Terra Thompson
20 July 2009, 6:00 PM
National Weather Center, Room 5820
120 David L. Boren Blvd.
University of Oklahoma
Norman, OK
Directions to the NWC (.pdf, 60 kb)
The first Phased Array Radar (PAR) adapted to observe weather via its unique scanning capabilities became operational in spring 2004 as part of the National Weather Radar Testbed (NWRT) located in Norman, Oklahoma. On 29 May 2004 the PAR observed a tornadic supercell as it traversed central Oklahoma, producing 10 tornadoes and significantly impacting Geary, OK and the greater Oklahoma City metropolitan area. This data set represents the first tornadic supercell observed using a volumetric rapid-scan 10 cm radar. The PAR radial velocity and reflectivity data are available every 20-30 seconds during a several hour period when the storm was less than 150 km from the radar. Using Ensemble Kalman Filtering (EnKF) data assimilation methodology (Dowell and Wicker 2009), the PAR observations are assimilated to generate storm-scale analyses representing the thermodynamic and kinematic structure of the tornadic supercell during a fifty-minute period. Results discuss the impacts of rapid scan data assimilation on the analyses. Comparing analyses produced from assimilating PAR data approximately every minute, with analyses produced from assimilating PAR data approximately every five minutes (i.e., more conventional weather radar observational frequencies), show that rapid scan data improves the analysis of the storm. The rapid scan data has the largest impact on the analyses early in the assimilation period. Results show that analyses produced from rapid scan PAR data yield quicker spin up times and better fit the radar observations. Results also show the difficulties of maintaining ensemble spread and the impacts on analyses. The sensitivities to the length of initialization time and microphysics parameters are discussed. This study also investigates a method that uses volumetric PAR data at five-minute intervals more than once, which produces analyses comparable to those from the rapid scan PAR data. These results suggest that the current approach used in EnKF algorithms to assimilate radar data may not be fully exploiting the available information. Finally, analyses produced using high and low time frequency PAR data are used to initialize forecasts. Results indicate that the forecast problem is not yet understood and model error leads to poor forecasts.
Back to NWC Seminar Series and Colloquia