🔭 astrophysics

JWST Advanced Deep Extragalactic Survey (JADES) Data Release 5: MIRI Coordinated Parallels in GOODS-S and GOODS-N

This paper presents the fifth data release of the JWST Advanced Deep Extragalactic Survey (JADES), detailing the reduction of ultra-deep and medium-depth MIRI coordinated parallel imaging in the GOODS-S and GOODS-N fields and releasing fully processed mosaics with forced photometry to advance the study of early galaxy populations.

Original authors: Stacey Alberts, Daniel J. Eisenstein, Andrew J. Bunker, Emma Curtis-Lake, Qiao Duan, Kevin Hainline, Ryan Hausen, Jakob M. Helton, Zhiyuan Ji, Benjamin D. Johnson, Jianwei Lyu, Jane Morrison, Pablo G.

Published 2026-01-23

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Stacey Alberts, Daniel J. Eisenstein, Andrew J. Bunker, Emma Curtis-Lake, Qiao Duan, Kevin Hainline, Ryan Hausen, Jakob M. Helton, Zhiyuan Ji, Benjamin D. Johnson, Jianwei Lyu, Jane Morrison, Pablo G. Perez-Gonzalez, George H. Rieke, Marcia Rieke, Pierluigi Rinaldi, Brant Robertson, Yang Sun, Sandro Tacchella, Christina C. Williams, Christopher N. A. Willmer, Zihao Wu

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

The Big Picture: A New Lens on the Early Universe

Imagine the James Webb Space Telescope (JWST) as a time machine that lets us look back at the very first galaxies that formed after the Big Bang. While the telescope's near-infrared cameras (NIRCam) have been great at taking "family photos" of these ancient galaxies, there was a missing piece of the puzzle: the Mid-Infrared Instrument (MIRI).

Think of MIRI as a special pair of glasses that sees heat and dust. It allows astronomers to see parts of the galaxy that are hidden from the other cameras, like the "soul" of the galaxy where stars are being born or where black holes are hiding. This paper is the instruction manual and the "photo album" for a specific set of deep-space photos taken by MIRI in two famous patches of sky: GOODS-S and GOODS-N.

The Mission: Two Types of "Long Exposures"

The team took two different kinds of pictures, which they call "parallels." In astronomy, a "parallel" observation is like taking a photo with one camera while the telescope is already pointed at a different spot for another camera. It's like taking a selfie with your phone while someone else is filming a movie with a cinema camera; you get extra footage without using extra time.

The "Ultra-Deep" Photos (The Long Haul):
- What they did: They stared at a tiny patch of sky (about the size of a grain of sand held at arm's length) for an incredibly long time—about 155,000 seconds (roughly 43 hours) in a single color filter (7.7 microns).
- The Analogy: Imagine trying to hear a whisper in a noisy room. If you listen for 10 seconds, you might hear nothing. But if you listen for 43 hours, you can hear the faintest whisper. This is what they did with light. They listened so long they could see galaxies that are 13.8 billion years old, including a record-breaking galaxy that existed just 300 million years after the Big Bang.
- The Result: They found a "Cosmic Rose" (a cluster of galaxies) and confirmed the existence of a galaxy so old it was previously only a theory.
The "Medium-Depth" Photos (The Wide Net):
- What they did: They took shorter pictures (5 to 15 hours) over a larger area, using three different color filters (7.7, 12.8, and 15 microns).
- The Analogy: If the Ultra-Deep photos are like zooming in on a single ant to see its legs, the Medium-Depth photos are like taking a wide-angle shot of the whole forest to see how the ants are moving around. This helps them understand how common these ancient galaxies are.

