Sea Temperature Forecast Model Based on PCA and LSTM
Full text can be downloaded on PC at the following address:
http://www.hyyb.org.cn/Magazine/Show.aspx?ID=3464
Reading Notes
Authors: Li Jingshi1 2 Kuang Xiaodi1 2 Li Qiong3 He Enye1 2 Zhang Yubai3 Yuan Chengyi4 Zhang Yanlin5
Affiliations: 1. National Marine Environmental Forecasting Center, Beijing 100081; 2. Key Laboratory of Marine Disaster Forecasting Technology, Ministry of Natural Resources, Beijing 100081; 3. Shandong Marine Disaster Reduction Center, Qingdao 266104; 4. Tianjin University of Science and Technology, Tianjin 300222; 5. Liaoning Provincial Natural Resources Service Center, Shenyang, Liaoning 110033
Classification Number: P731.31
Publication Year·Volume·Issue (Page Numbers): 2023·40·Issue 2 (1-10)
Abstract: Utilizing self-built buoy sea temperature observation data from Rongcheng and Haiyang stations, along with meteorological numerical forecast data from the regional atmospheric model Weather Research and Forecasting (WRF), this paper proposes a PCA-LSTM sea temperature forecasting model suitable for single-station sea surface temperature forecasting based on Principal Component Analysis (PCA) and Long Short-Term Memory (LSTM) neural network. This model can provide sea temperature forecasts for 24 to 120 hours, significantly improving forecasting accuracy compared to numerical and statistical models.
Keywords:Principal Component Analysis; Long Short-Term Memory Neural Network; Sea Temperature Forecast
Key Words:Principal Component Analysis; Long Short-Term Memory Neural Network; SST Forecast
Readers can also access the journal content through the public account menu
Enter the “Related Knowledge” section for the “Current Directory” and “Back Issue Browsing”
To view detailed journal content