PHANGS-ALMA: High-Res CO Observations

• PHANGS-ALMA observations are high-resolution CO (2–1) imaging of the molecular ISM in nearby star-forming galaxies, resolving individual giant molecular clouds.
• The survey employs advanced mosaicking with ALMA’s 12 m, 7 m, and Total Power arrays to capture detailed structure, kinematics, and feedback signatures at ~100 pc scales.
• The dataset enables robust statistical analyses of GMC demographics, star formation laws, and feedback processes, directly informing theoretical models and sub-grid prescriptions.

The PHANGS-ALMA (Physics at High Angular resolution in Nearby GalaxieS–Atacama Large Millimeter/submillimeter Array) observations constitute the preeminent high-resolution spectroscopic imaging survey of the molecular interstellar medium (ISM) in the local universe’s star-forming galaxy population. Mapping the ^12CO (J=2–1) transition at ~1″ (∼100 pc) resolution across ∼90 nearby disks, PHANGS-ALMA resolves GMC-scale structure, kinematics, dynamics, and feedback throughout the molecular gas that dominates present-day star formation. This dataset underpins advanced statistical studies of GMC properties, the star formation law, the impact of environment and feedback, and the dynamical interplay between molecular gas and galactic structure.

1. Survey Scope, Scientific Motivation, and Sample Selection

PHANGS-ALMA was designed to address the molecular gas context of star formation on sub-kiloparsec and GMC (~100 pc) scales across the full local “main sequence” disk-galaxy population (Leroy et al., 2021). The main targets comprise 75 ALMA-accessible galaxies with stellar mass $\log_{10}M_{\star}\gtrsim9.75$, star formation rate (SFR) above $10^{-2}\,M_\odot\,{\rm yr}^{-1}$, $i<75^{\circ}$, $\delta$ between $-75^{\circ}$ and $+25^{\circ}$, and $d\lesssim17\,{\rm Mpc}$; 15 additional galaxies extend the sample to lower-mass, low-sSFR, or nearby dwarf systems.

The scientific motivations include (i) quantifying GMC demographics, mass functions, and environmental dependence, (ii) measuring the efficiency and timescales of star formation across disk environments, (iii) probing dynamical phenomena (bars, arms, rings), and (iv) verifying predicted scaling relations and feedback signatures in resolved molecular gas. PHANGS-ALMA serves as the foundational molecular dataset for the PHANGS multi-wavelength campaign, aligning with HST, VLT-MUSE, AstroSat, and JWST programs (Leroy et al., 2021, Meidt et al., 2022).

2. Observational Strategy and Data Acquisition

ALMA's Band 6 receivers target the ^12CO(J=2–1) line at 230.538 GHz. Each galaxy is mosaicked over its star-forming disk, employing:

• The 12 m main array (compact configurations, baselines ∼13–300 m) for $\sim$1″ synthesized beams.
• The 7 m ACA for short spacings and extended structure.
• Total Power single-dish antennas for full flux recovery on the largest spatial scales (Leroy et al., 2021).

A typical mosaic comprises 10–450 pointings with integration times to reach 2.5 km s⁻¹ channels at $\sim$85 mK rms (per native beam), ensuring mass sensitivity to $2\times10^4\,M_\odot$ per beam (Leroy et al., 2021). The majority of targets achieve $<150$–180 pc linear resolution; four fields are observed to 26 pc scales. Data collection includes simultaneous mapping of emission-line and continuum windows where possible to maximize bandwidth and calibration coverage.

3. Data Reduction Pipeline, Calibration, and Imaging

Data processing is standardized through a dedicated pipeline (Leroy et al., 2021), involving:

1. Flagging, bandpass, amplitude, and phase calibration using ALMA's pipeline and standard calibrators.
2. Joint imaging of 12 m, 7 m, and Total Power cubes using CASA’s tclean with a hybrid multiscale CLEAN approach. Working channel width is typically rebinned to $\sim$2.5 km s⁻¹.
3. Primary-beam correction and construction of circular Gaussian beams for uniformity.
4. Flux recovery and fidelity restoration by "feathering" (Fourier domain combination) or tp2vis (pseudo-visibilities), ensuring flux conservation from cloud to kpc scales. After feathering, integrated flux is commonly within 5% of the Total Power reference over high-S/N emission (Leroy et al., 2021).

