ON THE FORECASTING POSSIBILITIES OF EMPIRICAL INFLUENCE FUNCTIONS

Abstract

The possibilities of using empirically derived influence functions for forecasting the upper-air contour-height distribution are explored. Such empirical influence functions are derived for an irregular three-dimensional grid of points over North America. A series of one hundred and twenty-five independent forecasts are made for periods of one, three and five days. The results indicate that empirical influence functions can be useful in forecasting the upper-air contour-height distribution.