Prospector prefers a younger age than the other three because of a recent burst in the SFH of this object.As a consequence of the burst, Prospector also estimates a lower SFR.The four estimates of the stellar mass posteriors exhibit significant overlap, though the median values differ by ± 0.2 dex.Future observations in the rest-optical with MIRI could break these degeneracies.
to vary to allow the uncertainty in metallicity to be included in the uncertainties on other parameters, though this parameter is not well-constrained with photometry alone.We note that Prospector and Cigale assume an error floor (of 5% and 10% of the flux, respectively), and that Prospector, Dense Basis and Bagpipes always fit the measured fluxes, while Cigale uses the flux errors as upper limits (by setting the flux and uncertainty equal to the measured flux error) when the fluxes are negative.
The marginalized posterior values of the inferred physical properties from Prospector are summarized in Table 1 and Fig. 4. We infer a stellar mass of log(M * /M ) = 8.5 +0.3  −0.4 .The attenuation in this galaxy is rather low with A V = 0.07 +0.22  −0.06 mag, though we stress that this is not well constrained because we only fit the rest-UV and it is degenerate with the slope of the attenuation law (which is variable in this fit).However, the low dust attenuation is in agreement with the measured UV spectral slope β = −2.47+0.09 −0.09 (measured using the same techniques as in Tacchella et al. 2022).This blue color implies little dust, though does not require extremely low metallicities (e.g.Finkelstein et al. 2012;Dunlop et al. 2013;Bouwens et al. 2014).Interestingly, this galaxy is about as blue as z ∼ 7 galaxies of similar mass (Finkelstein et al. 2012), implying little evolution in chemical enrichment between these two epochs for similar-mass galaxies.
We infer a SFR 10 (average of the past 10 Myr7 ) of 2.1 M yr −1 and the corresponding sSFR 10 is 10 −8.2 yr −1 .By looking at the posterior distribution of the SFH, it becomes apparent that the model for this galaxy had an episode of elevated star formation 10 − 20 Myr ago with a SFR of 8 +17 −7 M yr −1 , i.e. the SFR has been slightly decreasing in the recent 10 Myr.This explains the mass-weighted age of 18 +18 −9 Myr.This is also consistent with the half-mass formation time of dark matter halos at z ∼ 12 of a few tens of Myr (Tacchella et al. 2018).
These Prospector-based posterior distributions are consistent with the ones from Bagpipes, Cigale and Dense Basis (see Fig. 4), though the difference in age is large (age is defined at half-mass time, t 50 , which is close to the mass-weighted age).Bagpipes, Cigale and Dense Basis prefer higher age values (although the posterior distributions are also broader) with 67 +56 −37 Myr, 34 ± 20 Myr, and 64 +63 −54 Myr respectively.The SFH inferred from Dense Basis shows a recent burst of star formation in the last ∼ 30 Myr.The larger massweighted age comes from the long tail of low-level star formation in the galaxy leading up to the recent burst.
The spread in these results could be explained by differences in SFHs (e.g., non-parametric versus parametric), and also the lack of observational constraints in the restframe optical.These differences also lead to Prospector having the lowest SFR 10 value.
Several pre-JWST studies have focused on inferring SFHs and stellar ages of z ≈ 8 − 10 galaxies (e.g.Hashimoto et al. 2018;Laporte et al. 2021;Stefanon et al. 2022).Specifically, Tacchella et al. (2022) -using Prospector with the same bursty continuity prior -found a diversity of stellar ages, ranging from 10 Myr to 260 Myr, and stellar masses (10 9 − 10 11 M ), with more massive galaxies being older.In particular the galaxies at z ≈ 9 − 10 with stellar masses at the higher end and the older ages (t 50 ≈ 100 Myr) are consistent with being the descendants of Maisie's Galaxy.Recently, Naidu et al. (2022) inferred the properties of two galaxies at z ≈ 10.6 and z ≈ 12.4 (see also Castellano et al. 2022) with Prospector and a similar setup, allowing us to do a useful comparison.Their two galaxies have log(M * /M ) = 9.4 +0.3 −0.3 and 9.0 +0.3 −0.4 , SFR 50Myr = 12 +9 −4 M yr −1 and 7 +4 −3 M yr −1 , and t 50 = 111 +43 −54 Myr and 71 +33 −32 Myr, respectively.This is older than what we infer for our galaxy, though this age difference could be explained by the stellar mass difference, along with the higher preferred redshift for Maisie's Galaxy.Importantly, detailed stellar population analyses of early galaxies will advance significantly with JWST, in particular when including spectroscopic information.

