GOODS-ALMA 2.0: Wide-Field Dust Continuum Survey

• The survey is a blind ALMA Band 6 continuum study covering 72.42 arcmin² in GOODS-South that delivers a homogeneous census of dust-obscured star formation.
• It combines high- and low-resolution observations to robustly characterize millimeter continuum sources, providing precise measurements of fluxes, sizes, and star formation properties.
• The study reveals that dust-obscured star formation is systematically compact, offering key insights into galaxy evolution and the build-up of dense stellar cores.

Searching arXiv for GOODS-ALMA 2.0 and ASAGAO papers to ground the article in the relevant literature. GOODS-ALMA 2.0 is a blind ALMA Band 6 continuum survey of the GOODS-South field designed to obtain a homogeneous, contiguous census of dust-obscured star formation at 1.1 mm over cosmic noon and beyond. In the GOODS-ALMA 2.0 formulation, the survey covers a continuous area of 72.42 arcmin$^2$ in GOODS-South with combined high-resolution and low-resolution observations, yielding a synthesized beam of $0.447'' \times 0.418''$ and an average point-source sensitivity of $$\sigma_{1.1\,\mathrm{mm}} = 68.4\,\mu\mathrm{Jy}\,\mathrm{beam}^{-1}$$ (Gómez-Guijarro et al., 2021). Closely related work on the ASAGAO program, sometimes described as a GOODS-ALMA 2.0 survey in the 1.2 mm context, provides a complementary 26 arcmin$^2$ GOODS-South map with different survey geometry, depth, and science emphasis, including source counts, luminosity functions, and resolved morphology (Hatsukade et al., 2018, Fujimoto et al., 2018). Across these studies, GOODS-ALMA 2.0 denotes a transition from small-area or single-configuration millimeter surveys to a wide, uniform, multi-configuration framework for measuring source demographics, dust-continuum sizes, star-formation compactness, and the role of obscured systems in galaxy evolution.

1. Survey definition and observational architecture

GOODS-ALMA 2.0 is a blind 1.1 mm continuum survey covering 72.42 arcmin$^2$ in the GOODS-South field, defined at primary beam $\ge 20\%$ (Gómez-Guijarro et al., 2022, Gómez-Guijarro et al., 2021). The field center is given as $$\mathrm{RA} = 03^\mathrm{h}\,32^\mathrm{m}\,30^\mathrm{s}$$, $\mathrm{Dec} = -27^\circ\,48'\,00''$ (Gómez-Guijarro et al., 2022). The survey combines two ALMA Band 6 configurations in order to improve $uv$-coverage and recover both compact and extended emission (Gómez-Guijarro et al., 2021).

The high-resolution component, identified with GOODS-ALMA 1.0, used Cycle 3 observations with baselines $16.7\,\mathrm{m}/1808\,\mathrm{m}$, a synthesized beam of $0.447'' \times 0.418''$0, and rms $0.447'' \times 0.418''$1 (Gómez-Guijarro et al., 2022, Gómez-Guijarro et al., 2021). The low-resolution component used Cycle 5 observations with baselines $0.447'' \times 0.418''$2, a beam of $0.447'' \times 0.418''$3, and rms $0.447'' \times 0.418''$4 (Gómez-Guijarro et al., 2022, Gómez-Guijarro et al., 2021). The combined dataset reaches a beam of $0.447'' \times 0.418''$5 and average sensitivity $0.447'' \times 0.418''$6 (Gómez-Guijarro et al., 2021).

Each dataset is described as an 846-pointing mosaic with total on-source time $0.447'' \times 0.418''$7 per configuration (Gómez-Guijarro et al., 2021). The full $0.447'' \times 0.418''$8-coverage, $0.447'' \times 0.418''$9, was obtained by concatenating calibrated visibilities in CASA v5.6.1, and the continuum was imaged with TCLEAN using a dirty map and natural weighting, specifically to avoid beam-mismatch systematics when combining configurations (Gómez-Guijarro et al., 2021).

