Abstract:This paper comprehensively uses multi-source high-frequency detection data to analyze the formation causes and development of a localized severe convective event that occurred in Beijing on 27 August 2020, and the applicability of the observation data in nowcasting and warning of such severe convection. Utilizing the data from various high-frequency detection instruments including microwave radiometers, wind profile radars, S-band Doppler radars, X-band dual-polarization radars, and a three-dimensional lightning positioning system, we investigate the evolutionary characteristics and warning indicators during thunderstorm initiation, development and dissipation stages. The results demonstrate that the microwave radiometer-retrieved temperature, humidity and stability parameters (K-index, SI-index) show significant variations 30-120 min prior to the outbreak of severe convection, effectively indicating the energy accumulation. The enhanced vertical wind shear and low-level warm advection detected by wind profile radars can reflect dynamic lifting condition 55-120 min in advance. S/X-band radar observations of echo overhanging structures, centroid height variations, and dual-polarization parameters (ZDR, KDP) enable the identification of hail growth zones, achieving 12-37 min lead time for hail and thunderstorm gale warnings. Lightning jump signals (2σ algorithm) exhibit strong correlation with severe weather (hail and winds), with the first jump providing 52 min lead time for hail warning. Validation of other four severe convective cases in 2020 shows that synergistic application of multi-source observations can overcome the limitation of single-instrument detection, enabling dynamic tracking of environmental energy, storm structure and lightning activities, thereby providing valuable references for short-time forecasting and nowcasting.