A5103067648- Meteorological Phenomena and Simulations
- Tropical and Extratropical Cyclones Research
- Climate variability and models
- Precipitation Measurement and Analysis
- Spectroscopy and Chemometric Analyses
- Seismic Waves and Analysis
- Spectroscopy and Laser Applications
- Calibration and Measurement Techniques
- Cryospheric studies and observations
Institute of Atmospheric Physics
2024-2025
Chinese Academy of Sciences
2024-2025
University of Chinese Academy of Sciences
2024-2025
Beijing Academy of Artificial Intelligence
2025
Abstract The increasing volume of satellite data, particularly hyperspectral infrared combined with the real‐time monitoring requirements numerical weather prediction (NWP) systems, has heightened demand for computational efficiency and accuracy in radiative transfer models (RTM). Machine learning (ML) offers a promising approach, numerous studies on ML‐based RTM have emerged recently. However, existing RTMs were not end‐to‐end. Moreover, since label data do represent truth, trained loss...
Abstract Target observations have garnered significant attention owing to their successful applications in enhancing forecasting skills of extreme weather events, particularly tropical cyclone (TC) events. The key step implementing target observation is determine the sensitive area advance. Previous studies often obtained areas for TC by vertically integrating energy optimal perturbation and taking horizontal large energy, an attempt use it represent roughly sensitivity whole atmospheric...
Abstract The artificial intelligence (AI)‐based weather forecasting model named FuXi and its data assimilation (DA) system FuXi‐En4DVar has been developed for high‐efficiently high‐impact events such as tropical cyclones (TCs). Besides conventional observations, target observations are essential to further improve initial field accuracy then increasing event skills. identification of the sensitive area, where additional should be deployed, is key implementing observations. In this paper, a...
Abstract Recent machine learning (ML)‐based weather forecasting models have improved the accuracy and efficiency of forecasts while minimizing computational resources, yet still depend on traditional data assimilation (DA) systems to generate analysis fields. Four dimensional variational (4DVar) enhances model states, relying prediction propagate observation initial field. Consequently, fields from DA are not optimal for ML‐based models, necessitating a customized system. This paper...