A related but distinct GOODS-South ALMA survey, ASAGAO, mapped a contiguous $$\sigma_{1.1\,\mathrm{mm}} = 68.4\,\mu\mathrm{Jy}\,\mathrm{beam}^{-1}$$0 region, corresponding to $$\sigma_{1.1\,\mathrm{mm}} = 68.4\,\mu\mathrm{Jy}\,\mathrm{beam}^{-1}$$1 arcmin$$\sigma_{1.1\,\mathrm{mm}} = 68.4\,\mu\mathrm{Jy}\,\mathrm{beam}^{-1}$$2, at 1.2 mm in Band 6 with two tunings centered at 262.56 GHz and 253.56 GHz (Hatsukade et al., 2018). Its inclusion is important because several later GOODS-South analyses treat ASAGAO as part of the broader GOODS-ALMA 2.0 landscape, especially for source statistics and dusty-galaxy morphology. This suggests that the nomenclature around “GOODS-ALMA 2.0” has both a strict 1.1 mm definition and a broader GOODS-South ALMA survey context.

2. Detection strategy, catalog construction, and counterpart identification

The GOODS-ALMA 2.0 source catalog was built from a blind search on the combined dirty map using PyBDSF (Gómez-Guijarro et al., 2021). Detection thresholds were explored with pixel threshold $$\sigma_{1.1\,\mathrm{mm}} = 68.4\,\mu\mathrm{Jy}\,\mathrm{beam}^{-1}$$3 from 3.0 to 6.0 and island threshold $$\sigma_{1.1\,\mathrm{mm}} = 68.4\,\mu\mathrm{Jy}\,\mathrm{beam}^{-1}$$4, with one Gaussian per blob and $$\sigma_{1.1\,\mathrm{mm}} = 68.4\,\mu\mathrm{Jy}\,\mathrm{beam}^{-1}$$5 defined as

$$\sigma_{1.1\,\mathrm{mm}} = 68.4\,\mu\mathrm{Jy}\,\mathrm{beam}^{-1}$$6

Negative-image tests yielded $$\sigma_{1.1\,\mathrm{mm}} = 68.4\,\mu\mathrm{Jy}\,\mathrm{beam}^{-1}$$7 for $$\sigma_{1.1\,\mathrm{mm}} = 68.4\,\mu\mathrm{Jy}\,\mathrm{beam}^{-1}$$8 in the combined map, corresponding to 100% purity (Gómez-Guijarro et al., 2021). On that basis, 44 100%-pure sources with $$\sigma_{1.1\,\mathrm{mm}} = 68.4\,\mu\mathrm{Jy}\,\mathrm{beam}^{-1}$$9 define the main catalog (Gómez-Guijarro et al., 2021).

A supplementary prior-based catalog extends the sample below the pure blind threshold. Running blind detection down to $^2$0 yields $^2$1 real sources, and 44 new candidates with $^2$2 were validated by IRAC (3.6/4.5 $^2$3) or VLA (3 GHz) priors within $^2$4 and stellar mass $^2$5 (Gómez-Guijarro et al., 2021). The total GOODS-ALMA 2.0 catalog therefore contains 88 galaxies, split evenly between the main and supplementary components (Gómez-Guijarro et al., 2021, Gómez-Guijarro et al., 2022).

The catalog composition is summarized concisely below.

Component	Selection	Number of sources
Main catalog	$^2$6, 100% pure	44
Supplementary catalog	$^2$7, aided by priors	44
Total	Combined GOODS-ALMA 2.0 catalog	88

Among the 88 sources, 13 are optically dark or faint, specifically $^2$8- or $^2$9-band dropouts (Gómez-Guijarro et al., 2021). The final sample has median redshift $^2$0, spanning $^2$1, and a subset of 69 sources has Herschel detections (Gómez-Guijarro et al., 2022). In the supplementary analysis of faint GOODS-ALMA detections, 16 additional $^2$2 galaxies were identified in the 1.1 mm field using IRAC and VLA priors, including two galaxies undetected in HST/WFC3 $^2$3-band imaging to $^2$4 AB (Franco et al., 2020). That work also showed that the publicly released HST positions required both a global astrometric correction and a local distortion correction for reliable multiwavelength associations (Franco et al., 2020).

