Deep Learning-Based Correction of Decadal Predictions of the PDO and TAG Indices

Jivesh Dixit

doi:10.1109/LGRS.2026.3662257

Deep Learning-Based Correction of Decadal Predictions of the PDO and TAG Indices

2026-02-09

0

PRE

AI

J

Jivesh Dixit

H

Hariprasad Kodamana

S

S. Sandeep

K

Krishna AchutaRao

DOI：10.1109/LGRS.2026.3662257

Original

Original request for help

Save

Abstract

En 中文

Conventional multimodel ensemble averages (MME) from hindcasts, such as those from the Decadal Climate Prediction Project component A (DCPP-A), can suffer from diluted predictive skill due to the averaging of models with varying performance. This dilution limits the skillful prediction of key Decadal Climate Variability (DCV) modes, such as the Pacific Decadal Oscillation (PDO) and the Tropical Atlantic SST Gradient (TAG). We propose a Long Short-Term Memory (LSTM) network to correct the ensemble averages of each participating model and hence the MME. The LSTM is trained on individual model ensemble averages and the MME to learn nonlinear relationships, producing a more skillful time series for the PDO and TAG indices, relative to the conventional MME, as quantified by multiple performance metrics. These enhancements include superior temporal accuracy, reduced error, and a closer statistical correspondence in phase, timing, and magnitude with the observations. Our study demonstrates that deep learning can extract a more skillful signal from climate ensembles, offering an approach to more reliable decadal predictions.

Keywords:

Decadal climate variability (DCV)

Long Short-Term Memory (LSTM)

Pacific Decadal Oscillation (PDO)

Tropical Atlantic SST Gradient (TAG)

Deep Learning-Based Correction of Decadal Predictions of the PDO and TAG Indices

Abstract

Journal

Organization

Cited Papers

Citing Papers