Prediction of 3-D Ocean Temperature Based on Self-Attention and Predictive RNN

• Published in: IEEE geoscience and remote sensing letters Vol. 21; pp. 1 - 5
• Main Authors: Yue, Weihao, Xu, Yongsheng, Xiang, Liang, Zhu, Shanliang, Huang, Chao, Zhang, Qingjun, Zhang, Liqiang, Zhan, Xiangguang
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Climate change; Deep learning; Global climate; Global gridded temperature data with Barnes objective analysis (BOA-ARGO); Neural networks; Objective analysis; Ocean temperature; Oceans; Predictions; Predictive models; predictive recurrent neural network (PredRNN); Recurrent neural networks; self-attention; Spatial dependencies; Temperature data; Temperature distribution; Temperature fields; Temperature measurement; Temporal variations
• ISSN: 1545-598X; 1558-0571
• DOI: 10.1109/LGRS.2024.3358348

Predicting the 3-D ocean temperature field is a significant task that helps to understand global climate change and the state of ocean motion. Lots of numerical and data-driven models are used to predict ocean temperatures. However, these methods are restricted to the time-sequence prediction of discrete points or rely on convolutional layers to inefficiently capture local spatial dependencies for spatio-temporal prediction. In this letter, we propose a deep learning model named self-attention predictive recurrent neural network (SA-PredRNN) that combines attention mechanisms and predictive recurrent neural networks to capture global positional correlations and spatio-temporal features. Global gridded Argo temperature data with Barnes objective analysis (BOA-ARGO) are used to predict the future 3-D ocean temperature. The average root mean square errors (RMSEs) of the proposed model are promoted by at most 11% and 10%, which indicates that the SA-PredRNN model has a better performance than the other baseline models.