The ASAGAO 1.2 mm catalog construction followed a different workflow. Source extraction on the $^2$5-tapered map found 25 sources at $^2$6 and 45 at $^2$7, with spurious fractions estimated from negative peaks and completeness calibrated through 30,000 Monte Carlo point-source injections (Hatsukade et al., 2018). Optical and near-infrared counterpart matching used the ZFOURGE catalog after applying an astrometric correction of $^2$8 for Gaia DR1 tie-in, and 88% of $^2$9 sources had secure optical/IR/radio counterparts (Hatsukade et al., 2018).

3. Imaging products, fluxes, sizes, and morphological methodology

GOODS-ALMA 2.0 was explicitly designed to mitigate the common limitations of earlier millimeter surveys: discontinuous sky coverage, inhomogeneous sensitivity, and modest $\ge 20\%$0-coverage from single-array observations (Gómez-Guijarro et al., 2021). Its combined configuration permits robust characterization of both unresolved and resolved dust emission, with source sizes measured in the $\ge 20\%$1-plane via UVMODELFIT using a circular Gaussian model (Gómez-Guijarro et al., 2021, Gómez-Guijarro et al., 2022).

Fluxes were measured with 1.6''-diameter aperture photometry plus aperture correction from the local synthesized beam (Gómez-Guijarro et al., 2021). A source is considered unresolved when the fitted $\ge 20\%$2 falls below the Martí-Vidal et al. minimum resolvable size,

$\ge 20\%$3

and the effective radius is derived from the Gaussian FWHM as

$\ge 20\%$4

These definitions are central to the survey’s compactness analysis (Gómez-Guijarro et al., 2021).

For the main catalog, the median dust-continuum effective radius is $\ge 20\%$5, corresponding to $\ge 20\%$6 kpc (Gómez-Guijarro et al., 2021, Gómez-Guijarro et al., 2022). The high-resolution subset has $\ge 20\%$7 or $\ge 20\%$8 kpc, while low-resolution-only detections have $\ge 20\%$9 or $$\mathrm{RA} = 03^\mathrm{h}\,32^\mathrm{m}\,30^\mathrm{s}$$0 kpc (Gómez-Guijarro et al., 2021). For sources with $$\mathrm{RA} = 03^\mathrm{h}\,32^\mathrm{m}\,30^\mathrm{s}$$1 mJy, compact emission is described as prevailing, and sizes as extended as typical star-forming stellar disks are reported to be rare (Gómez-Guijarro et al., 2021).

The survey also quantified scaling behavior with redshift and stellar mass. Stacks above and below $$\mathrm{RA} = 03^\mathrm{h}\,32^\mathrm{m}\,30^\mathrm{s}$$2 and $$\mathrm{RA} = 03^\mathrm{h}\,32^\mathrm{m}\,30^\mathrm{s}$$3 gave

$$\mathrm{RA} = 03^\mathrm{h}\,32^\mathrm{m}\,30^\mathrm{s}$$4

and

$$\mathrm{RA} = 03^\mathrm{h}\,32^\mathrm{m}\,30^\mathrm{s}$$5

at fixed redshift and fixed mass, respectively (Gómez-Guijarro et al., 2021). These trends resemble the stellar-size evolution of late-type galaxies from optical measurements, but with $$\mathrm{RA} = 03^\mathrm{h}\,32^\mathrm{m}\,30^\mathrm{s}$$6 lower normalization (Gómez-Guijarro et al., 2021).