A two-track masking process creates “strict” (high-purity, $\sim$60% flux completeness) and “broad” (high-completeness, $\sim$98–100%) emission masks. This dual approach enables both high-fidelity GMC segmentation and robust global ISM property recovery. Three-dimensional noise modeling is performed at spatial and spectral scales to facilitate rigorous uncertainty estimates in subsequent moment maps.

4. Derived Data Products and Physical Quantities

PHANGS-ALMA delivers for every galaxy:

• Data cubes at the native beam (median 1.3″) and at fixed physical resolutions (e.g., 60, 90, 120, 150, 500 pc).
• Integrated intensity (moment-0), velocity (moment-1), and linewidth (moment-2) maps under both strict and broad masks.
• Ancillary noise, coverage, and uncertainty maps.

CO(2–1) surface brightness is converted to molecular gas mass surface density assuming $$\alpha_{\rm CO}=4.35\,M_\odot\,{\rm (K\,km\,s^{-1}\,pc^2)^{-1}}$$ and $R_{21}=0.65$ (Leroy et al., 2021, Ruiz-García et al., 2024). Stellar surface densities are derived from near-IR (3.4–3.6 μm) maps using color- and sSFR-dependent mass-to-light ratios, and SFRs from combined FUV and mid-IR indicators scaled to the GSWLC (Leroy et al., 2021). Delivered data products underpin creation of GMC catalogs and serve as direct inputs for analyses of molecular depletion timescales, star formation efficiency, and dynamic phenomena.

5. GMC Catalogs: Segmentation, Property Recovery, and Environmental Trends

Cloud segmentation uses the python implementation pycprops, an updated CPROPS algorithm (Rosolowsky et al., 2021, Zakardjian et al., 2023, Brunetti et al., 2022). The pipeline:

• Homogenizes cubes to common spatial (usually 90 pc) and noise (e.g., 75 mK/channel) scales.
• Applies dendrogram-based identification of significant local maxima and constructs “seeded watershed” segmentations.
• Completeness is evaluated by injecting synthetic clouds matched in mass, surface density, and virial parameter, with logistic regression fits yielding $M_{50}$ completeness thresholds (Rosolowsky et al., 2021).

For each cloud, the catalog reports beam- and sensitivity-corrected:

• CO luminosity ($L_{\rm CO}$), size (intensity-weighted spatial second moments), mass ($M_{\rm mol}= \alpha_{\rm CO} L_{\rm CO}$), velocity dispersion, virial mass, virial parameter ($\alpha_{\rm vir}$), surface density ($\Sigma_{\rm mol}$), and free-fall time ($t_{\rm ff}$).

Population-wide, clouds in disk environments exhibit median properties: size $R\sim 50$–70 pc, $M_{\rm mol}\sim10^{6}\,M_\odot$, $\Sigma_{\rm mol}\sim100\,M_\odot\,{\rm pc}^{-2}$, $\sigma_v\sim5$–7 km s⁻¹, $\alpha_{\rm vir}\sim1$–2, following Larson- and Heyer-type scaling relations (Rosolowsky et al., 2021, Zakardjian et al., 2023, Brunetti et al., 2022). Truncated power-law mass distributions exhibit steeper slopes and lower mass cutoffs in interarm regions ($\gamma\approx2.5$, $M_c\approx2.3\times10^6\,M_\odot$) versus spiral arms ($\gamma\approx2.3$, $M_c\approx5.8\times10^6\,M_\odot$) (Rosolowsky et al., 2021).

Central and barred regions show elevated velocity dispersions and virial parameters, likely due to streaming motions, increased external pressure, and blending with diffuse emission (Rosolowsky et al., 2021, Brunetti et al., 2022). Luminous IR mergers, e.g., NGC 3256, show systematically higher $\sigma_v$, $\Sigma_{\rm mol}$, $P_{\rm int}/k_B$, and lower $t_{\rm ff}$, indicating pressure-regulated, short-duration, high-efficiency star formation in these regimes (Brunetti et al., 2022).

