ALMA REBELS: Cosmic Dawn Galaxy Insights

• REBELS is a comprehensive survey targeting UV-bright, massive galaxies at z=6.5–9.5, characterizing their ISM, dust, star formation, gas, and metallicity.
• It employs full ALMA spectral scans of [CII] 158μm and [OIII] 88μm lines along with dust continuum observations to derive precise redshifts and ISM mass estimates.
• The program reveals rapid baryonic assembly and early dust/metal enrichment, establishing key empirical scaling relations and confirming significant obscured star formation at cosmic dawn.

The ALMA Large Program REBELS (Reionization Era Bright Emission Line Survey) represents the most comprehensive, statistically significant effort to characterize the interstellar medium (ISM), dust content, star formation, gas reservoirs, and metallicity in UV-bright, massive galaxies during the epoch of reionization ($6.5 < z < 9.5$). Executed using the Atacama Large Millimeter/submillimeter Array (ALMA), REBELS combines full spectral scans of far-infrared atom/ion cooling lines ([C II] $158\,\mu$m and [O III] $88\,\mu$m) and the dust continuum, systematically targeting 40 of the brightest Lyman-break galaxies over $\sim7$ deg$^2$ of legacy extragalactic fields. These observations provide precise spectroscopic redshifts, ISM mass measurements, obscured star-formation rates, and rest-optical nebular line data for the first statistically representative sample of massive galaxies in the EoR. REBELS, through its ALMA and JWST follow-up programs, has established empirical scaling relations, constraints on dust and metal enrichment, and direct evidence for rapid baryonic assembly and morphological complexity within 700 Myr of the Big Bang.

1. Survey Design, Target Selection, and Observational Strategy

The scientific goal of REBELS is to provide a census of the dust, gas, and metal content—and the star-formation activity—of the most massive, UV-luminous galaxies at $z=6.5-9.5$, thereby tracing the physical drivers of galaxy evolution at cosmic dawn (Bouwens et al., 2021, Rowland et al., 2024). Key elements include:

• Target Selection: 40 principal targets with $M_{\mathrm{UV}} < -21.3$ mag are pre-selected using deep ground-based and space (HST, Spitzer/IRAC) imaging, across COSMOS/UltraVISTA, VIDEO/UDS, and CANDELS fields. Each candidate is photometrically vetted via three independent SED-fitting codes, ensuring robust Lyman-break identification and redshift constraints, with final photometric redshifts $6.5 < z_{\mathrm{phot}} < 9.5$. Seven additional pilot sources are incorporated for enhanced statistics (Bouwens et al., 2021, Bowler et al., 2023).
• ALMA Observations:
  • Full spectral scans in Band 6 for [C II] $158\,\mu$m ($z \lesssim 8.5$) and Band 7/8 for [O III] $158\,\mu$0m ($158\,\mu$1). Observations are tuned per-target to achieve $158\,\mu$2 redshift likelihood coverage.
  • Integration times are tailored to achieve $158\,\mu$3 line sensitivity of $158\,\mu$4 at $158\,\mu$5, with synthesized beams $158\,\mu$6.
  • Rest-frame $158\,\mu$7m (and select $158\,\mu$8m) continuum is measured simultaneously with line scans, reaching $158\,\mu$9 continuum detection thresholds of $88\,\mu$0.
  • Calibration, imaging, and cleaning are performed in CASA, with custom pipelines for continuum subtraction and optimal map construction (Bouwens et al., 2021, Bowler et al., 2023, Inami et al., 2022).
• Sample Properties: REBELS spans $88\,\mu$1, star formation rates (SFRs) $88\,\mu$2 (unobscured), and $88\,\mu$3 to $88\,\mu$4 (Topping et al., 2022, Rowland et al., 17 Jan 2025, Dayal et al., 2022).

2. ISM Diagnostics: Dust Continuum, Molecular Gas, and Metallicity

REBEL’s unprecedented multi-wavelength data, including deep ALMA and JWST NIRSpec IFU follow-up (Fisher et al., 13 Nov 2025, Rowland et al., 17 Jan 2025), enable robust measurement of the ISM content and physical conditions:

• [C II] $88\,\mu$5m as a Redshift and Gas Mass Tracer:
  • [C II] is detected at $88\,\mu$6 in $88\,\mu$725 of 40 targets, delivering systemic redshifts with $88\,\mu$8 (Bouwens et al., 2021, Schouws et al., 2022).
  • Molecular gas mass is inferred via empirical calibration (Aravena et al., 2023):

  $88\,\mu$9

  yielding $\sim7$0, median gas-to-stellar mass ratio $\sim7$1, and depletion timescales $\sim7$2 Gyr (Aravena et al., 2023). - Dynamical mass estimates from resolved line widths and $\sim7$3 confirm high gas fractions ($\sim7$4; (Rowland et al., 2024, Hygate et al., 2023)).