ASAGAO extended morphological analysis beyond simple Gaussian sizes by fitting Sérsic profiles to stacked 1.2 mm emission. Using visibility stacking on 33 ASAGAO objects at $$\mathrm{RA} = 03^\mathrm{h}\,32^\mathrm{m}\,30^\mathrm{s}$$7, followed by Monte Carlo centering-error correction, the survey obtained

$$\mathrm{RA} = 03^\mathrm{h}\,32^\mathrm{m}\,30^\mathrm{s}$$8

at $$\mathrm{RA} = 03^\mathrm{h}\,32^\mathrm{m}\,30^\mathrm{s}$$9 (Fujimoto et al., 2018). In the rest-frame optical, 21 ASAGAO galaxies cross-matched to the van der Wel et al. HST catalog gave median

$\mathrm{Dec} = -27^\circ\,48'\,00''$0

supporting the interpretation that the FIR-emitting disk is embedded within a larger stellar disk (Fujimoto et al., 2018). ASAGAO also found that two-component fits preferred a point source plus disk over bulge plus disk, with the point source contributing $\mathrm{Dec} = -27^\circ\,48'\,00''$1 of the total 1.2 mm flux, interpreted as an AGN candidate (Fujimoto et al., 2018).

4. Number counts, resolved background light, and luminosity functions

A major objective of GOODS-ALMA 2.0 and ASAGAO is to quantify the faint millimeter-source population and its contribution to the cosmic infrared background and obscured star-formation history. In GOODS-ALMA 2.0, the per-source contribution to number counts is written as

$\mathrm{Dec} = -27^\circ\,48'\,00''$2

with purity, effective area, and completeness explicitly incorporated (Gómez-Guijarro et al., 2021). Completeness was calibrated through Monte Carlo injections over 30 fluxes and 11 sizes, yielding nearly 100% completeness for $\mathrm{Dec} = -27^\circ\,48'\,00''$3 mJy and lower completeness toward faint, extended sources (Gómez-Guijarro et al., 2021). Effective area from the combined noise map reached 100% area at $\mathrm{Dec} = -27^\circ\,48'\,00''$4 and 90% at $\mathrm{Dec} = -27^\circ\,48'\,00''$5 (Gómez-Guijarro et al., 2021).

The differential number counts span $\mathrm{Dec} = -27^\circ\,48'\,00''$6 mJy and are reported to be in good agreement with other ALMA studies (Gómez-Guijarro et al., 2021). A Schechter function fit,

$\mathrm{Dec} = -27^\circ\,48'\,00''$7

gives

$\mathrm{Dec} = -27^\circ\,48'\,00''$8

for the differential counts (Gómez-Guijarro et al., 2021). Integrating to $\mathrm{Dec} = -27^\circ\,48'\,00''$9 mJy yields

$uv$0

corresponding to $uv$1 of the COBE CIB at 1.13 mm (Gómez-Guijarro et al., 2021).

ASAGAO derived independent number counts at 1.2 mm using completeness- and reliability-corrected source weights,

$uv$2

and fitted the combined ALMA counts with a Schechter form characterized by

$uv$3

(Hatsukade et al., 2018). Integrating from $uv$4, ASAGAO found

$uv$5

and, relative to Planck’s $uv$6, obtained

$uv$7

This result places a quantitative lower bound on the fraction of the 1.2 mm background resolved into discrete sources (Hatsukade et al., 2018).

ASAGAO additionally constructed infrared luminosity functions in the redshift range $uv$8 using the $uv$9 method. At $16.7\,\mathrm{m}/1808\,\mathrm{m}$0, the fitted Schechter parameters are

$16.7\,\mathrm{m}/1808\,\mathrm{m}$1

and comparison with Koprowski et al. indicated positive $16.7\,\mathrm{m}/1808\,\mathrm{m}$2 evolution and negative $16.7\,\mathrm{m}/1808\,\mathrm{m}$3 evolution from $16.7\,\mathrm{m}/1808\,\mathrm{m}$4 (Hatsukade et al., 2018). This establishes that GOODS-South ALMA surveys are not only morphology experiments but also direct constraints on the evolving obscured luminosity function.

