A New Method of Characterizing Flow Patterns of Vortices and Detecting the Centers of Vortices in a Numerical Wind Field

Abstract

AbstractA vortex in a wind field is an important aspect of a weather system; vortices often result in hazardous weather, such as rainstorms, windstorms, and typhoons. As the availability of numerical meteorological data increases, traditional manual analysis no longer provides an efficient means of timely analysis of observed and predicted atmospheric vortices. Therefore, a method was proposed to automatically characterize flow patterns of vortices and to detect the centers of vortices in complex wind fields generated from numerical weather prediction (NWP) models. First, a statistical feature was developed to preliminarily filter regional wind data to obtain (anti)cyclonic vortices. Second, flow patterns of ideal axisymmetric wind fields were extracted by analyzing circular data related to wind directions. Third, for actual vortices in a complex wind field, a series of rules and deformation degree indices were constructed to retrieve the provisional centers of vortices. Fourth, the Ward hierarchical clustering algorithm was used to cluster these provisional centers, which were filled up by a dilation operation to cover the core region of the vortex. Finally, the vortices were classified as either cyclones or anticyclones based on their analyzed vorticity, and their global centers were precisely located. Experimental results show that the proposed preprocessing method was more effective than the traditional filtering method and that the features of the flow pattern were stable regardless of the variety in the resolution and scale. It was also proven that the proposed method can be further extended and applied to detecting typhoon centers, for which it was more effective than other currently used methods.