A Relationship between Lightning and Microphysics: An Application of GLM to Data Assimilation.

The occurrence of lightning in storms is a result of an electric field strong enough to break the dielectric strength of air allowing the discharge. The most accepted explanation for the development of such intense field is based on the charge exchange between the collision of ice particles in the cloud, especially graupel and ice crystals, and the subsequent transport of these charged particles into different regions forming different poles inside the cloud. This relationship between lightning and cloud microphysics is the main aspect on which this work is based on. The main goal is to elaborate averaged vertical profiles of polarimetric variables for different classes of lighting density according to the GLM grid and then, evaluate the potential use of these profiles for data assimilation in models of high spatial and temporal resolution. The data was based on a X-band polarimetric radar located in the city of Campinas-SP and on the Brazilian Network for the Detection of Atmospheric Discharge (GLM proxy). In addition to that, GLM pixels over the region were used to calculate the density of lightning occurrence. The main differences between each class averaged profiles for the four polarimetric variables Z_H, Z_DR, K_DP and ρ_HV was in the region between the surface and the melting layer. For more intense classes, it was observed signatures related with higher concentration of ice particles in altitudes, the presence of supercooled drops above the freezing level and with the occurrence of larger and more oblate raindrops. However, to better analyse the microphysical aspects of the cloud, a hydrometeor classification algorithm, based on the fuzzy logic, was used to elaborate a percentage distribution of the hydrometeors according to the density class. Observing the classification results is possible to notice, for example, the presence of graupel and ice crystals at higher elevations associated with more electric activity. It was also observed an increase in the distance between the region of higher concentration of these two particles in more intense systems, which is coherent with the presence of a stronger electric field and therefore greater number of lightning. Aiming to analyse the impact of the reflectivity profiles on the assimilation process as a possibility to indirectly correlated them with lighting information, it was made a study for the December 3^th 2016 case. Using the Weather Research and Forecast (WRF) model three different forecasts were generated: without assimilation, assimilating radar reflectivity and wind data and assimilating the reflectivity profiles defined as function of the lightning density. In general, the forecast obtained with assimilated profiles presented better results than without assimilation as expected. Comparing to radar data assimilation, the reflectivity profiles related to lighting density had a more significant improvement on the performance mainly after the first hour of forecasting, and they were also able to better allocate more intense precipitation systems. Considering the results obtained in this work and all the possibilities offered by the recently launched GLM, especially regarding its spatial coverage, it can be interesting to better understand and optimize the use of the mean profiles aiming to improve nowcasting.