Improving the Predictability of the Madden-Julian Oscillation at Subseasonal Scales with Gaussian Process Models

TL;DR

This paper introduces a Gaussian process-based model for MJO prediction that improves forecast accuracy and uncertainty quantification over neural network approaches, extending reliable probabilistic coverage.

Contribution

We develop a nonparametric Gaussian process model with covariance correction that enhances MJO forecast skill and uncertainty estimation at subseasonal scales.

Findings

Better prediction skill than neural networks for first five lead days

Extends probabilistic coverage by over three weeks

Provides calibrated uncertainty estimates

Abstract

The Madden--Julian Oscillation (MJO) is an influential climate phenomenon that plays a vital role in modulating global weather patterns. In spite of the improvement in MJO predictions made by machine learning algorithms, such as neural networks, most of them cannot provide the uncertainty levels in the MJO forecasts directly. To address this problem, we develop a nonparametric strategy based on Gaussian process (GP) models. We calibrate GPs using empirical correlations and we propose a posteriori covariance correction. Numerical experiments demonstrate that our model has better prediction skills than the ANN models for the first five lead days. Additionally, our posteriori covariance correction extends the probabilistic coverage by more than three weeks.