We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Forewarning method of downburst based on feature recognition and extrapolation.
- Authors
Wang, Xing; Wang, Jianhong; Miao, Chunsheng; Zeng, Kang
- Abstract
Currently, using common meteorological data for downburst identification and early warning lacks feasibility. Therefore, a method based on the Doppler weather radar is proposed in this paper to identify, track and predict the downburst. The method is outlined as follows. Firstly, by using the optical flow method, optical flow fields of the reflectivity factors in vertical section are analyzed. The temporal evolution rule of the thunderstorm core is found. Secondly, the Lagrangian mechanical model is used for function fitting of the thunderstorm-core top's sinking process. Lastly, by adopting histogram and the Bhattacharyya coefficient (BC), the statistical analysis is carried out to match and identify the positive and negative velocity pairs in the middle-level radial velocity field of the thunderstorm core. Moreover, a series of thresholds are also analyzed to achieve the intelligent forewarning of the downburst. In this paper, the local constraint algorithm featuring multi-layer iteration is introduced, which effectively improves the inapplicability issue in the traditional method of analyzing non-rigid moving targets such as echo. Furthermore, the measure based on the histogram and the BC are proposed to solve the image matching problem due to the asymmetric structure of positive and negative velocity pair images in the radial direction. The experiment results show that the method is able to identify the potential downburst at the early stage of the thunderstorm-core rising or sinking processes. Meanwhile, the descent velocity and time of the thunderstorm core reaching near surface can be estimated, and then, the intelligent identification and early warning of downbursts can be realized.
- Subjects
CONSTRAINT algorithms; EXTRAPOLATION; OPTICAL flow; DOPPLER radar; IMAGE registration; RADAR meteorology; CLUTTER (Radar)
- Publication
Natural Hazards, 2020, Vol 103, Issue 1, p903
- ISSN
0921-030X
- Publication type
Article
- DOI
10.1007/s11069-020-04018-4