Source Morphology
We derive the sizes of Maisie's Galaxy using two morphological fitting codes, GalfitM 8 (Häußler et al. 2013) and statmorph 9 (Rodriguez- Gomez et al. 2019).Gal-fitM is a modified version of Galfit 10 ( Peng et al. 2002;Peng et al. 2010), a least-squares fitting algorithm that finds the optimum Sérsic fit to a galaxy's light profile.We perform fits using GalfitM by allowing the Sérsic index to vary between 0.01 and 8, the magnitude of the galaxy between 0 and 45, and r half between 0.3 and 200 pixels (on our 0.03 pixel scale).As input, we use a 100×100 pixel cutout of the F277W science image, the segmentation map created by Source Extractor, and the empirical PSF measured from our CEERS2 pointing, which we allow GalfitM to oversample relative to the data by a factor of nine.We estimate the uncertainty on our fits by conducting a Monte 8 https://www.nottingham.ac.uk/astronomy/megamorph/ 9 https://statmorph.readthedocs.io/en/latest/ 10https://users.obs.carnegiescience.edu/peng/work/galfit/galfit.htmlCarlo analysis where we modify the input F277W science image to randomly vary the pixel-to-pixel noise, recompute the parameters, and then repeat this analysis 40 times.
Following this procedure, we measure a half-light radius of 3.0±0.12pixels (0.09 ± 0.0036 ), which corresponds to a physical size of 340 ± 14 pc at z = 11.8.We check these results using the standard configuration of Statmorph, a Python package developed to calculate the nonparametric morphology of galaxies as well as compute single Sérsic fits.Using the same images as input, we find a half-light radius of 2.9 pixels, in good agreement with the measurement from GalfitM.We repeat this measurement for the F200W filter and a stacked F200W+F277W image and find consistent results.The measured half-light radius of 3.0±0.1 pixels is significantly larger than that expected for a point-source (the median r h for our PSF stars is 1.95 ± 0.22 pixels in F277W), further ruling out a stellar origin for this source.

Comparison to Model Predictions
In Figure 5 we present predictions from a range of theoretical models, including the First Light And Reionisation Epoch Simulations (FLARES, Lovell et al. 2021;Vijayan et al. 2021;Wilkins et al. 2022), a suite of hydrodynamical cosmological zoom simulations; the large periodic volume hydrodynamical simulation Bluetides (Feng et al. 2016;Wilkins et al. 2017); the Delphi (Dayal et al. 2014(Dayal et al. , 2022) ) and Santa Cruz SAM (Yung et al. 2019(Yung et al. , 2020) ) semi-analytical models, the semi-empirical UniverseMachine (Behroozi et al. 2020), Mason et al. (2015), and Behroozi & Silk (2015) models.For the FLARES, Delphi, and Behroozi & Silk (2015) models, we show both the attenuated and un-attenuated (intrinsic) predictions.These predictions were made by interpolating and integrating either the binned or Schechter luminosity functions across z = 15 → 12 taking account of the areal size of the CEERS observations.Almost all of these models predict an expected source density much less than one, making the observation of even a single object at this redshift and magnitude surprising and potentially hinting at significant differences between the physical assumptions in these models and the real early universe.
The exception is the Behroozi & Silk (2015) model, which extrapolated galaxy formation to high redshifts by assuming that the ratio between galaxies' sSFRs and their host halos' specific accretion rates remained constant, which they showed was equivalent to assuming that galaxies' stellar masses are proportional to a power of their host halo masses.This model was constrained only with observational data at z ≤ 8, and predicted no change in the slope of the CSFR relation with redshift at z > 8.As a result, it predicted many more high-redshift galaxies than later models that were constrained to match z ∼ 9 and ∼ 10 data from HST that suggested more rapid declines in the number densities of early galaxies.We caution against over-interpretation, as the current sample contains only a single object with a consequently large Poisson error in addition to additional uncertainty due to cosmic variance.Nevertheless, if confirmed, the existence of this object places informative constraints on galaxy formation models in this epoch.

