Explainable Graph Neural Networks for Observation Impact Analysis in Atmospheric State Estimation

TL;DR

This paper introduces an explainable graph neural network approach to analyze the impact of various observational data on atmospheric state estimation, improving understanding and optimization in weather forecasting.

Contribution

It integrates meteorological data into a GNN framework with explainability methods to quantify observation importance in atmospheric state estimation.

Findings

Effective visualization of observation importance.

Enhanced understanding of observation impact.

Potential for optimizing observational data collection.

Abstract

This paper investigates the impact of observations on atmospheric state estimation in weather forecasting systems using graph neural networks (GNNs) and explainability methods. We integrate observation and Numerical Weather Prediction (NWP) points into a meteorological graph, extracting $k$-hop subgraphs centered on NWP points. Self-supervised GNNs are employed to estimate the atmospheric state by aggregating data within these $k$-hop radii. The study applies gradient-based explainability methods to quantify the significance of different observations in the estimation process. Evaluated with data from 11 satellite and land-based observations, the results highlight the effectiveness of visualizing the importance of observation types, enhancing the understanding and optimization of observational data in weather forecasting.