Numerical modeling of tropospheric atmosphere based on GCN-GRU

This paper presents a novel approach for numerical modeling of the tropospheric atmosphere using a Graph Convolutional Network combined with Gated Recurrent Units (GCN-GRU).The GCN-GRU model integrates spatial and temporal correlations in atmospheric data to enhance the accuracy of tropospheric waveguide prediction.By leveraging the strengths of GCN in capturing spatial dependencies and GRU in handling temporal sequences, this hybrid model demonstrates significant improvements over traditional methods such as the Weather Research and Forecasting (WRF) model.The study focuses on the application of GCN-GRU to predict atmospheric waveguide events over the East China Sea and South China Sea, achieving a prediction accuracy improvement of over 13.56% compared to conventional models.The results highlight the potential of GCN-GRU for high-precision atmospheric numerical prediction and waveguide discrimination.