Abstract
The efficiency with which molecular gas is converted into stars remains an important open question in galaxy evolution, particularly in understanding what sets the star formation efficiency locally and how it varies across different environments. This is especially relevant in luminous infrared galaxies (LIRGs), where high gas densities, turbulence, and strong radiation fields shape the interstellar medium (ISM). In these systems, the physical conditions of molecular gas, i.e., temperature, density, pressure, and dynamical state, can vary dramatically across environments shaped by starbursts, active galactic nuclei (AGN), and galaxy interactions, complicating the interpretation of observables and the derivation of star formation properties. A key challenge is to understand how these local ISM conditions regulate star formation and how they combine to produce the global scaling relations observed in galaxies. In particular, it remains unclear how variations in gas excitation, optical depth, and gas kinematics influence the molecular Kennicutt Schmidt relation and the inferred connection between gas and star formation in merging systems.
In this dissertation, I address these questions by combining spatially resolved observations of molecular gas and star formation across multiple scales and merger stages in local LIRGs. First, I analyze the molecular Kennicutt–Schmidt relation in the early-stage merger Arp 240 at ∼ 500 pc resolution. I find that distinct star formation regimes coexist within the system, with high surface brightness regions following a near-linear relation and more diffuse regions exhibiting significantly flatter slopes. These differences are driven in part by spatial offsets between molecular gas and star formation tracers, demonstrating that a single power-law relation breaks down at subkiloparsec scales for this galaxy pair.
I then extend this analysis to a sample of sixteen LIRGs spanning the merger sequence. While individual galaxies show significant diversity in their star formation relations, this diversity likely reflects the role of gas inflows, dynamical evolution, and feedback in shaping the ISM. Despite these variations, the ensemble relation remains close to linear, indicating that global scaling laws emerge from the combined contribution of different local environments.
Finally, I investigate the excitation of molecular gas through spatially resolved CO spectral line energy distributions (SLEDs) in NGC 7469, spanning transitions from CO(1–0) to CO(8–7) at ∼120 pc resolution. I find strong environmental variations in excitation, with the nucleus exhibiting the highest excitation and the largest relative contribution from the mid- and high-J transitions, while the circumnuclear star-forming ring, a prominent kiloparsec-scale structure of intense star formation, occupies an intermediate regime and the outer disk shows lower excitation. The relative contribution of mid-J to low-J emission decreases with galactocentric radius, with a secondary enhancement associated with the star-forming ring, and the highest-J transitions are only detected in the brightest regions. These results indicate that galaxy-integrated measurements reflect a luminosity weighted superposition of multiple excitation states tied to specific galactic environments, rather than a single set of physical conditions, and highlight the importance of spatially resolving molecular gas when interpreting star formation and gas mass estimates.
These results provide quantitative, spatially resolved evidence that star formation in LIRGs depends on local ISM conditions that vary across environments and merger stages, and illustrate how global star formation relations arise from the aggregation of these distinct regimes. This work provides new insight into the physical origin of these relations and into the interpretation of molecular gas and star formation in both nearby and unresolved high-redshift galaxies.
