Stochastic–dynamical rainfall simulation based on weather radar volume scan data

The authors present a new conceptual approach to rainfall simulation for hydrologic model studies. The simulation method addresses needs of operational hydrometeorological systems and the model formulation uses the characteristics of operationally available space–time observations of rainfall. The primary state of the simulation model is the vertically integrated rainwater content derived from radar reflectivity. The simulation model components represent the formation of surface rainfall from cloud column rainwater and cloud moisture influx. The mean behavior of the model states is explicitly derived from weather radar observations, and a stochastic formulation is used to represent the behavior of the residual fluctuations. Parameter estimation using observed radar reflectivity from three convective storms in Florida is discussed along with a simulation example based on parameters derived from the observed rain events.