Statistical Single-Station Short-Term Forecasting of Temperature and Probability of Precipitation: Area Interpolation and NWP Combination

Abstract

Two statistical single-station short-term forecast schemes are introduced and applied to real-time weather prediction. A multiple regression model (R model) predicting the temperature anomaly and a multiple regression Markov model (M model) forecasting the probability of precipitation are shown. The following forecast experiments conducted for central European weather stations are analyzed: (a) The single-station performance of the statistical models, (b) a linear error minimizing combination of independent forecasts of numerical weather prediction and statistical models, and (c) the forecast representation for a region deduced by applying a suitable interpolation technique. This leads to an operational weather forecasting system for the temperature anomaly and the probability of precipitation; the statistical techniques demonstrated provide a potential for future applications in operational weather forecasts.