Correction and Interpretation Technology for Wind Speed Forecast Based on Deviation Analysis

Based on the 10 m wind data of Shengsi, Dinghai, Putuo and Daishan stations of Zhoushan, Zhejiang Province, this paper analyzes the EC (European Centre for Medium-Range Weather Forecasts) 10 m wind forecast product deviation of 2019-2020. The revised model is constructed accordingly, and the data of 2021 are used to independently test the model. The research results show that: ① the absolute values of the actual every 3-h forecast errors at the stations from 0 to 33 hours are relatively stable. ②The errors of mean wind forecast at different times have a univariate linear relationship with stable coefficient. ③The relationship between the mean and extreme wind at the stations is associated the terrain shielding, wind force and underlying surface. The unitary quadratic model can reflect the relationship more accurately. ④ The average absolute errors of mean wind at the index stations are reduced to 1.16 m/s, 0.86 m/s, 0.94 m/s and 0.94 m/s respectively; the improvement percentage are 12%, 62%, 17% and 10%, respectively; and the average absolute errors of extreme wind are 2.12 m/s, 2.58 m/s, 2.69 m/s and 2.86 m/s respectively. In general, the application effect of the correction and interpretation technology is satisfactory and can be applied to operational wind speed forecast.