Abstract: Accurate El Ni?o-Southern Oscillation (ENSO) predictions are important for preventing extreme weather and climate events. Although deep learning has made significant progress in the field of ENSO prediction, existing models still have limitations in capturing complex dynamic spatial and temporal correlations. To address the above problems, the study uses global SST data provided by the UK Met Office Hadley Centre, and proposes an ENSO prediction model based on the Dual Spatial Temporal Network (DualST) architecture, which innovatively designs a Spatiotemporal Graph Convolutional Network (ST-GCN) that learns the connectivity between ENSO spatiotemporal data through a novel edge computing module and converts it to graph data, enabling the graph convolutional network to accurately capture the complex spatiotemporal dynamic characteristics of ENSO. In addition, the advantage of the attention mechanism, which is good at capturing long-range dependencies, is exploited to improve the long-term prediction accuracy of ENSO. The experimental results show that DualST significantly outperforms the other state-of-the-art deep learning models compared in terms of Ni?o 3.4 index prediction correlation coefficients and can achieve 20 months of effective ENSO prediction, in addition to a more moderate increase in MAE and RMSE. Thus, it is shown that the DualST model effectively modelled the dynamic connectivity between SST fields through innovative modules such as edge calculation, which significantly improved the long-term prediction performance of ENSO.