Research

My research focuses on spatially resolved star formation and stellar feedback processes in galaxies. I use cosmological zoom-in simulations, and develop analytic theories, in order to understand how local star formation connects to the properties of galactic disks and their environments. I am especially interested in how stellar feedback, in the form of supernovae, drives turbulence in the interstellar medium, and outflows of warm and cold material into the near-circumgalactic medium. I also work to produce synthetic observations from simulations to help differentiate between various star formation theories, validate empirical scalings, and determine the observational consequences of various feedback physics.


Light from the FIRE: CO, CII and H alpha Emission in Cosmological Simulations (Coming soon!)

A remaining challenge in connecting simulations with observations lies in how theorists typically report physical quantities taken directly from their simulations, while observers must derive the underlying properties from, e.g., line-widths and luminosities by making a chain of assumptions about molecular and atomic processes, and the dynamical state of the gas. Addressing this, I have developed a pipeline for producing synthetic observations of cosmological zoom-in simulations to directly compare spatially resolved observables predicted by simulations with observations. In the near future, I will be exploring the signatures of star formation, stellar feedback (e.g., cosmic rays, supernovae, etc.), and gas physics models in resolved galaxy scaling relations, probing various observables, like CO or C+ emission, to do so.


Swirls of FIRE: Connecting Turbulence and Feedback

Seeing the awesome power of supernova explosions, there is a natural question of how star formation is connected with the turbulence in the ISM. I studied how gas velocity dispersions and star formation rates relate in FIRE-2 disk galaxies, and was especially interested in the extent to which star formation is able to drive kpc-sized patches of disks away from marginal stability (i.e., Toomre's Q~1). The galaxies are never far from stability, and understanding why this occurs and how quickly local patches can regulate themselves, was a specific focus. I looked into the predictions of Orr et al. 2019 regarding delay times between star formation and (supernova) feedback, finding evidence for `on-off' cycles of star formation in the disks.

See the paper here: Orr et al. 2020, MNRAS


Do Galaxy Disks Breathe?

In the feedback-regulated model of star formation, the dissipation of supersonic turbulence (the primary support of cold and dense gas in galaxy disks) is balanced by the injection of stellar feedback momentum from ionizing radiation and supernovae. This is generally thought of as being in a "static equilibrium". However, the injection of supernova feedback is inherently bursty, time-dependent, and lags relative to star formation itself. As a result, galaxy disks may be expected to "breathe" to some extent. How much do typical disk galaxies breathe? I developed a simple analytic model to understand how star formation can drive itself to ~dex variations on 10 Myr timescales.

See the paper here: Orr et al. 2019, MNRAS 486, 4, 4724-4737


What FIREs Up Star formation? Kennicutt-Schmidt on FIRE

Cosmological simulations have given us a unique ability to understand how scaling relations in star formation emerge on kpc-scales in galaxies. I investigated the Kennicutt-Schmidt relation in the FIRE simulations- specifically, looking at how and at what spatial and temporal scales the relation breaks down, and its dependencies on redshift, metallicity, and gas surface density and star formation rate tracers. As well, I looked at what sets the extent of the star-forming disks in our galaxy simulations, finding that they were circumscribed by the limits of gravitational fragmentation in hot, ionized gas.

See the paper here: Orr et al. 2018, MNRAS 478, 3, 3653-3673


Stacking is Hacking

High-redshift observations of galaxies are difficult, and have low signal-to-noise. And so, observers often stack many spatially-resolved observations of similar galaxies to get a result. How does stacking affect inferred average star formation rate profiles? Turns out, quite a bit. High-redshift galaxies are very bursty in their star formation, and as a result stacking can bias our inferences of how those galaxies and their star formation rate profiles evolve.

See the letter here: Orr et al. 2017, ApJ 849:L2