Numerous wartime meteorological researches have contributed greatly to our understanding of the mechanism of atmospheric processes, but any resultant improvement in weather forecast accuracy to date has been highly questionable. Wartime experiences have emphasized that the usual complement of personnel in a weather station is wholly inadequate and incapable of analyzing and digesting the abundant data in the form of surface observations, rasonde-rawin soundings, pilot reports and weather reconnaissance reports now available to the meteorologists. Although meteorologists are not agreed as to its operational feasibility, the concept of divorcing analysis from forecasting was demonstrated to be, in all major operational efforts during the War, the only practicable solution to the problem of making maximum use of available data. The establishment of key analysis centers manned by expert analysts who devote their entire energies to the analysis of weather data in chart form which are transmitted by facsimile to the field forecaster has proved to be the best and only method for handling the wealth of available weather information and has resulted in a maximum efficiency in the preparation of forecasts. A principle was established that forecasting responsibility must be decentralized and given to those meteorologists immediately associated with the particular forecasting problem. The recognition and application of the principle of separation of analysis from forecasting will result in an immediate increase in forecast efficiency and accuracy and will create the only firm basis upon which private meteorology can be established.
* Paper read at the joint AMS-Inst. Aeron. Sci. meeting in New York, January, 1947.