Abstract
A technique is described that adds diabatic forcing from stratiform precipitation to a vertical normal-mode initialization of a mesoscale model. The technique uses observed precipitation amounts and cloud-top height estimations with analyzed thermodynamic and kinematic fields to vertically distribute diabatic heating that arises from stratiform precipitation. Simulation experiments reveal the importance of incorporating this heating into the initialization. An adiabatic initialization recovered about 65%–75% of the maximum upward vertical motions, whereas a diabatic initialization, with respect to stratiform precipitation, recovered nearly all the original vertical motions. A real-data case study is presented using combined rain gauge-satellite precipitation analyses with cloud-top heights estimated from Geostationary Operational Environmental Satellite infrared brightness temperatures. The short-term precipitation forecasts from a diabatically initialized model, with respect to stratiform precipitation, demonstrate improvement over forecasts from an adiabatically initialized model.
