ACE2: Accurately learning subseasonal to decadal atmospheric variability and forced responses

TL;DR

ACE2 is a large-scale machine learning emulator that accurately reproduces atmospheric variability and responses over 80 years, capturing emergent phenomena and climate trends, with some limitations in sensitivity to specific external forcings.

Contribution

The paper introduces ACE2, a novel autoregressive machine learning model capable of simulating atmospheric variability across multiple timescales with high accuracy and stability.

Findings

Reproduces tropical cyclones, MJO, and stratospheric warmings.

Accurately models atmospheric response to El Niño and temperature trends.

Operates efficiently, simulating ~1500 years per day.

Abstract

Existing machine learning models of weather variability are not formulated to enable assessment of their response to varying external boundary conditions such as sea surface temperature and greenhouse gases. Here we present ACE2 (Ai2 Climate Emulator version 2) and its application to reproducing atmospheric variability over the past 80 years on timescales from days to decades. ACE2 is a 450M-parameter autoregressive machine learning emulator, operating with 6-hour temporal resolution, 1{\deg} horizontal resolution and eight vertical layers. It exactly conserves global dry air mass and moisture and can be stepped forward stably for arbitrarily many steps with a throughput of about 1500 simulated years per wall clock day. ACE2 generates emergent phenomena such as tropical cyclones, the Madden Julian Oscillation, and sudden stratospheric warmings. Furthermore, it accurately reproduces the atmospheric response to El Ni\~no variability and global trends of temperature over the past 80 years. However, its sensitivities to separately changing sea surface temperature and carbon dioxide are not entirely realistic.