Probabilistic Weather Forecasting with Hierarchical Graph Neural Networks

TL;DR

This paper introduces Graph-EFM, a hierarchical graph neural network model for probabilistic weather forecasting that efficiently generates large ensembles and accurately captures forecast uncertainty, outperforming some deterministic models.

Contribution

The paper presents a novel hierarchical graph neural network model that enables fast, probabilistic weather forecasts with uncertainty quantification, a significant advancement over existing deterministic models.

Findings

Achieves lower or comparable errors to deterministic models.

Effectively captures forecast uncertainty in ensemble predictions.

Enables fast sampling of large, spatially coherent forecast ensembles.

Abstract

In recent years, machine learning has established itself as a powerful tool for high-resolution weather forecasting. While most current machine learning models focus on deterministic forecasts, accurately capturing the uncertainty in the chaotic weather system calls for probabilistic modeling. We propose a probabilistic weather forecasting model called Graph-EFM, combining a flexible latent-variable formulation with the successful graph-based forecasting framework. The use of a hierarchical graph construction allows for efficient sampling of spatially coherent forecasts. Requiring only a single forward pass per time step, Graph-EFM allows for fast generation of arbitrarily large ensembles. We experiment with the model on both global and limited area forecasting. Ensemble forecasts from Graph-EFM achieve equivalent or lower errors than comparable deterministic models, with the added benefit of accurately capturing forecast uncertainty.