A 4D-Var study on the potential of weather control and exigent weather forecasting

Four-dimensional variational data assimilation is a well-establishcd operational technique whereby it background estimate of the atmosphere is optimally blended with observations, Subject to the constraints of the model dynamics and the uncertainties of the information presented to the system. We extend the usual approach by applying a modified version of the Penn State/NCAR fifth-generation mesoscale model (MM5) 4D-Var to find the smallest temperature increments required to minimize the wind damage over Southern Florida during hurricane Andrew of 1992. The increments calculated by 4D-Var in this experiment created imbalances and asymmetries. As the storm resymmetrizes, at the end of the 4D-Var interval the model storm is largely weakened in situ. The amount of energy required to effect these changes is large. An alternate objective measure of the size of the increments could be formulated in terms of the likelihood of occurrence with respect to the estimated error characteristics of the model background field and the observations. A possible operational technique is presented whereby the likelihood of weather events of consequence is estimated-both subjectively and objectively.