Retrieval and Assimilation of Storm Characteristics from Both In-Cloud and Cloud-to-Ground Lightning Data to Improve Mesoscale Model Forecasts

摘要

To improve the accuracy of regional weather forecasts, we (1) obtained and operated a lightning mapping system that detects all types of lightning to provide data for this project, (2) quantified and tested relationships between lightning and other storm properties that will be useful for assimilation, and (3) developed techniques for assimilating data from all types of lightning into COAMPS. Observational data analysis and storm simulations showed that total lightning flash rates were correlated with a storms mass and volume of graupel, updraft mass flux through the mixed phase region, and the volume of updraft exceeding 10 m/s. Gridded lightning data were assimilated into COAMPS by nudging the trigger function of the Kain-Fritsch subgrid-scale convective parameterization. In a test case from the central United States in July 2000, assimilation of lightning data greatly improved the surface moisture, the intensity and location of surface cold pools, and the location of deep convection at the time of forecast initialization. The best results were obtained when convection was completely suppressed where no lightning was observed.