Retrieval of flow and thermodynamic fields from single Doppler radar for prediction model initialization

In order to utilize data from the new US weather radar network system to initialize prediction models, it is necessary to rapidly retrieve the most important meteorological fields not directly measured by the radars--that is, cross-beam wind velocities, pressure, and temperature. New techniques for parameter retrieval from single Doppler radar use the equations of atmospheric dynamics, and combine the retrieval with data assimilation into prediction models. Adjoint and forward assimilation schemes have been developed and tested. In application to convective cloud simulations both techniques show promise, but they have not yet been applied to real data.