Comparative Analysis Of Two Lightning Forecast Methods Based On The WRF Model

Thunderstorm processes occurred in Nanjing have been simulated by WRF in this paper. The lightning forecast model, which is based on atmospheric stratification characteristics and the distribution of water into particles in the thunderstorm was built to make lightning forecast in the Nanjing area. The model was made a further improve of the forecast level and improve the lightning warning accuracy to forecast lightning density.First, two thunderstorm processes occurred in Nanjing on May27,2008and July7,2009have been simulated by using WRF in this paper. The grid point data of two simulate processes around Nanjing is selected for two groups of modeling and samples respectively. Under the lightning positioning data of Jiangsu province for reference. Then four convective parameters selected into the model, combined with model sample data to establish Logistic thunderstorm forecast model, and check the prediction effect of experience formula by test examples. The results show that the simulation of WRF model can well present the thunderstorm process. The shape of strong echo area is similar to reality, but has some differences to posit its center. When choosing the area of example, the deviations have been taken into consideration to select the example data of modeling and test for set up the predict model, including the checking part. The accuracy of this model reach67.32%, real skill score (TS) achieve to0.466, which present good forecast effect. The high spatial and temporal resolution of WRF model provides higher accuracy to obtain meteorological information before thunderstorm coming. Thus, to increase the precision of model simulation can improve the effect of thunderstorm potential forecast by WRF model.Then, in order to enrich the forecast means and inspection simulation results in different resolution of WRF model. In this paper, we introduce two index, which reflect the distribution of hydrometeors particle in the thunderstorm cloud:F index and Lightning Potential Index(LPI). And use of3km,1km horizontal resolution to simulate the Nanjing area of thunderstorms. The best resolution of lightning forecast obtained from the comparative analysis of the actual simulation effects in both resolution and the use of F-index and Lightning Potential Index(LPI)distribution. Moreover, use the advantage of the F-index and LPI in the lightning forecast, we establish an empirical formula which contains both index called THREAT index. By a simulation of a thunderstorm occurred in Nanjing on July7,2009, we found that Logistic lightning forecast formula well forecast areas of lightning, however, higher rate of false and the THREAT lightning prediction formula with good results in forecasting lightning density but have some underreporting phenomenon of lightning.