Comparisons to Extrapolations from Lower Redshift
We are now only just getting our first glimpse into this epoch with the first JWST data.Nonetheless, we can compare our observed number density to a few recent observations.We calculate a rough number density for M U V = −20.3galaxies assuming a top-hat selection function over 11.5 < z < 12.5.This is overly simplistic, and does not account for incompleteness (which, although this is a >10σ detection, certainly is non-unity due to our stringent selection criteria).Nonetheless it is illustrative of a rough number density.We find a maximum volume over the CEERS first-epoch area of 5.0 × 10 4 Mpc 3 , for a number density for our singular source of 2.0 +4.6  −1.7 × 10 −5 Mpc −3 (where the uncertainties are Poisson based on our detection of one object).
We illustrate this number density in Figure 6.Our derived number density is not inconsistent with a variety of observational constraints at z ∼ 10, as well as recent results at z ∼ 12-13.The solid gray line shows the predicted z = 12 Schechter function from Finkelstein (2016), which is extrapolated from an empirical fit to observations at z = 4-8, assuming smooth redshift evolution.Interestingly, our rough number density measurement is above, though consistent within 1σ, with this prediction, which would support its assumption of a smooth decline in the luminosity function (and in the corresponding star-formation rate density).Our results are also consistent with the smoothly-evolving doublepower law (DPL) model at z = 12 from Finkelstein & Bagley (2022), again supporting a smooth decline in the rest-UV luminosity function to z > 10.
As noted by several previous studies the bright-end of the luminosity function at z ≥ 9 exhibits an interesting excess over predicted levels (e.g.Bowler et al. 2020;Rojas-Ruiz et al. 2020;Morishita et al. 2018;Bagley et al. 2022;Finkelstein et al. 2022).While our survey area does not yet probe the volume densities needed to reach these brighter potential z = 12 galaxies, if the high-redshift luminosity function follows a DPL form, the forthcoming 0.6 deg 2 COSMOS-Web survey (PIs Kartaltepe & Casey) should be able to discover this population.In combination with the full Cycle 1 slate of surveys, including the completed CEERS imaging, it will afford a more complete view of the z = 12 universe.

