A Thunderstorm Recognition Method Based on Dual-polarization Weather Radar Volume Scan Data

Abstract

The invention relates to a thunderstorm identification method based on dual-polarization weather radar volume scan data. The method includes the identification of combined reflectivity strong echo area, CAPPI horizontal scale screening, echo intensity constraint, CAPPI particle type constraint and other processes. In order to obtain the optimal identification parameters, the radar echoes of 312 thunderstorm cells were identified and tested, and the reflectivity factors of different strong echo areas at different temperature stratification heights (0°C, -10°C, -15°C) of CAPPI were tested Recognition threshold (30dBZ, 35dBZ, 40dBZ) and graupel area scale (1km 2 , 2km 2 , 3km 2 ) on the recognition results. The results show that in the combined reflectivity, 40dBZ is used as the identification threshold of the strong echo area, the reflectivity factor identification threshold of the CAPPI 0°C stratified height layer strong echo area is 30dBZ, and the screening threshold of the graupel particle area is 2km 2 When , the optimal thunderstorm identification result can be obtained, the identification accuracy is over 91%, the identification false alarm rate is only 6.9%, and the critical success index is 85.3%.

Description

technical field [0001] The invention relates to the field of radar, in particular to a thunderstorm identification method based on dual-polarization weather radar volume scan data. Background technique [0002] Due to the short life cycle and small range of thunderstorms, and the small spatiotemporal expansion and low horizontal and vertical resolution of operational numerical forecast models, thunderstorm early warning and forecasting is one of the more difficult tasks in weather forecasting, and it is also a difficulty and challenge faced by meteorological departments. challenge. Due to the suddenness and locality of the occurrence and development of strong convective weather, thunder and lightning warnings rely more on real-time monitoring means of weather radar. [0003] The current technical methods for identifying thunderstorms based on weather radar mainly include the following aspects: [0004] (1) Establish the distribution relationship between echo top height, ve...

AI Technical Summary

Problems solved by technology

[0007] According to the analysis of the applied statistical results of the above methods, the accuracy of thunderstorm identification is generally not more than 80%, and it is difficult to achieve accurate early warning of thunderstorms

Images