🔭 astrophysics

JWST Advanced Deep Extragalactic Survey (JADES) Data Release 5: Photometric Catalog

JADES Data Release 5 presents comprehensive photometric catalogs derived from 35 deep JWST and HST imaging mosaics in the GOODS fields, featuring advanced methodologies for source detection, deblending, multi-aperture photometry, uncertainty modeling, and photometric redshift estimation to supersede previous releases.

Original authors: Brant E. Robertson, Benjamin D. Johnson, Sandro Tacchella, Daniel J. Eisenstein, Kevin Hainline, Stacey Alberts, Santiago Arribas, William M. Baker, Andrew J. Bunker, Alex J. Cameron, Stefano Carniani

Published 2026-01-23

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Brant E. Robertson, Benjamin D. Johnson, Sandro Tacchella, Daniel J. Eisenstein, Kevin Hainline, Stacey Alberts, Santiago Arribas, William M. Baker, Andrew J. Bunker, Alex J. Cameron, Stefano Carniani, Courtney Carreira, Jacopo Chevallard, Chiara Circosta, Emma Curtis-Lake, A. Lola Danhaive, Qiao Duan, Eiichi Egami, Ryan Hausen, Jakob M. Helton, Zhiyuan Ji, Roberto Maiolino, Pablo G. Pérez-González, Dávid Puskás, Marcia Rieke, Pierluigi Rinaldi, Fengwu Sun, Yang Sun, Hannah Übler, James A. A. Trussler, Natalia C. Villanueva, Lily Whitler, Christina C. Williams, Christopher N. A. Willmer, Chris Willott, Zihao Wu, Yongda Zhu

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the universe as a massive, dark library filled with billions of books (galaxies). For years, we could only read the titles of the books on the top shelves. But with the James Webb Space Telescope (JWST), we finally have a pair of super-powered glasses that let us see the tiny, dusty books hidden in the deepest, darkest corners of the library, some of which were written just 300 million years after the universe began.

This paper is the "Librarian's Guide" for the JADES Data Release 5 (DR5). It doesn't just show you the pictures; it explains exactly how the team built a massive, organized catalog of every single object they found in two specific deep-sky neighborhoods (called GOODS-North and GOODS-South).

Here is how they did it, explained simply:

1. The Detective Work: Finding the Needles in the Haystack

The images from the telescope are incredibly deep, meaning they show very faint objects. But they are also crowded, like a busy city street at night where streetlights (bright stars) and neon signs (nearby galaxies) make it hard to see the small, dim houses (distant galaxies) in between.

The "Detection" Image: To find these faint objects, the team stacked up many long-wavelength images (like looking through a red filter) to create a single, super-bright "detection map." This is like turning up the volume on a quiet song so you can hear the melody.
The "Deblending" Image: Once they found a blob of light, they needed to know if it was one big galaxy or three small ones huddled together. They used sharper, shorter-wavelength images (like a high-resolution zoom lens) to separate them.
The "Deblending" Algorithm: Think of this as a smart robot that looks at a messy pile of overlapping shadows and says, "Okay, this shadow belongs to the tree, and this one belongs to the car." The paper describes a new, custom-built robot that is very good at untangling these messy overlaps without accidentally tearing the objects apart.

2. Measuring the Objects: The "Gaussian" Recipe

Once the team isolated a galaxy, they needed to measure its size and shape.

The Analogy: Imagine trying to measure the size of a cloud. Is it the whole fluffy shape, or just the dense center?
The Method: The team used a new, fast mathematical trick called "Gaussian regression." Instead of just guessing the edge of the galaxy, they fit a smooth, bell-shaped curve to the light, like fitting a perfect cookie cutter to a blob of dough. This tells them exactly where the galaxy starts and stops, allowing them to measure its "Kron" size (a standard astronomical way to define a galaxy's total area).

3. The "Common-PSF" Magic: Leveling the Playing Field

Telescopes don't see everything with the same sharpness. Some filters (colors) make stars look like sharp points; others make them look like fuzzy blobs.

The Problem: If you compare a sharp photo to a blurry one, your measurements will be wrong.
The Solution: The team created "Common-PSF" mosaics. Imagine taking all the photos and running them through a special blender that makes every single image look exactly as blurry as the blurriest one. Now, every galaxy looks the same size and shape across all colors, making it fair to compare them.

