About me
I am an applied mathematician and/or a climate scientist, often morphing daily from one into the other as I criss-cross the street. My research lies at the intersection of Ellis and 58th street, but strives to tell the truth everywhere else.
Beginning in January 2027, I will be starting a new role as Assistant Professor in the School of Complex Adaptive Systems at Arizona State University. I am extremely excited to begin this new chapter, beyond grateful to the many incredible mentors and collaborators that have brought me here, and eager to pay it forward.
I completed my PhD in Computational and Applied Mathematics (CAM) at the University of Chicago (and as a visitor at NYU), advised by Jonathan Weare. Subsequently I worked for three years as a postdoctoral associate at MIT in the department of Earth, Atmospheric, and Planetary Sciences, advised by Paul O'Gorman. In the past year I have been in a second postdoctoral appointment back at my Alma Mater of UChicago, jointly between the Data Science Institute and the department of Geophysical Sciences, advised by Pedram Hassanzadeh and Dorian Abbot. I study the dynamics and predictability of extreme weather events and climate tipping points, using frameworks and methods borrowed from statistical mechanics, reliability engineering, and statistics/data science/machine learning/artificial intelligence/latest synonymous buzzphrase.
Extreme weather events and climate tipping points are enormously consequential, but very hard to study through direct simulation, precisely because they are rare. My overarching goal is to get the most out of the available data, and generating the highest-value synthetic data possible by combining theoretical reasoning, numerical simulation, and data-driven emulation to characterize extreme events. My PhD focused on Sudden Stratospheric Warming (SSW), a spectacular breakdown of the stratospheric polar vortex that can wreak havoc on the Chicago winter. My postdoctoral work has focused closer to Earth, on extreme midlatitude precipitation and tracer turbulence, while making heavy use of model hierarchies to distill the thorny algorithmic challenges that plague us whenever trying to apply algorithms out-of-the-box to new applications.