Star formation, from cold giant molecular clouds to diverse population of stars, is a complex process involving a wide variety of physical processes. In this work, we constrain the link between the gas-star formation cycle and several secular and environmental probe of galaxies. Specifically, we quantify the spatial correlation between molecular gas and star-forming regions for 49 nearby galaxies using the ALMA and narrowband-Hα imaging from the PHANGS survey. At the resolution (150 pc) at which the individual molecular clouds and star-forming regions can be identified, we find that molecular clouds and star-forming regions do not necessarily coexist. The decoupled molecular clouds and star-forming regions are a signature of evolutionary cycling and feedback of the star formation process. Therefore, the impact of galactic-scale conditions and environments must be considered for a complete understanding of how stars form in galaxies and how this process influences the evolution of the host galaxies.

Interactions and mergers have been known as key scenarios to enhance global star formation rates and to lower the metal content of galaxies. However, little is known on how interactions affect the spatial distribution of gas metallicities. Thanks to the SDSS-IV MaNGA survey we are able to statistically constrain the impact of interactions across the optical distributions of galaxies. In this study, we compare the radial distribution of the ionized gas metallicity from a sample of 329 interacting objects – covering different interaction stages – with a statistical robust control sample. Our results suggest that galaxies close to coalesce tend to have flat, lower metallicities than non-interacting star-forming galaxies.

Observations and simulations have now reached the point where the giant molecular cloud (GMCs) populations can be studied over a whole galaxy. This is immensely helpful for understanding star formation. Yet, are these two groups really comparing the same objects? While simulators work in 6D (x, y, z, vx, vy, vz) position-position-position (PPP) space, observers see 2 + 1D (RA, Dec, vlos) projected properties along the line of sight, identifying clouds in position-position-velocity (PPV) space. In this research we generated PPP and PPV data for a high-resolution simulated galaxy and compared the identified clouds in both data sets. The results show that 70% of the clouds have a single counterpart in each data structure. Cloud boundaries of these clouds are indeed the same. Scatter of the derived cloud properties (radius and velocity dispersion) between PPP and PPV are typically within a factor of two. However, this small scatter can make it difficult to determine if a cloud is truly gravitationally bound.

Our new compilation of interferometric CO data suggests that nuclear and extended molecular gas disks are common in the final stages of mergers. Comparing the sizes of the molecular gas disk and gas mass fractions to early-type and late-type galaxies, about half of the sample show similar properties to early-type galaxies, which have compact gas disks and low gas mass fractions. We also find that sources with extended gas disks and large gas mass fractions may become disk-dominated galaxies.