4. Counting the Light: Photometry

Now that they know where the galaxies are and how big they are, they count the light.

Circular Apertures: They put a series of invisible circles of different sizes over every galaxy and counted the light inside. It's like measuring how much water is in a bucket at different depths.
Curve of Growth: This is a new feature in this release. Instead of just one measurement, they measured how the light builds up as you make the circle bigger and bigger, step-by-step. It's like watching a sponge soak up water; you can see exactly when the sponge is fully saturated. This helps astronomers understand the galaxy's structure.

5. The "Uncertainty" Map: Knowing How Sure We Are

In science, it's not enough to say "we found a galaxy." You have to say "we found it, and we are 95% sure it's real."

The Challenge: The telescope images have "noise" (static) that isn't random; it's connected, like a ripple in a pond. If you measure a small area, the noise is different than if you measure a large area.
The Solution: The team built a new mathematical model that acts like a weather map for uncertainty. It predicts exactly how much "static" there is for any size of measurement in any part of the image. This ensures that when they say a galaxy is faint, they aren't just guessing; they have a precise error bar.

6. Guessing the Distance: Photometric Redshifts

Since the galaxies are too far away to measure their distance with a ruler, the team has to guess based on their color.

The Analogy: Think of a distant siren. As it moves away, the pitch drops (redshift). Similarly, as galaxies move away, their light stretches and turns redder.
The Method: The team compared the colors of the galaxies to a library of "template" galaxy colors. They found the best match to estimate how far away the galaxy is. They did this twice: once using small, sharp circles (good for finding the galaxy) and once using the "Common-PSF" blurry images (good for measuring the total light accurately).

7. The Final Product: The Catalog

The result of all this work is a massive spreadsheet (the catalog) containing about 500,000 objects.

For every single object, the catalog lists:
- Where it is (coordinates).
- How bright it is in 35 different colors (filters).
- How big it is.
- How far away it is likely to be.
- A "quality flag" telling you if the measurement might be tricky (e.g., if it's right next to a bright star or near a chip gap in the camera).
- A "hash code" that tells you exactly which telescope programs contributed to the data for that specific pixel.

Summary

This paper is the instruction manual for the JADES DR5 catalog. It explains how the team took raw, messy telescope data, cleaned it up, separated the tangled galaxies, measured them with extreme precision, and organized them into a public library that anyone can use to study how the universe grew from its infancy to today. They didn't just take a picture; they built a highly accurate, searchable map of the early universe.

Technical Summary: JADES Data Release 5 Photometric Catalog

Problem and Context
The James Webb Space Telescope (JWST) has revolutionized the study of the early universe, extending the observable horizon to the first 300 million years of cosmic time. However, the sheer depth and complexity of JWST imaging in the Great Observatories Origins Deep Survey (GOODS) North and South fields present significant challenges for source detection, deblending, and photometric measurement. The James Webb Advanced Deep Extragalactic Survey (JADES) aims to provide a uniform, deep dataset to study galaxy formation and evolution. While previous releases (e.g., DR3) provided foundational catalogs, the integration of new data from JADES, the Next Generation Deep Extragalactic Exploratory Public (NGDEEP) survey, and various other public programs (MIRI Deep Imaging Survey, JEMS, PEARLS, etc.) necessitates a comprehensive update. The primary technical challenge addressed in this paper is the construction of a robust, public photometric catalog (Data Release 5, or DR5) that accurately handles the complex point-spread function (PSF) structure of JWST, correlated noise in heterogeneous mosaics, and the deblending of crowded, faint sources across 35 space-based filters.

