Atmosphere Ocean Science Colloquium

---

DED on arrival: Data, Extremes, and Dynamics

Speaker: Edwin P. Gerber, NYU Courant

Location: Warren Weaver Hall 1302

Date: Wednesday, October 22, 2025, 3:30 p.m.

Synopsis:

In a presentation geared towards newly arrived members of our community, I will discuss two research projects led by former PhD students. In both cases, the goal was to use data-driven methods (statistical and machine learning models) to explore the dynamics of extreme events in our atmosphere.  In the first, the thesis work of Justin Finkel, we use short trajectories from a large ensemble weather prediction dataset to probe extreme events in the stratosphere.  I’ll discuss how the statistics of rare events can be devilishly hard to estimate without a tremendous amount of data, and how we used this “ensemble of opportunity”  (academese for model integrations done for another purpose) to estimate the statistics of extremely rare Sudden Stratospheric Warming events using a general mathematical framework for exploring rare events.   The second portion of the talk will highlight work by Huan Zhang to understand atmospheric blocking events, a (somewhat) rare synoptic pattern of variability associated with extreme weather.  Here, a deep learning predictor (a convolutional neural network, CNN) was trained to predict the persistence of blocking events.  While the CNN had impressive skill, our true goal was to use it to back out dynamical insights into persistent blocking patterns and explore how similar predictions could be made with scarce reanalysis data with the use of transfer learning.   Both of these projects were the fruit of a collaboration with Jonathan Weare and Dorian Abbot (Chicago), and I hope to encourage our PhD students and postdocs to collaborate with our colleagues in the math department and beyond.