Extratropical Atmospheric Circulation Response to ENSO in Deep Learning Pacific Pacemaker Experiments

TL;DR

This study uses a deep learning climate emulator to analyze atmospheric responses to ENSO, revealing amplified teleconnection responses and biases in extreme weather simulation, emphasizing the need for rigorous validation of such models.

Contribution

First application of coupled deep learning climate emulator in Pacific pacemaker experiments to assess ENSO teleconnections and biases.

Findings

Emulator captures internal atmospheric variability realistically.

Amplified teleconnection response to ENSO observed.

Biases in simulating atmospheric extremes identified.

Abstract

Coupled atmosphere-ocean deep learning (DL) climate emulators are a new frontier but are known to exhibit weak ENSO variability, raising questions about their ability to simulate teleconnections. Here, we present the first Pacific pacemaker (PACE) experiments using a coupled DL emulator (DLESyM) to bypass this weak variability and isolate the atmospheric response to observed ENSO forcing. We find that while the emulator realistically captures internal atmospheric variability, it produces a significantly amplified forced teleconnection response to ENSO. This amplified response leads to biases in simulating extremes, notably an overestimation of atmospheric blocking frequency and duration with the underestimation of peak intensity. Our findings underscore that coupled DL climate models require in-depth and physically-grounded validation, analogous to traditional numerical models, to build confidence in their use for physical climate analysis.