Tracking severe weather storms in Doppler radar images

We describe an automatic storm-tracking system to help Doppler radar system to detect severe storms such as thunderwith the forecasting of severe storms. In this article, we present the storms and tornadoes. The Doppler radar generates intensity and concepts of fuzzy point, fuzzy vector, fuzzy length of a fuzzy vector, radial velocity images, examples of which are shown in Figures and fuzzy angle between two nonzero fuzzy vectors, that are used in 1 and 2. These images are preprocessed to remove pixels correour tracking algorithm. These concepts are used to overcome some sponding to lines, bars, and writing. Intensity interpolation is of the limitations of our previous work, where fixed center-of-mass used to represent the values of removed pixels. Typically, there storm centers did not provide smooth tracks over time, while at the are 0–20 potential severe storms per image. same time, their detection was very threshold sensitive. Our algorithm The recognition and tracking of storms in these radar images uses region splitting with dynamic thresholding to determine storm is currently performed manually by human experts, and the task masses in Doppler radar intensity images. We represent the center of a hypothesized storm using a fuzzy point. These fuzzy storm cenis time-consuming. To improve the efficiency and quality of ters are then tracked over time using an incremental relaxation algoweather forecasting, AES is interested in developing an automatic rithm. We have developed a visualization program using the X11 storm-tracking system for use in their operations. Toward this Athena toolkit for our storm visualization tool. The algorithm was end, we have developed a tracking program with visualization tested on seven real radar image sequences obtained from the Atmocapabilities that uses a hypothesize and verify model to detect spheric Environment Service radar station at King City, Ontario, Canstorms in radar images and construct storm tracks. We first hyada. We can obtain storm tracks that are long and smooth and which pothesize storm masses in the Doppler radar intensity images. closely match an expert meteorologist’s perception. q 1998 John Then we verify the correctness of these hypothesized storms by Wiley & Sons, Inc. Int J Imaging Syst Technol, 9, 201–213, 1998 tracking them over time. If a hypothesized storm can be tracked over a desired number of frames, we conclude that the storm is a valid storm and we record its track. When all potential storms