基于多普勒雷达径向速度图的风暴信息提取-模式识别与智能系统专业论文.docxVIP

ABSTRACT With the development of the science and technology，especially with the progress of radar detecting technology，the weather forecasting has stronger tirneliness and the higher accuracy，however，批nowωnnot predict summer rainstorm especially shorttirne rainstorm in tirne and its accuracy remains to irnprove. With the popularity of a generation of new weather radar，the researchers both at 句 home and abroad fmd that strongωnvective storm happens together w?th mesocyclones that are closely related to Spouts，heavy rain， strong winds and haiI and 咱 other severe weather. Recognition of mesocyclones is irnportant to the timeliness and accuracy of predicting tornado ，rainstorm， and hail.ln addition，recognition of adverse winds plays a significant role in forecasting the rainstorm and severe wind，especially short-tirne rainstorms. In this paper，we have done the research with digital image processing ，Doppler radar and the knowledge relevant to the weather characteristics combined with meteorological expertise and apply MATLAB as softwareωprocess the information of convective rainstorms.ηle study includes as following aspects: ln statistical characteristics analysis of mesocyclones ，we obtain various shear information of speed pairs with the relation among the speed pairs existed in the mesocyclones. We also try to get the relation between the shear value of characteristic quantity in mesocyclones and all convective rainstorms with the method of characteristics analysis. Moreover，by analyzing the shear value to the tirne，the prediction of convective rainstorms is realized. 唱In recognition on mesocyclone ，we advance the algorithm of district inclusion with the knowledge related ωthe digital image processing and graphic feature of 唱 adverse-wind region and distinguish the adverse winds from radial velocity map of Doppler radar success民llly. 飞 All abovementioned algorithms have been carried out by