Statistical Analysis of Lightning and Radar Data for Lightning Forecast

This study has two main objectives; firstly statistical analysis of total lightning and conventional radar data; secondly application of the statistical relationship to numerical weather prediction (NWP) for lightning forecast and evaluation of NWP performance. In the previous studies, the relationship between lightning activity and radar reflectivity are well documented in different regions. Firstly, the relationship is verified to summertime convective thunderstorm in Japan of 2015. For this purpose, conventional radar data, which has 1 km horizontal resolution and 15 vertical layer in 15 km altitude every 10 minutes, and total lightning data are used. 1351 convective cells were observed in summer 2015 in Kanto plain in Japan, and the correlations between radar parameters (maximum radar reflectivity, radar reflectivity at -10 C degree, radar echo volume) and lightning activity are computed. As a result of this study, volume of 40 dBZ radar echo above 0 C degree level shows the best correlation to lightning activity. Next, to evaluate lightning forecast performance, the relationship is applied to NWP result, which has 1-km horizontal resolution and 12-hour forecast. There are no significant correlations between estimated radar parameters from NWP and observed lightning activity. On the other hand, vertical velocity volume in NWP shows slightly better correlation to lightning activity. The reason for these result maybe the NWP physics is not precise enough, especially in cloud microphsyics of large ice particles (e.g. hail or graupel) for summertime thunderstorms.