Post-disaster investigation and WRF model reconstruction of April 27th tornado in Guangzhou based on muti-source data fusion

This research investigates the Guangzhou "4·27" tornado event through the integration of multi-source data, including Unmanned Aerial Vehicle (UAV) imagery and ground-based damage surveys. A total of 471 characteristic points along six survey routes were systematically recorded and analyzed. The Weather Research and Forecasting (WRF) mesoscale atmospheric numerical model was employed to simulate and validate this extreme weather event. Results demonstrate that the tornado exhibited a predominantly west to southeast-by-east trajectory, spanning approximately 8.2 km in length with a mean path width of 0.7 km and maximum width reaching 1 km. The damage characteristics and assessment criteria for different Enhanced Fujita (EF) scale levels were systematically summarized. This synthesis led to the determination of the tornado's intensity as EF3, clarifying the relationship between the degree of structural damage and the EF rating. Detailed statistical analysis of 431 affected structures revealed damage severity distributions: 60.32% undamaged, 7.42% lightly damaged, 10.21% moderately damaged, 3.71% considerably damaged, and 18.33% severely damaged. Critical structural vulnerabilities were identified, particularly in metal roof connections, prompting specific wind-resistant design recommendations including enhanced joint detailing and optimized roof geometries. Wind speed estimations derived from two collapsed utility poles yielded gust velocities of V ₙ=74.59 m/s and V ₙ=79.77 m/s, respectively. Numerical simulations using the KF-Eta cumulus parameterization scheme successfully replicated the extreme meteorological conditions, with the WRF model accurately capturing characteristic features including hook echo pattern, storm trajectory, and enhanced storm-relative helicity centers that showed strong consistency with observational data.