PHANGS: High-Res Physics in Nearby Galaxies

• PHANGS survey is a coordinated multi-observatory campaign that maps molecular cloud lifecycles and star formation in nearby galaxies.
• It combines ALMA, HST, VLT/MUSE, JWST, and AstroSat data to constrain gas dynamics, stellar feedback, and dust physics at 5–50 pc scales.
• Custom processing pipelines and uniform, panchromatic mosaics provide robust empirical calibrations for advancing ISM and galaxy evolution models.

The Physics at High Angular Resolution in Nearby GalaxieS (PHANGS) Survey is a coordinated, multi-observatory legacy campaign targeting detailed, multi-phase, cloud-scale mapping of star formation, gas dynamics, ISM structure, stellar feedback, and dust physics in a representative population of nearby galaxies. By integrating high-resolution data from ALMA (CO molecular line cubes), HST (UV–optical imaging), VLT/MUSE (optical IFU spectroscopy), JWST (near- and mid-IR imaging), and, more recently, ancillary programs with AstroSat UVIT, PHANGS systematically connects the physical processes driving the baryon cycle at ∼5–50 pc scales to global properties and morphologies across the local galaxy population.

1. Scientific Rationale and Survey Design

PHANGS was designed to address the following core questions: What regulates the formation, evolution, and destruction of molecular clouds? How does clustered star formation proceed and disrupt its environment? What is the empirical lifecycle of gas and dust from dense clouds to stars and back? To this end, the survey targeted a volume-limited subset (D ≲ 23 Mpc) of nearby, main-sequence, low-inclination, star-forming spirals, spanning stellar masses ≳10^9.3 M_⊙, SFRs from 0.1–10 M_⊙ yr^–1, Hubble types S0–Sd, and diverse barred/non-barred morphologies (Lee et al., 2021, Hassani et al., 2023).

Observational strategy prioritized spatial resolution sufficient to resolve the full lifecycle of GMCs, H II regions, star clusters, and feedback-induced substructures (<150 pc, typically 5–50 pc in the nearest systems). The main data assets include:

• ALMA CO(2–1) mosaics (50–150 pc resolution) tracing molecular gas masses, kinematics, cloud segmentation (Lang et al., 2020).
• HST UV–optical five-band imaging (WFC3/UVIS: F275W–F814W, ≲5–15 pc) identifying star clusters and associations, measuring age/mass/reddening distributions (Lee et al., 2021, Turner et al., 2021).
• VLT/MUSE IFU datacubes (0.6″≲0.5–20 pc) mapping H II regions, nebular diagnostics, stellar absorption (Lee et al., 2022).
• JWST/NIRCam+MIRI 8-band images (2–21 μm, 5–50 pc) probing embedded clusters, PAHs, and dust continuum (Lee et al., 2022, Williams et al., 2024, Sutter et al., 2024).
• HST narrow-band Hα imaging (2–10 pc), JWST PAH bands, and AstroSat/UVIT FUV/NUV mosaics round out the core (Chandar et al., 24 Mar 2025, Hassani et al., 2023).

2. Multi-Observatory Data Products and Pipeline Architecture

PHANGS developed and deployed custom processing and homogenization pipelines to ensure uniformity across the diverse data sources. For JWST, the pjpipe wrapper augments the official calibration pipeline with optimized background subtraction, cross-band astrometry, destriping, and PSF convolution, delivering 8-band mosaics (F200W–F2100W) at matched resolution (0.9″ FWHM ≈10–50 pc), pixel-aligned to ALMA/HST/MUSE grids. Astrometric registration achieves relative uncertainties <0.05″ (TW HST–NIRCam–MIRI) (Williams et al., 2024).

HST cluster catalog construction involves DOLPHOT PSF photometry, multi-aperture concentration index (MCI) selection, CNN/visual classification, and SED fitting using CIGALE (BC03 SSP, Chabrier IMF, full Bayesian inference with physical priors) (Turner et al., 2021, Lee et al., 2021, Hannon et al., 2023). ALMA CO(2–1) cubes are imaged and flux-calibrated with high-fidelity mask algorithms and joint 12m+7m+TP uv-plane merging, enabling matched-beam analysis with MUSE/HST (Lang et al., 2020). MUSE nebular line maps are used to identify, mask, and quantify H II regions, correct for [N II] blending in Hα images, and to provide local metallicity for the conversion of CO intensity to H_2 surface density (Chandar et al., 24 Mar 2025).

3. Star Formation, Cloud Lifecycle, and Stellar Cluster Demographics

PHANGS characterizes the spatially resolved distribution of molecular gas, ionized gas, and young stars, distinguishing key evolutionary phases—quiescent GMCs, embedded SF, revealed H II regions, exposed clusters—within the sub-kpc structure of spiral disks (Pan et al., 2022, Lee et al., 2022). A pixel-based classifier assigns sightlines as “CO-only,” “overlap,” or “Hα-only,” deriving phase fractions:

• f_CO ≈ 36%, f_overlap ≈ 30%, f_HII-only ≈ 20% at 150 pc resolution (median over 49 galaxies) (Pan et al., 2022)
• Timescales for molecular gas visibility (at 150 pc) vary systematically: t_gas ~ 20–25 Myr in high-M_* disks, ~10–15 Myr in intermediates, ≲5 Myr in low-M_* systems, controlling for Hα lifetimes t_Hα ~ 5–10 Myr.

