Testing and Evaluation of Radar Reflectivity and GOES-R Lightning Data Assimilation Capabilities in JEDI LETKF for RRFS-Like Configurations during the 2024 Summer Hydrometeorology Testbed Forecast Experiment

Both radar reflectivity and GOES-R geostationary lightning mapper (GLM) flash extent density (FED) data provide valuable information for storm analysis and forecasting. In recent years, OU/CAPS developed capabilities for directly assimilating radar and GOES-R lightning FED within the Joint Effort for Data assimilation Integration (JEDI) data assimilation (DA) framework, targeting the UFS-based Rapid Refresh Forecasting System (RRFS) version 2. To evaluate the impacts of radar reflectivity and GLM FED DA on convective-scale analysis and forecasting, in support of the 2024 NOAA Hydrometeorology Testbed (HMT) Flash Flood and Intense Rainfall (FFaIR) Forecast Experiment and the UFS-R2O project, we conducted ensemble DA and forecast experiments using the JEDI LETKF DA system coupled with FV3-LAM at 3-km horizontal grid spacing over the CONUS domain for five weeks in summer 2024. These experiments were compared with quasi-operational RRFSv1 forecasts and a control experiment without assimilating radar or lightning data.

Initial evaluation results show that assimilating radar and lightning observations, in addition to conventional observations, improves short-range precipitation forecasts. The impact of radar reflectivity assimilation (ZDA) is particularly pronounced, with its 1-hour accumulated precipitation forecasts being somewhat superior to quasi-operational RRFSv1 (which assimilates radar data using GSI DA system) and significantly better than both the control (NODA) and lightning DA (LDA) experiments. For accumulated precipitation forecasts over the first 6, 12 and 24 hours, ZDA's performance is comparable to RRFSv1 and outperforms NODA and LDA though the difference narrows over time. The advantage of LDA over NODA is smaller than ZDA and sustained for only about 6 hours. More detailed evaluations will be presented at the conference.