As a first step to achieving this goal, these groups have been using an Observing System Simulation Experiment (OSSE) approach prior to working with actual radar data. For the most part, though not entirely, these OSSE studies have made a perfect model assumption in which model error does not play a role; thus, caution is advised in the interpretation of the results. Nonetheless, a consensus appears to be emerging that the use of Ensemble Kalman filter (EnKF) assimilation of both radial velocity and reflectivity observations from radars with rapid, adaptive scanning capability yields the best results in terms of minimal analysis and forecast errors.
This presentation will discuss some of the issues addressed by the NWC group in these OSSE studies, including: determination of advantages that can be gained by use of adaptive scanning and more frequent observational updating; exploration of the impact of variations in cloud microphysical parameters in the presence or absence of model error; studies of the impact of assimilated polarimetric radar data on the fidelity of storm scale analyses; optimization of the cutoff radius for covariance localization to mitigate sampling errors; inclusion of correlated errors in the error model; and tests of single vs. multiple radars in a mesoscale radar network on storm scale analyses. Recommendations for conduct of future convective-scale OSSE studies will be made.