Estimation of Extreme Wind Speed for Large-Scale Synoptic Wind Based on Mean Upcrossing Rate of Mean Wind Speed Time Series Derived from Cumulative Distribution and One-Lag Autocorrelation

The estimation of extreme wind speed is critical for the structural wind-resistance design. The probability distribution of extreme wind speed under large-scale synoptic wind condition is often derived from observational wind speed data. In cases where the observation period is limited, short-term data must be utilized to accomplish this estimation. The combination of the mean upcrossing rate and Poisson assumption is popular for deriving extreme values from short-term time series data. However, its application on the wind speed is complicated by the difficulty of obtaining the mean upcrossing rate of the mean wind speed time series which is not a time-differentiable Gaussian random process. This paper addresses a general method to model the mean upcrossing rate of the mean wind speed time series concisely, utilizing the cumulative distribution function and the one-lag normalized autocorrelation function. It makes the extreme wind speed estimation from the mean wind speed time series both efficient and convenient. Utilizing long-duration mean wind speed time series data generated numerically, the effectiveness of the proposed method was verified. To evaluate the performance of the proposed method with respect to short-term mean wind speed time series, observational data encompassing a range of observation periods were systematically analyzed.