CONCLUSIONS
We present the results from a search for ultra-highredshift galaxy candidates from the first epoch of NIR-Cam imaging from the JWST CEERS survey.We use imaging from both the short and long-wavelength cameras over four pointings, covering 34.5 sq.arcmin in the F115W, F150W, F200W, F277W, F356W, F410W and F444W filters, reaching m ∼ 29 (5σ) in the deepest bands.We measure photometry using Source Ex-tractor, with an emphasis on robust measurements of colors, total fluxes, and uncertainties.
We estimate photometric redshifts with the EAZY software package, including new blue templates designed to better recover the colors of very distant galaxies.We develop iteratively a set of conservative selection criteria to select candidate galaxies at z ≥ 12.We find one candidate galaxy satisfying stringent non-detections (SNR <1.5) in all dropout bands, and detected at >10σ in the detection bands.
This object, dubbed Maisie's Galaxy, has a photometric redshift of 11.8 +0.2 −0.3 , and was found in the CEERS2 field.We explored all known potential sources of contamination, including instrumental effects, systematic biases in the analysis, and contamination by lowerredshift galaxies or Galactic stars.We find that none of these alternative explanations can account for the observations, leaving us with the conclusion that it is a robust z ∼ 12 galaxy candidate.
We explore the physical properties of this unexpected galaxy.As might be expected for such an early epoch, this galaxy is blue, with a UV spectral slope β = −2.5, consistent with low levels of dust attenuation.Stellar population modeling with multiple codes are in agreement that this source has a modest stellar mass of log (M/M ) ∼8.5, with a high log sSFR of −8.2 yr −1 .The mass-weighted age of Maisie's Galaxy is young, with a median of ∼20 Myr, though stellar populations as old as 100 Myr (z f orm ≥ 14) cannot be ruled out.The galaxy candidate is significantly resolved, with r h = 3.0 ± 0.1 pixels, for a physical size of ∼340 pc at z ≈ 12.
We compare the abundance of this single galaxy both to model predictions and previous observations.We find that the presence of this source is unexpected based on most model predictions, though given our sample size the tension is modest at best.However, both semiempirical models and empirical extrapolations which assume a smooth decline in the SFR density at z > 8 predict volume densities of such z ∼ 12 sources in agreement with our observations.Should more such sources be found in early JWST surveys, it would provide further evidence against accelerated decline SFR density scenarios.
Such a galaxy population would also present challenges for a variety of dark matter models with suppressed power on small scales, such as fuzzy dark matter (e.g.Sullivan et al. 2018), and possibly even for standard ΛCDM models.Additionally, the presence of this galaxy ∼370 Myr after the Big Bang may be consistent with redshifted 21-cm absorption at z ∼ 18 reported by the Experiment to Detect the Global Epoch of Reioniza-tion Signature (EDGES Bowman et al. 2018) presumed to be caused by light from the first stars.
We caution the reader that this galaxy is a candidate.While we have exhausted multiple avenues to explore whether its presence in our data could be caused by instrumental effects, whether our measurement techniques were biased, or whether its colors could be consistent with lower-redshift sources, the "gold standard" of distance measurements is spectroscopic confirmation.Such confirmation should be possible in modest exposure times with the NIRSpec and/or MIRI spectrographs on board JWST.The combination of larger samples being compiled by JWST Cycle 1 programs, including the remainder of CEERS, COSMOS-Web (PIs Kartaltepe & Casey), JADES (PIs Rieke & Ferruit), PRIMER (PI Dunlop), PEARLS (Windhorst et al., in prep) and NGDEEP (PIs Finkelstein, Papovich, & Pirzkal) coupled with subsequent spectroscopic followup will further illuminate the earliest phases of galaxy formation.
We acknowledge that the location where this work took place, the University of Texas at Austin, sits on indigenous land.The Tonkawa lived in central Texas and the Comanche and Apache moved through this area.We pay our respects to all the American Indian and Indigenous Peoples and communities who have been or have become a part of these lands and territories in Texas, on this piece of Turtle Island.We thank the entire JWST team, including the engineers for making possible this wonderful over-performing telescope, the commissioning team for obtaining these early data, and the pipeline teams for their work over the years building and supporting the pipeline.The authors acknowledge the Texas Advanced Computing Center (TACC) at The University of Texas at Austin for providing HPC and visualization resources that have contributed to the research results reported within this paper.We thank Brendan Bowler, Caroline Morley, Jim Dunlop and Mike Boylan-Kolchin for helpful conversations, and thank the anonymous referee for constructive comments.(Carnall et al. 2018), Cigale (Burgarella et al. 2005;Noll et al. 2009;Boquien et al. 2019), Dense Basis (Iyer & Gawiser 2017;Iyer et al. 2019), Drizzle (Fruchter & Hook 2002), eazy (Brammer et al. 2008), GalfitM (Peng et al. 2010;Häußler et al. 2013), Photutils (Bradley et al. 2020), Prospector (Johnson et al. 2021), SciPy (Virtanen et al. 2020), Source Extractor (Bertin & Arnouts 1996), Statmorph (Rodriguez-Gomez et al. 2019), STScI JWST Calibration Pipeline (jwst-pipeline.readthedocs.io)

