A meteorological model for forecasting the wind-profile between 1.0 and 18.0 kilometers

A meteorological model for making short period wind-profile forecasts in the tropospheric and lower stratospheric regions has been formulated and extensively tested. A general purpose analog computer was used for the initial development and subsequent modifications of the model. However, the resulting digital suitability of the final form of the model permitted verification, as well as input data processing, with an IBM 7094. Testing of the model, under varying meteorological conditions, was conducted for two widely separate geographical areas centered on Bedford, Massachusetts and Fort Worth, Texas. Initial testing utilized routine upper air data and, as expected, the quality, frequency, and spatial density of these data severely affected the applicability of the model. The primary limiting factor is the restricting but necessary assumption of steady state boundary conditions based on finite differencing of local observations which are seldom spatially and temporally reqresentative. Subsequent testing showed that the model, if used in conjunction with the National Meteorological Center's three level model for the Northern Hemisphere, will give acceptable meso-scale upper air wind forecasts. It is concluded that this model or any other simplified upper air model (that is, one that does not require a dense reporting grid and inclusion of meteorological parameters other than wind) cannot, within itself, be practically useful