5. Physical properties, compact star formation, and “starbursts in the main sequence”

The GOODS-ALMA 2.0 sample has been used to derive star formation rates, gas fractions, depletion timescales, dust temperatures, and structural parameters from UV-to-mm data (Gómez-Guijarro et al., 2022). For sources with Herschel detections, the infrared SED was modeled with Stardust using stellar templates, AGN torus templates, and Draine & Li (2007) plus Draine (2014) dust models (Gómez-Guijarro et al., 2022). Total star formation rates are defined as

$16.7\,\mathrm{m}/1808\,\mathrm{m}$5

with

$16.7\,\mathrm{m}/1808\,\mathrm{m}$6

for a Salpeter IMF (Gómez-Guijarro et al., 2022). Gas masses are estimated from the metallicity-dependent gas-to-dust ratio, $16.7\,\mathrm{m}/1808\,\mathrm{m}$7, gas fraction is $16.7\,\mathrm{m}/1808\,\mathrm{m}$8, depletion time is $16.7\,\mathrm{m}/1808\,\mathrm{m}$9, and the star-formation surface density is

$0.447'' \times 0.418''$00

(Gómez-Guijarro et al., 2022).

Within the 69 Herschel-detected sources, 23 systems were identified as “starbursts in the main sequence,” defined by $0.447'' \times 0.418''$01 within the main sequence, $0.447'' \times 0.418''$02, but with $0.447'' \times 0.418''$03 below the $0.447'' \times 0.418''$04 scatter of the Tacconi et al. (2018) depletion-time relation (Gómez-Guijarro et al., 2022). Relative to typical main-sequence galaxies at the same mass and redshift, these objects show median properties

$0.447'' \times 0.418''$05

compared with average values of $0.447'' \times 0.418''$06 in gas fraction and $0.447'' \times 0.418''$07 in dust temperature for typical main-sequence systems (Gómez-Guijarro et al., 2022). They are also very massive: among galaxies with $0.447'' \times 0.418''$08, $0.447'' \times 0.418''$09, or $0.447'' \times 0.418''$10, are classified as starbursts in the main sequence (Gómez-Guijarro et al., 2022).

The role of compactness is explicit. The star-forming area is defined as $0.447'' \times 0.418''$11, and an optical comparison area $0.447'' \times 0.418''$12 is defined from the expected optical disk area at the galaxy’s $0.447'' \times 0.418''$13 and $0.447'' \times 0.418''$14 (Gómez-Guijarro et al., 2022). At fixed mass and redshift, increasing compactness of the 1.1 mm emission relative to the optical disk, expressed as larger $0.447'' \times 0.418''$15, correlates with lower $0.447'' \times 0.418''$16, lower $0.447'' \times 0.418''$17, and higher $0.447'' \times 0.418''$18, with Spearman $0.447'' \times 0.418''$19 (Gómez-Guijarro et al., 2022). Starbursts in the main sequence occupy the extreme of these relations (Gómez-Guijarro et al., 2022).

Stacked visibilities reinforce the compactness contrast. The $0.447'' \times 0.418''$20-stack of the 23 starbursts in the main sequence yields $0.447'' \times 0.418''$21 kpc after correction to the median sample $0.447'' \times 0.418''$22, $0.447'' \times 0.418''$23, whereas the remaining main-sequence galaxies stack to $0.447'' \times 0.418''$24 kpc (Gómez-Guijarro et al., 2022). This difference links the survey’s size measurements directly to physical regulation of star formation.

A related mass-selected analysis based on ASAGAO and ZFOURGE found that ALMA-detected ZFOURGE galaxies at median redshift $0.447'' \times 0.418''$25 generally follow the star-forming main sequence, but exhibit systematically larger infrared excess, $0.447'' \times 0.418''$26, than K-selected galaxies without ALMA detections even at similar redshifts, masses, and star formation rates (Yamaguchi et al., 2020). This implies that the consensus stellar-mass versus IRX relation from rest-frame-UV-selected galaxies cannot fully predict ALMA detectability in stellar-mass-selected samples (Yamaguchi et al., 2020).

