Data Science Webinar Series

We are pleased to announce the US CLIVAR Data Science Webinar Series. Organized by the Data Science Working Group, the webinar series will feature experts in Earth science, statistics, and computer science with the specific goal of fostering collaboration across these disciplinary boundaries. Our general mission is to accelerate the understanding, adoption, and further development of modern data science tools (artificial intelligence, machine learning, and statistics) for the analysis of large-to-massive climate data sets. The webinars will occur biweekly, with 30min talks and 15min for Q&A from the online audience. The talk and Q&A will be posted on the US CLIVAR Youtube channel as part of the US CLIVAR’s mission to serve the broader climate community and society. Upcoming webinars can be found on our Google Calendar.

If you are interested in attending the biweekly seminar, please fill out this Google Form and we will add you to our webinar's mailing list.

To learn more about the different US CLIVAR webinar series and instructions on how to join an upcoming webinar, click here. 

Next webinar: January 25, 2021 @ 3pm ET/12pm PT

Presenter: Karthik Kashinath (Lawrence Berkeley National Lab)

Talk: Physics-informed machine learning for weather and climate science

Abstract: Machine learning (ML) provides novel and powerful ways of accurately and efficiently recognizing complex patterns, emulating nonlinear dynamics, and predicting the spatio-temporal evolution of physical systems.

In the first part of the talk I will focus on how ML can be used for detecting and tracking extreme weather patterns in large climate datasets for highly precise climate analytics. In the second part of the talk I will describe a novel physics-informed deep learning approach for space-time super-resolution of turbulent flows of relevance to weather and climate modeling. I will conclude with some approaches to incorporating physics and domain knowledge into ML models.