Statistical characteristics of storm cells and centroid-based probability nowcasts by tracking error
Journal
Atmospheric Research
Journal Volume
329
Start Page
108548
ISSN
01698095
Date Issued
2026-01
Author(s)
Chung, Kao-Shen
Lee, Tsai-Jung
Wang, Bing-Zhang
Ke, Chieh-Ying
Tsou, Yi-Hao
Chen, Shin-Gan
Lin, Ping-Yu
Huang, Treng-Shi
Abstract
This study investigated the characteristics of summer storm cells over Taiwan, focusing on the tracking error statistics generated by the Storm Cell Identification and Tracking algorithm (SCIT). The primary objective is to understand the quantitative details of the currently lacking statistical features of storm cells and develop storm-based nowcasting techniques to improve early warning capability. Radar data from Wu-Fen Shan and Chi-Gu, collected from May to August over 8 years (2011–2018), were analysed. The study classified the statistics into synoptic and weak synoptic days to examine the influence of prevailing weather systems, and also analysed the characteristics of northern and southern Taiwan using different radars. The durations of summer storm cells detected by the SCIT algorithm in Taiwan mostly are within an hour. The main differences in statistical features under different conditions depend on the distribution location of storm cells. Complex terrain significantly contributes to storm cell deceleration. Storm cells exhibit larger tracking errors when detected by radars with longer scanning intervals or under conditions of higher movement speeds. Based on error statistics from different groups, two nowcasting techniques: the Potential Track Area for Storm (PTAS) and the Probability of Storm Tracking (PoST), were developed. PTAS assists forecasters in visualizing error regions for tracking forecasts, maintaining a 0.7 Probability of Detection (POD) score in 1 h forecasts, especially effective during typhoon circulations. PoST, complementary to PTAS, quantifies the probability values of each region threatened by storm cells. These products can be quickly updated with radar scans during real-time application, providing decision-makers with effective and timely warning information to enhance public safety during severe weather events.
Subjects
Nowcasting
Statistical analysis
Storm cells
Tracking error
Publisher
Elsevier Ltd
Type
journal article