6. Evolutionary interpretation, star-formation histories, and links to quenching

The GOODS-ALMA 2.0 dataset has been used to test whether compact dusty systems are transient post-starburst objects or a structurally distinct component within the main sequence. Using CIGALE v2022.1 with non-parametric star-formation histories implemented through the “sfhNlevels” module, the last gigayear SFHs of 65 GOODS-ALMA galaxies with full UV–mm coverage were reconstructed in seven logarithmically spaced lookback-time bins (Ciesla et al., 2022). To quantify recent evolutionary direction in the SFR–$0.447'' \times 0.418''$27 plane, the analysis defined the SFR gradient $0.447'' \times 0.418''$28 as the angle between the evolutionary vector and the horizontal axis: $0.447'' \times 0.418''$29 Values near $0.447'' \times 0.418''$30 indicate constant specific SFR, $0.447'' \times 0.418''$31 a strong recent starburst, and $0.447'' \times 0.418''$32 rapid recent quenching (Ciesla et al., 2022).

For the 14 starbursts in the main sequence and 51 normal main-sequence GOODS-ALMA galaxies, the median SFR gradients are similar over 100 Myr, 300 Myr, and 1 Gyr. At 100 Myr, the medians are $0.447'' \times 0.418''$33 for normal main-sequence galaxies and $0.447'' \times 0.418''$34 for starbursts in the main sequence; at 300 Myr, $0.447'' \times 0.418''$35 and $0.447'' \times 0.418''$36; and at 1000 Myr, $0.447'' \times 0.418''$37 and $0.447'' \times 0.418''$38, respectively (Ciesla et al., 2022). No starburst-in-the-main-sequence galaxy shows $0.447'' \times 0.418''$39 over any interval (Ciesla et al., 2022). The two distributions are statistically indistinguishable, with KS $0.447'' \times 0.418''$40 (Ciesla et al., 2022).

These results were used to evaluate three scenarios. A compaction scenario involving violent inflow to the top of the main sequence was argued to be disfavored because starbursts in the main sequence show low gas fractions and no past high-burst signature. A post-starburst passage through the main sequence was argued to be ruled out by the absence of large negative $0.447'' \times 0.418''$41. A “mild compaction/feeding within the MS” scenario, in which centrally concentrated star formation builds compact cores while galaxies remain inside the main sequence, was supported by the positive or flat gradients (Ciesla et al., 2022). This suggests that main-sequence scatter contains structurally heterogeneous galaxies with similar self-regulated recent growth.

The ASAGAO morphology analysis points in a compatible direction. The radial surface-density profile of the total SFR, built from stacked HST/F606W and ALMA data, shows that $0.447'' \times 0.418''$42 dominates over $0.447'' \times 0.418''$43 at $0.447'' \times 0.418''$44 kpc, indicating a heavily dust-enshrouded central kiloparsec (Fujimoto et al., 2018). Under a constant-SFR assumption over either the gas-depletion time $0.447'' \times 0.418''$45 Gyr or the cosmic interval $0.447'' \times 0.418''$46 Gyr from $0.447'' \times 0.418''$47, the shrinkage of $0.447'' \times 0.418''$48 toward values typical of compact quiescent galaxies can be reproduced, but the transformation from $0.447'' \times 0.418''$49 to $0.447'' \times 0.418''$50 cannot (Fujimoto et al., 2018). The stated implication is that additional dynamical processes, such as dissipation or merging, are required to build the spheroidal stellar profiles of $0.447'' \times 0.418''$51 quiescent galaxies (Fujimoto et al., 2018).

