Research on Sea Surface Temperature Prediction and El Niño Phenomenon Prediction Based on Graph Neural Networks

As the impact of global climate change on human society and the natural environment becomes increasingly significant, accurately predicting the El Niño phenomenon has become crucial. The El Niño phenomenon is a natural event characterized by abnormal warming of sea surface temperatures in the Pacific Ocean, which has profound effects on global weather patterns and climate systems. Traditional forecasting models, such as statistical and physical models, perform well in short-term predictions but show a significant decline in accuracy as the time span increases. This study explores the application of the GraphSAGE model based on Graph Neural Networks (GNN) using sea surface temperature (SST) data to predict El Niño phenomena. By modeling SST data as a graph structure, GraphSAGE captures the latent relationships between grid nodes, improving long-term prediction accuracy. Experimental results show that compared to traditional machine learning methods, the GraphSAGE model demonstrates higher predictive accuracy over extended time spans.