Application of Clustering Algorithm in Lightning Early Warning for Four Supercell Strong Convective Processes
It selects four supercell strong convective processes that occurred in Xingtai on May 3,2020,the three-dimensional lightning data is divided at 6 minutes intervals.The DBSCAN clustering algorithm is used to remove discrete points,and the K-means clustering algorithm is used to perform clustering analysis on the three-dimensional lightning data.It selects the K value with the highest contour coefficient and compare it with the radar echo mosaic data to identify four supercells and calculate the clustering center and clustering maximum radius.Trend projection is used to predict the motion trajectories of the four supercells.Analysis shows that the DBSCAN clustering algorithm can effectively delete discrete points and has strong operability.The K-means algorithm for four supercells has good consistency in clustering centers and strong echo areas above 30 dBZ,and can obtain the movement trajectory of clustering centers and the maximum radius of clustering.Using data from nearly three time periods for trend projection,with the minimum MSE,this method has reference value for the release of lightning early warning signals.
国家科技期刊平台
NSTL
万方数据
