An Operational and Real‐Time Severe Weather Nowcasting System Based on Lightning Jumps: Introduction and Evaluation

In this study, the effectiveness of integrating the lightning jump (LJ) detection method with radar‐based storm tracking and extrapolation techniques for operational nowcasting of severe convective weather events in the Beijing–Tianjin–Hebei region is investigated. On the basis of severe weather observations, lightning data from the Beijing Lightning Location System, and multiband radar composite products, convective storms are systematically identified, tracked, and forecasted. The 2‐sigma (2σ) LJ algorithm is applied to detect rapid increases in lightning activity indicative of storm intensification. Analyses of two convective case studies involving multiple severe weather events from April 28, 2023, and June 12, 2024, illustrate that the LJ‐based nowcasting method can be used to accurately predict the occurrence of hail, gales, and heavy rainfall (HR). A statistical evaluation (2023–2024) of LJ‐based warnings indicated systematic improvements in predictions, with 2‐h extrapolation‐based forecasts consistently outperforming 1‐h predictions. A relatively high probability of detection (POD; 73.2% in 2024, +8.3% in 2023) and a low false alarm rate (FAR; 42.4% in 2024, 38.1% in 2023) were achieved for 2‐h forecasts. The critical success index (CSI) peaked at 47.6% in 2024. The lead time increased to 47.8 min (+3.7%) for the 1‐h forecasts and 71.6 min (+6.1%) for the 2‐h forecasts. The overall forecasting accuracy was highest for HR events (86.8%), intermediate for gale events (78.5%), and lowest for hail events (28.6%). This suggests that the storm identification method presented in this paper achieves optimal classification performance for heavy precipitation and strong winds. However, improvements are needed for hail classification. These findings underscore the operational value of integrating LJ algorithms into real‐time severe weather monitoring systems to enhance the information available to forecasters and extend severe weather warning lead times. This method can also serve as a reference for severe weather nowcasting in other regions.