Research

Climate models, AI systems, and the people using them.

Climate modeling

I’m interested in how physical models and machine learning can help us understand the climate, and how to evaluate what those models tell us.

My public project Morro explores forecasts and simulated cloud-seeding interventions. It compares perturbed and unperturbed weather-model runs, with a dashboard for examining the differences.

Human–computer interaction

I also work in human–computer interaction, with an interest in evaluating AI systems and making them safer to use. I’m part of UC San Diego’s Data Science Teaching and Learning Lab.

Some of that interest shows up in my software: Arnold asks people to review a meal entry before saving it, and Wendy separates drafting an inquiry from sending one.

Earlier research

At UC Santa Barbara’s 2024 Summer Research Academies, I worked with Haeum Lee and Simran Karkhanis on Soil and Solar: Analyzing the Social Barriers to Agrivoltaics Implementation in the United States.

Our capstone talk is listed in the 2024 program (page 11).

From models to decisions

The questions I keep returning to are practical: What can we check? What does a result actually show? How much should someone trust it?

For public code and demonstrations, see my projects. For research conversations, email me.