6. Multi-Scale ISM Structure, Dynamical Analysis, and Feedback

PHANGS-ALMA data provide a uniform baseline for ISM structuring investigations:

• Filamentary web statistics measured from JWST/MIRI imaging correspond quantitatively to the turbulent Jeans length $\lambda_J=\sigma^2/(G\,\Sigma)$ computed from local ALMA-derived $\Sigma_{\rm mol}$ and $\sigma$ on $\lesssim$150 pc scales (Meidt et al., 2022).
• The observed filament spacing, $\lambda_{\rm fil}$, consistently agrees with $\lambda_J$ across diverse disks, supporting gravitational fragmentation as the organizing principle for ISM structure on scales of several hundred pc, distinct from the Toomre length ($\lambda_T$) or vertical scale height ($h$).
• Quantitative analysis of rotational dynamics via harmonic decomposition in 150 pc rings enables robust recovery of disk position angles, inclinations, and high-resolution rotation curves for 67 galaxies, validating the use of CO (2–1) as a dynamical mass tracer down to $R\sim500$ pc (Lang et al., 2020).
• Barred galaxies are analyzed for dynamical resonances by combining CO kinematics with Spitzer-based stellar potentials. The mean ratio of bar corotation radius to bar length is $\langle\mathcal{R}\rangle=1.12\pm0.39$, favoring “fast” bars and supporting models of angular-momentum transport and secular evolution (Ruiz-García et al., 2024).

Feedback studies using cross-matched PHANGS-ALMA, HST, and MUSE catalogs locate and characterize $\sim$100–300 pc scale molecular superbubbles, finding their expansion properties and energetics consistent with supernova-driven “snowplough” models (momentum-conserving expansion). Only ~10% of injected stellar mechanical energy couples to molecular gas kinetic energy, supplying key calibrations for sub-grid SN feedback prescriptions (Watkins et al., 2023).

At $\sim$100 pc scales, GMCs associated with HII regions in MUSE/ALMA overlays show modest, quantifiable perturbations—higher CO peak brightness and mass—which likely reflect pre-supernova feedback processes (photoionization, radiation pressure) altering the host GMC’s immediate environment (Zakardjian et al., 2023).

7. CO Line Ratios, Excitation, and Conversion Factors

PHANGS-ALMA’s coverage of CO(2–1) is supplemented by matched-beam CO(1–0) and (3–2) datasets to derive low-J line ratios (Leroy et al., 2021). Key findings include:

Ratio	Mean	Median	16–84% Range	Trends
R_{21}	0.65	0.61	[0.50–0.83]	↑ center, SFR; ↓ radius
R_{32}	0.50	0.46	[0.23–0.59]	↑ center, SFR; ↓ radius
R_{31}	0.31	0.29	[0.20–0.42]	↑ center, SFR; ↓ radius

All ratios show central enhancements (~0.2–0.3 dex), positive correlation with star-formation rate surface density, and an anti-correlation with radius. Non-LTE modeling indicates typical excitation conditions: $T_{\rm kin}=10$–20 K, $n_{\rm H_2}=300$–$10^3$ cm⁻³, and $\tau_{1-0}=4$. R_{21} variations imply a corresponding anti-correlation, $\alpha_{\rm CO}\propto R_{21}^{-0.9}$, entailing that a central rise in $R_{21}$ of 0.2 dex translates into a comparable drop in $\alpha_{\rm CO}$ (Leroy et al., 2021).

• (Leroy et al., 2021) PHANGS-ALMA: Arcsecond CO(2-1) Imaging of Nearby Star-Forming Galaxies
• (Leroy et al., 2021) PHANGS-ALMA Data Processing and Pipeline
• (Rosolowsky et al., 2021) Giant Molecular Cloud Catalogues for PHANGS-ALMA: Methods and Initial Results
• (Watkins et al., 2023) Quantifying the energetics of molecular superbubbles in PHANGS galaxies
• (Zakardjian et al., 2023) The impact of HII regions on Giant Molecular Cloud properties in nearby galaxies sampled by PHANGS ALMA and MUSE
• (Brunetti et al., 2022) Extreme giant molecular clouds in the luminous infrared galaxy NGC 3256
• (Meidt et al., 2022) PHANGS--JWST First Results: ISM structure on the turbulent Jeans scale in four disk galaxies observed by JWST and ALMA
• (Leroy et al., 2021) Low-J CO Line Ratios From Single Dish CO Mapping Surveys and PHANGS-ALMA
• (Ruiz-García et al., 2024) Dynamical resonances in PHANGS galaxies
• (Lang et al., 2020) PHANGS CO kinematics: disk orientations and rotation curves at 150 pc resolution

PHANGS-ALMA establishes the reference dataset for cloud-scale molecular ISM physics, SFR laws, and feedback in nearby disks, supporting cross-survey studies, calibration of theoretical models, and the formulation of sub-grid prescriptions in galaxy simulations.