The Research On Lightning Activities And Potential Forecasting Over Liaoning Province

Based on the data of thunderstorms at14stations in Liaoning province from1981to2010and the data of lightning form2007to2011in Liaoning, the climate variation trends of thunderstorms and the temporal distribution of lightning was analyzed by means of climatic trends, wavelets analysis and assurance ratios. The result shows that the result shows that the annual number of thunderstorm days has a decreasing trend. Decreasing trends mainly exist in autumn, but they are not significant in spring and summer. The variation periods are stable and exist across the whole region when6-10a and14a, the number of thunderstorm days distributes in periods of larger concentration in Liaoning. Under different assurance ratios, the initial dates of thunderstorms occur in the second pentad of May for an assurance ratio of90%, the late April for50%.While the ending dates of thunderstorms occur the first pentad of November for an assurance ratio of90%, the late October for50%. The ratio of negative lightning is93.19%, and the ratio of positive lightning is6.81%. The prone period of thunderstorms is June to August, the same as the analysis of the data of thunderstorms. The intensity of positive and negative lightning is mainly concentrated in10～40kA. The prominent maxima of lightning density was find to the junction of plain and mountain, which reflects that thunderstorm activities were well related to the surface characteristics.A neural network-based scheme to do a multivariate analysis for forecasting the occurrence of thunderstorm is presented in Shenyang using sounding data and lightning location system data. Well correlate nine sounding factors are selected as the predictors, then all the input factors were normalized, output data are adopted to two-stage category so that the BP network with double hidden layers was established and the independent samples can be tested in it. The results indicate that,in the same condition, compared with single hidden layer BP network, double hidden layers BP network shows its advantage on solving classification problem. Compared with multivariate statistics regression algorithm, the neural network algorithm obtain higher thunderstorm forecasts score and More reliable results. It showed good nonlinear processing ability in the thunderstorm forecasts base on sounding.Trying the forecasting experiments of6hours and3hours and the cause of error analysis, the results indicate that,the6hour of hidden layers BP network forecasting model is more suitable for Shenyang.