Use of Perfect Method to Minimum and Maximum Temperature Prediction

Convenient statistical post-processing approaches such as Perfect Prognosis Method (PPM), Model Output Statistics (MOS) and Kalman Filtering are introduced. Furthermore, by using PPM, surface minimum and maximum temperature prediction equations have been developed while measures and causes of errors have been recognized. Developing of these equations are based upon the linear regression analysis between 850-700 hPa and 700-500 hPa thickness as predictor (independent variable) and maximum temperature as predictand (response ariable), from the period 1986-1995. Obtained results express significant correlation between independent and response variables. In implementing of those regression models, thickness quantities from MM5 numerical model output have been used. But, due to the significant error of MM5 on predicting altitude of 850 hPa level over Tehran, one can not use 850-700 hPa thickness as predictor variable. By substituting 700-500 hPa thickness in computations, the difference between predicted and real values decreases highly.