Field Tests and Measurement Accuracy of Automatic Weather Stations (AWS)

Based on the computed correlation coefficients and the error of the differences between the tested values from AWS and the conventional observations, the reliability of the comparative tests were demonstrated. A method of calculating the standard deviation, qualified rate and accuracy of the AWS measurement results were suggested. The hidden periodic analysis and autocorrelation model were used to compute the AWS measurement data. The results show clearly that some measurement parameters from AWS are better than the man-made observations