Figure 1 .
Figure1.Top) 1.8 × 1.8 cutout images centered on the position of Maisie's Galaxy in the non-PSF-matched images.This source exhibits the hallmark colors of a distant galaxy -no discernible flux in a dropout band (we show stacked F606W+F814W as well as F115W images; the circle has a radius of 0.3 ) and a significant detection in the bluest detection band (F200W in this case).Very faint flux is visible (at ∼2.8σ significance when measured in a 0.2 diameter aperture) in F150W, which drives the redshift to z ∼ 12.The wide wavelength range of NIRCam allows this source to be well-detected in multiple filters, and in the imaging alone it is clear this source exhibits a blue spectral shape.Bottom) Same ordering as the top, for sky-uncertainty maps constructed from the variance of the readout noise, all using a linear scale from 0.33 to 3× the robustly-measured sky standard deviation in each band.The patchiness of the uncertainties is due to loss of exposure time when cosmic-rays are detected and rejected in the multiple readouts or in outlier rejection when combining the dithered exposures (the 2× larger original pixel scale of the long-wavelength channels results in larger patches than the short-wavelength channels).The uncertainty arrays show no excess in rejected pixels near the candidate galaxy.

Figure 2 .
Figure1shows 1.8 cutout images of this source at multiple wavelengths.This source shows the expected pattern for a high-redshift galaxy, with no significant flux in multiple dropout bands, with robust flux in redder bands.The very sharp break between F115W and F200W is consistent with a redshift of z > 11 (the faint detection in F150W pushes the redshift solution to z ∼ 12 as opposed to higher redshift).The advantage of JWST is clear here, as this source is well-detected in all five NIRCam filters redward of the break.This multi-band detection essentially eliminates the possibility of a spurious nature.Of note is that while persistence from previous observations affected several HST programs (see discussion inFinkelstein et al. 2022 andBagley et al. 2022), CEERS observed with the bluest fil-

Figure
Figure3.Left) The circles denote our fiducial photometry, with green and red denoting HST/ACS+WFC3 and NIRCam instruments, respectively.This SED exhibits the hallmark shape of a high-redshift galaxy, with several non-detections in blue filters, followed by significant detections with a blue spectral slope.The arrows denote 1σ upper limits.The F115W-F200W break color is >1.9 mag (2σ), which is sufficient to rule out all low-redshift solutions.The blue curve shows the best-fitting eazy model at z = 11.8.The orange curve shows the result if we force eazy to find a solution at z < 7.This model is unable to match the amplitude of the break as well as the slope redward of the break, and is correspondingly ruled out at high confidence (χ 2 low−z = 54.5, compared to 15.8 for the z = 11.8 solution).Right) Photometric redshift probability distribution functions for Maisie's Galaxy.The thick blue curve shows the fiducial PDF from eazy, which exhibits no low-redshift solution and a peak at z = 11.8 +0.2 −0.3 (the purple curve shows the eazy result without HST photometry, which prefers z = 12.0 and has a tail to z = 14).The remaining curves show the results from independent runs with Prospector, Bagpipes, Cigale and Dense Basis (see §5).All results significantly prefer a z > 12 solution, with all four codes finding best-fit redshifts nearly identical to eazy (z = 11.8 +0.2 −0.3 , z = 11.6 +0.2 −0.2 , z = 11.8 +0.4 −0.4 , and 11.9 +0.4 −0.4 , respectively).
Figure3.Left) The circles denote our fiducial photometry, with green and red denoting HST/ACS+WFC3 and NIRCam instruments, respectively.This SED exhibits the hallmark shape of a high-redshift galaxy, with several non-detections in blue filters, followed by significant detections with a blue spectral slope.The arrows denote 1σ upper limits.The F115W-F200W break color is >1.9 mag (2σ), which is sufficient to rule out all low-redshift solutions.The blue curve shows the best-fitting eazy model at z = 11.8.The orange curve shows the result if we force eazy to find a solution at z < 7.This model is unable to match the amplitude of the break as well as the slope redward of the break, and is correspondingly ruled out at high confidence (χ 2 low−z = 54.5, compared to 15.8 for the z = 11.8 solution).Right) Photometric redshift probability distribution functions for Maisie's Galaxy.The thick blue curve shows the fiducial PDF from eazy, which exhibits no low-redshift solution and a peak at z = 11.8 +0.2 −0.3 (the purple curve shows the eazy result without HST photometry, which prefers z = 12.0 and has a tail to z = 14).The remaining curves show the results from independent runs with Prospector, Bagpipes, Cigale and Dense Basis (see §5).All results significantly prefer a z > 12 solution, with all four codes finding best-fit redshifts nearly identical to eazy (z = 11.8 +0.2 −0.3 , z = 11.6 +0.2 −0.2 , z = 11.8 +0.4 −0.4 , and 11.9 +0.4 −0.4 , respectively).

