The 2023 hurricane season is witnessing unusually rapid hurricane intensification due to warmer ocean temperatures, exemplified by Hurricane Milton's rapid growth and impact on Tampa.
Effiziente Helmholtz-Vorbedingung für die Kompressible Euler-Gleichungen.
Implicit solvers for atmospheric models are accelerated via Helmholtz preconditioning, ensuring stability and efficiency in solving compressible Euler equations.
Modeling precipitation evolution with global deterministic motion and local stochastic variations improves prediction accuracy.
Using Graph Neural Networks to analyze the impact of observations on atmospheric state estimation.
The author employs conditional generative adversarial networks (CGANs) and convolutional neural networks (CNN) to post-process convection-allowing model forecasts, resulting in skillful severe weather predictions with improved Brier Skill Scores. The approach combines deep generative models with neural networks to enhance severe weather prediction accuracy.
The author presents a novel preconditioner for the compressible Euler equations, showcasing improved stability and efficiency compared to existing methods.
The author presents a Transformer-based model for nowcasting radar image sequences using satellite data, aiming to bridge the gap between ground- and space-based observations for more accurate weather prediction.
The authors present Adas, a novel data assimilation model for global weather variables, combined with FengWu to create the first end-to-end AI-based global weather forecasting system. This system demonstrates stable long-term operation and superior performance in real-world scenarios.
The author proposes the SFTformer, a Spatial-Frequency-Temporal correlation-decoupling Transformer, to effectively model radar echo dynamics by decoupling spatial morphology and temporal evolution. The model outperforms existing methods in short-term precipitation forecasting.