Nowcasting method and system of thunder cloud cluster based on boundary recognition and tracer technique

Abstract

The invention discloses a nowcasting method and system of a thunder cloud cluster based on the boundary recognition and tracer technique. The method performs the cloud cluster boundary recognition and topology process to establish the relationship between cloud cluster life timing and clan pedigree by using the pattern recognition technique. The method comprises the steps of performing the boundary recognition on the pre-processed radar data; respectively recognizing the life timing of each cloud cluster and the information of moving direction, velocity, area and strong center and the like ofeach cloud cluster through the six judging factors of the quadtree matching analysis factor, the overlap factor, the area factor, the circum-rectangle factor, the outline comprehensive factor and thelocal similarity determination factor; and performing linear extrapolation on the moving direction, velocity, area and the strength of the cloud cluster. The preliminary result indicates that the system and the method of the invention can realize better recognition and extrapolation forecast of the thunder cloud cluster.

Description

technical field [0001] The invention relates to a short-term nowcasting and early warning method and system for a meteorological business department, in particular to a method and system for thunderstorm cloud identification, tracking and early warning technology based on high temporal and spatial resolution radar data. Background technique [0002] The identification, tracking and early warning technology of thunderstorm cloud based on high spatial and temporal resolution radar data is one of the most important nowcasting and early warning technologies for thunderstorm cloud. The main methods include cross-correlation (TREC) method and single centroid method. The cross-correlation method divides the entire data area into several small areas, calculates the correlation coefficient between the radar echo images at adjacent times, and determines the area in the image at the adjacent time by the maximum correlation coefficient, thereby determining the average motion of the echo;...

AI Technical Summary

Problems solved by technology

The cross-correlation method divides the entire data area into several small areas, calculates the correlation coefficient between the radar echo images at adjacent times, and determines the area in the image at the adjacent time by the maximum correlation coefficient, thereby determining the average motion of the echo; this method is widely used in China It has been widely used, but its disadvantage is that it cannot accurately divide the physical boundaries of cloud clusters, and cannot obtain the characteristics of individual storms

The monomer centroid method first identifies the monomer, and calculates the center of mass position, volume and projected area and other characteristics, and then performs monomer matching and tracking on the scanning data at two moments before and after, to separate the storm monomer, and then through multiple moments Extrapolation and early warning of tracking results; the disadvantage of this method is that it focuses on individual thunderstorms, and must separate the connected cloud clusters as much as possible to improve the accuracy of predictions. The splitting and merging of thunderstorms in the process of activity is not fully considered, and the influence of thunderstorm groups is ignored. activity characteristics

Images