A simple technique for using radar data in the dynamic initialization of amesoscale model

A simple technique for using radar reflectivity to improve model initializa tion is presented. Unlike previous techniques, the scheme described here do es not infer rain rates and heating profiles from assumed relationships bet ween remotely sensed variables and precipitation rates. Rather, the radar d ata are only used to tell the model when and where deep moist convection is occurring. This information is then used to activate the model's convectiv e parameterization scheme in the grid elements where convection is observed . This approach has the advantage that the convective precipitation rates a nd heating profiles generated by the convective parameterization are compat ible with the local (grid element) environment. The premise is that if conv ection is forced to develop when and where it is observed during a data ass imilation period, convectively forced modifications to the environment wilt he in the correct locations at the model initial forecast time and the res ulting forecast will be more accurate. Three experiments illustrating how the technique is applied in the simulati on of deep convection in a warm-season environment are presented: a control run in which no radar data are assimilated, and two additional runs where radar data are assimilated for 12 h in one run and 24 h in the other. The r esults indicate that assimilating radar data can improve a model's descript ion of the mesoscale environment during the preforecast time period, thereb y resulting in an improved forecast of precipitation and the mesoscale envi ronment.