Overview

I am primarily a computational astrophysicist studying how galaxies grow and evolve across cosmic time. A lot of my work focuses on decoding the life stories of galaxies - how they form stars, interact with their environments, and change over billions of years. In this, studying galaxies is similar to studying the rise and fall of other complex systems like cities or forests - there are processes acting on many different scales in space and time that affect them, and understanding their interplay gives us insights into why they are so diverse when observed as a population. For a city, this could mean its location (is it near a river or trade route), the demographics of its population (what is the median age of the workforce? are people settling there or leaving?), and its self-regulatory processes (how do prices/taxes go up/down in response to various factors?). You can find my talks about galaxy evolution at a public level here or at a conference/colloquium-like level here or here (more talks in the travel & outreach pages).

When it comes to galaxy evolution, my research interests broadly span the following areas:

• What are the physical processes that regulate star formation in galaxies? What scales do they act on and what physical imprints do they leave in their stellar populations?
• How can we robustly infer the physical properties of galaxies from their multiwavelength observations?
• What can cosmological simulations tell us about how feedback and baryon cycling are imprinted on present-day galaxy populations?
• How can theoretical models of star formation as a stochastic process help bridge theory / simulations and observations?
• How do galaxies form at the earliest epochs of the universe? How is star formation different at these epochs under more extreme conditions, how do the first supermassive black holes form?
• What is the relation between galaxies and their globular cluster populations today? What clues can looking at star clusters in the distant past provide?
• What factors explain the observed diversity of galaxies today, and how is that tied to their past?

More generally, I like to find interesting use-cases for astrostatistics and machine learning in noisy, heterogeneous datasets, and am a big fan of developing interpretable or physically-motivated ML methods, or using AI for model building. This ranges from studies like Katachi (which shows which parts of an input galaxy image drive a networks prediction about its past star formation activity) to GP-SFH (which uses physical prescriptions in a Gaussian process kernel to infer star formation variability in a population of galaxies).

Over the last five-ish years, I have also been interested in the problem of keeping up with the rapidly increasing astronomy literature, and have been looking for better methods to keep up with it, ranging from literature survey and knowledge discovery to hypothesis generation. The recent introduction of large language models in this area has revolutionized what we can do, and I am excited to see where we will go from here!