Progress in and Outlook for Objective Severe Convective Weather Nowcasting Technology

This article provides an overview of the progress in monitoring and nowcasting techniques for severe convective weather (SCW), as well as the development of operational nowcasting systems. It focuses on summarizing the progress of monitoring and nowcasting techniques using deep learning (DL) models based on multi-source data and points out the challenges and opportunities. Based on multi-source observations such as those from dual-polarization weather radars and geostationary satellites, the monitoring capabilities of SCW types and intensities, convective initiation, identification and tracking of convective storm cells are further improved using storm structural feature recognition, fuzzy logic, DL, and other technologies. The application of deep generative models has significantly improved the accuracy and lead time of SCW nowcasting. The performance of China Meteorological Administration’s SWAN3.0 (Severe Weather Analysis and Forecasting) system has been steadily improving and has been widely used in operations in China. In the future, it is necessary to make full use of fine observation and numerical forecast data at the hundred-meter resolution to enhance the understanding of the mechanisms of SCW at the meso-γ- and microscales. Physical-informed artificial intelligence (AI) models, as well as large scale AI models should be developed for nowcasting to continuously improve the monitoring and nowcasting capabilities, leading to the full utilization of forecasters' comprehensive judgment role and enhancing their ability to predict extreme weather.