• Dust Continuum and Temperatures:

  • 16/42 galaxies are detected in the $\sim7$5m continuum; dust continuum detections correlate with redder UV slopes.
  • Dust masses $\sim7$6 and inferred temperatures $\sim7$7 K are derived using a combined line+continuum method (Sommovigo et al., 2022).
  • Infrared luminosities span $\sim7$8, corresponding to SFR$\sim7$9/yr, with a median $^2$0 for $^2$1K (Inami et al., 2022, Barrufet et al., 2023).
• Metallicity with JWST IFU:
  • NIRSpec IFU spectroscopy reveals $^2$2 (up to near-solar), and ionization parameters $^2$3 (Rowland et al., 17 Jan 2025).
  • At $^2$4, the mass–metallicity relation ($^2$5; normalization $^2$6) is already in place (Rowland et al., 17 Jan 2025).

3. Star Formation: Rates, Histories, and Obscured Fractions

The combination of ALMA (dust + fine-structure lines) and JWST (rest-optical nebular lines) constrains both obscured and unobscured star formation, as well as star formation histories (SFHs):

• Total, Obscured, and Unobscured SFRs:
  • SFR$^2$7 from UV luminosity: $^2$8 $^2$9
  • SFR$z=6.5-9.5$0 from $z=6.5-9.5$1: $z=6.5-9.5$2
  • The typical fraction of obscured star formation $z=6.5-9.5$3 rises with stellar mass, $z=6.5-9.5$4 for $z=6.5-9.5$5 (Algera et al., 2022, Bowler et al., 2023).
  • In the most massive systems (e.g., REBELS-25), $z=6.5-9.5$6 (Hygate et al., 2023, Rowland et al., 2024).
• Specific SFR and SFHs:
  • Median sSFR for $z=6.5-9.5$7 is $z=6.5-9.5$8 (constant SFH fits), decreasing to $z=6.5-9.5$9 for non-parametric SFHs, with an evolutionary scaling $M_{\mathrm{UV}} < -21.3$0 (Topping et al., 2022).
  • JWST-ALMA samples show steeply rising SFHs ($M_{\mathrm{UV}} < -21.3$1 Myr) implying SFRs can be overestimated by $M_{\mathrm{UV}} < -21.3$2 if constant SFH is assumed (Fisher et al., 13 Nov 2025).
  • Updated SFR--luminosity calibrations for rising SFHs: $M_{\mathrm{UV}} < -21.3$3 yr$M_{\mathrm{UV}} < -21.3$4 erg$M_{\mathrm{UV}} < -21.3$5 s Hz (Fisher et al., 13 Nov 2025).
• Obscured Cosmic Star Formation Rate Density:
  • Integration of the IR luminosity function yields $M_{\mathrm{UV}} < -21.3$6, $M_{\mathrm{UV}} < -21.3$7 of UV-based estimates but representing $M_{\mathrm{UV}} < -21.3$8 of the total at $M_{\mathrm{UV}} < -21.3$9 (Barrufet et al., 2023, Algera et al., 2022).

4. ISM Morphology, Kinematics, and Scaling Relations

A central result from REBELS is the direct spatial characterization of gas, dust, and stellar emission, including kinematic decomposition at kiloparsec to sub-kiloparsec scales:

• Morphology and Sizes:
  • Stacked [C II] effective radius at $6.5 < z_{\mathrm{phot}} < 9.5$0 is $6.5 < z_{\mathrm{phot}} < 9.5$1 kpc; the [C II] region is $6.5 < z_{\mathrm{phot}} < 9.5$2 larger than both dust continuum ($6.5 < z_{\mathrm{phot}} < 9.5$3 kpc) and rest-UV emission ($6.5 < z_{\mathrm{phot}} < 9.5$4 kpc) (Fudamoto et al., 2022).
  • No significant evolution in [C II] size with redshift over $6.5 < z_{\mathrm{phot}} < 9.5$5, indicating morphologically gas-dominated systems persist through the EoR (Fudamoto et al., 2022).
  • Spatial offsets of $6.5 < z_{\mathrm{phot}} < 9.5$6 between rest-UV and FIR emission peaks indicate decoupled phases of star formation; clumpy morphologies and bar-like structures are observed in select high-resolution targets (REBELS-25; (Rowland et al., 2024)).
• Kinematics:
  • REBELS-25, at $6.5 < z_{\mathrm{phot}} < 9.5$7, shows a cold, rotationally supported disk (V$6.5 < z_{\mathrm{phot}} < 9.5$8/$6.5 < z_{\mathrm{phot}} < 9.5$9 = $158\,\mu$0; rotational velocity $158\,\mu$1km s$158\,\mu$2, dispersion $158\,\mu$3 km s$158\,\mu$4) with a near-exponential surface-brightness profile ($158\,\mu$5) (Rowland et al., 2024).
  • The occurrence of dynamically cold disks at $158\,\mu$6 challenges standard models predicting predominantly turbulent, merger-driven morphologies at such epochs (Rowland et al., 2024, Hygate et al., 2023).
• Outflows and Feedback:
  • [C II] line profiles in REBELS-25 and others show high-velocity wings ($158\,\mu$7 km s$158\,\mu$8), possibly indicative of strong star-formation-driven outflows with mass outflow rates $158\,\mu$9 yr$z \lesssim 8.5$0 (Hygate et al., 2023).
• [O III] and [C II] as ISM Probes:
  • REBELS detects [O III] $z \lesssim 8.5$1m in all ALMA–JWST follow-up galaxies, establishing an empirical $z \lesssim 8.5$2–SFR relation valid from local to $z \lesssim 8.5$3 (Algera et al., 19 Sep 2025).
  • [O III]/[C II] ratios ($z \lesssim 8.5$4) are elevated relative to local dwarfs at fixed metallicity and ionization parameter, with burstiness (as traced by $z \lesssim 8.5$5) found to be the primary driver (Algera et al., 19 Sep 2025, Leeuwen et al., 4 Aug 2025).

