Reconciling extragalactic star formation efficiencies with theory: insights from PHANGS

New extragalactic measurements of the cloud population-averaged star formation efficiency per free-fall time, ϵ_ff, from PHANGS show little sign of a theoretically predicted dependence on the gas virial level and weak variation with cloud-scale gas velocity dispersion. We explore ways to bring theory into consistency with the observations, particularly by highlighting systematic variations in internal density structure that must accompany an increase in virial parameter typically found toward denser galaxy centers. To introduce these variations into conventional turbulence-regulated star formation models, we adopted three adjustments, all motivated by the expectation that the background host galaxy has an influence on the cloud scale: (1) We incorporate self-gravity and an internal density distribution that contains a broad power-law (PL) component and resembles the structure observed in local resolved clouds; (2) We allow the internal gas kinematics to include motion in the background potential and let this regulate the onset of self-gravitation; (3) We assume that the distribution of gas densities is in a steady state for only a fraction of a cloud free-fall time. In practice, these changes significantly reduce the efficiencies predicted in multi-free-fall (MFF) scenarios compared to purely lognormal probability density functions (PDFs) and tie efficiency variations to variations in the slope of the PL α. We fit the model to PHANGS measurements of ϵ_ff to identify the PL slopes that yield an optimal match. These slopes vary systematically with galactic environment in the sense that gas that sits furthest from virial balance contains fractionally more gas at high density. We relate this to the equilibrium response of gas in the presence of the galactic gravitational potential, which forces more gas to high density than characteristic of fully self-gravitating clouds. Viewing the efficiency variations as originating with time evolution in the PL slope, our findings would alternatively imply coordination of the cloud evolutionary stage within environment. With this “galaxy regulation” behavior included, our preferred “self-gravitating” multi-freefall sgMFF models function similarly to the original, roughly “virialized cloud” single-free-fall models. However, outside the environment of disks with their characteristic regulation, the flexible MFF models may be better suited.