The Challenge: Cleaning Up the "Static"

Taking these photos wasn't easy. The telescope's detectors are like sensitive microphones that pick up "static" from the universe and the telescope itself. The paper details how the team fixed three major problems:

The "Ghost" Problem (Persistence):
- The Issue: If a very bright star shines on the detector, it leaves a "ghost" image behind, like a camera flash leaving a spot on your retina. This ghost can linger and mess up the next photo, even if the bright star isn't there anymore.
- The Fix: The team created a custom "eraser." They manually identified which pixels got "burned" by bright stars and told the computer to ignore those specific spots in future photos, effectively masking the ghosts so they don't ruin the picture.
The "Background Noise" Problem:
- The Issue: The mid-infrared sky isn't truly black; it has a faint, glowing background (like the glow of city lights at night).
- The Fix: They built a "super-background" map. By stacking many photos together and removing the actual galaxies, they figured out exactly what the background noise looked like and subtracted it, leaving only the clear, crisp images of the galaxies.
The "Cosmic Ray" Problem:
- The Issue: Space is full of high-energy particles (cosmic rays) that hit the detector like tiny bullets, creating random bright spots. Because they took such long exposures, almost every pixel on the camera was hit by a cosmic ray at some point.
- The Fix: They took so many photos of the same spot (over 100 times per area) that they could mathematically filter out the "bullets." If a pixel was bright in only one photo but dark in the other 100, they knew it was just a cosmic ray and ignored it.

The Result: A New Data Release

This paper announces JADES Data Release 5. Think of this as the team uploading their raw, cleaned-up, and organized photo album to a public library (the MAST archive) for anyone to use.

What's in the box: They are releasing the final, high-quality maps of the sky in three different mid-infrared colors.
Why it matters: These maps are the deepest mid-infrared images ever taken. They provide the "rest-frame near-infrared" view of early galaxies, which is crucial for understanding how heavy elements (metals) and dust formed in the early universe.
The "Forced Photometry": They also added a special feature where they measured the brightness of every single galaxy they found with the other cameras (NIRCam) using these new MIRI maps. It's like having a complete ID card for every galaxy, listing its size, brightness, and age.

Summary

In short, this paper says: "We took the deepest, most detailed mid-infrared photos of the early universe ever made. We developed special cleaning tools to remove ghosts, background noise, and cosmic ray scratches. Now, we are giving these cleaned-up maps to the world so scientists can study the very first galaxies that ever existed."

Technical Summary: JWST Advanced Deep Extragalactic Survey (JADES) Data Release 5: MIRI Coordinated Parallels in GOODS-S and GOODS-N

Problem and Context
The James Webb Space Telescope (JWST) has revolutionized the study of early galaxy evolution by accessing rest-frame optical and near-infrared wavelengths. However, critical spectral features, such as H $\alpha$ , H $\beta$ , and [OIII] emission lines, shift into the observed mid-infrared (mid-IR) at redshifts $z \gtrsim 6.5$ . While the Near-Infrared Camera (NIRCam) has led the charge in building a new view of the early Universe, the Mid-Infrared Instrument (MIRI) is required to probe these rest-frame optical features and constrain stellar populations, dust content, and active galactic nuclei (AGN) at high redshifts. Previous analyses indicated that without mid-IR constraints, stellar mass and star formation rate (SFR) estimates for high-redshift galaxies could be overestimated by factors of 2–3 due to emission line boosting in shorter wavelength bands. Furthermore, deep mid-IR imaging is necessary to identify unique populations, such as "Little Red Dots" and obscured AGN, and to resolve hot dust structures.

Despite the scientific necessity, deep MIRI surveys face practical challenges: the instrument's small field-of-view ( $\sim$ 2.3 arcmin $^2$ ) and the steeply rising thermal background at wavelengths $>15$ $\mu$ m limit the area that can be surveyed efficiently. Additionally, data reduction for deep MIRI imaging presents specific technical hurdles, including the management of warm/hot pixels, cosmic ray impacts on long integrations, and artifacts caused by persistence from saturating sources.

Methodology
This paper presents the data reduction and release of the MIRI coordinated parallels component of the JADES program, conducted in the GOODS-S and GOODS-N fields. The survey design comprises two main components:

Ultra-Deep Parallel (GS-Deep): A 2 $\times$ 2 mosaic in the F770W filter over $\sim$ 10 arcmin $^2$ in GOODS-S, achieving a total on-source exposure time of 155 ks (43 hours) per pointing. This was observed in parallel with NIRCam ultra-deep imaging centered on the Hubble Ultra Deep Field (HUDF).
Medium-Depth Parallels: Surveys in both GOODS-N (GN-Medium) and GOODS-S (GS-Medium) covering non-contiguous areas of $\sim$ 26 arcmin $^2$ total. These utilized F770W, F1280W, and F1500W filters with exposure times ranging from 5 to 15 ks.

