loading page

Oceanic harbingers of Pacific Decadal Oscillation predictability in CESM2 detected by neural networks
  • Emily M Gordon,
  • Elizabeth A Barnes,
  • James Wilson Hurrell
Emily M Gordon
Colorado State University, Colorado State University

Corresponding Author:[email protected]

Author Profile
Elizabeth A Barnes
Colorado State University, Colorado State University
Author Profile
James Wilson Hurrell
Colorado State University, Colorado State University
Author Profile

Abstract

Predicting Pacific Decadal Oscillation (PDO) transitions and understanding the associated mechanisms has proven a critical but challenging task in climate science. As a form of decadal variability, the PDO is associated with both large-scale climate shifts and regional climate predictability. We show that artificial neural networks (ANNs) predict PDO persistence and transitions on the interannual timescale. Using layer-wise relevance propagation to investigate the ANN predictions, we demonstrate that the ANNs utilize oceanic patterns that have been previously linked to predictable PDO behavior. For PDO transitions, ANNs recognize a build-up of ocean heat content in the off-equatorial western Pacific 12-27 months before a transition occurs. The results support the continued use of ANNs in climate studies where explainability tools can assist in mechanistic understanding of the climate system.
16 Nov 2021Published in Geophysical Research Letters volume 48 issue 21. 10.1029/2021GL095392