Climate Analytics Lab

Improving climate projections with machine learning

Teaching

SIO Courses

SIO(c) 209 - Deep Learning for Environmental/Geo Science

This hands-on graduate course introduces the application of machine learning in the environmental sciences. It covers the main classes of supervised and unsupervised learning, from neural-network regression and classification through object detection, physics-informed neural networks, clustering and dimensionality reduction, to generative models and large language models. Students get practical experience training and validating models on real-world environmental data, including climate model emulation with ClimateBench, and finish with a final project. Last taught in Spring 2026.

  • Course notes are in our online course book, which is still being refined.
  • The Spring 2026 notes, slides and schedule are on GitHub.

HDSI Courses

DSC 80 - Practice and Application of Data Science

This course provides an in-depth exploration of data science principles and techniques, bridging lower- and upper-division coursework while covering algorithms, statistics, machine learning, visualization, and data systems, with prerequisites DSC 30 and DSC 40A, and priority given to DS25 majors.

  • Materials from the Winter 2025 offering are on GitHub.

DSC 200 - Data Science Programming

Programming for data science in the age of coding agents. Most working code is now produced by steering AI agents rather than written by hand, so this course focuses on the judgment around the code: turning ambiguous requirements into precise specifications, choosing data structures, algorithms and architectures with the right trade-offs, validating that generated code and its outputs are actually correct, and knowing when a project is done. It covers Python and the PyData ecosystem (NumPy, pandas, xarray, matplotlib) at the level needed to read, verify and steer code, along with algorithms and data structures, working with coding agents, and software engineering fundamentals, and ends with a team project.

  • The Fall 2026 schedule, lectures and AI policy are on GitHub. The syllabus of record and quizzes are on the course Canvas page.

DSC 180A/B - Data Science Capstone

Data science capstone course. Students work in teams on a year-long climate data science project, currently on differentiable climate model emulation in JAX. Project management, communication, and teamwork skills are emphasized.

  • Please see the course page here for details.