Forecasting hail using a one-dimensional hail growth model inline within WRF

The one-dimensional hail model developed as part of the HAILCAST algorithm was inserted into the Weather Research and Forecasting model (WRF) as part of the Air Force Weather Agency (AFWA) diagnostics package. When WRF is run at convection-allowing resolution, vertical profiles of updraft, temperature, and liquid and ice water content are passed from WRF to the 1D hail model which in turn integrates a set of Lagrangian ordinary differential equations to calculate hail size. Modifications have been made to the 1D hail model, including variable riming density, soaking, and the initialization of five frozen hail embryos of different sizes in the updraft at the first model level above the freezing level. The latter hail embryo specification replaces the former assumption of a single liquid embryo inserted at the cloud base. The five final hail sizes resulting from the assumed embryos are stored, and their mean and standard deviation are recorded in each WRF output file. The 2D fields of HAILCAST-diagnosed hail size are output to aid evaluation of the potential of forecasted deep convection to develop severe hail.

This algorithm has been implemented operationally as part of the AFWA Mesoscale Ensemble Prediction System. Hail forecasts using this algorithm were also evaluated against storm reports and the WSR-88D Maximum Expected Size of Hail (MESH) product as part of the NOAA Hazardous Weather Testbed. Sample forecast hail trajectory variables and object-based verification statistics of the hail forecast product using the MESH product will be presented.