Stellar structures, molecular gas, and star formation across the PHANGS sample of nearby galaxies

TL;DR

PHANGS demonstrates that stellar structures imprint a strong, coherent organisation of molecular gas and star formation across 74 nearby galaxies by constructing detailed morphological masks from Spitzer 3.6 μm images. Using ALMA CO(2-1) data and multi-wavelength SFR tracers, the study finds centres to host the highest densities and typically the shortest depletion times, while spiral arms and bars show substantial internal diversity but do not universally boost star formation efficiency. The molecular Kennicutt–Schmidt relation remains nearly linear across environments ($N\approx 1$), though the central regions exhibit the largest offsets; arm/interarm contrasts yield typical density ratios of a few. Overall, morphology governs gas organization, with the impact on global star formation efficiency being nuanced and time-variable, underscoring the need for environment-aware analyses in galaxy evolution.

Abstract

We identify stellar structures in the PHANGS sample of 74 nearby galaxies and construct morphological masks of sub-galactic environments based on Spitzer 3.6 micron images. At the simplest level, we distinguish centres, bars, spiral arms, interarm and discs without strong spirals. Slightly more sophisticated masks include rings and lenses, publicly released but not explicitly used in this paper. We examine trends using PHANGS-ALMA CO(2-1) intensity maps and tracers of star formation. The interarm regions and discs without strong spirals dominate in area, whereas molecular gas and star formation are quite evenly distributed among the five basic environments. We reproduce the molecular Kennicutt-Schmidt relation with a slope compatible with unity within the uncertainties, without significant slope differences among environments. In contrast to early studies, we find that bars are not always deserts devoid of gas and star formation, but instead they show large diversity. Similarly, spiral arms do not account for most of the gas and star formation in disc galaxies, and they do not have shorter depletion times than the interarm regions. Spiral arms accumulate gas and star formation, without systematically boosting the star formation efficiency. Centres harbour remarkably high surface densities and on average shorter depletion times than other environments. Centres of barred galaxies show higher surface densities and wider distributions compared to the outer disc; yet, depletion times are similar to unbarred galaxies, suggesting highly intermittent periods of star formation when bars episodically drive gas inflow, without enhancing the central star formation efficiency permanently. In conclusion, we provide quantitative evidence that stellar structures in galaxies strongly affect the organisation of molecular gas and star formation, but their impact on star formation efficiency is more subtle.

Figures (11)

• Figure 1: Spitzer IRAC $3.6$$\mu$m images illustrating the morphological structures considered in our environmental masks. Sect. \ref{['Sec:masks']} explains the mask construction scheme in detail. Bulges and discs are defined on photometric decompositions of near-infrared images (2015ApJS..219....4S for S$^4$G galaxies; 2004MNRAS.355.1251L or new fits otherwise). The sizes of bars, rings and lenses are defined visually and their ellipticity is measured via ellipse fitting (2015AA...582A..86H for S$^4$G, with additional measurements from the literature). Spiral arms are identified as peaks on unsharp-masked $3.6$$\mu$m images followed by log-spiral fits in polar coordinates, the width being assigned based on CO emission (2015AA...582A..86H and new measurements). The galaxies shown are, from left to right, NGC 2775, NGC 628, NGC 1300, NGC 3627, NGC 3351 and NGC 4457; they are all displayed using an arcsinh stretch.
• Figure 2: Notation used in the 'simple' masks, where each pixel is uniquely assigned to a dominant environment. The background image is the Spitzer$3.6$$\mu$m map of NGC 1300 and the different colours denote different environments. Several of these environments can be grouped together for further simplicity, as indicated in the bottom-right diagram and in Table \ref{['table:masknotation']}.
• Figure 3: Strategy followed to assign emission to environments illustrated with NGC 3627. For the census of area, $M_{\rm mol}$, and SFR as presented in Sect. \ref{['Sec:piechart']}, we consider all the emission within the mask footprint of each environment (working at our native resolution of ${\sim}1\arcsec$). For the analysis presented in Sects. \ref{['Sec:surfdens']}--\ref{['Sec:contrasts']}, we turn to measurements on kpc-sized hexagonal apertures, so that we implicitly average over the star-forming cycle and SFRs are more robustly estimated.
• Figure 4: Stacked bar charts showing the relative distribution of area, integrated molecular gas mass, and integrated star formation rates in each of the environments that we consider in this paper and across the entire PHANGS--ALMA sample of galaxies. These measurements consider the full resolution of the data as explained in Sect. \ref{['Sec:piechart']} and are limited to the ALMA field of view.
• Figure 5: Molecular gas and star formation rate surface densities as a function of galactocentric radius (normalised to $R_{25}$) for kpc-scale measurements across the PHANGS--ALMA sample of galaxies. The colours correspond to the different environments that we analyse in this paper. For clarity, the middle and right panels display alternative versions of the plots showing contours of data point density for different environments, highlighting their relative offsets (contours encompass 30%, 50%, and 80% of all data points in each category).