Font Size: a A A

The Research On Characteristics And Forecasting Of Lightning Activity In Shaoyang

Posted on:2014-12-20Degree:MasterType:Thesis
Country:ChinaCandidate:M W ZhouFull Text:PDF
GTID:2250330425984768Subject:Lightning science and technology
Abstract/Summary:PDF Full Text Request
The data obtained by lightning location system, Doppler radar, radiosonde and NCEP re-analysis were analyzed to study the characteristics of lightning activity, the lightning warning indicator. The lightning nowcasting program can give lightning occurrence probability. It is mainly concerned with three aspects as it follows:(1) Based on the lightning monitoring data from lightning location system from2008to2010in Shaoyang, the temporal and spatial distributions characteristics of lightning were analyzed. The results indicate that lightning is most negative and usually occurs on13:00-19:00from March to September. The frequency of cloud-to-ground lightning decreases from the northeast to the southwest of Shaoyang. The biggest density of lightning appears in summer, the smallest one occurs in winter. Furthermore, the distribution of the current intensity is contrary to lightning frequency. That is to say the lightning take places less/more in some areas where the lightning current density is relatively large/smaller. Lightning current curve of positive CG, negative CG and total flash shows a single peak changes.(2) Based on the data of the lightning location and thunderstorm days in Shaoyang area(110.5-111.5°E,26.5-27.5°N), the thunderstorms samples and non-thunderstorms samples is in all1007.80%samples are used as the training samples. With the help of the NCEP historical re-analysis data,48physical quantities are calculated.11significant predictors that are greater than0.3are selected by relevance analysis. support vector machine method, Logistic regression analysis and Bayes classification respectively can give thunderstorm potential forecasting model in Shaoyang area. Out of the three methods tested, it is found that SVM gives best prediction performance with TSS, which shows that the model can provide certain reference value for thunderstorms potential prediction in Shaoyang area.(3)11thunderstorm cells and9non-thunderstorm cells occurred in summer in Shaoyang are studied. The percentage of the echo top above0℃,-10℃and-20℃level with reflectivity exceed30dBZ,35dBZ and40dBZ on20cases. It is found that the biggest difference between thunderstorm cells and non-thunderstorm cells is the reflectivity exceed40dBZ. The relationship of the echo area and lightning activity of the echo area above30dBZ,40dBZ and50dBZ with0℃,-10℃and-20℃level are also analyzed. It shows that the echo area above-10℃level with reflectivity exceed40dBZ(S40-10) is the highest correlation coefficients with CG lightning frequency in all temperature levels. The40dBZ echo top and the echo area above-10℃level with reflectivity exceed40dBZ are more indicative for the end of the lightning activity. Through analyzing thunderstorms in Shaoyang, a warming method of the beginning and end of lightning activity is obtained. Finally,15thunderstorm cases are tested the accuracy of the forecast method and the accuracy is respectively80%and83%.
Keywords/Search Tags:lightning activity regularity, support vector machine, thunderstormpotential forecast, the end of lightning activity, nowcasting
PDF Full Text Request
Related items