MIGHTEE/COSMOS-3D: The discovery of three spectroscopically confirmed radio-selected star-forming galaxies at z=4.9-5.6

This study presents the discovery and spectroscopic confirmation of three high-redshift ($z=4.9$–$5.6$) radio-selected star-forming galaxies using MIGHTEE and JWST COSMOS-3D data, demonstrating that their dust-independent radio emission traces intense star formation rates consistent with the main sequence and potentially triggered by mergers, thereby opening a new window for studying early Universe star formation without dust bias.

The Gist

Imagine the early Universe as a bustling, dusty construction site. For decades, astronomers have been trying to take a clear photo of the workers (stars) building the first galaxies. But there's a problem: thick clouds of cosmic dust are everywhere, blocking our view and hiding the true amount of work being done. It's like trying to count how many bricks are being laid in a factory through a thick fog.

Usually, astronomers use light from stars (optical/UV) or heat from dust (infrared) to count these bricks. But if the dust is too thick, the count is wrong.

Enter the "Radio Flashlight."

This new paper is like finding a special flashlight that can see through the fog. Radio waves don't care about dust. They pass right through. The authors of this study used a powerful radio telescope called MeerKAT (part of the MIGHTEE survey) to scan a patch of sky called COSMOS. They were looking for "High-Redshift Radio Sources" (HzRSs)—galaxies so far away they are seen as they were when the Universe was very young (about 1 billion years old).

The Big Discovery: Finding the "Quiet" Builders

In the past, when astronomers looked for these ancient radio galaxies, they only found the "rock stars" of the universe: massive, violent black holes (Active Galactic Nuclei or AGN) shooting out huge jets of energy. These are like the construction sites with loud, roaring cranes that are impossible to miss.

But this team was looking for something different: the quiet builders. They wanted to find galaxies where the radio signal came only from normal star formation, not from a crazy black hole. They were looking for the "crossover point"—the moment where a galaxy is just making stars so intensely that it starts glowing in radio waves, even without a black hole.

They found three.

These three galaxies are located at a time when the Universe was just 4.9 to 5.6 billion years old (wait, that's not right, the paper says redshift 4.9-5.6, which is actually when the universe was roughly 1 billion years old). They are incredibly faint in radio waves compared to the "rock star" black holes, but they are there.

The Detective Work: Putting the Puzzle Together

Finding these three was like finding three specific needles in a cosmic haystack. Here is how they did it:

1. The Radio Net: They cast a wide net using radio data to find faint signals.
2. The "Guessing Game" (Photometry): They looked at the same spots using the James Webb Space Telescope (JWST) and older telescopes. By looking at the colors of the light, they guessed the galaxies were very far away (high redshift).
3. The "ID Check" (Spectroscopy): This is the most important step. They used JWST's super-powerful "slitless spectroscopy" (a way of splitting light into a rainbow to see specific chemical fingerprints). They looked for a specific fingerprint called H-alpha, which is a signature of young, hot stars.
   • Analogy: Imagine trying to identify a person in a crowd. You see a silhouette (radio), you guess their height and hair color (photometry), but you need to hear them say their name (spectroscopy) to be 100% sure.
   • Result: Three galaxies said their names clearly. They were confirmed to be at redshift 4.9–5.6.

What Are These Galaxies Like?

The paper reveals some fascinating details about these three cosmic "babies":

• They are Starbursting: These aren't just normal galaxies; they are going through a "teenage rebellion" phase. They are forming stars at a rate of 100 to 1,800 suns per year. For comparison, our Milky Way only makes about 1 or 2 suns a year. They are in a massive, chaotic burst of creation.
• They are Messy (and Merging): Two of the galaxies look like they are in the middle of a cosmic dance, or even a crash. They have clumpy, irregular shapes, suggesting two galaxies are merging. This collision is likely the trigger for their massive star formation.
• No "Monster" Black Holes: A big question was: "Is there a giant black hole hiding inside, powering the radio waves?" The answer seems to be no. The galaxies look like extended, fluffy disks or clumps, not like a single bright dot (which a black hole would look like). The radio waves are coming from the stars themselves.
• The "Steep" Signal: The radio waves from these galaxies have a "steep" spectrum.
  • Analogy: Think of a radio station. As the signal travels through the expanding Universe, it loses energy. At these great distances, the "Cosmic Microwave Background" (the afterglow of the Big Bang) acts like a giant sponge, soaking up the energy of the electrons creating the radio waves. This makes the signal drop off quickly at higher frequencies, creating a "steep" slope. This confirms they are indeed very far away and very young.

