Discovering Spatial Correlations of Earth Observations for weather forecasting by using Graph Structure Learning

TL;DR

This paper introduces CloudNine-v2, a graph neural network approach with adaptive structure learning to better capture dynamic spatial correlations in Earth observations, significantly improving weather prediction accuracy.

Contribution

It proposes an adaptive structure learning method for spatiotemporal graph neural networks to effectively model dynamic spatial correlations in weather data.

Findings

Achieved up to 15% RMSE reduction over existing models.

Outperformed baselines in high variability regions.

Validated on real-world East Asian atmospheric data.

Abstract

This study aims to improve the accuracy of weather predictions by discovering spatial correlations between Earth observations and atmospheric states. Existing numerical weather prediction (NWP) systems predict future atmospheric states at fixed locations, which are called NWP grid points, by analyzing previous atmospheric states and newly acquired Earth observations. However, the shifting locations of observations and the surrounding meteorological context induce complex, dynamic spatial correlations that are difficult for traditional NWP systems to capture, since they rely on strict statistical and physical formulations. To handle complicated spatial correlations, which change dynamically, we employ a spatiotemporal graph neural networks (STGNNs) with structure learning. However, structure learning has an inherent limitation that this can cause structural information loss and over-smoothing problem by generating excessive edges. To solve this problem, we regulate edge sampling by adaptively determining node degrees and considering the spatial distances between NWP grid points and observations. We validated the effectiveness of the proposed method (CloudNine-v2) using real-world atmospheric state and observation data from East Asia, achieving up to 15\% reductions in RMSE over existing STGNN models. Even in areas with high atmospheric variability, CloudNine-v2 consistently outperformed baselines with and without structure learning.