Abstract : This paper systematically analyzes the influence of different observation times on the statistical results of wind speed averages and wind speeds in various wind directions by using hourly wind speed and direction observation data in 2007—2023 from the Dianbai National Benchmark Climate Station . Methods such as linear regression analysis , t-test of mean values , and correlation coefficients are adopted to evaluate the impact of different observation times on long-term wind data statistics . The results are as follows:1) The average annual and monthly wind speeds obtained from four times of observation (Ve4 , V4 ) are lower , while those from three times of observation (Ve3 , V3 ) are closer to the mean wind speeds from 24 times of observation ( V24 ) . 2 ) When different wind directions are considered , there are certain differences between the wind speed averages calculated by three times of observation (Ve3 , V3 ) and V24 in some prevailing wind directions as the result of randomness and regularity of wind direction . In four observation times , the average wind speeds of V4 have more stable results with a smaller difference from V24 , while the wind speed error in non-prevailing wind directions is larger generally. 3 ) Over the past 17 years , both the monthly and annual average wind speeds in the Dianbai region have shown a decreasing trend , with a more pronounced decrease in autumn and winter. There is no significant difference in the linear trend results of the average wind speeds calculated by five times of observation . Overall , when the historical wind speed and direction data from national meteorological stations are used , the impact of observation times on data accuracy must be fully considered .