About

Hi! I’m Julian, a fourth-year Ph.D. student in Environmental Science and Engineering at Caltech. I work with Tapio Schneider as a member of the Climate Modeling Alliance, where I develop and improve weather and climate models and mathematical tools to support them. During a 6 month internship at Google X during H1 2026, I lead a project on AI for precipitation forecasting.

At CliMA, I’m developing a new framework for diagnosing and improving model calibration in the non-differentiable slow forward model setting. I’ve been applying this framework to CliMA’s atmosphere model, ClimaAtmos.jl, to develop more efficient loss functions for use in global calibrations. I’m also developing a new closure for clouds focusing on the idea that machine learning closures trained originally with gradient descent can be further improved by learning online to longer-running statistics.

Beyond research, I coordinate weekly trail runs for the Caltech Alpine Club in the beautiful San Gabriel Mountains, where I get to enjoy both the outdoors and the awesome local running community.

Before Caltech, I completed my Applied Mathematics degree at Harvard University in 2023, where I had the opportunity to work on diverse projects spanning high-performance computing, climate modeling, and statistical methods. I collaborated with Marine Denolle on seismic waveguide analysis using Julia and AWS, with Mimi Hughes, Nathaniel Johnson, and Kai-Chih Tseng on snow drought climatology, and with Kelly McConville on forest carbon estimation. I also worked briefly as a software engineer at Coolant, developing machine learning tools for carbon stock quantification.