On Statistical Forecasting of the Atmospheric Temperature for Month

Abstract

Irregular nature of the meteorological data requires obviously the probability-statistical data processing methods, but these methods didn’t find the significant application in meteorology. In this paper we investigate statistical algorithm based on Kolmogorov’s theory random sequences extrapolation for forecasting of atmospheric temperature. On the many real forecasts we compare the accuracy of statistical forecast with deterministic numerical and simple climatological forecasts when forecasting is performed for a month ahead. This analysis shows advantages and disadvantages as statistical as numerical forecasts.