Figure 4 .
Figure 4. Left) Plot showing our fiducial photometry of Maisie's Galaxy alongside best-fit SED models from the SED fitting code Prospector (red line; fiducial, see Table1), Bagpipes (black), Cigale (green dotted) and Dense Basis (blue dotted).Right) Posterior distributions of the key stellar population properties from all four codes.The panels show stellar mass, mass weighted age, dust attenuation and SFR averaged over the last 10 Myrs.The vertical dotted lines indicate the mean of the posteriors.Posteriors of attenuation are consistent between all four codes.Prospector prefers a younger age than the other three because of a recent burst in the SFH of this object.As a consequence of the burst, Prospector also estimates a lower SFR.The four estimates of the stellar mass posteriors exhibit significant overlap, though the median values differ by ± 0.2 dex.Future observations in the rest-optical with MIRI could break these degeneracies.

Figure 5 .
Figure5.Theoretical predictions from a range of simulations in the recent literature.The left panel shows the predicted number of sources at m < 27.3 (the brightness of our source) and z > 12 over our survey area of 34.5 sq.arcmin.The vertical axis and the values above each bar give the number predicted.Dark (light) shading denotes the value derived from models with (without) dust attenuation applied.The right panel shows these same theoretical predictions, now showing the cumulative number as a function of apparent magnitude.The bulk of these models predict that m ∼ 27 galaxies at z 12 are not highly likely, though theBehroozi & Silk (2015) model, which has no accelerated decline in the cosmic SFR density at z > 8, has the least tension.However, our detection of one source has a large Poisson (and cosmic variance) uncertainty (gray shading in the left panel; error bar in the right), so strong conclusions cannot yet be made.

Figure 6 .
Figure 6.A view on the luminosity function at z ≥10.The shaded light blue regions show observational constraints at z ∼ 10 (Bagley et al. 2022; Finkelstein et al. 2022), while the thin line shows the z = 9 DPL luminosity function from Bowler et al. (2020).The remaining points show z > 12 results from this work (triangle), the ground-based work of Harikane et al. (2022), and the recent JWST work of Naidu et al. (2022).The thick lines show empirical luminosity function models which evolve smoothly with redshift, with the solid line denoting a Schechter function evolved to z = 12 (Finkelstein 2016), and the dashed line a DPL evolved to z = 12(Finkelstein & Bagley 2022).The constraints placed by our observations on the faint-end of the luminosity function are consistent with a smooth decline out to z ∼ 12, though both our observations and those of brighter galaxies do lie above the bright-extension of these smoothly-declining functions.The shaded box shows the parameter-space reached by the upcoming COSMOS-Web survey, which will probe the very bright end at these redshifts.
from NASA through STScI ERS award JWST-ERS-1345.D. B. and M. H.-C. thank the Programme National de Cosmologie et Galaxies and CNES for their support.RA acknowledges support from Fondecyt Regular 1202007.

Table 1 .
Properties of Maisie's Galaxy