A Thunderstorm Core Identification and Tracking Method Based on Hybrid Clustering Algorithm

Abstract

The invention discloses a thunderstorm nuclear identification and tracking method based on a hybrid clustering algorithm. The specific steps are as follows: use deployed lightning monitoring points to detect and record ground lightning data, and preprocess the recorded ground lightning data, and divide them into various categories The lightning data set of the time period; using the GPS clock synchronization technology and the time difference of the arrival time difference of the radiation pulse of the lightning electric field change at each site, the spatial positioning coordinates of the lightning are obtained through the arrival time difference algorithm; for the lightning positioning data obtained in the above steps, use DBSCAN Algorithm and KMEANS algorithm to obtain the thunderstorm core centroid coordinate position, lightning frequency and the correlation among thunderstorm core of lightning location data. The experiment proves that this method can accurately reflect the change trend of lightning in thunderstorm weather, and achieve a good effect of thunderstorm nucleus identification and thunderstorm movement tracking.

Description

technical field [0001] The invention belongs to the field of lightning monitoring and relates to a thunderstorm nucleus identification and tracking method based on a hybrid clustering algorithm. Background technique [0002] With the development of electronic technology and computer technology, thunderstorm lightning activity monitoring has developed from traditional lightning location observation to complete recording of the detailed characteristics of lightning activity during the entire thunderstorm life history. On this basis, various lightning data based on the evolution of thunderstorm life history can be developed product. Lightning activity is an indicator of the strength of thunderstorm convective activity. Compared with meteorological radar detection of thunderstorm cloud precipitation particles, the potential of diagnosing the timeliness and accuracy of strong convective activity has attracted more and more attention. Monitoring of thunderstorm convective activit...

AI Technical Summary

Problems solved by technology

However, when applied to the clustering of lightning data, the result is a cluster, not a "center", and the existing noise points cannot be distinguished

The key to the KMEANS algorithm is the selection of the K value. If the distribution of lightning location data is too scattered and aggregated according to a fixed K value, the position of the centroid obtained may be far from the actual position.

Images