Research On Now-casting Technique For Severe Convective Weather In Ningbo

The article's aim is to improve the accuracy and efficiency of now-casting convective weather from the weather forecaster's point. Analysising on the doppolar datum from 2003 to 2008,the research concerning on these ones:the quality control of non-meteorological radar echoes; The way to make TREC flow; Artificial intelligence to identify, track and extrapolation storms; To build the Ningbo severe convective weather model and with the model how to use multiple radar products on the same screen to quickly qualitative analysis and now-casting severe weather. The important content and ideas are as follows:1) Improve the algorithm for identification and elimination of normal propagation ground clutters(NPGC) and abnormal propagation ground clutters(APGC). Statistical analysis of non-meteorological radar echoes in Ningbo for four years shows that these clutters mainly are NPGC and APGC.These clutter echoes show discontinuous and discrete characteristics on reflectivity products, radial velocity values and spectrum width values. Based on these echo characteristics and combined with the traditional clutter identification of forecaster series multi-dimensional linear approximation of the clutter identification and filtering for clutter suppression algorithm are desiigned.190 cases have been tested with the techniques under weather conditions including clear sky, showers and deep convections. Results show that it works well, and has good performance to remove the non-meteorological radar echoes completely and accurately within the range of 150 km. with clutter more than 83% recognition rate and error rates of diffusion controlled precipitation less than 3%.2)The algorithm of TREC(Tracking Radar Echoes by Correlation) on significant area is given in the article. The result displays on the vector grid. It put forward the application of a mainstream analysis algorithms to revise error. At last the final results of TREC effect and the related issues are discussed.3) Improvement and development of thunderstorms objective identification and tracking algorithm:A new algorithm is put forward to identify,track and extrapolate severe storm echoes intelligently, which takes advantages of traditional methods such as correlation tracking algorithm of CTREC and centric tracking technique of TITAN. The centric tracking algorithm is applied to decide the movements of entire echoes, while the correlation tracking algorithm is utilized to match echo blocks in consecutive time series which based on the significant region. The stream field of marked echo blocks has also been calculated to judge echoes'local development, in order to see their mergence and split. In the mean time, simplified polygons are applied to decide the boundaries of echo blocks, which is theoretically thought to have more analogical echo shape than that of TITAN. On the basis of the linear approximation, hourly exploration of echoes has been carried out, and verification has been done simultaneously. By continuously valuating 20 cases with 1396 hours of radar base datum, the method seems do better while storms are big and long life, but for those of short life storms or storms with merging process, it doesn't perform very well.4)According to relationship between the precipitation particle phase and temperature stratification curve, the heights layer of 0℃and -20℃are very important to determine the type of severe convective weather. By using this theoretical basis, A simple and practical Ningbo severe convective model is built.A new nowcasting idea by analyzing multiple radar products on the same screen to quickly forecasting and identifying severe convective weather is put forward. This method used in the business in the past two years are proved obviously effective.