Using radar reflectivity to evaluate the vertical structure of forecast convection

Forecasts are evaluated by using concepts from object-based approaches to compare the three-dimensional simulated reflectivity field with the reflectivity field as observed by radar.

Results from analysis of both schemes reveals that forecasts generally perform well near the surface but differ considerably aloft both from observations and from each other. Forecasts are found to contain too many convective cores that are individually larger than in the observations, with at least double the number of observed convective cores reaching the midtroposphere (i.e., 4-8 km). Although the number of cores is overpredicted, WSM6 cores typically contain lower simulated reflectivity values than the observations, and the regions of highest reflectivity values do not extend far enough vertically.