HST-based cluster catalogs for 38 galaxies yield hundreds of thousands of clusters and associations. Mass functions are typically dN/dM ∝ M^–2, with plausible evidence of mass-independent disruption. The inclusion of Hα photometry enables precise age-dating, breaking age–extinction degeneracies, and calibrating cluster formation efficiencies (Γ) (Turner et al., 2021, Chandar et al., 24 Mar 2025).

Recent application of deep transfer learning (VGG19-BN, ResNet18 with ImageNet weights) now enables automated, high-purity morphological classification (classes 1–4, resolved vs. multi-peaked vs. noncluster), at ∼60–80% validation accuracy across the full PHANGS-HST sample (Hannon et al., 2023).

4. ISM Structure, Feedback, and Baryon Cycling

GMC properties and feedback-driven structures are revealed by combining ALMA, HST, JWST, and MUSE. Superbubbles associated with recent star formation are identified as closed CO shells with coincident young clusters. Expansion velocities (v_exp ≈ 5–25 km/s), radii (30–330 pc), and swept-up masses (10^5–7×10^7 M_⊙) are extracted. Kinematic ages (t_dyn ≈ 3–7 Myr) agree with cluster ages, supporting a SN blast-wave scenario (η ≈ 0.25 in t_dyn=ηR/v_exp), with SN energy coupling efficiency ε ≈ 10% into molecular shells (Watkins et al., 2023, Barnes et al., 2022).

JWST mid-IR imaging further unveils PAH and dust–mass fraction variations, revealing “holes” corresponding to H II regions (q_PAH drops by ~40% in H II vs. neutral gas). H II regions and larger superbubbles follow “Swiss-cheese” topologies carved by clustered feedback (R_voids ≳ 1000 pc, v_exp ~ 20–50 km/s), matching the predictions of modern galaxy simulations incorporating clustered SN feedback (Barnes et al., 2022, Sutter et al., 2024).

5. Dust, PAH Emission, and ISM Phase Diagnostics

JWST NIRCam/MIRI imaging in 2–21 μm filters enables mapping of key PAH bands (3.3, 7.7, 11.3 μm) at 10–50 pc resolution. Band ratios F335M/F1130W, F1130W/F770W, and their environmental trends are used to infer PAH size and ionization state as functions of radiation field, molecular fraction, and Hα intensity (Chastenet et al., 2022, Sandstrom et al., 2023). A robust proxy for the PAH mass fraction is provided by R_PAH = (F770W + F1130W)/F2100W, with q_PAH[%] ≈ 0.33 R_PAH + 1.3 for typical interstellar radiation fields (Sutter et al., 2024).

PAHs are found to be rapidly destroyed within H II regions (q_PAH ≈ 2–3%), yet remain abundant elsewhere in disk ISM (q_PAH ≈ 4–6%), confirming their central role in ISM cooling, chemistry, and photoelectric heating. The linear mapping between F1130W PAH brightness and total gas surface density down to Σgas ≈ 2 M⊙ pc^–2 enables cloud-scale imaging of CO-dark, low-density regions, filling a critical gap not accessible with standard CO or HI tracing (Sandstrom et al., 2022).

6. Gas Kinematics, Bars, and Secular Evolution

ALMA CO (2–1) kinematic maps provide 150 pc–scale rotation curves for 67 disks, with kinematic orientations tightly matching photometric axes and enabling robust dynamical mass and inner mass concentration estimates (Lang et al., 2020). Bar corotation radii and dynamical resonances (CR, ILR, OLR) are determined by gravitational torque methods combining Spitzer 3.6 μm stellar maps with CO molecular gas. For barred galaxies, the average ratio ℛ = R_CR/R_bar ≈ 1.12 ± 0.39 signifies predominantly fast bars, with gas torques driving inflow inside corotation and gas piling up at resonances associated with rings. The coupling between bar dynamics and ISM structure is manifest in systematic variations in excitation, star formation, and gas morphology across galactic environments (Ruiz-García et al., 2024, Lee et al., 17 Jul 2025).

7. Panchromatic Synergy and Legacy Science Impacts

PHANGS provides a uniquely panchromatic, spatially resolved database for empirical and theoretical studies of star formation laws, cluster formation and disruption, feedback-driven structure formation, dust/gas chemistry, PAH evolution, and secular dynamics in disk galaxies. The public release of aligned, flux-calibrated mosaics (FITS format) for all imaging and spectral products—including full error propagation, PSF-matched multi-band cubes, and value-added catalogs—enables joint, cross-phase analyses at the ∼5–50 pc scale.

PHANGS continues to serve as a reference dataset for testing ISM models, dust and PAH prescriptions, star cluster formation scenarios, feedback parameterizations in hydrodynamic and semi-analytic simulations, and the calibration of multi-wavelength SFR and dust-mass diagnostics (Lee et al., 2022, Williams et al., 2024, Hassani et al., 2023). The Cycle 2 expansion with JWST will extend the sample to 74+ galaxies, further solidifying PHANGS as the keystone nearby-galaxy microphysics survey for the post-JWST era.