A later JWST+ALMA analysis of 33 GOODS-ALMA 2.0 galaxies at $0.447'' \times 0.418''$52 extends this interpretation. It reports that the 1.1 mm-selected sample shows a $0.447'' \times 0.418''$53 steeper decrease of size with increasing wavelength than the field star-forming population, and that at 4.4 $0.447'' \times 0.418''$54 these galaxies align with the quiescent size–mass locus rather than the field star-forming one (Bodansky et al., 25 Jul 2025). The rest-NIR surface-brightness profiles are described as remarkably similar to those of quiescent galaxies in the inner regions, which was interpreted as evidence that dusty star-forming galaxies have already built up stellar mass in a severely dust-obscured core (Bodansky et al., 25 Jul 2025). This suggests a strong evolutionary connection between compact dusty systems and subsequent quiescent descendants.

7. Scientific significance, scope, and common points of confusion

The scientific significance of GOODS-ALMA 2.0 lies in the combination of contiguous area, homogeneous sensitivity, two-configuration $0.447'' \times 0.418''$55-coverage, and the integration of deep ancillary data in GOODS-South. The survey increased the sample size of robust millimeter detections from the earlier high-resolution-only census to 88 galaxies and established that 1.1 mm dust emission is generally compact, with a median $0.447'' \times 0.418''$56 of $0.447'' \times 0.418''$57 kpc (Gómez-Guijarro et al., 2021). For bright sources with $0.447'' \times 0.418''$58 mJy, it concluded that compact dust continuum emission prevails and that dust disks as large as typical star-forming stellar disks are rare (Gómez-Guijarro et al., 2021).

At the population level, the survey showed that optically dark or faint sources form a non-negligible component of the millimeter-selected sample, with 13 such objects in the 88-source catalog (Gómez-Guijarro et al., 2021). The supplementary prior-assisted GOODS-ALMA work similarly found that lowering the threshold with robust multiwavelength priors reveals galaxies that are, on average, twice as large in the FIR and of lower stellar mass than the original main-catalog sample, while still remaining significantly more compact in the FIR than in the optical (Franco et al., 2020). This indicates that compact dusty star formation is not restricted to the most extreme bright submillimeter galaxies.

A recurrent point of confusion concerns the relationship between GOODS-ALMA 2.0 and ASAGAO. The strict GOODS-ALMA 2.0 survey is the blind 1.1 mm GOODS-South program covering 72.42 arcmin$0.447'' \times 0.418''$59 with two array configurations (Gómez-Guijarro et al., 2021, Gómez-Guijarro et al., 2022). ASAGAO, by contrast, is a 1.2 mm ALMA Band 6 survey over 26 arcmin$0.447'' \times 0.418''$60 in GOODS-South, with its own catalog, source counts, and morphological analysis (Hatsukade et al., 2018, Fujimoto et al., 2018). The surveys are scientifically adjacent rather than identical. Their results are nevertheless strongly complementary: both find compact dust-obscured star formation, both operate in GOODS-South with extensive ancillary data, and both support the view that obscured star formation is centrally concentrated relative to the stellar distribution.

Another important interpretive issue is whether compact dusty galaxies are simply merger-driven outliers above the main sequence. GOODS-ALMA 2.0 does not support a single, universal answer. A substantial fraction of its galaxies lie within the main sequence scatter despite short depletion times, low gas fractions, high dust temperatures, and compact star-forming regions (Gómez-Guijarro et al., 2022). The SFH analysis further argues that these “starbursts in the main sequence” are not obviously fading post-starbursts, but systems with recent evolution similar to other main-sequence galaxies (Ciesla et al., 2022). A plausible implication is that compact star formation can be a long-lived or recurrent mode of self-regulated growth rather than only a brief excursion.

Taken together, GOODS-ALMA 2.0 and the associated GOODS-South ALMA analyses provide a benchmark empirical framework for studying dusty star-forming galaxies from source detection and number counts to structural evolution and pre-quenching physics. Their central result is not merely that millimeter-selected galaxies are dusty, but that the dust-obscured star formation itself is systematically compact, physically consequential, and tightly linked to the assembly of dense central stellar structures (Gómez-Guijarro et al., 2021, Gómez-Guijarro et al., 2022, Fujimoto et al., 2018, Bodansky et al., 25 Jul 2025).