The Relationship Between Lightning And Ice-phase Particles Based On The TRMM Satellite And Its Application In Lightining Data Assimilation

The forecast of strong convection has always been a difficult problem in weather forecasting.Lightning flash is closely related to cloud dynamic and microphysical processes,and can be a good indicator of convective activity,so more and more attention has been paid to the research of lightning data assimilation in recent years.In this paper,the relationships between total lightning flash rate and ice-phase particles(i.e.graupel,ice and snow)on the scale of 10 km were constructed based on the observations of lightning imaging sensor(LIS),precipitation radar(PR),and passive microwave imaging(TMI)onboard the TRMM satellite,then two experiments of lightning data assimilation based on the relationship between lightning and ice particles was conducted by the mesoscale weather research and prediction model(WRF),whose result were compared with another lightning data assimilation method by adjusting humidity profile.The main conclusions are presented as follows:(1)It is found that the consistency between total ice-phase particles calculated from radar reflectivity with lightning is better than that of TMI products.The total ice phase particle density calculated by PR at 10 km grid scale is linearly related to the lightning frequency at different temperature levels(-10?~-40?)with a good correlation(most of the correlation coefficients are greater than 0.4).The coefficients vary slightly in land and ocean areas,which means the relationship between both of them is relatively common in land and ocean.The total ice phase particle density calculated from PR can be converted into the density of different water particles according to the statistical results of the percentile of different water component in the TMI product.(2)Two case studies(one was in the Huang-Huai Valleys on May 9,1999,the other was in the Jiangsu Province on August 5,2010)were used to test the relationships between total lightning flash and ice-phase particle.The water vapor field and the cold pool in initial fields have been effectively adjusted,and horizontal wind field in the convective area get an obvious cyclonic growth,which is more conducive to the development of convection,and the forecast of radar reflevtivity and precipitation has been improved.The Equitable Threat Score(ETS)of radar reflectivity at different thresholds between 15~35 dBZ in the results based on the ice-phase particle method are greater than those in the results based on the humidity profile method(except the 35 dBZ in 1999 case)and the control experiments.(3)The method based on ice-phase particles generates a weaker prediction of precipitation,while the method based on the humidity profile gets stronger precipitation in the 2010 case,which shows that both these two methods is not very stable for the precipitation forecast.There are also significant differences in the spatial and temporal effects of these two assimilation methods on the simulation results.In the method based on ice-phase particles,some adjustments in the areas where no lightning occurs by the relaxation coefficient in the horizontal and vertical directions can be seen.In addition,it is found that this method can affect the distribution of water vapor in a larger temperature range,and present a faster trigger and a longer duration for convection.But,in the method based on the humidity profile,the assimilation range is limited to the area where lightning occurs and the height levels,and assimilation results are subject to the control experimental,convection initiates slowly and lasts shortly.