1. A methodology for identifying southwest vortices in China.
- Author
-
Yuan, Chenhu, Qiao, Panjie, Wang, Xiaojuan, Liu, Wenqi, Feng, Guolin, Zhao, Ning, and Zhang, Yongwen
- Subjects
- *
GEOPOTENTIAL height , *VORTEX methods , *FALSE alarms , *RESEARCH personnel , *ALGORITHMS - Abstract
The Southwest Vortex (SWV) is a mesoscale closed low-pressure system in Southwest China that significantly influences regional heavy rainfall. Understanding SWV characteristics and formation mechanisms is crucial for research in this field, with effective identification as the foundation. Currently, Chinese researchers lack a unified standard for SWV identification. Despite the availability of high spatiotemporal resolution reanalysis data, which provides a superior foundation for studying SWVs, existing methods struggle to fully exploit these datasets. Therefore, this study proposes PVSIA (Geopotential-Height-Field and Vector-Wind-Field based SWV Identification Algorithm), a method that integrates the geopotential height field and wind field characteristics. The PVSIA employs an innovative calculation method for wind vector rotation properties, enhancing identification speed and overcoming challenges posed by complex terrain. Furthermore, this study utilizes the ERA5 reanalysis dataset from 2019 to 2021 to systematically compare the performance of the PVSIA method with traditional geopotential height field methods in identifying SWVs, employing key evaluation metrics such as hit ratio, false alarm rate, and missing report rate. During this period, the PVSIA method achieved a hit rate of 94.25% in identifying SWVs, with a false alarm rate of 5.20% and a missing report rate of 5.74%. In contrast, the geopotential height field method, while achieving a higher hit rate of 98.85%, exhibited a significantly higher false alarm rate of 70.54%. In contrast, the PVSIA method maintains a high hit rate while significantly reducing the false alarm rate, demonstrating a more stable and reliable identification performance. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF