A New Algorithm for Automated Mesocyclone Detection by Using Wavelet Transforms

A New mesocyclone detection algorithm has been developed, which is an application of the two-dimensional continuous wavelet transforms, allowing radar observations to be transformed to a particular frequent range. The convolution method has been used to detect locations of vortices under the assumption of the Rankine vortex model. The vorticity and divergence of the vortex have been investigated by the variational method. This study shows that the wavelet filtered data gives much more precise results. With this method, a vortex can be detected and its vorticity estimated, even when that vortex is disturbed by the another vortex flow or by white noise. Furthermore, observational experiments in this study provide vertical distributions of vorticity and divergence of mesocyclones.