MR-GNF: Multi-Resolution Graph Neural Forecasting on Ellipsoidal Meshes for Efficient Regional Weather Prediction

TL;DR

MR-GNF is a lightweight, physics-aware graph neural network that efficiently performs multi-scale regional weather forecasting directly on ellipsoidal Earth meshes, matching or surpassing heavier models at low computational cost.

Contribution

This paper introduces MR-GNF, a novel multi-resolution graph neural network for weather prediction that captures 3-D structures and operates efficiently on multi-scale Earth meshes.

Findings

Achieves +6 to +24 hour forecasts for temperature, wind, and precipitation.

Operates with only 1.6 million parameters and under 80 GPU-hours.

Matches or exceeds the performance of larger regional AI systems.

Abstract

Weather forecasting offers an ideal testbed for artificial intelligence (AI) to learn complex, multi-scale physical systems. Traditional numerical weather prediction remains computationally costly for frequent regional updates, as high-resolution nests require intensive boundary coupling. We introduce Multi-Resolution Graph Neural Forecasting (MR-GNF), a lightweight, physics-aware model that performs short-term regional forecasts directly on an ellipsoidal, multi-scale graph of the Earth. The framework couples a 0.25{\deg} region of interest with a 0.5{\deg} context belt and 1.0{\deg} outer domain, enabling continuous cross-scale message passing without explicit nested boundaries. Its axial graph-attention network alternates vertical self-attention across pressure levels with horizontal graph attention across surface nodes, capturing implicit 3-D structure in just 1.6 M parameters. Trained on 40 years of ERA5 reanalysis (1980-2024), MR-GNF delivers stable +6 h to +24 h forecasts for near-surface temperature, wind, and precipitation over the UK-Ireland sector. Despite a total compute cost below 80 GPU-hours on a single RTX 6000 Ada, the model matches or exceeds heavier regional AI systems while preserving physical consistency across scales. These results demonstrate that graph-based neural operators can achieve trustworthy, high-resolution weather prediction at a fraction of NWP cost, opening a practical path toward AI-driven early-warning and renewable-energy forecasting systems. Project page and code: https://github.com/AndriiShchur/MR-GNF