5. Dust Properties, Attenuation Laws, and Obscuration Scaling

REBELS provides the first robust calibration of dust attenuation in massive high-$z \lesssim 8.5$6 galaxies:

• Obscured Fraction and IRX–$z \lesssim 8.5$7:
  • Dust-obscured star formation comprises $z \lesssim 8.5$8 for $z \lesssim 8.5$9 (Bowler et al., 2023, Algera et al., 2022).
  • At fixed $158\,\mu$00, IRX is a factor of $158\,\mu$01 lower at $158\,\mu$02 versus $158\,\mu$03 main-sequence relations, reflecting lower dust covering fractions or production efficiency (Bowler et al., 2023, Barrufet et al., 2023).
  • The IRX–$158\,\mu$04 relation of REBELS+ALPINE sample is consistent with the local Calzetti starburst law ($158\,\mu$05 for $158\,\mu$06), with no significant ($158\,\mu$07) deficit down to $158\,\mu$08 (Bowler et al., 2023).
• Dust Temperature Evolution:
  • REBELS galaxies have median $158\,\mu$09 K; a physical scaling $158\,\mu$10 is established, with the spread at fixed $158\,\mu$11 set by gas column density and metallicity (Sommovigo et al., 2022).
  • More obscured, metal-poor galaxies exhibit higher $158\,\mu$12 due to a lower dust-to-gas ratio for a given SFR (Sommovigo et al., 2022).
• Dust Production and Retention:
  • Semi-analytic modeling finds core-collapse SNe dominate REBELS dust budgets; ISM grain growth ($158\,\mu$13 Myr) contributes only $158\,\mu$14 (Dayal et al., 2022).
  • Dust-to-stellar mass ratios are $158\,\mu$15 at $158\,\mu$16. Outliers with up to $158\,\mu$17 may require underestimated $158\,\mu$18, overly low assumed $158\,\mu$19, or rare, transient, starburst-driven enrichment (Dayal et al., 2022).

6. Implications for Galaxy Evolution at Cosmic Dawn

Synthesis of REBELS results supports a paradigm of rapid, gas-driven assembly and early metal/dust enrichment in massive galaxies at $158\,\mu$20:

• Baryonic Assembly:
  • Gas-rich, dynamically settled disks, complex ISM morphologies, and high sSFRs evidence rapid baryon accretion and efficient star formation within $158\,\mu$21 Myr of the Big Bang (Rowland et al., 2024, Hygate et al., 2023, Aravena et al., 2023).
• ISM Diversity and Main Sequence Offset:
  • REBELS galaxies frequently lie on or above the $158\,\mu$22 main sequence; outliers such as REBELS-25 are precursors to massive quiescent galaxies observed at $158\,\mu$23 (Hygate et al., 2023, Rowland et al., 17 Jan 2025).
  • High metallicities ($158\,\mu$24) indicate rapid enrichment and efficient retention (Rowland et al., 17 Jan 2025).
• Obscured Star Formation and Dust Evolution:
  • $158\,\mu$25 of star formation is already obscured by dust—requiring SFRD estimates at $158\,\mu$26 to adopt empirical, mass-dependent corrections (Algera et al., 2022, Bowler et al., 2023).
  • Declining IRX at high-$158\,\mu$27 supports the role of low dust-to-gas ratios and geometrical biases in emerging galaxies (Bowler et al., 2023).
• ISM Tracers and Line Diagnostics:
  • [O III] 88 $158\,\mu$28m is validated as a robust SFR tracer from $158\,\mu$29 to $158\,\mu$30, whereas [O III]/[C II] ratios reflect the prevalence of bursty, low-metallicity, high ionization regions in early systems (Algera et al., 19 Sep 2025).

REBELS serves as a foundational dataset for refinement of early galaxy formation models, providing both benchmarks for simulations and high-priority targets for deep JWST follow-up. Its results underscore the necessity of combining UV, FIR, and rest-optical (nebular line) measurements to fully capture early galaxy assembly, baryonic cycling, and the timing of dust and metal enrichment at cosmic dawn.