The data processing pipeline builds upon the official JWST Calibration Pipeline (V1.16.1) but incorporates several custom steps to optimize the dataset:

Persistence Correction: A new step was introduced to address artifacts caused by persistence from saturating sources. The authors manually flagged saturated pixels in time-ordered uncalibrated (UNCAL) files and propagated these masks to subsequent exposures (including different filters) to prevent "donut" artifacts and negative persistence signals in the final mosaics.
Warm/Hot Pixel Masking: Custom routines were developed to identify and mask warm/hot pixels by median stacking contemporary exposures, as these defects increase over time and are not always captured in the standard bad pixel masks.
Background Subtraction: The authors employed a "super background" subtraction technique. This involves constructing a robust background model by masking sources and image edges across multiple independent pointings and taking a pixel-wise median. This method mitigates "shadowing" effects caused by small dither steps and accounts for temporal variations in the mid-IR sky background.
Astrometry: Astrometric corrections were applied pointing-by-pointing, registering the data to the JADES/NIRCam catalog (itself tied to GAIA) with a typical accuracy of 0.01 $''$ , or to HST/CHARGE catalogs where NIRCam overlap was absent.

Key Results

Data Release: The fully reduced JADES/MIRI mosaics are released as part of JADES Data Release 5 (DR5). The release includes 36.4, 25.8, and 22 arcmin $^2$ in the F770W, F1280W, and F1500W filters, respectively.
Depth and Sensitivity: The GS-Deep parallel reaches a 5 $\sigma$ point source sensitivity of 19 nJy (28.2 AB mag) at F770W. The medium-depth surveys reach sensitivities ranging from 90 nJy (26.5 AB) to 510 nJy (24.6 AB).
Performance vs. Predictions: The achieved depths are approximately 1.3 to 2.5 times deeper than predictions from the Exposure Time Calculator (ETC). The authors attribute this to the effectiveness of their custom background subtraction and the high redundancy of the dither patterns.
Cosmic Ray Mitigation: Despite using long integration times (up to $\sim$ 1300 s) that theoretically expose $\sim$ 80% of pixels to cosmic rays, the final mosaics show no significant degradation in depth compared to surveys using shorter ramps. The high redundancy (114 exposures per pointing) and the custom processing pipeline effectively mitigate the impact of cosmic rays.
Scientific Validation: The paper highlights the utility of these data through specific examples, including the detection of a $z \sim 14$ galaxy (JADES-GS-z14) boosted by [OIII]+H $\beta$ emission in the F770W filter, and the placement of strong upper limits on a fainter $z \sim 14$ candidate (JADES-GS-z14-1), constraining its nebular emission properties.

Significance
The paper claims that this dataset represents a critical window into the rest-frame near-infrared and optical regimes of the Epoch of Reionization and beyond. By providing ultra-deep and medium-depth mid-IR imaging with robust photometric constraints, the JADES/MIRI parallels enable:

Accurate Physical Characterization: Correcting stellar mass and SFR estimates for high-redshift galaxies that were previously biased by emission line contamination in NIRCam bands.
Discovery of Rare Populations: Facilitating the study of extremely red galaxies, obscured AGN, and "Little Red Dots" through their unique mid-IR spectral energy distributions.
Resolved Analysis: Leveraging MIRI's spatial resolution ( $\sim$ 0.2 $''$ –0.85 $''$ ) to resolve structures such as hot dust in Lyman- $\alpha$ leakers.
Synergy with NIRCam: The coordinated parallel strategy ensures that deep mid-IR data is available for the vast majority of sources detected in the deep NIRCam surveys, providing a comprehensive multi-wavelength view of early galaxy populations.

The authors emphasize that while the area covered is small, the depth and quality of these mosaics, combined with the growing number of ultra-deep MIRI spectroscopic programs, constitute the current best resource for understanding the physical properties of the earliest galaxies.

The Big Picture: A New Lens on the Early Universe

The Mission: Two Types of "Long Exposures"

The Challenge: Cleaning Up the "Static"

The Result: A New Data Release

Summary

More like this