Why Does This Matter?

This is a "first" in astronomy. Before this, we could only see the "loud" black holes in the early Universe. We couldn't see the "quiet" but intense star-forming galaxies because they were too faint for older radio telescopes.

By using the MIGHTEE survey (which is incredibly deep and sensitive) and JWST, the authors have opened a new window. They proved that we can now find galaxies powered only by star formation in the early Universe, without the dust getting in the way.

The Takeaway:
Imagine you were trying to study the history of a city. For years, you could only see the skyscrapers (black holes) because they were the only things visible through the fog. Now, you have a new tool that lets you see the entire city, including the busy factories and construction sites (star-forming galaxies) that were hidden before. This paper gives us our first clear look at how the earliest galaxies were building themselves, independent of the dust that usually hides them.

Technical Summary

Here is a detailed technical summary of the paper "MIGHTEE/COSMOS-3D: The discovery of three spectroscopically confirmed radio-selected star-forming galaxies at z = 4.9–5.6".

1. Problem Statement

High-redshift radio sources (HzRSs, $z > 4.5$) have historically been dominated by "radio-loud" Active Galactic Nuclei (AGN) with powerful jets. While radio observations offer a dust-independent probe of star formation (SF), previous surveys lacked the sensitivity to detect HzRSs at lower luminosities ($L_{1.3 \text{ GHz}} \lesssim 10^{24} \text{ W Hz}^{-1}$) where star formation is expected to dominate over AGN activity.

• The Gap: There is a lack of spectroscopically confirmed, radio-selected star-forming galaxies (SFGs) at $z > 4.5$. Most high-$z$ radio detections are AGN, and low-luminosity SFGs are difficult to identify due to the extreme depth required in radio continuum data.
• The Challenge: Distinguishing between low-luminosity AGN and extreme starbursts at these redshifts is difficult without spectroscopic confirmation and multi-wavelength constraints. Additionally, the "Little Red Dots" (LRDs) discovered by JWST remain ambiguous regarding their power sources (AGN vs. compact starbursts), with no confirmed radio counterparts yet.

2. Methodology

The authors utilized a multi-wavelength approach combining deep radio continuum data with advanced optical/IR imaging and spectroscopy.

• Radio Data: Used the MIGHTEE (MeerKAT International GHz Tiered Extragalactic Exploration) survey Data Release 1 (DR1) at 1.3 GHz in the COSMOS field. This provides high sensitivity ($\sim 2.4 \mu\text{Jy beam}^{-1}$) and resolution ($5.2''$).
• Candidate Selection:
  • Cross-matched MIGHTEE sources with ground-based LBG (Lyman-Break Galaxy) catalogues ($z \sim 3.5-5.5$) from Adams et al. (2023) and the COSMOS2025 catalogue (based on JWST NIRCam/MIRI and ground-based data).
  • Applied Spectral Energy Distribution (SED) fitting using LePhare and BAGPIPES to identify robust high-redshift candidates ($z_{\text{phot}} > 4.5$) and remove low-redshift interlopers.
  • This process yielded 18 HzRS candidates overlapping with the COSMOS-3D footprint.
• Spectroscopic Confirmation:
  • Utilized JWST COSMOS-3D, a Cycle 3 Large Programme providing wide-field slitless spectroscopy (WFSS) in the F444W filter.
  • Extracted 1D and 2D spectra using grizli to search for emission lines.
  • Confirmed redshifts via the detection of the H$\alpha$ emission line (rest-frame 6563 Å, observed at $\sim 3.9-4.5 \mu$m).
• Physical Property Derivation:
  • Calculated Star Formation Rates (SFRs) using three independent tracers: SED fitting (averaged over 10 Myr and 100 Myr), dust-corrected H$\alpha$ flux, and 1.3 GHz radio luminosity.
  • Analyzed morphology using JWST NIRCam images and Sérsic profile fitting to distinguish between AGN-dominated point sources and extended star-forming structures.
  • Measured radio spectral indices ($\alpha$) by combining MIGHTEE (1.3 GHz) and VLA-COSMOS (3 GHz) data.

