The JADES Origins Field: A New JWST Deep Field in the JADES Second NIRCam Data Release

TL;DR

The paper presents the JWST JADES Origins Field (JOF) as a cornerstone deep field in GOODS-S, detailing the Cycle 2 deep medium-band imaging (15 NIRCam filters) and the Cycle 3 NIRCam WFSS spectroscopy that together push the redshift frontier toward $z>15$. It describes the Cycle 1 Coordinated Parallel setup, the observing strategies that maximize high-redshift discovery while mitigating interloper contamination, and the ultra-deep NIRSpec/MOS programs targeting the HUDF and JOF footprint. The results include the spectroscopic confirmation of a $z=14.18$ galaxy (JADES-GS-z14-0), non-detection of $z>15$ sources in Year 1, and a robust demonstration that the medium-band approach markedly improves high-redshift candidate validation, supported by dense-shutter spectroscopy efforts that yield hundreds of redshifts. The Year 1 data release for JOF provides deep imaging and catalogs across 15 filters, with comprehensive data reduction, astrometric alignment, and photometric redshifts, enabling a rich resource for studying early galaxy formation and planning future JWST observations.

Abstract

We summarize the properties and initial data release of the JADES Origins Field (JOF), the longest single pointing yet imaged with the James Webb Space Telescope (JWST). This field falls within the GOODS-S region about 8' south-west of the Hubble Ultra Deep Field (HUDF), where it was formed initially in Cycle 1 as a parallel field of HUDF spectroscopic observations within the JWST Advanced Deep Extragalactic Survey (JADES). This imaging was greatly extended in Cycle 2 program 3215, which observed the JOF for 5 days in six medium-band filters, seeking robust candidates for z>15 galaxies. This program also includes ultra-deep parallel NIRSpec spectroscopy (up to 91 hours on-source, summing over the dispersion modes) on the HUDF. Cycle 3 observations from program 4540 added 20 hours of NIRCam slitless spectroscopy and F070W imaging to the JOF. With these three campaigns, the JOF was observed for 380 open-shutter hours with NIRCam using 15 imaging filters and 2 grism bandpasses. Further, parts of the JOF have deep 43 hr MIRI observations in F770W. Taken together, the JOF is one of the most compelling deep fields available with JWST and a powerful window into the early Universe. This paper presents the second data release from JADES, featuring the imaging and catalogs from the year 1 JOF observations.

Figures (9)

• Figure 1: The layout of data sets in the GOODS-S field most immediate to this paper, overlaid on a background F160W image from HST CANDELS Whitaker19. JADES DR1 and DR2 are shown in green and grey, using the F356W exposure map. The parallel imaging in 1210 is the deepest portion; program 3215 extends this with 6 medium-bands. The 3215 secondary field location lies to the northwest; it overlaps other JADES NIRCam imaging from program 1286 (not shown). The NIRSpec MSA footprints for the two 3215 pointings are shown. We also display the footprints of the HUDF ACS field Beckwith06, FRESCO grism Oesch23, and JEMS medium-band Williams2023, as these are immediately supportive of the target selection for the 3215 NIRSpec observations. There are many other powerful data sets in this region, not shown for brevity!
• Figure 2: A demonstration of the ability of medium-bands to starkly differentiate between $z>15$ galaxies and $z\sim5$ interlopers. Starting from the photometry of the CEERS $z\approx16$ candidate Donnan23 scaled fainter to $S/N=7$ in F277W and F356W in our survey field, we perform fits to the Cycle 1 wide-band photometry, first for $z\sim16$ solutions and then for $z\sim5$ solutions, using the Prospector galaxy spectrum modeling code Johnson21. With the Cycle 1 bands alone, one can get acceptable fits at either redshift ($\Delta\chi^2=0.6$); these best-fit smoothed spectra are plotted. We then predict the medium-band photometry from each model. These results clearly distinguish the two hypotheses: the F182M and F210M bands indicate a well-localized break, while the F250M and F300M bands completely exclude the emission lines that the $z\sim5$ solution requires. Our photometric depth yields $\Delta\chi^2=43$, an overwhelming confirmation.
• Figure 3: We compare two filter strategies in their ability to reject a low-redshift hypothesis for a high-redshift galaxy. The first strategy is the Cycle 1 strategy of program 1210, focused on wide bands. The second strategy is the wide+medium-band strategy adopted with programs 1210 & 3215. We have rescaled the errors from the first strategy so as to match the total exposure time of the second, i.e., 2.68 times what program 1210 actually provides. In both, we take the high-redshift fit from Figure \ref{['fig:mb']} and shift it in redshift, holding the F277W signal-to-noise ratio constant (at 7, for the wide+medium strategy). At each true redshift, we fit with Prospector and report the $\Delta\chi^2$ between the best low-redshift false solution and the best high-redshift true solution. This quantifies the rejection of the low-redshift hypothesis. We find that at $z>15$, the wide+medium strategy provides a clear advantage, nearly 4 times better $\Delta\chi^2$ at fixed exposure time compared to strategy of a typical wide-band survey.
• Figure 4: F444W magnitude versus photometric redshift for the galaxies for which shutters were allocated for NIRSpec observations in program 3215. The color coding indicates the exposure time with the prism for each target.
• Figure 5: Layout of JADES Cycle 3 observations in the GOODS-South field. Cycle 1 JADES NIRCam imaging observations are shown as the background image. The JOF regions are shown with solid magenta line. The spectral coverage region of the Cycle 3 NIRCam WFSS GRISMC observations are shown as green shaded regions (for complete spectral coverage at 3--5 ) and solid orange lines (for partial spectral coverage across 2.4--5 ), respectively. The dispersion direction of GRISMC is indicated as the gray arrow. To the west of the NIRCam footprints, the coordinated MIRI imaging parallel observations are indicated by the solid orange (F560W), red (F770W and F1000W) and maroon lines (F1500W and F2100W), respectively. For comparison we also plot the footprints of CANDELS HST F160W imaging Whitaker19, SMILES MIRI multi-wavelength imaging (Cycle-1 GTO 1207; alberts2024, riekeg2024), and FRESCO NIRCam imaging and F444W GRISMR WFSS Oesch23.