Analysis Of Severe Thunderstorm In Nanjing On July 6, 2009

Thunderstorm is a kind of local convective weather system with lightning strikes. As there are abundant precipitations and more frequent lightning activity in Nanjing, thunderstorms may cause invaluable historical and cultural heritage losses as well as economic losses. Therefore, the effective analysis and application of various observation data is of great importance in improving the knowledge of thunderstorms, statistically verifying thunderstorm warning index, analyzing the correlation between convective precipitation and lightning, and ultimately forecasting thunderstorms disasters.Based on the statistical analysis, wavelet de-noising, multifractal detrended fluctuation analysis (MF-DFA), and multi-scale wavelet edge detection methods, the characteristics of severe thunderstorm in Nanjing on July 6,2009 are studied via the observation data obtained by lightning location system, atmospheric electric field sensor, and other data such as sounding data, satellite and radar data. The result shows that:(1) The lightning location system(LLS) of Jiangsu province showed that from 23:00 on July 6 to 11:00 on July 7, there were 5213 flashes in Nanjing,116 of which were positive flashes and 5097 were negative flashes. Positive lightning current amplitude mainly vibrates from 20kA to 40kA and negative lightning current amplitude mainly vibrates from-10kA to-40kA(2) Verifying the identification criteria'the 40dBZ radar echo reaches -10'C level aloft' in this process of thunderstorm in Nanjing, and find this criteria can well identity the lightning location area.23 of the chosen 30 samples, that is,77% of the experimental data are perfectly identified.(3) Using the methods of wavelet to analyse the atmospheric electric field data, we found that the low frequency part of atmospheric electric field signal can help to highlight the main trend of atmospheric electric field.(4) Time series of electric field data show multi-fractal behavior before the strong lightning discharges happen in Nanjing district, and the signal singularity is obvious. Comparing the discrepancies of MF-DFA analysis in different developing stages, we can try to make spectral width (Δα) as thunderstorm identification criteria. WhenΔα>0.5, we can make a thunderstorm warning; whenΔα>1, there will be strong discharges in 10 minutes. (5) Multi-scale wavelet edge detection algorithm can effectively identify convective cloud echo region, then convective precipitation obtained by echo inversion is analyzed with lightning. Results show that rainyields per flash (RPF) ranks from 7.47×107 to 4.46×108kg, average value is 1.72×108kg; non-linear curve fitting equation of precipitation intensity(I, mm/6min) and lightning frequency(F, flashes/6min) is I=1.728lnF+1.122, correlation coefficient is 0.8044; non-linear curve fitting equation of total rainfall(Q, kg/6min) and lightning frequency(F, flashes/6min) is Q=2.372ln F+3.378, correlation coefficient is 0.7355.