3. Key Contributions

• First Spectroscopic Confirmation: This work presents the first spectroscopically confirmed radio-selected SFGs at $z > 4.5$ that are not dominated by AGN.
• Probing the Low-Luminosity Regime: The study successfully accesses a radio luminosity regime ($L_{1.3 \text{ GHz}} \approx 2\text{--}5 \times 10^{24} \text{ W Hz}^{-1}$) previously unexplored for high-$z$ sources, demonstrating that radio surveys can now detect star formation-dominated galaxies at these epochs.
• Validation of Multi-Tracer SFRs: It demonstrates broad agreement between SFRs derived from dust-independent radio emission, H$\alpha$ (optical), and SED fitting, validating radio as a robust dust-free SFR tracer in the early Universe.
• Morphological Insights: The discovery of complex, multi-component morphologies in these high-$z$ radio sources suggests that merger-induced starbursts are a viable mechanism for driving extreme star formation at $z \sim 5$.

4. Key Results

The study identified and confirmed three HzRSs:

1. MGTDR1 J100000.46+022256.2 ($z_{\text{spec}} = 5.54$)
2. MGTDR1 J095940.32+022331.5 ($z_{\text{spec}} = 5.20$)
3. MGTDR1 J100034.18+015733.8 ($z_{\text{spec}} = 4.90$)

Physical Properties:

• Radio Luminosity: All three sources have $L_{1.3 \text{ GHz}} \approx 2\text{--}5 \times 10^{24} \text{ W Hz}^{-1}$, which is at least two orders of magnitude fainter than previously known HzRSs.
• Star Formation Rates: SFRs range from $\sim 100$ to $1800 , M_\odot \text{ yr}^{-1}$.
  • SFRs derived from H$\alpha$ and radio luminosity are consistent with SED-derived SFRs (averaged over 100 Myr).
  • The sources lie on or 0.5–1.0 dex above the star-forming main sequence at $z=4\text{--}6$, indicating they are undergoing recent starbursts.
• Morphology:
  • None show a dominant point-source component (ruling out AGN dominance in UV/optical).
  • Two sources (J10000+02225 and J09594+02233) exhibit complex, multi-component morphologies (clumps), suggesting merger-triggered starbursts.
  • The third (J10003+01573) shows a disk-like morphology with a faint bulge.
• Spectral Indices: All three sources exhibit steep radio spectra ($\alpha_{1.3}^{3.0} \approx 0.96\text{--}1.68$). This steepening is attributed to Inverse-Compton (IC) scattering losses of relativistic electrons against the Cosmic Microwave Background (CMB), which scales as $(1+z)^4$ and dominates synchrotron losses at $z \gtrsim 3$.
• Detection Rate: Out of 18 photometric candidates, 3 were confirmed with H$\alpha$. This $\sim 17\%$ detection rate aligns with theoretical expectations given the different timescales probed by H$\alpha$ ($\sim 10$ Myr) versus radio emission ($\sim 100$ Myr).

5. Significance

• Dust-Independent Star Formation: These observations provide a critical, dust-unbiased window into star formation in the early Universe ($z > 4.5$), complementing UV and IR tracers which are often heavily obscured or limited by confusion.
• AGN vs. Starburst Discrimination: The study clarifies that at these low radio luminosities, star formation can be the primary power source, challenging the assumption that all high-$z$ radio sources are AGN. This has implications for understanding the "Little Red Dots" phenomenon.
• Cosmological Context: The results support the scenario where the most extreme starbursts at high redshift are often triggered by mergers.
• Future Prospects: This work lays the foundation for using deep radio surveys (like MIGHTEE and future SKA) combined with JWST spectroscopy to trace the assembly of the first massive structures and the evolution of the cosmic star formation history without the biases introduced by dust extinction.