Methodology
The JADES DR5 catalog is constructed from 35 space-based imaging mosaics combining JWST/NIRCam, JWST/MIRI, HST/ACS, and HST/WFC3 data, totaling approximately 1,250 hours of JADES imaging plus extensive public observations. The methodology follows a rigorous pipeline:

Detection and Deblending:
- Detection Images: Constructed as inverse-variance-weighted signal-to-noise ratio (SNR) mosaics using long-wavelength NIRCam filters (and MIRI/HST where NIRCam is absent). Noise mosaics are derived from pipeline ERR images but filtered to prevent artificial SNR spikes caused by pixel-level noise variations.
- Deblending Images: Created from short-wavelength NIRCam filters ( $\lambda \approx 2\mu$ m) to leverage higher resolution.
- Algorithms: A custom segmentation-based approach identifies sources. Initial blended segments are refined using morphological operations (filling holes, erosion, dilation). A deterministic deblending algorithm identifies local SNR peaks in the deblending image, assigning pixels based on iso-significance contours. This process includes specific heuristics to recover faint satellites via edge detection (Scharr filter) and extremely red sources that may be missed in blue deblending images.
- Curation: An interactive web tool (FitsMap) allows for manual review and correction of spurious fragments (e.g., shredded diffraction spikes), with changes propagated via a pixel-level source ID restoration method to ensure consistency across data versions.
Source Characterization:
- Centroiding: Determined from SNR-weighted barycenters within local dominance regions.
- Gaussian Regression: A novel fast two-dimensional Gaussian regression method is applied to detection image profiles to determine source sizes, orientations, and unscaled Kron radii. This provides a more direct estimate of source properties than standard moment calculations.
Photometry and Uncertainty Modeling:
- Apertures: Forced circular aperture photometry (radii $r = 0.1''$ to $0.5''$ ) and ellipsoidal Kron photometry are performed on both native-resolution and common-PSF mosaics.
- PSF Handling: Empirical PSFs are derived for HST, while model PSFs (using STPSF) are generated for JWST NIRCam and MIRI. Common-PSF mosaics are created by convolving subregion mosaics to a target F444W PSF, accounting for varying position angles and diffraction spikes.
- Uncertainty Framework: A new pixel-level regression framework models photometric uncertainties as a function of aperture size and local mosaic properties. This accounts for correlated noise in heterogeneous mosaics by fitting a power-law scaling ( $\sigma \propto N^\beta$ ) to random aperture flux distributions, using the pipeline variance image as a template for single-pixel uncertainty.
Photometric Redshifts:
- Generated using the EAZY code with template-based fitting. Two catalogs are produced: one using small-aperture (CIRC1) photometry on unconvolved images (optimized for high SNR and dropout selection) and one using Kron photometry on common-PSF images (optimized for unbiased colors and total flux).

Key Contributions and Results

Catalog Scope: The release includes photometry, curve-of-growth measurements, and redshifts for $\sim$ 500,000 sources across 35 filters.
Improved Deblending: The custom deblending algorithm successfully isolates compact, distant sources and separates them from foreground interlopers and complex interacting systems, outperforming previous methods in crowded fields.
Uncertainty Modeling: The new regression framework for uncertainty scaling provides a more accurate representation of photometric errors in heterogeneous JWST mosaics compared to previous methods that assumed uniform noise properties.
Photometric Consistency: Comparisons with the JADES DR3 catalog show flux differences typically better than 1%, with the largest discrepancies occurring in the bluest filters (F090W).
Redshift Accuracy: Validation against a compilation of $\sim$ 12,500 spectroscopic redshifts yields a normalized mean absolute deviation (NMAD) of $\sigma_{NMAD} \approx 0.028-0.029$ for GOODS-S and $0.044-0.045$ for GOODS-N. Catastrophic outliers are significantly reduced at $z > 6$ due to the Lyman- $\alpha$ break shifting into NIRCam coverage.
Data Accessibility: The catalogs are released via the Mikulski Archive for Space Telescopes (MAST) and an interactive FitsMap interface, including a "bithash" integer field that encodes the specific JWST programs contributing to each source's data, facilitating proper citation.

Significance
The paper positions the JADES DR5 photometric catalog as the superseding standard for the GOODS fields, replacing previous releases. It provides the necessary infrastructure for studies ranging from the assembly of stellar mass and the growth of structure to the identification of rare objects at both high and low redshift. By introducing robust methods for handling correlated noise, complex PSFs, and source deblending, the release enables precise measurements of galaxy properties from the local universe to the epoch of reionization. The authors emphasize that the catalog is designed to be a foundational resource for the community, supporting statistical analyses of galaxy populations and the discovery of faint sources, while the inclusion of curve-of-growth data allows users to define custom flux measurements tailored to specific scientific needs.