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.

本研究通过使用图神经网络（GNNs）和可解释性方法，研究了观测对天气预报系统中的大气状态估计的影响。我们将观测和数值天气预报（NWP）点整合到一个气象图中，提取以 NWP 点为中心的 k 跳子图。采用自监督的 GNNs 来聚合这些 k 跳半径内的数据，估计大气状态。通过基于梯度的可解释性方法，评估了来自 11 个卫星和陆地观测数据，结果突出了可视化观测类型重要性的有效性，增强了对天气预报中观测数据的理解和优化。