Ian Baxter (University of Chicago, USA) will give this seminar, organized by the ELI-C department. 

This seminar will be held exclusively via Zoom. Join the meeting here. 

Abstract

In recent decades, regional long-term trends have emerged under continued greenhouse gas emissions. 

Many of these signals of forced changes in the Earth System were broadly predicted using the first general circulation models, including amplified Arctic warming. However, as anthropogenic emissions have continued, current models have been shown to struggle in capturing the sign or strength of many regional changes that have been seen in observations and reanalyses. 

I will discuss tools that we are employing to probe the key processes driving these trends and how these processes are represented in Earth System Models. One such tool is atmospheric nudging, where we “replay” reanalysis winds or temperature over the satellite era in the Community Earth System Model (CESM). With this we can quantify the importance of large-scale circulation trends for regional moisture changes over the satellite era. 

We then combine nudging with water isotope tagging to show the primary pathways by which water vapor has been transported into regions with discrepant trends, which has highlighted the role of land surface processes. By constraining dynamics in the model, questions can be asked of model parameterizations and coupling which may be leading to model-observation discrepancies. 

Briefly, I will also discuss the use of AI models or atmospheric emulators, which replace the physics with machine learning approaches, in better understanding discrepant regional trends. I’ll show that partitioning in the responses to carbon dioxide and sea surface temperatures by the AI models determines their ability to reproduce observed regional trends and whether these trends are generated for what we think are the right reasons.

The guest 

Ian Baxter is a postdoctoral researcher at the University of Chicago working with Tiffanny Shaw on large-scale climate variability and change. He has obtained his PhD from the University of California, Santa Barbara in Arctic climate modeling with a focus on sea ice supervised by Qinghua Ding in 2024.