Storm Tracking Algorithm Development Based On The Three-dimensional Radar Image Data

thunderstorms, hail and tornadoes often happens, which has larger damage to human beings' production and life. The Doppler radar system based on the modern computer technology and information-processing technology is the most effective means for the accurate tracking, reasonable forecasting of severe weather. Therefore, the Doppler radar systems have been gradually built in the regions where severe weather often happens. Much experience has accumulated in the practical application. The key of making the Doppler radar more effective is that the data processing of the radar images about the severe weather is researched and explored.The means and theories of the target tracking in the modern computer version have been introduced. The tracking and forecasting are archived by processing the Doppler radar images. In this thesis, the method is discarded that the centroid of the storm is used to track the storms, instead of using some points which are in the storms to track the storms. The risk of tracking storms is dispersed through it. Because when the movement trend of the centroid of storm has a great error, the former maybe results in the failure to track the storms. But to the latter, it can succeed in tracking the storms so long as the movement trends of the most points in the storms are good. To the forecast of the storm centroid position, the linear extrapolation and Optical Flow method are adopted. It not only makes the forecast result accord with the actual movement trend of the storms, but also makes the forecast result match the change trend of the shape.Via analyzing and processing the radar data of the summer from Beijing and Tianjin, the method is proved to be effective. It can track the storms exactly and forecast the storm position reasonably. In the last section of this thesis, the statistical data about the storms properties are computed. By analyzing these data, some useful results are obtained. The method of the storm tracking and forecasting in the thesis is archived on the VC++ development platform, in the Windows XP.