利用阿拉善右旗1960—2015年整点风速数据，分析了日、月、年风速累积概率分布。结果表明：WeibullCum分布函数在描述风速累积概率分布时具有不同时间尺度的普适性。在多年平均逐日和逐月风速累积概率分布函数中，4个常数项存在周期性变化规律，而在1960—2015年逐年分布函数中，4个常数项分别具有周期性变化规律和减小趋势。对比2016—2020年不同时间尺度起沙风速频率统计结果和利用风速累积概率分布函数计算的结果发现，分布函数对月和年起沙风的频率具有很高的预测精度，而对日起沙风频率的预测能力不足，这是由日时间尺度上风的不确定性造成的。

Based on the hourly wind speed data of Alxa Right Banner from 1960 to 2015， the cumulative probability distribution of wind speed in different time scales （day， month and year） was statistically analyzed. The results show that WeibullCum distribution function is universally applicable in different time scales when describing the cumulative probability distribution of wind speed. In the cumulative probability distribution functions of multi-year averaged daily and monthly wind speed， four constant terms in WeibullCum function changes periodically. In the functions of yearly from 1960 to 2015， two constant terms change periodically and the other two terms change in a decreasing trend. According to the statistics of daily， monthly and yearly sand-driving wind frequencies from 2016 to 2020， and the probabilities of sand-driving wind in the same period， the distribution function has high prediction accuracy for monthly and yearly sand-driving wind occurrence frequencies， but it doesn't work for daily sand-driving wind occurrence frequencies resulted from the great uncertainty of wind in the daily time scale.