Research on Pattern Recognition of Wind Speed Variation by Fusing Large-Scale GNSS-R Data

Global climate anomalies are becoming more and more frequent, and the demand for ocean wind field monitoring is increasing. In this study, we propose a wind speed variation pattern recognition method by fusing large-scale GNSS-R data, and realize the accurate recognition of six typical wind speed patterns through time-frequency feature extraction, spatial distribution pattern analysis and multi-scale wind speed feature modeling. The improved CNN-LSTM model significantly improves the accuracy of wind speed pattern recognition, especially in the early warning of extreme weather events. The research results provide a new idea to deepen the knowledge of the complex ocean wind field dynamics, which is of great significance to marine weather forecasting and climate change research.