Skillful Precipitation Nowcasting Using Physical‐Driven Diffusion Networks

Accurate and timely precipitation nowcasting is essential for numerous applications including emergency services, infrastructure management, and agriculture. Recently, deep learning (DL) techniques have shown promise in enhancing nowcasting capabilities. This study introduces a novel Physical‐Driven Diffusion Network (PDDN) model that leverages both radar and numerical weather prediction (NWP) data to improve the accuracy and physical consistency of precipitation nowcasts. Our approach integrates the strengths of data‐driven DL techniques with physics‐based NWP models. The PDDN model utilizes latent diffusion models and autoencoders within a two‐stage architecture to predict future radar images, incorporating the Weather Research and Forecasting (WRF) model data to enhance understanding of atmospheric dynamics. Our results demonstrate significant improvements over traditional models, particularly in short‐term forecasting up to 6 hr. This research highlights the potential of combining advanced machine learning techniques with conventional meteorological data, offering new directions for enhancing the accuracy and reliability of weather forecasting. Plain Language Summary Traditional methods for predicting rainfall often lose accuracy over time, especially in fast‐changing weather conditions. Deep learning techniques have recently shown promise in improving these predictions. To enhance rainfall nowcasting, we developed the Physical‐Driven Diffusion Network (PDDN). This new model combines radar data and numerical weather prediction (NWP) data to improve the accuracy and consistency of forecasts. By integrating advanced machine learning with physics‐based models, the PDDN excels in predicting rainfall for up to 6 hr. Our results show that the PDDN significantly outperforms traditional methods, providing more accurate and reliable short‐term weather forecasts. Key Points Our PDDN model outperforms baseline models, exhibiting high proficiency in predicting precipitation up to 6 hr ahead The integration of physical data from WRF enhances the model’s understanding of atmospheric dynamics and improves forecasting accuracy The novel design of latent diffusion models in a two‐stage architecture enables more accurate and robust precipitation nowcasts